Well Cementing

Second Edition

Editors
Erik B. Nelson and Dominique Guillot
Schlumberger
225 Schlumberger Drive
Sugar Land, Texas 77478
www.slb.com

Technical Editor: Jennifer Smith

Production Manager and Creative Director: Martha Dutton
Production: Don Daigle, Daigrafiks

Copyright © 2006, Schlumberger, All rights reserved.

No part of this book may be reproduced, stored in a retrieval system, or transcribed in any form
or by any means, electronic or mechanical, including photocopying and recording, without the
prior written permission of the publisher. While the information presented herein is believed
to be accurate, it is provided “as is” without express or implied warranty.

06-CE-053
ISBN-13: 978-097885300-6
ISBN-10: 0-9788530-0-8
Preface
C
Contents

P-1 Primary cementing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

P-2 Remedial cementing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
1
2
P-3 Technological advances . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
P-3.1 Cement additives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
P-3.2 Temperature modeling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
P-3.3 Standardization procedures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
P-3.4 Deepwater drilling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
P-3.5 Cement sheath evaluation tools . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
P-3.6 Mechanical properties of set cement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
P-4 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
P-5 Acronym list . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
Chapter 1 Implications of Cementing for Well Production and Performance
1-1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
1-2 Zonal isolation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
1-2.1 Index of zonal isolation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
1-3 Water and gas influx . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
1-4 Impact on artificial-lift operations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
1-5 Matrix stimulation and hydraulic fracturing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
1-6 Logging . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
1-7 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
1-8 Acronym list . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
C h a p t e r 2 Chemistry and Characterization of Portland Cement
2-1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
2-2 Chemical notation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
2-3 Portland cement manufacture . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
2-3.1 Raw materials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
2-3.2 Raw material preparation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
2-3.3 Heat treatment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
2-3.4 Cooling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
2-3.5 Grinding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
2-3.6 Storage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
2-4 Hydration of the clinker phases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
2-4.1 Hydration of the silicate phases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
2-4.1.1 Preinduction period . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
2-4.1.2 Induction period . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
2-4.1.2.1 Impermeable hydrate layer theory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
2-4.1.2.2 Electrical double-layer theory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
2-4.1.2.3 Nucleation of Ca(OH)2 theory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33

Well Cementing ■ Contents iii

 2-4.1.2.4 Nucleation of C-S-H theory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
2-4.1.3 Acceleration and deceleration periods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
2-4.1.4 Diffusion period . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
2-4.2 Hydration of the aluminate phases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
2-4.2.1 Tricalcium aluminate . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
2-4.2.2 Calcium aluminoferrite . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36
2-5 Hydration of Portland cements—The multicomponent system . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36
2-5.1 Volume changes during setting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
2-5.2 Effect of temperature . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
2-5.3 Flash set and false set . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
2-5.4 Effects of aging . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
2-5.5 Influence of alkalis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
2-5.6 Influence of surface area . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
2-5.7 Sulfate resistance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40
2-6 Advanced techniques for monitoring and modeling cement hydration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40
2-6.1 FTIR spectroscopy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41
2-6.2 XRD techniques . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42
2-6.3 Computer modeling of Portland cement hydration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
2-7 Classification of Portland cements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44
2-7.1 Classification criteria . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44
2-7.2 API and ISO classification systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
2-8 Acronym list . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
C h a p t e r 3 Cement Additives and Mechanisms of Action
3-1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49
3-2 Variability of additive response . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49
3-3 Accelerators . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50
3-3.1 Examples . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50
3-3.2 Calcium chloride—mechanisms of action . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
3-3.2.1 Effects of calcium chloride on the hydration of principal Portland cement phases . . . . . . . . . . . . 51
3-3.2.2 Change in C-S-H structure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52
3-3.2.3 Diffusion of chloride ions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52
3-3.2.4 Change in aqueous phase composition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52
3-3.3 Secondary effects of calcium chloride . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52
3-3.3.1 Heat of hydration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52
3-3.3.2 Slurry rheology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
3-3.3.3 Compressive strength development . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
3-3.3.4 Shrinkage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
3-3.3.5 Permeability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
3-3.3.6 Sulfate resistance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
3-3.4 Chloride-free accelerators . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
3-4 Retarders . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54
3-4.1 Lignosulfonates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54
3-4.2 Hydroxycarboxylic acids . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55
3-4.3 Saccharide compounds . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56
3-4.4 Cellulose derivatives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56
3-4.5 Organophosphonates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56
3-4.6 Inorganic compounds . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58
3-5 Extenders . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58
3-5.1 Clays . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59

iv Well Cementing
 3-5.2 Sodium silicates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60
3-5.3 Pozzolans . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61
3-5.3.1 Diatomaceous earth . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62
3-5.3.2 Fly ashes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63
3-5.3.3 Commercial lightweight cements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63
3-5.3.4 Silica . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64
3-5.4 Lightweight particles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64
3-5.4.1 Expanded perlite . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64
3-5.4.2 Gilsonite . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64
3-5.4.3 Powdered coal . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66
3-5.4.4 Microspheres . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67
3-5.4.5 Nitrogen . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69
3-6 Weighting agents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69
3-6.1 Ilmenite . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70
3-6.2 Hematite . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70
3-6.3 Barite . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70
3-6.4 Manganese tetraoxide . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70
3-7 Dispersants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71
3-7.1 Chemical composition of cement dispersants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71
3-7.2 Effect of dispersants on the rheology of cement slurries . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73
3-7.3 Mechanisms of action . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74
3-7.4 Factors affecting the response of cements to dispersants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77
3-7.5 Particle settling and free water . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78
3-7.6 Prevention of free water and slurry sedimentation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79
3-8 Fluid-loss control agents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80
3-8.1 Particulate materials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80
3-8.2 Water-soluble polymers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81
3-8.2.1 Natural polymers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83
3-8.2.1.1 Cellulose derivatives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83
3-8.2.1.2 Galactomannans . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84
3-8.2.2 Synthetic polymers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84
3-8.2.2.1 Nonionic synthetic polymers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84
3-8.2.2.2 Anionic synthetic polymers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85
3-8.2.2.3 Cationic synthetic polymers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87
3-9 Lost-circulation prevention agents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88
3-9.1 Bridging materials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88
3-9.2 Thixotropic cements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89
3-10 Miscellaneous cement additives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89
3-10.1 Antifoam agents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89
3-10.2 Strengthening agents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89
3-10.3 Radioactive tracing agents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90
3-10.4 Mud decontaminants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90
3-11 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90
3-12 Acronym list . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90
Chapter 4 Rheology and Flow of Well Cement Slurries
4-1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93
4-2 Basic rheological principles and calculations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93
4-2.1 Types of flow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93
Laminar flow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93

Contents v
 Turbulent flow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94
4-2.2 Basic definitions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94
4-2.3 Rheological models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95
Newtonian fluids . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95
Non-Newtonian fluids . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96
Power-law fluids . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96
Bingham plastic fluids . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97
Herschel-Bulkley fluids . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97
Other rheological models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98
4-2.4 Shear-rate range encountered in a wellbore . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98
4-2.5 Laboratory determination of rheological properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100
Pressurized equipment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101
4-2.5.1 Experimental procedure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101
4-2.5.2 Gel strength . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101
4-2.5.3 Data analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102
Example . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102
Comments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103
4-3 Limitations of field practices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 104
4-3.1 Principle and flow equations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 104
4-3.2 Flow of model fluids in coaxial cylinder viscometers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105
4-3.3 Application to standard oilfield equipment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 106
4-3.4 Newtonian and narrow-gap approximations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 107
4-3.5 What about the rheological model? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108
4-3.6 End effects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 109
4-3.7 Homogeneity of the cement slurry and wall slip . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 110
4-3.8 Particle migration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 112
4-3.9 Testing slurries containing large particles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113
4-3.10 Foamed cement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113
4-3.11 Cement rheology measured with different equipment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 114
4-4 Pressure, temperature, and time dependency . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115
4-4.1 Temperature dependency . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115
4-4.2 Pressure dependency . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 117
4-4.3 Time dependency under shear . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 117
4-4.4 Gel strength measurements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 118
4-5 Flow of spacers and cement slurries in the wellbore . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 119
4-5.1 Generic laminar flow equations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 119
4-5.1.1 Pipe flow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 120
4-5.2 Narrow concentric annular flow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 124
4-5.3 Concentric annular flow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127
4-5.4 Circulation efficiency . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 129
4-6 Friction pressure calculations for all flow regimes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 130
4-6.1 Pipe flow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 130
End of laminar flow regime . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 132
End of the transition regime . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 133
Example of application to a power-law fluid . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 134
Example of application to nonpower-law fluids . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 134
4-6.2 Concentric annular flow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 135
Example of application to a power-law fluid . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 135
Example of application to nonpower-law fluids . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 136
4-6.3 Effect of pipe eccentricity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 137

vi Well Cementing
4-7 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 142
4-8 Acronym list . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 142
Chapter 5 Mud Removal
5-1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 143
5-2 Well preparation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 145
5-2.1 Borehole quality . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 145
5-2.2 Circulation and displacement efficiency . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 146
5-2.3 Mud conditioning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 148
5-2.4 Modeling mud circulation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 148
5-2.4.1 Laminar flow in a concentric annulus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 149
5-2.4.2 Turbulent flow in a concentric annulus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 151
5-2.4.3 Influence of string eccentricity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 151
5-2.4.4 Effect of casing movement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 155
5-2.4.5 Discussion of theoretical models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 158
5-2.4.6 Removing gelled mud and mud filtercake during the circulation process . . . . . . . . . . . . . . . . . . . . . . 158
5-2.4.6.1 Gelled mud . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 159
5-2.4.6.2 Mud filtercake . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 159
5-2.4.6.3 Solids beds . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 160
5-2.4.6.4 Removal of mud filtercakes and solids beds . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 161
5-2.4.7 Measuring mud-circulation efficiency in the field . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 161
5-2.5 Mud circulation—conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 163
5-3 Mud displacement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 163
5-3.1 Field practices and observations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 164
5-3.2 Problem description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 164
5-3.3 Laminar displacement of mobile mud in concentric vertical annuli . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 165
5-3.3.1 Experimental validation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 168
5-3.4 Modeling of mud removal in eccentric annuli . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 169
5-3.5 Laminar mud displacement criteria . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 174
5-3.6 Turbulent flow and erosion mechanisms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 175
5-3.7 Experimental investigations of immobile mud removal . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 179
5-3.8 Field studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 182
5-3.9 Effect of casing movement and casing hardware . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 182
5-3.10 Mud displacement in other oilfield applications—Summary and conclusions . . . . . . . . . . . . . . . . . . . . . . . . 183
5-4 Drilling fluids, spacers, and washes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 183
5-4.1 Drilling fluids . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 184
5-4.2 Preflushes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 184
5-4.2.1 Washes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 184
5-4.2.2 Spacers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 185
5-5 Cleaning-fluid sequence—current practices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 187
5-6 Other cement-placement problems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 187
5-7 Qualitative recommendations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 188
5-8 Acronym list . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 189
C h a p t e r 6 Cement-Formation Interactions
6-1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 191
6-2 Fluid loss . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 191
6-2.1 Static filtration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 192
6-2.2 Dynamic filtration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 193
6-2.3 Fluid loss in primary cementing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 194
6-2.3.1 Fluid-loss criteria . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 194

Contents vii
 6-2.3.2 Effects of fluid loss on cement-slurry properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 195
Slurry volume . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 196
Slurry density . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 196
Hydrostatic pressure increase . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 197
Slurry thickening time . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 197
Slurry viscosification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 197
Dynamic placement pressure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 198
Concluding remarks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 198
6-2.3.3 Annular restriction and bridging . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 198
6-2.3.4 Mud removal . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 199
6-2.3.5 Formation damage and formation invasion depth . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 199
6-2.3.6 Pressure transmission during WOC . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 200
6-2.3.7 Comparison between various phenomena and conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 201
6-2.4 Fluid loss during remedial cementing operations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 201
6-2.5 Field measurements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 201
6-3 Lost circulation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 202
6-3.1 Consequences of lost circulation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 202
6-3.2 Classification of the severity of losses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 203
6-3.2.1 Seepage losses (1–10 bbl/hr [1.5 m3 to 15 m3/hr]) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 203
6-3.2.2 Partial losses (10–100 bbl/hr) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 203
6-3.2.3 Severe losses (100–500 bbl/hr) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 203
6-3.2.4 Total losses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 203
6-3.3 Classification of lost circulation zones . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 203
6-3.3.1 Unconsolidated formations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 204
6-3.3.2 Highly permeable or depleted formations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 204
6-3.3.3 Natural fractures or fissures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 204
6-3.3.4 Induced fractures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 204
6-3.3.5 Cavernous and vugular formations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 204
6-3.4 Lost circulation while drilling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 204
6-3.4.1 Bridging agents in the drilling fluid . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 206
Granular LCMs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 206
Fibrous LCMs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 207
Lamellar LCMs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 207
Mixed LCMs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 207
6-3.4.2 Surface-mixed systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 209
6-3.4.2.1 Cement plugs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 209
6-3.4.2.2 High-fluid-loss squeezes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 209
6-3.4.2.3 Polymer-base crosslinked systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 209
6-3.4.2.4 Other surface-mixed systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 210
6-3.4.3 Downhole-mixed systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 212
6-3.5 Methods for controlling massive lost circulation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 214
6-3.5.1 “Floating mudcap” method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 214
6-3.5.2 Drilling “blind” . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 214
6-3.5.3 Using specialized drilling fluids . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 214
6-3.5.3.1 Foamed and aerated mud systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 214
6-3.5.3.2 Mixed metal hydroxide and mixed metal oxide systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 214
6-3.6 Lost circulation during cementing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 215
6-3.6.1 Downhole pressure reduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 215
6-3.6.2 Preflushes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 216
6-3.6.3 LCMs for cement slurries . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 216

viii Well Cementing

 6-3.6.4 Thixotropic cement systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 217
6-3.6.5 Foamed cement systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 217
6-3.7 Lost circulation design software . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 217
6-3.8 Lost circulation—conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 218
6-4 Cement-formation bonding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 219
6-4.1 Cement-to-casing bonding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 220
6-4.2 Cement-to-formation bonding (sandstone and limestone) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 222
6-4.3 Properties of clays . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 224
6-4.4 Interactions between drilling fluids and shales . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 226
6-4.5 Interactions between cement slurries and shales . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 227
6-4.6 Bonding between cements and shales treated with various drilling fluids . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 229
6-4.7 Cement-to-formation bonding studies performed outside of the petroleum industry . . . . . . . . . . . . . . . . . . 230
6-4.8 Cement-to-formation bonding conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 231
6-5 Cementing across evaporite zones . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 231
6-6 Gas hydrates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 232
6-7 Acronym list . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 232
C h a p t e r 7 Special Cement Systems
7-1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 233
7-2 Thixotropic cements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 233
7-2.1 Clay-base systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 234
7-2.2 Calcium sulfate-based systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 234
7-2.3 Aluminum sulfate/iron (II) sulfate system . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 235
7-2.4 Crosslinked cellulose polymer systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 235
7-3 Expansive cement systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 235
7-3.1 Ettringite systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 236
7-3.2 Salt cements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 236
7-3.3 Aluminum powder . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 237
7-3.4 Calcined magnesium oxide . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 237
7-4 Freeze-protected cements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 237
7-5 Salt cement systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 239
7-5.1 Salty water as mixing fluid . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 239
7-5.2 Salt as a cement additive . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 240
7-5.3 Cementing across shale and bentonitic clay formations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 240
7-5.4 Cementing across massive salt formations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 240
7-6 Latex-modified cement systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 242
7-6.1 Behavior of latexes in well cement slurries . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 243
7-6.2 Early latex-modified well cement systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 243
7-6.3 Styrene butadiene latex systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 243
7-7 Cements for corrosive environments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 244
7-7.1 Cements for chemical waste disposal wells . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 244
7-7.2 Cements for enhanced oil recovery by CO2 flooding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 245
7-8 BFS systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 245
7-8.1 Production, composition, and hydration of BFS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 245
7-8.2 BFS-Portland cement blends . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 246
7-8.2.1 Squeeze cementing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 247
7-8.2.2 Salt-zone cementing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 247
7-8.2.3 CO2 resistant cements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 247
7-8.2.4 High-pressure, high-temperature gas wells . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 247

Contents ix
 7-8.2.5 Foamed cements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 248
7-8.2.6 Steam injection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 248
7-8.2.7 Treatment of construction cements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 249
7-8.3 Cementitious drilling fluids . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 250
7-8.3.1 Mud-to-cement process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 250
7-8.3.2 Mud-to-cement performance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 250
7-9 Engineered particle-size distribution cements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 251
7-9.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 251
7-9.2 Engineered particle-size concept . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 251
7-9.3 EPS slurry design . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 252
7-9.3.1 Slurry density . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 253
7-9.3.2 Slurry rheology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 253
7-9.3.3 Slurry stability: free water and segregation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 253
7-9.3.4 Additives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 254
7-9.4 Performance of EPS cement systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 254
7-9.4.1 Low-density systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 254
7-9.4.2 High-density systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 255
7-10 Ultralow-density cement systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 256
7-10.1 Microsphere and EPS systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 256
7-10.2 Foamed cements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 256
7-10.2.1 Foam stability and structure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 256
7-10.2.2 Foamed cement design . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 258
7-10.2.2.1 Cement, foaming agents, stabilizers, and additives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 258
7-10.2.2.2 Foamed cement properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 258
7-11 Flexible cements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 262
7-11.1 Slurry density reduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 262
7-11.2 Flexible particles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 262
7-11.3 Elastomeric composites . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 264
7-11.4 Fibers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 264
7-12 Microfine cements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 265
7-13 Acid-soluble cements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 265
7-14 Chemically bonded phosphate ceramics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 265
7-15 Storable cement slurries . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 266
7-16 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 266
7-17 Acronym list . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 268
Chapter 8 Mechanical Properties of Well Cements
8-1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 269
8-2 Basic concepts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 269
8-2.1 Stress . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 269
8-2.2 Strain . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 271
8-3. Cement behavior . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 271
8-3.1 Stress-strain curve . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 271
8-3.2 Linear elasticity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 272
8-3.3 Thermoelasticity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 273
8-3.4 Poroelasticity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 274
8-3.5 Plastic behavior . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 276
8-3.6 Creep of cement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 276
8-3.7 Set-cement strength and set-cement failure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 277
8-3.8 Influence of confining pressure on set-cement behavior and failure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 278

x Well Cementing
 8-3.9 Postfailure behavior of set cement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 278
8-3.10 Shrinkage and expansion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 279
8-4 Mechanical behavior of a cement cased wellbore . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 280
8-4.1 State of stress in the cement sheath . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 280
8-4.2 Modeling the cement sheath using thermoelasticity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 280
8-4.3 Influence of wellbore pressure increase . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 282
8-4.4 Influence of wellbore temperature increase . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 283
8-4.5 Influence of wellbore pressure and temperature decrease . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 284
8-4.6 Casing support, perforation and fracturing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 285
8-4.7 Tectonics, subsidence, and formation creep . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 285
8-5 Guidelines for cement design . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 287
8-6 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 288
8-7 Acronym list . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 288
Chapter 9 Annular Formation Fluid Migration
9-1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 289
9-2 Practical consequences of gas migration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 289
9-3 Physical process of gas migration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 290
9-3.1 Root causes for gas migration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 290
9-3.2 Gas migration categories . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 290
9-3.2.1 Immediate gas migration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 291
9-3.2.2 Short-term gas migration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 292
9-3.2.3 Long-term gas migration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 292
9-3.3 Factors affecting gas migration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 292
9-3.3.1 Fluid loss . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 292
9-3.3.2 Gel strength development . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 294
9-3.3.3 Cement shrinkage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 296
9-3.3.4 Permeability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 298
9-3.3.5 Free fluid (free water) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 298
9-3.3.6 Mud removal . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 299
9-3.3.7 Microannulus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 299
9-3.3.8 Cement sheath mechanical failure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 299
9-4 Predicting short-term gas migration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 299
9-4.1 Gas flow potential . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 300
9-4.2 Expanding the factors for gas flow potential . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 300
9-4.2.1 Formation factor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 300
9-4.2.2 Hydrostatic factor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 301
9-4.2.3 Mud removal factor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 301
9-4.2.4 Slurry performance number . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 301
9-4.3 Recent developments in gas migration prediction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 302
9-4.3.1 Formation parameter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 302
9-4.3.2 Mud removal parameter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 303
9-4.3.3 Pressure decay limit parameter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 303
9-5 Theoretical strategies for combating short-term gas migration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 304
9-6 Practical solutions for combating gas migration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 305
9-6.1 Physical techniques . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 305
9-6.2 Compressible cements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 307
9-6.3 Low-permeability cement slurries . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 308
9-6.4 Fluid loss and free-fluid control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 309
9-6.5 Thixotropic cement slurries . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 309

Contents xi
 9-6.6 Surfactants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 309
9-6.7 Right-angle-set cements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 309
9-6.8 Expanding cements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 310
9-6.9 Flexible cements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 310
9-7 Laboratory testing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 310
9-7.1 Large-scale testing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 311
9-7.2 Small-scale gas migration testers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 311
9-7.3 Gel strength testing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 315
9-7.3.1 Pressure drop tubes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 315
9-7.3.2 Vane rheometers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 315
9-7.3.3 Rotating-paddle devices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 315
9-7.3.4 Acoustic devices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 316
9-8 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 316
9-9 Acronym list . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 317
C h a p t e r 1 0 Thermal Cements
10-1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 319
10-2 High-temperature chemistry of Portland cement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 319
10-3 High-alumina cement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 322
10-4 Class J cement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 322
10-5 Calcium aluminosilicate systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 323
10-6 Calcium phosphate systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 324
10-7 Deep oil and gas wells . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 325
10-7.1 Thickening time and initial compressive strength development . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 325
10-7.2 Cement-slurry rheology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 326
10-7.3 Cement-slurry density . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 326
10-7.4 Fluid-loss control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 326
10-7.5 Long-term performance of cements for deep wells . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 326
10-8 Geothermal well cementing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 328
10-8.1 Well conditions associated with geothermal wells . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 329
10-8.2 Performance requirements and design considerations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 330
10-8.2.1 Portland cement–based geothermal well cement compositions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 330
10-8.2.2 Alternate geothermal well cement compositions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 334
10-9 Thermal recovery wells . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 334
10-9.1 Steam recovery wells . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 335
10-9.2 In situ combustion wells . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 338
10-10 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 340
10-11 Acronym list . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 341
C h a p t e r 1 1 Cementing Equipment and Casing Hardware
11-1 Cementing materials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 343
11-1.1 Cement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 343
11-1.2 Water . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 343
11-1.3 Dry cement additives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 343
11-1.4 Liquid cement additives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 344
11-2 Basic equipment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 344
11-2.1 Cement and dry additive blending . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 344
11-2.2 Transportation of bulk materials or blends to the wellsite . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 346
11-2.3 Wellsite storage of cement or blends . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 347
11-2.4 Metering of water . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 348
11-2.5 Liquid additive storage and mixing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 349

xii Well Cementing

 11-2.5.1 Liquid additive metering system with metering tanks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 349
11-2.5.2 Liquid-additive metering system without metering tanks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 350
11-2.6 Surge tanks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 350
11-2.7 Cement mixing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 351
11-2.7.1 Conventional jet mixer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 351
11-2.7.2 Recirculation jet mixer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 352
11-2.7.3 Recirculation mixer without conventional jets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 352
11-2.7.4 Cement mixing units . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 354
11-2.8 High-pressure pumps . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 355
11-2.8.1 Convertibility . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 355
11-2.8.2 Hydraulic horsepower . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 355
11-2.8.3 Versatility . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 355
11-2.8.4 Maximum flow rate and pressure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 356
11-2.8.5 Drive . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 356
11-2.9 Controls and instruments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 356
11-2.10 Steel flowhoses and cement head . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 358
11-3 Cementing units . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 359
11-4 Introduction to casing hardware . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 362
11-5 Casing hardware . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 362
11-5.1 Guide shoes and float shoes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 362
11-5.1.1 Nose materials and shapes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 364
11-5.2 Float equipment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 365
11-5.3 Valve materials and types . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 368
11-5.4 Surge pressure and float equipment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 370
11-5.5 Automatic fill-up equipment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 371
Casing insert equipment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 372
Large-bore autofill float equipment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 372
Other autofill float equipment valves . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 374
11-5.6 Differential fill equipment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 374
11-5.7 Inner string cementing equipment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 375
Stab-in . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 375
Screw-in . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 376
Latch-in . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 377
11-5.8 Shoe and collar shell design . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 377
11-5.9 Cementing plugs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 378
11-5.10 Stage equipment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 381
Liner stage tool options . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 386
11-5.11 Annular packoff equipment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 387
11-5.12 Cement displacement enhancement tools . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 389
Torque- and drag-reduction centralizers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 394
Other casing-mounted cementing devices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 395
11-6 Liners . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 397
11-6.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 397
11-6.2 Liner types . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 397
11-6.3 Reasons for installing a liner . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 398
11-6.4 Principal components of a liner . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 398
11-6.4.1 Surface cementing equipment (Fig. 11-91[1]) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 399
11-6.4.2 Liner setting tool assembly (Fig. 11-91[2]) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 400
11-6.4.3 Liner setting sleeve/liner top packer assembly (Fig. 11-91[3]) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 403
11-6.4.4 Liner hanger . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 405

Contents xiii
 11-6.4.4.1 Predicting liner hanger capacities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 407
11-6.4.4.2 Liner hanger types . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 408
11-6.4.5 Shoe track assembly . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 415
11-6.5 Displacement plugs for liner applications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 416
11-6.6 Liner selection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 419
11-7 Remedial cementing tools . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 421
11-7.1 Cased hole squeeze packers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 421
11-7.2 Cased hole bridge plugs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 424
11-7.3 Tubing testers and unloaders . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 426
11-7.4 Cementing using inflatable service tools . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 427
11-7.5 Inflatable service packers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 427
11-7.6 Circulating valve . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 428
11-7.7 Inflatable bridge plugs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 428
11-7.8 Through-tubing electric wireline setting tool . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 429
11-7.9 Inflatable cement retainer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 429
11-7.10 Software for inflatable tools . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 430
11-8 Through-tubing inflatable products . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 431
11-8.1 Through-tubing inflatable retrievable packer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 431
11-8.2 Through-tubing inflatable retrievable bridge plug . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 432
11-8.3 Through-tubing inflatable permanent bridge plug . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 432
11-8.4 Through-tubing inflatable permanent cement retainer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 433
11-8.5 Through-tubing inflate design software . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 433
11-9 Acronym list . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 433
11-10 Suggested reading . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 433
C h a p t e r 1 2 Primary Cement Job Design
12-1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 435
12-2 Problem analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 435
12-2.1 Depth and dimensional data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 435
12-2.2 Wellbore environment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 440
12-2.3 Temperature regime . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 441
12-3 Slurry selection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 443
12-3.1 Slurry density . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 443
12-3.2 Compressive strength and mechanical properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 443
12-3.3 Well temperature . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 444
12-3.4 Cement additives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 444
12-3.5 Cement-slurry design . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 444
12-4 Placement mechanics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 445
12-4.1 Spacers and chemical washes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 445
12-4.2 Casing centralization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 445
12-5 Well control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 446
12-6 U-tubing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 448
12-7 Example of job design procedure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 448
12-7.1 Well conditions and cement job considerations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 448
12-7.2 Slurry formulation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 452
12-7.3 Mud removal strategy and spacer formulation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 452
12-7.4 Job design . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 452
12-8 Preparing for the job . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 457
12-9 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 457
12-10 Acronym list . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 458

xiv Well Cementing

Chapter 13 Primary Cementing Techniques
13-1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 459
13-2 Classification of casing strings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 459
13-2.1 Conductor pipe . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 460
13-2.2 Surface casing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 461
13-2.3 Intermediate casing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 463
13-2.4 Production casing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 465
13-2.5 Combination casing strings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 465
13-3 Cement placement procedures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 466
13-3.1 Cementing through drillpipe (stab-in or inner-string cementing) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 466
13-3.2 Grouting (top-up cementing) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 468
13-3.3 Single-stage cementing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 469
13-3.3.1 Mud conditioning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 469
13-3.3.2 Bottom wiper plug . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 469
13-3.3.3 Displacement procedures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 469
13-3.4 Multistage cementing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 470
13-3.4.1 Conventional two-stage cementing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 472
13-3.4.1.1 Cementing the first stage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 472
13-3.4.1.2 Cementing the second stage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 472
13-3.4.2 Continuous two-stage cementing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 473
13-3.4.3 Three-stage cementing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 474
13-4 Liners . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 475
13-4.1 Running a liner . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 475
13-4.2 Liner cementing procedure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 476
13-4.2.1 Mud removal . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 476
13-4.2.2 Regular liner cementing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 477
13-4.2.3 Planned squeeze . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 482
13-4.2.4 WOC for liners . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 482
13-4.2.5 Tieback liners . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 482
Tieback sleeve . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 483
Tieback sealing nipple . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 483
13-5 Special offshore techniques . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 484
13-5.1 Conventional subsea plug system . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 484
13-5.2 Innovative system for launching plugs with subsea wellheads . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 486
13-5.2.1 Dart launcher . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 486
13-5.2.2 Subsea tool . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 486
13-5.3 Deepwater cementing considerations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 487
13-6 Cementing techniques and new enablers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 487
13-6.1 Extended-reach wells . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 487
13-6.2 Solid expandable tubulars . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 489
13-6.3 Multilaterals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 491
13-6.4 Casing drilling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 493
13-7 Operational considerations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 495
13-7.1 Calculations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 495
13-7.1.1 Slurry volume . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 495
13-7.1.2 Displacement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 495
13-7.2 Hole condition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 495
13-7.3 Temperature . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 496
13-7.4 Pressure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 497
13-7.5 Quality control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 497

Contents xv
 13-7.6 Casing movement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 497
13-7.7 Cement job monitoring . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 498
13-7.8 Casing operational sequence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 498
13-7.8.1 Stationary rig (e.g., land rig, jackup rig) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 498
13-7.8.2 Floating rig (e.g., semisubmersible, drillship) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 500
13-8 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 500
13-9 Acronym list . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 501
C h a p t e r 1 4 Remedial Cementing
14-1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 503
14-1.1 Overview of remedial cementing operations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 503
14-1.2 Methodology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 504
14-2 Problem identification—Well problems cured by remedial cementing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 505
14-2.1 Well construction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 505
14-2.1.1 Lost circulation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 505
14-2.1.2 Sidetrack and directional drilling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 506
14-2.1.3 Protective plugs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 507
14-2.1.4 Well construction evaluation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 507
14-2.1.5 Circulating squeeze . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 508
14-2.1.6 Block squeeze . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 508
14-2.1.7 Top of liner squeeze . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 508
14-2.2 Production/injection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 509
14-2.2.1 Before stimulation treatments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 509
14-2.2.2 Sustained casing pressure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 509
14-2.2.3 Casing and packer leaks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 509
14-2.2.4 Production and injection control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 510
14-3 Plug cementing—Tools and techniques . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 512
14-3.1 Balanced plug . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 513
14-3.2 Dump bailer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 513
14-3.3 Two-plug method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 514
14-3.4 Inflatable packers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 514
14-3.5 Umbrella-shaped membranes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 514
14-3.6 Inside blowout preventer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 515
14-3.7 Coiled tubing placement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 515
14-4 Plug cementing—Placement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 515
14-4.1 Plug location . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 515
14-4.2 Contamination . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 516
14-4.2.1 Mechanical devices to prevent or minimize contamination . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 516
Bridge plugs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 516
Tailpipe or stinger . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 517
Diverter tool . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 517
Mechanical plugs, darts, and balls . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 517
14-4.2.2 Interface stability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 517
14-4.2.3 Displacement volume for a balanced plug . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 518
14-4.2.4 Inadequate mud removal . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 520
14-4.2.5 Pipe removal and reversing out . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 520
14-4.3 Cleanup . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 520
14-4.4 Pressure and mechanical-strength constraints . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 520
14-5 Plug cementing—Cement-slurry design . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 520
14-6 Cement-plug evaluation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 521

xvi Well Cementing

14-7 Squeeze cementing—Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 521
14-8 Squeeze cementing—Perforation plugging theory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 522
14-9 Squeeze cementing tools and techniques . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 525
14-9.1 Low-pressure squeeze . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 527
14-9.2 High-pressure squeeze . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 527
14-9.3 Running squeeze method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 528
14-9.4 Hesitation squeeze method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 528
14-9.5 Bradenhead squeeze (no packer) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 529
14-9.6 Squeeze-tool placement technique . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 529
14-9.6.1 Retrievable squeeze packer method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 529
14-9.6.2 Drillable cement retainer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 530
14-9.7 Perforation washing tool . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 530
14-9.8 Squeeze with coiled tubing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 531
14-10 Squeeze cementing—Cement-slurry design . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 532
14-10.1 Slurry selection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 533
14-10.2 Fluid-loss control and filtercake development . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 533
14-10.3 Rheology and sedimentation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 533
14-10.4 Density . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 533
14-10.5 Size of particles in the slurry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 534
14-10.6 Thickening time and temperature . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 534
14-10.7 Chemical resistance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 534
14-10.8 Economics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 535
14-11 Job design . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 535
14-11.1 Volume selection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 535
14-11.2 Spacers and washes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 535
14-11.3 Displacement volume for a squeeze operation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 536
14-11.4 Pressure and mechanical strength checks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 536
14-12 Injection test . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 537
14-12.1 Determination of injection pressure and rate . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 537
14-12.2 Determination of void volume behind casing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 537
14-12.3 Determining the nature of flow paths behind casing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 538
14-13 Basic squeeze-job procedures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 538
14-14 Monitoring squeeze jobs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 540
14-15 Cleanup after a squeeze job . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 540
14-16 Squeeze job evaluation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 541
14-16.1 Temperature log . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 541
14-16.2 Positive and negative pressure tests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 541
14-16.3 Acoustic logs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 542
14-16.4 Production changes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 542
14-16.5 Cement hardness . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 542
14-16.6 Radioactive tracers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 542
14-17 Reasons for squeeze cementing failures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 542
14-17.1 Improper slurry selection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 542
14-17.2 Excessive final squeeze pressure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 542
14-17.3 Plugged perforations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 542
14-17.4 Improper packer location . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 543
14-18 Noncementing solutions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 543
14-19 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 543
14-20 Acronym list . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 543

Contents xvii
C h a p t e r 1 5 Cement Job Evaluation
15-1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 549
15-2 Hydraulic testing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 549
15-2.1 Pressure testing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 549
15-2.2 Inflow testing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 551
15-2.3 Tests through perforations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 551
15-3 Temperature, nuclear, and noise-logging measurements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 551
15-3.1 Temperature logging . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 551
15-3.1.1 Cement hydration detector . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 551
15-3.1.2 Communication indicator (pump-in temperature log) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 552
15-3.2 Nuclear logging . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 554
15-3.3 Boron-pulsed neutron logs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 554
15-3.4 Oxygen activation logs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 554
15-3.5 Noise logging . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 555
15-4 Acoustic logging measurements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 555
15-4.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 555
15-4.2 Quality control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 555
15-4.2.1 Measurement repeatability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 555
15-4.2.2 Calibration summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 556
15-4.3 Acoustic properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 556
15-4.3.1 Definitions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 556
15-4.3.2 Acoustic properties of formations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 557
15-4.3.3 Acoustic properties of cements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 558
15-4.3.4 Acoustic properties of fluids . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 559
15-4.4 Sonic cement evaluation techniques . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 561
15-4.4.1 History . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 561
15-4.4.2 The CBL . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 562
15-4.4.2.1 Description of the conventional bond-logging tool . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 562
15-4.4.2.2 Acoustics of the bond-log measurement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 562
15-4.4.2.3 Description of the full acoustic wave display: VDL . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 563
15-4.4.2.4 CBL/VDL: Qualitative interpretation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 566
15-4.4.2.5 Quantitative data from the acoustic wave . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 567
15-4.4.2.6 CBL attenuation rate . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 568
15-4.4.2.7 CBL bond index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 569
15-4.4.2.8 Bond-log presentation formats . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 572
15-4.4.2.9 Quality control of CBL/VDL logs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 574
15-4.4.2.10 Influence of well parameters on the CBL . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 576
15-4.4.2.11 Influence of cement job parameters on the CBL . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 578
15-4.4.2.12 Influence of post-job events on the CBL . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 578
Thermal expansion or contraction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 579
Mechanical expansion or contraction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 579
Mechanical fatigue . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 579
15-4.4.2.13 CBL/VDL examples . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 579
Limitations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 583
15-4.4.2.14 Pad-type sonic tools for cement evaluation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 583
15-4.5 Ultrasonic pulse-echo cement evaluation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 585
15-4.5.1 History . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 585
15-4.5.2 Advantages and disadvantages of ultrasonic tools . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 585
15-4.5.3 Tools . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 585
15-4.5.4 Principle of ultrasonic cement evaluation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 586

xviii Well Cementing

 15-4.5.5 Measurements and processing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 586
15-4.5.6 Response to materials in the annulus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 589
Gas . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 589
Liquid . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 589
High-quality cement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 589
Cement with channels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 590
Microannulus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 590
Mud layer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 592
Thin cement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 592
15-4.5.7 Other factors affecting tool response . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 593
Casing shape . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 593
Tool eccentering . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 593
Rugose casings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 593
Heavy muds . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 593
Pipe coating . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 593
Casing thickness . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 593
15-4.5.8 Basic interpretation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 593
15-4.5.8.1 Impedance thresholds . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 593
15-4.5.8.2 Statistical approaches . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 594
15-4.5.9 Log presentation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 595
Borehole fluid data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 595
Quality control and auxiliary data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 595
Casing data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 597
Cement/annular data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 597
CBL/VDL data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 597
Openhole and cementing data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 597
15-4.6 Quality control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 597
15-4.6.1 Combined interpretation examples . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 598
Good cement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 600
Mud channel . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 601
Cement top . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 602
Low-density cement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 602
Mud layer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 604
Wet microannulus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 605
Gas channel and microannulus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 605
Various materials with microdebonding processing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 606
Patchy bond (microdebonding) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 607
Gas-affected low-density cement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 608
Casing shape . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 608
Deposits . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 609
Reflections from an outer casing string . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 610
15-5 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 612
15-6 Acronym list . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 612
Appendix A Digest of Rheological Equations
Table A-1. Definitions of the Main Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 613
Table A-2. Main Equations for Laminar Flow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 614
Table A-3. Definition of the Local Power-Law Parameters and Friction Factor/Reynolds Number Relationships . . . . 615
Table A-4. Newtonian Fluid Relationships . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 616
Table A-5. Power-Law Fluid Relationships . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 617

Contents xix
Table A-6a. Bingham Plastic Fluid Relationships . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 619
Table A-6b. Approximations for Bingham Plastic Fluids when ψ << 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 620
Table A-7. Herschel-Bulkley Fluid Relationships . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 621
Table A-8. Dimensionless Breakthrough Time and Circulation Efficiency in Laminar Flow . . . . . . . . . . . . . . . . . . . . . . . . 622
A-1 Application of the local-power approach to Bingham plastic fluids . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 623
A-2 Conversion factors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 625
Appendix B Laboratory Testing, Evaluation, and Analysis of Well Cements
B-1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 627
B-2 Relationship between the API and the ISO . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 627
B-3 Sample preparation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 628
B-4 Performance of conventional cement slurries . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 628
B-4.1 Slurry preparation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 628
B-4.2 Slurry density . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 630
B-4.3 Thickening time . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 630
Primary cementing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 632
Squeeze cementing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 634
B-4.4 Fluid loss . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 635
B-4.5 Strength . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 636
B-4.5.1 Compressive strength . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 636
B-4.5.2 Sonic strength . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 637
B-4.6 Free fluid and slurry sedimentation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 638
B-4.7 Rheological measurements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 639
B-4.8 Static gel strength . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 640
B-4.9 Expansion and shrinkage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 641
B-4.10 Gas migration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 643
B-4.11 Permeability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 643
B-5 Spacers and chemical washes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 644
B-5.1 Compatibility . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 644
B-5.2 Cleaning effect . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 644
B-6 Mechanical properties testing of well cements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 646
B-6.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 646
B-6.2 Sample preparation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 646
B-6.3 Cement static mechanical properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 647
B-6.3.1 Uniaxial unconfined compression test . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 647
B-6.3.2 Confined triaxial compression test . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 650
B-6.3.3 Brazilian test . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 650
B-6.4 Dynamic mechanical properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 651
B-6.4.1 Dynamic compressive strength . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 651
B-6.4.2 Dynamic elastic constants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 651
B-6.5 Thermal properties of set cement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 652
B-6.5.1 Thermal conductivity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 653
B-6.5.2 Coefficient of linear thermal expansion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 653
B-7 Cement characterization and analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 653
B-7.1 Chemical characterization of Portland cement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 654
B-7.2 Physical characterization of neat cement and cementing materials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 654
B-7.3 Analysis of dry-blended cements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 656
B-7.4 Chemical characterization of set cement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 656
B-7.5 Analysis of cement mix water . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 656
B-8 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 657
B-9 Acronym list . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 657

xx Well Cementing
A p p e n d i x C Cementing Calculations
C-1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 659
C-2 Cement-slurry properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 659
C-2.1 Specific gravity of Portland cement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 659
C-2.2 Absolute and bulk volumes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 660
C-2.3 Concentrations of additives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 660
C-2.4 Slurry density and yield . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 661
C-2.4.1 Special additives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 661
Salt . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 661
Fly ash . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 662
Bentonite . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 663
Weighting agents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 664
Water requirement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 664
C-3 Primary cementing calculations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 664
C-3.1 Annular volumes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 664
C-3.2 Density, yield, and mix water . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 665
C-3.3 Displacement volume to land the plug . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 665
C-3.4 Pump pressure to land the plug . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 665
C-3.5 Hydrostatic pressure on the formation (fracture and pore pressure) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 665
C-3.6 Example well calculations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 666
C-3.7 Pressure to lift the casing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 667
C-4 Plug balancing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 668
C-4.1 Equations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 669
C-4.2 Example calculations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 669
C-5 Squeeze cementing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 670
C-5.1 Example calculations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 670
C-5.1.1 Forces on the casing at the packer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 670
C-5.1.2 Maximum surface pressure safely applied to the annulus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 672
C-5.1.3 Maximum allowable squeeze pressure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 672
C-5.1.4 Maximum allowable pressure when reversing out . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 672
C-6 Calculations for foamed cement jobs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 672
C-7 Acronym list . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 677
C-8 Suggested reading . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 677
Nomenclature . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 7 9
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 9 9
Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 5 1

Contents xxi
Seventeen years ago, following the success of Reservoir
Stimulation (edited by M.J. Economides and K.G. Nolte),
Schlumberger decided to produce a companion work
Acknowledgments
Erik B. Nelson and Dominique Guillot—Schlumberger

Sandra Aldana
Joseph Ayoub
Robert Beirute
Schlumberger
Schlumberger
A
Beirute Consulting LLC
concerning well cementing technology. The result, Well
Cementing, was published in 1990. The positive response Glen Benge ExxonMobil
from the industry has been very gratifying and humbling. Steve Bissonnette Schlumberger
In late 2002, Schlumberger decided to produce a second Ben Bloys Chevron
edition. It is our sincere hope that this updated textbook
Martial Brangetto Total
will be considered a worthy successor. During the past 3
years, we have become deeply indebted to many people J. Ernest Brown Schlumberger
and organizations without whose generous assistance D.G. (Jerry) Calvert Consultant
this project could never have been completed. Bob Carpenter Chevron
We greatly appreciate the positive attitude and
Bill Carruthers LaFarge North America
patience of Schlumberger management and would like
to express our thanks to Sandra Aldana, Mark Andersen, John Cook Schlumberger
Hervé Anxionnaz, Louise Bailey, Iain Cooper, Mark K.M. Cowan Shell
Corrigan, Glenda de Luna, Keith Dismuke, Paul Howard, Doug Elliott ProTechnics
Stuart Jardine, Roger Keese, Gerry Kennedy, Geoff
Bernard Fraboulet Total
Maitland, and Ali Mazen.
The production manager and creative director, Martha Hwawen Gai BP
Dutton, shepherded this project with dedication and Craig Gardner Chevron
perseverance. To give the textbook a consistent “look,” Greg Garrison Schlumberger
Martha and her team modified much of the graphic
Ashley Hibbert Consultant
material submitted by the authors. In many cases they
worked miracles, transforming very rough drawings into Daryl Kellingray BP
clear and coherent illustrations. This new edition is in Joseph Levine U.S. Minerals Management
color, and we hope the result will enhance the clarity of Service, Department of the
the presented material. Interior
Our copy editor, Jennifer Smith, was involved Julien Massonier Schlumberger
throughout the duration of the project—from the initial
James Metson Schlumberger
manuscript drafts to the final layout. Her meticulous
attention to grammar, composition, and style greatly Larry Moran ConocoPhillips
improved the readability of each chapter. In addition, Eric Moulin LaFarge—Laboratoire Central
Jennifer was responsible for maintaining a consistent de Recherche
symbolic nomenclature throughout the textbook and John Patterson ConocoPhillips
verifying the accuracy of the references.
Dave Struthers Schlumberger
This textbook benefited substantially from the techni-
cal assistance of people who reviewed the material and Lawrence Weber Chevron
suggested corrections, additions and changes. We wish Jeff Wendel Texas Industries
to express deep gratitude to our Editorial Review Board. Ray Wydrinski BP

Well Cementing ■ Acknowledgments xxiii

 We are most grateful to many publishing companies Micromeritics
and organizations that gave us permission to reproduce OFI Testing Equipment, Inc.
tables and figures from their publications.
Oil & Gas Journal
American Petroleum Institute (API)
Pegasus Vertex
ASTM International
Petroleum Society of CIM
Baker Atlas
Portland Cement Association
Baker Oil Tools
ProTechnics
BJ Services, Inc.
Ray Oil Tool Company Ltd.
Cambridge University Press
Royal Society of Chemistry
Canadian Well Logging Society
Society of Petroleum Engineers (SPE)
Cement Testing Equipment, Inc.
Thomas Telford Ltd.
Chandler Engineering, LLC
Weatherford International
Elsevier
WOODCO USA
Fann Instrument Company
Finally, special thanks go to Gretchen Gillis, advisory
Geothermal Education Office editor of the Oilfield Review magazine. Gretchen read
Gulf Publishing Company every word of the manuscript and made innumerable
Horwood Publishing Ltd. suggestions that greatly improved the clarity and flow
of the chapters. Her participation in this project far
International Organization for Standardization (ISO)
exceeded the call of duty.
IOP Publishing Ltd.
John Wiley & Sons, Inc.
Malvern Instruments
Materials Research Society

xxiv Well Cementing

The first edition of Well Cementing, published in 1990,
presented a comprehensive review of the scientific fun-
damentals, engineering considerations, and operational
D.G. (Jerry) Calvert—Consultant

PPreface

Complete cement
procedures associated with the processes collectively sheath with no mud
known as cementing. Such processes include primary or gas channels
cementing (casings and liners) and remedial cementing
(plugs and squeezes).
Since 1990, the cementing industry has made signifi-
cant progress in many areas, including
■ new cement additives

■ improved temperature modeling

■ standardized testing procedures Cement bonded

■ deepwater cementing to formations
■ enhanced cement-evaluation tools and procedures
Cement bonded
■ mechanical properties of set cement and long-term
to casing
durability.
This second edition is an updated reference that
reflects these advances. However, before discussing Oil or gas
recent progress, let’s review some of the basic cementing pay zone
objectives.

P-1 Primary cementing Fig. P-1. Objectives of primary cementing.

Primary cementing is the process of placing cement in
the annulus between the casing and the formations
exposed to the wellbore. This subject is presented in The basic process for accomplishing a primary
Chapter 13. Since its inception in 1903, the major objec- cementing job uses the two-plug method for pumping
tive of primary cementing has always been to provide and displacement. This method was first used in 1910 in
zonal isolation in oil, gas, and water wells, i.e., to exclude shallow wells in California. After the well reaches the
fluids such as water or gas in one zone from oil in desired depth, the drillpipe is removed, and a larger-
another zone in the well. To achieve this objective, a diameter string of casing is run to the bottom of the well.
hydraulic seal must be created between the casing and At this time, the drilling mud is still in the wellbore. This
cement and between the cement and the formations, mud must be removed and replaced with a cement
while at the same time preventing fluid channels in the slurry. The most common process to accomplish this is
cement sheath (Fig. P-1). This requirement makes pri- the two-plug cementing method (Fig. P-2). To prevent
mary cementing the most important operation per- contamination with mud, two plugs isolate the cement
formed on a well. Without complete isolation in the well- as it is pumped down the casing. Sufficient cement
bore, the well may never reach its full producing slurry is pumped into the casing to fill the annular
potential. column from the bottom to at least the top of the pro-

Well Cementing ■ Preface 1

ductive zones. Typically, cement slurry is brought to Although wells are drilled deeper today (30,000 ft
higher locations to exclude other undesirable fluids from [9,100 m] or more), technology has advanced, and
the wellbore, to protect freshwater zones, and to protect cementing practices have changed, the basic two-plug
the casing from corrosion. The cementing process is method described above remains predominant.
completed when a pressure increase at the surface indi-
cates the top plug has reached the landing collar or float
collar and displacement with mud or water is termi- P-2 Remedial cementing
nated. The well is shut in for a time to allow the cement
Remedial cementing, presented in Chapter 14, consists
to harden before completion work or drilling to a deeper
of two broad categories: squeeze cementing and plug
horizon begins.
cementing. Squeeze cementing is the process of placing
a cement slurry into the wellbore under sufficient
hydraulic pressure to partially dehydrate or expel water
Cement head from the cement slurry, leaving a competent cement that
Cementing will harden and seal all voids (Fig. P-3). Plug cementing
unit (Fig. P-4) is the placement of a limited volume of cement
at a specified location inside the wellbore to create a
solid seal or plug.
Remedial cementing operations are performed for
various reasons: to repair faulty primary cementing jobs,
alter formation characteristics, repair casing problems,
and abandon wells. Both operations require as much
Displacing mud technical, engineering, and operational experience as
any primary cement job.
Top plug

Mud being displaced

Bottom plug

Float collar

Spacer or wash

Centralizer

Cement

Casing shoe

Fig. P-2. Typical primary cementing job.

2 Well Cementing
 Microannulus Casing leak Repair of channel Top of liner Old perforation Squeeze through
squeeze repair behind casing squeeze squeeze gravel pack

Fig. P-3. Various squeeze cementing techniques.

P-3 Technological advances

Since the early days, many advances have been made in
Displacement fluid all of the disciplines associated with cementing. Special
Portland cements, manufactured expressly for use in
well cementing, allow the industry to use cement sys-
tems that are tailored for the conditions encountered
downhole. A wide variety of chemical additives makes it
Spacer fluid
possible to place durable cement slurries in wellbore
environments ranging from permafrost to geothermal
reservoirs. To achieve optimal cement placement and
zonal isolation, improved techniques have been devel-
oped to condition the wellbore before a primary cement
Cement
job. Such techniques also reduce the need for remedial
cementing. Equipment and techniques have been devel-
oped to properly monitor all cement-job parameters.
This allows timely decisions during a job to increase the
probability of success. Finally, many tools and tech-
niques have been developed to evaluate the quality of
the cement job, allowing the operator to make informed
Mud decisions regarding future operations. Since 1990,
progress has been especially significant in the areas dis-
Fig. P-4. Plug cementing.
cussed below.

Preface 3
P-3.1 Cement additives For many years, the industry has used cement addi-
The cement service companies have introduced multi- tives that reduce the cement-slurry density. Most of
purpose additives that simultaneously act as a retarder, these additives absorb water, allowing the addition of
fluid-loss additive, and dispersant. We have also seen the water to the slurry without solids segregation. In most
introduction of additives that are applicable to a much cases, as the slurry density decreases, the compressive
wider range of well conditions. Such additives reduce strength of the set cement also decreases. Consequently,
the number of cement additives required to perform a the lowest slurry density at which water-absorbing addi-
cement job. The reader will find a complete discussion of tives can be used is about 12 lbm/gal [1,440 kg/m3].
cement additives in Chapter 3. During the 1980s, hollow microspheres and foamed
cements were developed to allow the preparation of
competent cements at densities as low as 8 lbm/gal
[960 kg/m3]. Foamed cements have become especially
popular; however, their preparation requires special
equipment at the wellsite. During the 1990s, the major
cement service companies developed competent 8- to
12-lbm/gal cement systems that were not gas entrained.
At the other extreme, additives that increase the
slurry density (weighting agents) have been improved.
Traditionally, such additives have been dry-blended with
the cement. For many years, the industry has wanted a
weighting agent that could be added directly to the mix
water. This approach would give the operator much more
flexibility in controlling the slurry density during the job
and reduce the need to prepare custom cement blends.
During the past 10 years, this goal has been realized and
such a material is being used both on land and offshore.

P-3.2 Temperature modeling

Well cementing requires an accurate knowledge of
downhole temperatures for proper slurry design and
placement. The bottomhole static temperature is a start-
ing point for slurry design. Prejob tests, such as deter-
mining the thickening time, are based on the bottom-
hole static temperature. The thickening time test is
performed at the bottomhole circulating temperature at
a given casing point. The initial bottomhole circulating
temperatures were developed from temperature gradi-
ents derived from geothermal maps and a very limited
number of data points collected by an operator in the
early 1950s. The initial American Petroleum Institute
(API) temperature tables were based on a temperature
gradient of 1.5°F/100 ft [0.8°C/30 m]. A series of data
points was taken in the mid-1960s, and a set of API
tables was introduced in the mid-1970s covering tem-
perature gradients from 0.9° to 1.9°F/100 ft [0.5° to
1°C/30 m] over a range of depths.

Fig. P-5. Cement additives allow the use of Portland cement in a

wide range of wellbore environments.

4 Well Cementing
 then, operators and cement service companies have
developed downhole tools to obtain more accurate tem-
perature data. In the past 10 years, the major cement
service companies have introduced temperature-model-
ing programs that more accurately predict downhole
temperatures for cement-slurry design. The reader is
referred to Chapter 12 and Appendix B for a complete
discussion of this topic.

P-3.3 Standardization procedures

In the early 1980s, the petroleum industry embarked on
a major effort to merge the API standard recommended
practices with the ISO documents. It should be clarified
that the ISO standards incorporate standards originat-
ing from various countries. Joint cooperation between
operators, cementing service companies, and manufac-
turers has allowed the conversion of many of the current
API documents pertaining to well cementing to active
ISO documents. Two primary examples follow.
■ API Specification 10A—Specification for Cements
and Materials for Well Cementing—has also been
issued as ISO-10426-1, Petroleum and natural gas
industries - Cements and materials for well cement-
ing - Part 1: Specification.
■ API Specification 10B—Recommended Practice for
Testing Well Cements—has been issued as ISO-10426-
2, Petroleum and natural gas industries - Cements
Fig. P-6. Computer modeling. and materials for well cementing - Part 2: Testing of
well cements.
In addition to the temperature modeling programs, The cementing industry continues to address cement
standard API/International Organization for Stand- and cement-testing standards. A full presentation of
ardization (ISO) publications (for example, API RP 10B, cement testing is given in Appendix B.
Recommended Practice for Testing Well Cements, and
ISO 10426-2, Petroleum and natural gas industries -
Cements and materials for well cementing - Part 2:
Testing of well cements) now include equations to cal-
culate bottomhole circulating temperatures for casing-,
liner-, squeeze-, and plug-cementing operations. Since

Preface 5
Fig. P-7. Laboratory testing.

P-3.4 Deepwater drilling very quickly to this need, and appropriate slurries have
The last 10 years have seen deepwater drilling increase been developed. The testing procedures have been com-
considerably. The shallow-hole sections of the wellbore pleted and published as ISO document 10426-3,
have required the development of special cement slur- Petroleum and natural gas industries - Cements and
ries and testing procedures to properly reflect the ser- materials for well cementing - Part 3: Testing of deep-
vice conditions. The cement service industry responded water well cement formulations.

Fig. P-8. Major deepwater hydrocarbon provinces (red).

6 Well Cementing
P-3.5 Cement sheath evaluation tools made with acoustic tools. Nevertheless, the proper inter-
Ultrasonic cement-sheath evaluation has continued to pretation of cement logs still requires extensive infor-
expand. Modern tools may be used in larger-diameter mation about the well and the cement properties. Very
casing, oil-outside-phase drilling fluids, and high-density low-density cement sheaths (foamed or otherwise) still
drilling fluids. Similar improvements have also been pose a challenge.

Pipe Microannulus

Cement
Casing weld

Mud channel
in cement

Gas
invasion

Perforations

Well-centered casing,
good bond

Eccentered casing

Washout Pipe against

formation

Mud left
in washout

Fig. P-9. Processing the log information requires teamwork between the operator, cement service company,
and logging company. Cement sheath evaluation is presented in Chapter 15.

Preface 7
P-3.6 Mechanical properties of set cement Based on the data collected, the major cement ser-
For many years, uniaxial compressive strength and ten- vice companies developed modeling programs that
sile strength were the primary mechanical properties determine the most appropriate cement-slurry design
considered in well cementing. In rare cases, the flexural for a given set of well parameters. These programs
strength was determined. During the early 1980s, scien- require much more detailed information than was gath-
tists and engineers began to estimate the cement-sheath ered in the past. There is also a task group in the API
integrity over the life of the well. Early on, researchers that is charged with developing a set of common test
determined that well operations, completion practices, procedures to determine the desired mechanical prop-
and production following the cement job had major erties of the cement sheath.
effects on cement-sheath integrity. Since then, joint pro- State-of-the-art modeling programs have shown that
jects between the major cement service companies and some of the conventional cement slurries have a high
many of the major operators have been performed to failure probability. The time window may be weeks,
better understand the mechanical properties of well months, or years. This becomes a very important risk
cements. The principal properties are listed below. factor, because the life of a given well could be extended
if it is properly cemented in accordance with its produc-
■ Young’s modulus
tion capability and history. The mechanical properties of
■ Poisson’s ratio cements were given only limited attention in the first
■ Plastic parameters edition of this textbook. In this second edition, an entire
■ Friction angle chapter is devoted to the subject (Chapter 8).
■ Flexural strength

■ Tensile and compressive strength (triaxially loaded)

P-4 Conclusions
In a short preface, it is impossible to cover all of the
Young’s modulus, E ␴a important technological advances. However, from this
Tensile discussion, it is clear that significant progress has been
strength made during the past decade. It should be noted that
Δεa additional work is continuing in the area of well cement-
ing. Our industry has met the challenges to improve, and
Axial will continue to do so in the future. Our goal is 100% job
εa
stress, ␴a ␴ success, regardless of the type of cement job.
(tension) E = εa
a

␴a P-5 Acronym list

Axial strain, εa (tension) API American Petroleum Institute
Δr r ␴a ISO International Organization for Standardization

Δεa
Poisson‘s ratio ν = – Δr
Δεa
εa

␴a

Fig. P-10. Mechanical properties of set cement.

8 Well Cementing
 Erik B. Nelson and Dominique Guillot—Schlumberger

The first edition of Well Cementing was published in

1990. Since then, cementing technology has advanced
rapidly on many fronts, and Schlumberger decided to
Introduction

I
Chapter 2 is concerned with the central unifying
theme of this textbook—Portland cement. The physical
and chemical properties and the performance of this
produce an updated edition. remarkable material are crucial to every facet of well
Well cementing technology is an amalgam of many cementing technology. This chapter presents (in a well
interdependent scientific and engineering disciplines, cementing context) a review of the manufacture, chem-
including chemistry, geology, and physics and petro- ical composition, hydration chemistry, and classification
leum, mechanical, and electrical engineering. Each is of Portland cements. The updated discussion reflects an
essential to achieve the primary goal of well cement- improved understanding of aluminate-phase hydration,
ing—durable zonal isolation. By preparing this second and also includes a section concerning advanced tech-
edition, the authors have aspired to produce a compre- niques for monitoring and modeling cement hydration.
hensive and updated reference concerning the applica- Well cementing exposes Portland cement to condi-
tion of these disciplines toward cementing a well. tions far different from those anticipated by its inventor.
Like the first edition, this textbook comprises five Cement systems must be designed for placement under
principal sections. The first section (Chapter 1) illus- conditions ranging from below freezing in permafrost
trates how the quality of the hydraulic seal provided by zones to greater than 1,000°F [538°C] in some thermal
the cement sheath can affect well performance. The recovery wells. After placement, the cement systems
second section (Chapters 2 through 11) presents infor- must preserve their integrity and provide zonal isolation
mation that must be considered during the design phase during the life of the well. It has only been possible to
of a cementing treatment. Various aspects of cement job accommodate such a wide range of conditions through
execution are covered in the third section (Chapters 12 the development of additives that modify the behavior of
through 14). The fourth section (Chapter 15) addresses Portland cements for individual well requirements. The
cement job evaluation. The fifth section contains appen- impressive array of cement additives used in the well
dices that present information about rheology, cement cementing industry is discussed in Chapter 3. The chem-
testing, and cementing calculations. ical nature of the various classes of additives is
In the preface, D.G. Calvert states that “primary described, and typical performance data are provided.
cementing (is) the most important operation performed In addition, building upon the material presented in
on a well.” Indeed, from operational experience, few Chapter 2, the mechanisms by which the additives oper-
would dispute that no other event has a greater effect on ate are also explained.
the production potential of a well. Yet it is interesting to The rheology of well cement systems is examined in
note that very little work has been published regarding Chapter 4. A review of the relevant rheological models
the quantification of zonal isolation from a reservoir and concepts is presented, followed by discussions of
engineering point of view. In Chapter 1, common reser- flow in pipes and annuli, the effects of particle-laden
voir engineering concepts are used to derive a theoreti- fluids, foamed cement rheology, and equipment for rhe-
cal index of zonal isolation, which can be used to calcu- ological characterization.
late the maximum tolerable cement sheath permeability The rheological behavior of a cement slurry must be
(matrix and interfacial). The index of zonal isolation optimized to effectively remove drilling mud from the
concept is subsequently applied to typical wellbore sce- annulus. The appropriate cement slurry design is a func-
narios, and the results further underscore the critical tion of many parameters, including the wellbore geome-
importance of cement sheath integrity. Chapter 1 also try, casing hardware, formation integrity, drilling fluid
provides several examples of consequences suffered by characteristics, and presence of spacers and washes. A
operators when adequate zonal isolation is not attained. large amount of theoretical and experimental work con-

Well Cementing ■ Introduction 9

cerning mud removal has been performed since 1940, Annular fluid migration has been a topic of intense
yet this subject remains controversial today. Chapter 5 is interest and controversy for many years, and a thorough
a review of the work performed to date, contrasting the review is presented in Chapter 9. This complex phenom-
opposing viewpoints and distilling some mud removal enon may occur at any time during well construction,
guidelines with which the majority of workers in this production, and abandonment, and it has long been rec-
field would agree. ognized as one of the most troublesome problems of
The interactions between cement systems and the the petroleum industry. The causes and consequences
formations with which they come into contact are impor- of fluid migration are discussed, and theoretical and
tant topics. Such interactions encompass four principal experimental models are described. In addition, meth-
effects—fluid loss, formation damage, bonding, and lost ods to predict and solve fluid migration problems are
circulation. It is generally acknowledged that an inap- presented.
propriate level of fluid-loss control is often responsible The physical and chemical behavior of well cements
for primary and remedial cementing failures. In addi- changes significantly at high temperatures and pres-
tion, invasion of cement filtrate into the formation may sures; consequently, special guidelines must be followed
be damaging to production. Chapter 6 discusses static to design cement systems that will provide adequate
and dynamic fluid loss, the deposition of cement filter- casing protection and zonal isolation throughout the life
cakes on formation surfaces, and the influence of a pre- of so-called “thermal wells.” In addition, the presence of
viously deposited mudcake on the fluid-loss process. corrosive zones and weak formations must frequently be
Recent work concerning cement-to-formation bonding is considered. Thermal cementing encompasses three
also highlighted. Another section of Chapter 6 reviews principal types of wells—deep oil and gas wells, geo-
methods for preventing or correcting lost circulation. thermal wells, and thermal recovery (steamflood and
Because lost circulation is best attacked before the fireflood) wells. In Chapter 10, each scenario is dis-
cementing process is initiated, the treatment of this cussed separately, because the cement system design
problem during drilling is also presented. parameters can differ significantly. The chemistry of
During the past 16 years, the cementing industry has thermal cements, including newer systems that are not
rapidly expanded the number of special cement systems based on Portland cement, is also presented, and data
that address problems such as slurry fallback, lost circu- are provided to illustrate the long-term performance of
lation, and microannuli and salt-formation, permafrost, typical systems.
deepwater, and corrosive-well environments. These Proper mixing and placement of well cements rely
technologies are described in Chapter 7. The more upon the application of electrical and mechanical tech-
notable systems discussed in the chapter include nology. Chapter 11 focuses on cementing equipment and
foamed cements, engineered particle-size distribution casing hardware. In line with the trend toward deeper
cements, flexible cements, cements containing blast-fur- wells and more severe working environments, this tech-
nace slag, and cementitious drilling fluids. The composi- nology has become increasingly sophisticated, and the
tions of the cement systems (several of which do not equipment has become more flexible in application and
involve Portland cement) are explained, and typical per- more reliable in operation. First, an extensive discussion
formance data are provided. is presented concerning the various types of equipment
Until recently, the well cementing industry was for bulk handling, storage, cement mixing, and pumping.
largely concerned with only one mechanical property of In addition, the special considerations for onshore and
set cement—compressive strength. Today, we recognize offshore cementing, as well as cementing in remote loca-
that compressive strength alone is an insufficient indi- tions, are examined. The second section of this chapter
cator of a cement system’s ability to provide zonal isola- is a discussion of the wide variety of casing hardware
tion throughout the lifetime of a well and after aban- (float equipment, cementing plugs, stage tools, central-
donment. Consequently, a significant effort is under way izers, scratchers, liners, etc.) and explains how these
to better understand the mechanical response of cement devices work. This presentation is supported by an
systems to downhole conditions. Chapter 8 discusses the extensive series of illustrations.
industry’s current level of understanding. In this chap- Chapters 2 through 11 contain information the engi-
ter, basic rock-mechanics concepts are presented in neer must consider when designing a cement system or
the context of well cements, followed by an examination choosing the proper equipment for the cementing treat-
of how these concepts may be applied to design appro- ment. Sophisticated computer programs are available to
priate cement systems for the anticipated downhole perform most job-design tasks; nevertheless, this has not
environment. diminished the need for simple engineering common

10 Well Cementing
sense. The methodology by which an engineer may sys- describes many devices and techniques that were not
tematically develop an optimal cement job design is dis- available when the first edition was published. Appendix
cussed in Chapter 12. Examples of job design procedures C is a presentation of common calculations for slurry
are also presented. design, primary and remedial cementing, and foamed
Chapter 13 is a presentation of primary cementing cementing. Most of these calculations are performed
techniques. This chapter explains primary cementing today by software applications; nevertheless, this mater-
terminology, the classification of casing strings, and the ial has been included for the reader’s reference.
special problems associated with the cementation of each It is important to mention that this edition of Well
type of string. The cementing of large-diameter casings, Cementing differs from its predecessor in terms of
multilaterals, expandable tubulars, horizontal wells, and nomenclature. The text generally conforms to symbol
liners, as well as stage cementing, is also covered. guidelines published by the Society of Petroleum
Chapter 14 is devoted to remedial cementing tech- Engineers (SPE), and the editors have endeavored to
niques—plug cementing and squeeze cementing. The give symbols a consistent meaning throughout the entire
chapter begins by describing the types of well problems textbook. Consequently, the reader will notice that the
that can be cured by remedial cementing. The next sec- symbols in many equations differ from those that
tion discusses plug cementing techniques and includes appeared in the original referenced publications. Also,
information about special tools and cement design since this book encompasses many disciplines, symbol
guidelines. The third section presents squeeze cement- conflicts frequently occurred. To avoid confusion, this
ing from both a theoretical and practical point of view. textbook presents a comprehensive nomenclature list
Placement techniques such as low- and high-pressure and, where symbol conflicts exist, the meaning of the
squeezes, Bradenhead squeezes, and hesitation symbols in particular chapters is noted. For example, in
squeezes are described. Finally, common misconcep- rock mechanics, the symbol E refers to Young’s modulus
tions about remedial cementing, reasons for failure, and while, for cement job evaluation, E refers to acoustic
evaluation of remedial cementing jobs are discussed. waves. Therefore, E denotes Young’s modulus in Chapter
After a well has been cemented, the results are often 8 and the peak amplitude of the acoustic wave arrival in
evaluated to check whether the objectives have been Chapter 15.
reached. Chapter 15 is a comprehensive presentation of As stated earlier, this textbook was written to provide
the available techniques to perform such evaluations. the reader with updated technical information concern-
These include hydraulic testing; nondestructive meth- ing well cementing. Since work to produce this book
ods such as temperature, nuclear, or noise logging; and began in late 2002, virtually all aspects of cementing
acoustic and ultrasonic cement logging. The theoretical technology have continued to advance at a rapid pace;
basis of each technique is discussed, the measuring consequently, we were obliged to continually revise and
devices are described, and the interpretation of the update many chapters until press time. While this has
results is explained. The interpretation discussion is been somewhat exasperating for the authors, it is a
supported by many illustrations. strong indication of the industry’s continuing commit-
Three appendices are included in this textbook to ment to the improvement of well cementing technology.
supplement the material covered in the chapters. We have attempted to present the material in a logical
Appendix A is a digest of rheological equations com- and easily understandable form and to reduce the aura
monly used in well cementing, presented in a tabular of mystery that seems to be associated with many
format. Appendix B is an updated examination of labo- aspects of this technology. It is our fervent hope that this
ratory cement testing procedures and the equipment new edition of Well Cementing will be a useful addition
commonly used to perform such tests. This discussion to the reader’s reference library.

Introduction 11
1-1 Introduction
Implications of Cementing
for Well Production and Performance

Production optimization begins with a good completion,

and a good completion depends on the integrity of the
Carl T. Montgomery—ConocoPhillips 1
surface casing vent flows and 814 wells with gas migra-
tion problems in Alberta, Canada (Alberta Energy and
Utilities Board, 1999).
primary cement job. About 15% of primary cement jobs The primary reason for these failures is improper bal-
fail, costing the oil and gas industry an estimated ancing of the pressures, which allows gas and fluid
USD 450 million annually in remedial cementing work influx into the cement-filled annulus during the primary
(Newman et al., 2001). Figure 1-1 shows the percentage cement job, and movement of pipe, cement, or both in
of wells experiencing sustained casing pressure (SCP) the wellbore during depletion. Methods to improve pri-
versus age for the 22,000 wells in the U.S. Gulf of Mexico mary cementing practices and prevent zonal-isolation
(personal communication, J. Levine, 2003). This means failures will be discussed in later chapters. In this intro-
that there are 8,000 to 11,000 wells in the Gulf of Mexico ductory chapter, the effect of cement job quality on the
with sustained SCP. In 1999, there were 3,810 wells with long-term performance of a well will be outlined.

60

50

40

Wells affected
by SCP 30
(%)

20

10

0
0 3 6 9 12 15 18 21 24 27 30
Age of well (years)

Fig. 1-1. Gulf of Mexico wells with SCP (from personal communication, J. Levine, 2003).
Figure from U.S. Minerals Management Service.

Well Cementing ■ Chapter 1 Implications of Cementing for Well Production and Performance 13
1-2 Zonal isolation pipe/cement interface fails. Several production prac-
Complete and durable zonal isolation is the foremost tices can cause debonding:
■ the gradual pressure decrease as a well is produced
goal of the cement job. During the life of a producing
oil and gas well, the quality of the cement job has a ■ casing movement as subsidence occurs
direct impact on the economic longevity of the well. ■ cement shrinkage with time
From the time the well is first produced until the well is ■ temperature and pressure fluctuations
abandoned, appropriate cement-slurry design and
■ stimulation practices, such as hydraulic fracturing.
placement techniques will affect well productivity, both
physically and economically. If allowed to set undis- Figure 1-3 shows images from a 4-arm caliper log in a
turbed, Portland cement systems of normal density well that has experienced severe casing movement. The
(≈16.0 lbm/gal or 1,930 kg/m3) usually exhibit extremely cement in the region where movement occurred was
low matrix permeability. The literature quotes values in destroyed, but a good primary cement job kept the undis-
the microdarcy range. However, during the productive turbed region intact and isolated.
life of a well, the cement is subjected to various severe The third condition, called “shear failure,” typically
conditions that can affect the longevity of this low matrix results in complete failure of the cement sheath. Shear
permeability. The first condition, termed “cracking,” is failure is normally caused by effective-stress increases
caused by thermal or pressure fluctuations in the well around a wellbore caused by rock subsidence and move-
caused by the production process. For example, gas ment as the reservoir is depleted. This effect can also be
wells are subjected to large variations in drawdown pres- caused by vibrations from downhole pumps or gas-lift
sure and temperature as the gas demand changes. operations.
Depending on the magnitude and frequency of these Any of these conditions will result in flow paths in the
production variables, the casing and cement sheath form of discrete conductive fractures in the cement, or
expand and contract in different ways (Chapter 8). This microannuli. These paths, and their effective widths,
causes stress gradients that gradually crack cement, create cement permeabilities that far exceed the intrin-
with the subsequent loss of cement integrity (Fig. 1-2). sic permeability of the undisturbed cement. Even a
The second condition, termed “debonding,” occurs small microannulus results in a large effective perme-
when the bond between either the cement/rock or the ability along the cement sheath.

Fig. 1-2. Photograph of set cement that has undergone thermal fracturing.

14 Well Cementing
 rcsg rwb

Reservoir 1 (pres)1

k* ΔL1

h, k, p

Well
ΔL2

Reservoir 2 (pres)2

Fig. 1-3. Casing deformation caused by formation movement.

Cement
sheath

1-2.1 Index of zonal isolation

Fig. 1-4. Typical well completion configuration. The symbols h, k,
Zonal isolation is the most important function of cement; and p are respectively height, permeability, and pressure, and rcsg
however, the minimum necessary zonal isolation is not and rwb are respectively the outer casing radius and the radius of
often quantified. In 1990, Economides proposed a simple the wellbore.
way to attempt such an estimate. One can compare the
producing rate of the active layer with the contributions
rwb = wellbore radius
of an overlying or underlying formation through
sres = skin factor
breaches in the cement sheath. Consider the typical
μres = viscosity (cp).
completion configuration shown in Fig. 1-4. In the
middle is a perforated interval with two other potentially For a gas well, the analogous equation is
producing intervals (one above and one below) sepa-
⎡
( )2 − ( pf bh ) ⎤⎥⎦
2
rated by some relatively impermeable layers, of thick- kres hres ⎢ pres
ness ΔL1 and Δ L2, respectively. ⎣
For simplicity, let us consider steady-state flow into qres = (1-1b)
⎛ r ⎞
the well from the producing layer. The equation describ- 1424μ res Z res Tres ⎜ ln res + sres ⎟
ing this rate for a radial oil reservoir is easily derived ⎝ rwb ⎠
from Darcy's law and is given below in U.S. units.
where

qres =
(
kres hres pres − pf bh ) (1-1a)
qres = flow rate (Mscf/D)
Tres = reservoir temperature (°R)
⎛ r ⎞ Zres = gas deviation factor.
141.2 Bresμ res ⎜ ln res + sres ⎟
⎝ rwb ⎠ Note that, for gas, the flow rate is proportional to the
pressure squared.
where Crossflow from the adjoining formations into the pro-
Bres = formation volume factor ducing layer is likely to occur, because a pressure gradi-
hres = thickness (ft) ent is formed between them. The flow rate is propor-
kres = permeability (mD) tional to the vertical permeability of the producing layer.
pfbh = flowing bottomhole pressure (psi) For flow into the producing layer from another forma-
pres = reservoir pressure (psi) tion, the largest vertical pressure gradient would be at
qres = flow rate (stb/D) the cement sheath, which must have at least as low a
rres = reservoir radius

Chapter 1 Implications of Cementing for Well Production and Performance 15

permeability as the barrier layers. From the geometry Table 1-1. Well and Reservoir Data for Oil Flow Along
shown in Fig. 1-4, the flow area through the cement the Cement Sheath
sheath is equal to rwb 0.406 ft (93⁄4-in. OD)

⎡
( ) ( )
2 2⎤ rcsg 0.328 ft (77⁄8-in. OD)
A = π ⎢ rwb − rcsg ⎥ . (1-2)
⎣ ⎦ pres 3,000 psi

Darcy’s law can be applied along the cement annulus. Bres 1.1 bbl at reservoir conditions/stb
Thus, from the generalized expression
μres 1 cp
kA Δp ΔL1 20 ft
q= , (1-3)
μ ΔL pfbh 1,000 psi
and replacing A as given by Eq. 1-2, an expression giving
the flow rate (in U.S. units) through the cement sheath
can be obtained. 100

( ) ( ) ⎤⎥⎦ ⎡⎣( pres )1 − p fbh ⎤⎦

⎡ 2 2
k * π ⎢ rwb − rcsg 10

qleak = ⎣
( )1 (μ res )1 ( Δ L1)
(1-4)
141.2 Bres 1
Oil flow
rate
where (stb/D) 10–1
( pres)1 = adjoining reservoir pressure
qleak = flow rate (Mscf/D).
10–2
Analogous expressions to Eq. 1-4 can be readily
derived for the flow of gas or water. In the case of gas,
the expression is 10–3
0.01 0.1 1 10 100 1,000

( ) ( ) ( ) ( )
⎡ 2 2 ⎤⎧ 2 2⎫ Annulus equivalent permeability (mD)
k∗π ⎢ rwb − rcsg ⎥ ⎨ ⎡ pres ⎤ − p fbh ⎬
⎣ ⎦⎩ ⎣ 1⎦
⎭.
qleak =
( )
Fig. 1-5. Oil flow rate along the cement sheath for a range of annu-
1424μ1 Zres T Δ L1 lar equivalent permeabilities.
(1-5)
Equations 1-4 and 1-5 provide the oil and gas flow
1,000
rate, whether through the cement sheath matrix perme-
ability, through a microannulus formed between the
cement and casing, or through the cement and the for- 100
mation. The equivalent permeability value, k*, can be
related to width of the microannulus, as will be shown 10
later. Gas flow
Using Eq. 1-4, the leakage rate through the cement rate
sheath can be estimated for various values of equivalent (Mscf/D) 1
permeability. Table 1-1 contains some typical values
from reservoir and well data. The distance between the 10–1
target reservoir and an adjoining formation, ΔL1, is
taken as equal to 20 ft. Figure 1-5 is a graph of the
steady-state oil flow rate for a range of equivalent per- 10–2
0.01 0.1 1 10 100 1,000
meability, using the data in Table 1-1. Figure 1-6 is an
analogous example of a gas well, using the data in Table Annulus equivalent permeability (mD)
1-2 and Eq. 1-5. For an ideal, low-permeability cement
sheath bonded to the casing and formation, the leakage Fig. 1-6. Gas flow rate along the cement sheath for a range of
annular equivalent permeabilities.
rates are negligible. When a microannulus exists at the
cement-formation interface, the situation is completely
different.

16 Well Cementing
Table 1-2. Well and Reservoir Data for Gas Flow ratio of the flow rate into the well from the intended for-
Along the Cement Sheath mation to the flow rate through the cement is defined as
the index of zonal isolation (IZI) and is given in Eq. 1-8.
rwb 0.406 ft (93⁄4-in. OD)

rcsg 0.328 ft (77⁄8-in. OD)

IZI =
qres
=
( )
kres hres Δ L1
2 ⎤⎛ ⎛ r ⎞ ⎞
( ) ( )
pres 3,000 psi qleak ⎡ 2
k *π ⎢ rwb − rcsg ⎥ ⎜ ln ⎜ res ⎟ + s ⎟
pfbh 1,000 psi ⎣ ⎦ ⎝ ⎝ rwb ⎠ ⎠
μres 0.025 cp (1-8)
Zres 0.95
Interestingly, all variables that distinguish Eq. 1-1a
Tres 640°R (for oil and water) and Eq. 1-1b (for gas) are the same as
those evident in Eq. 1-4 (for oil and water) and Eq. 1-5
ΔL1 20 ft
(for gas). Thus, the IZI expression as given by Eq. 1-8 is
valid for any fluid.
Equation 1-8 can provide the quantification of zonal
In a consistent unit system, the microannulus perme- isolation. It can be used either to calculate the required
ability is given by cement equivalent permeability to provide a desired
flow-rate ratio or, for a given cement permeability, what
w2 the flow rate would be through the cement sheath from
k= , (1-6) adjoining layers. As discussed earlier, the equivalent
12
cement permeability, k*, is an equivalent permeability
where w is the annulus width. value, resulting either from the presence of a microan-
However, the equivalent annular permeability, defined nulus or from an unnaturally high cement-matrix per-
by Eq. 1-5, is meability. The latter may be caused by the disruptive
effects of fluid invasion as the cement changes from
rwb w3 liquid to solid (Chapter 9). While a large matrix perme-
k * = kcem + (1-7a) ability within the cement sheath is unlikely (for the mag-
( ) ( )
⎡ 2 2⎤
12 ⎢ rwb − rcsg ⎥ nitudes shown in Figs. 1-5 and 1-6), a large equivalent
⎣ ⎦ permeability can result from a relatively small microan-
nulus width.
because the relevant parameter is the product of the Equation 1-8 can also be used as an evaluation tool to
permeability and the cross-sectional area. When detect flow through the cement sheath. If a vertical-
expressing permeabilities in millidarcies, radii in feet, interference or a multilayer test is performed and the
and width in inches, Eq. 1-6 becomes reservoir is well defined, then crossflow through the

(4.54 × 10 ) r
adjoining low-permeability layers may be calculated
9 3
* wb w (Ehlig-Economides and Ayoub, 1986). As a result, the
k = kcem + . (1-7b) ideal flow rate from the targeted interval can be calcu-
⎡
( ) ( )
2 2⎤
r
⎢ wb − rcsg ⎥
lated.
⎣ ⎦ Deviations from this value can be attributed to flow
through an imperfect cement sheath, assuming that the
Using the wellbore geometry given in Table 1-1 and cement-sheath leak was not active when the vertical
assuming a microannulus width of 0.002 in. [51 μm], interference test was run. Using Eqs. 1-1a or 1-1b, the
the second term of the right-hand side of Eq. 1-7b is permeability of the cement can then be extracted. The
equal to 2.54 × 10–13 m2 or 258 mD. Therefore, in the net flow rate through the perforated interval is
presence of a microannulus, cement permeability can
be a negligible portion of the equivalent permeability.
And, as can be seen from Figs. 1-5 and 1-6, the leakage q = qres + ∑ qcf + ∑ qleak , (1-9)
rates are then substantial.
where
These leakage rates can be compared to the flow rate
from the producing formation. For simplicity, we will qres = lateral reservoir flow rate
now assume that in Eq. 1-4, the fluid and the reservoir Σqcf = the sum of crossflow contributions through
pressure in the adjoining layer are the same as in the the barrier
producing formation. Dividing Eq. 1-1a by Eq. 1-4, the

Chapter 1 Implications of Cementing for Well Production and Performance 17

Σqleak = the sum of contributions through the cement The quantified IZI then becomes an important vari-
sheath. able to control. For tight reservoirs, if only absolute
Figure 1-7 is a graph for an example well using an 80- contributions or losses from or into adjoining formations
acre spacing, a skin effect equal to 5, and a wellbore are of concern, then a low IZI can be tolerated. However,
radius of 0.406 ft. The product k reshresΔ L1 is graphed it should be remembered, especially when influx of
on the abscissa while the equivalent cement permeabil- foreign fluids such as gases, water, or oil of different
ity is graphed on the left ordinate. On the right ordinate physical properties is evident, the minimum tolerable IZI
is the equivalent path width squared that would result may be very high and contingent on the production facil-
in a similar flow rate. Two curves are offered: one for ities at the wellhead. In such cases, even more stringent
50% and another for 100% of the q⁄qleak ratio (IZI). As can IZI requirements may be necessary in tight, thinly sepa-
be seen, the cement permeability requirements and rated formations, as implied in Eq. 1-8.
the need for more zonal isolation become more critical The ability of cement to withstand the expansion and
for lower-permeability producing formations that are contraction of the tubing string is directly related to the
separated by thin barriers. In both cases, the product composition of the original cement. As wells are cycled,
k reshresΔ L1 becomes small, requiring a small cement- cement experiences fatigue. This can be overcome to a
sheath permeability. The low product is not a problem certain extent through the use of flexible cements, as
if only the innate matrix permeability of the cement described in Chapters 7 and 8.
sheath is considered. For most cements, this permeabil- The effects of cement failure on the long-term viabil-
ity is less than 0.01 mD. ity of a well or even a field can be drastic. A poor cement
However, the presence of a continuous microannulus job in a pressure-maintenance or enhanced oil recovery
can severely aggravate the situation. The width cubed of well (water, miscible injection fluid, polymer, steam,
the microannulus is graphed on the right ordinate of etc.) allows injection fluid to be introduced into zones in
Fig. 1-7. For a typical reservoir (kres = 4 mD, hres = 50 ft, which it is not needed. This can result in severe and
Δ L1 = 50 ft, resulting in k reshresΔ L1 = 104) with a flow costly environmental, drilling, and longer-term produc-
rate ratio of 50, the microannulus width must be less tion problems. When this occurs, the injection fluid will
than 0.0014 in. (35 μm), which corresponds to an equiv- normally migrate to the highest-permeability region,
alent permeability of 86 mD. These calculations clearly allowing a “short circuit” to form in the system. This
demonstrate the extreme importance of obtaining an results in poor sweep efficiency, wasted injection fluid,
intimate bond between the cement sheath and casing and shortened well life. A short circuit to adjacent wells
and formation surfaces. can also result in a breach to surface. A secondary effect

10,000 3.1 × 10–7

1,000 3.1 × 10–8

100 3.1 × 10–9

IZI = 50
10 3.1 × 10–10
Equivalent Slot width
permeability cubed
IZI = 100
(mD) 1 3.1 × 10–11 (in.3)

10–1 3.1 × 10–12

10–2 3.1 × 10–13

10–3 3.1 × 10–14

1 10 10 2 10 3 10 4 10
5 10
6 107

kreshresΔL1 (mD-ft2)
Fig. 1-7. Example of the IZI concept.

18 Well Cementing
is the overpressuring of shallow sands, resulting in uncontrolled water invasion can have numerous nega-
drilling hazards and blowout issues when additional tive consequences. Corrosion of tubulars can occur,
field development is needed. shortening the useful life of the well. The water content
of the produced fluids may increase, raising the lifting
and disposal costs. In addition, contamination of fresh-
1-3 Water and gas influx water aquifers is a serious environmental consequence.
During the production period, the sudden appearance of
excessive amounts of water, gas, or both in the annulus
is a strong indication of cement failure. The effects of 1-4 Impact on artificial-lift operations
this failure are numerous and include lost production The purpose of artificial lift is to increase flow from the
revenue, hazardous operations, supercharging of shallow reservoir by lowering the bottomhole pressure at the
water sands, and hazardous conditions around the well- production zone. This reduced pressure increases the
head. Table 1-3 offers a quick diagnostic tool to deter- stress on the cement. If the cement job is poor, channels
mine the causes of gas influx (personal communication, can form that connect with zones other than the desired
L. Moran, 2003). Goodwin and Crook (1990) provide a producing interval.
good discussion of why the cement sheath fails in areas Communication with other zones can result in the
where high-temperature or high-pressure fluctuations production of unwanted fluid, either water or gas, which
occur in the well. The main mechanism they describe for will affect the lift equipment. The lift equipment may be
cement-sheath failure under excessive temperature and undersized for the greater amount of fluid or gas. Unless
pressure is the expansion of the casing both diametri- corrective actions are taken, the desired bottomhole
cally and circumferentially. The circumferential expan- pressure cannot be achieved, resulting in lower than
sion creates a shearing force at the cement-casing inter- expected production rates. In severe cases, a poor
face, causing failure at the interface or fracturing of the cement job that communicates with a water zone or a
cement sheath radially from the inner casing surface to deeper aquifer can create water coning, seriously reduc-
the outer casing or formation surface. Laboratory testing ing the ultimate recovery of the resource. Many types of
showed that all cement systems exposed to casing lift equipment subject the produced fluids to relatively
expansion exhibited failure, ranging from the creation of high-shear conditions. If water is introduced into the
stress fractures running either parallel or perpendicular system, water-in-oil emulsions can form, drastically
to the inner casing surface to total structural failure, in reducing the pump efficiency. If the introduced water is
which the cement was reduced to putty. The point at incompatible with the formation water, scale can pre-
which failure occurred and the nature of the failure were cipitate, reducing pump efficiency and eventually caus-
functions of specific parameters of the cement, includ- ing pump blockage.
ing elasticity and compressive strength. Their work, pre- Likewise, depending on the type of lift equipment
sented in Chapter 8, also provides a simple model to cal- employed, communication into a gas zone creates prob-
culate the magnitude of casing expansion that can lead lems. Free gas decreases the performance of all the
to cement failure. pumps used to lift fluid. Excessive gas can reduce run
In addition to gas, failed cement can allow water to times, cause frequent shutdowns, and, in the worst case,
invade the annulus. Although not as dangerous as gas, cause sufficient cavitation to destroy the pump.

Table 1-3. Gas Influx Diagnostic Table†

Symptom Cause
Gas flows immediately. The well is underbalanced or a gas zone exists above the cement top.

Gas appears 3 to 12 hr after the cement job. The cement lost hydrostatic pressure during the setting process and
allowed gas influx.

Gas appears in a day to a few days after the cement job. A microannulus formed between the cement and casing or formation.

Gas appears months to years after the cement job. The well suffered a loss of hydrostatic pressure in a channeled mud
column, the cement is highly permeable, or the cement was damaged
during the production process.
† From personal communication, L. Moran, 2003

Chapter 1 Implications of Cementing for Well Production and Performance 19

1-5 Matrix stimulation and hydraulic sand. The interpretation of this log is that the cement
fracturing failed during the third stage of the hydraulic fracturing
treatment. This indicates that the cement-to-pipe or
The cement-to-pipe bond is a very important considera- cement-to-formation bond strength was insufficient.
tion when matrix stimulation or hydraulic fracturing Parcevaux and Sault (1984) showed that there is no
treatments are attempted. It is critical that the cement apparent correlation between the cement compressive
sheath provide zonal isolation above and below the zone strength and the shear-bond strength. Furthermore,
of interest. Unfortunately, and surprisingly, this is an they determined that the shear-bond strength ranges
area of research that has not received sufficient atten- from 1,000 psi (about 7 MPa) for standard cement to
tion. Handin (1965) and Banthia (2003) attempted to 1,800 psi [about 12 MPa] for cements containing bond-
characterize the “strength” of well cements at downhole enhancing agents (BA) as shown in Fig. 1-9. These
pressure and temperature conditions. Handin (1965) values imply that, for many reservoirs in which the ten-
characterized the compressive strength of cements and sile strength of the rock is greater than 1,000 psi [about
determined the ultimate strength at failure. He con- 7 MPa], the cement-to-pipe bond will fail first, resulting
cluded that oilwell cements become ductile, even under in a microannulus along the pipe. This has major impli-
low effective confining pressures. Because of the magni- cations for the loss of stimulation fluids (acid, scale
tude of ultimate compressive strengths at normal system inhibitors, fracturing fluids, etc.) as well as the migra-
densities, these cements have mechanical constitutive tion of reservoir fluids following the treatment. In such
properties similar to that of sedimentary rocks under a situation, remedial cementing would be indicated.
similar confining conditions. Banthia studied the perfor-
mance of fiber-reinforced cement systems and found
that the toughness parameters were strongly influenced
by the load conditions. 1-6 Logging
The perforating process can shatter the cement Many types of cased-hole wireline logs depend on good
sheath in the immediate vicinity of the perforating- cement-to-casing and cement-to-rock bonds to provide
charge detonation. Later, when a fluid is injected into a accurate data. The absence of good bonding causes
reservoir at a pressure greater than the fracture gradi- erratic tool response and makes interpretation difficult.
ent, such as during hydraulic fracturing or water injec- In the case of sonic logs, the wave train will be reflected
tion operations, the failure mechanism is tensile, and back to the tool in a manner termed “ringing,” which
the cement sheath is potentially subjected to two phe- masks the sonic response of the reservoir. This same
nomena. The first is fracture propagation within the response, used at a higher frequency, is the principal
cement sheath beyond the perforations, and the second technique used in cement bond logs to determine where
is debonding of the cement sheath from the pipe. In good cement coverage exists. If a microannulus is pre-
either case, the net result is the creation of a microan- sent in the cement, the cased-hole resistivity log can also
nulus. Depending on the formation strength compared be misinterpreted. For example, if brine has migrated
to the cement and the net pressure created during the into a cement microannulus in a hydrocarbon-bearing
treatment, the fracture created in the cement can have zone, the resistivity log would indicate low resistivity,
a width of the same order as the induced fracture in the and the zone would be interpreted as water-saturated.
rock. Pulsed neutron tools have very shallow depths of investi-
Figure 1-8 shows a tracer log of a well following a gation. If the wellbore has a poor cement job or thick
hydraulic fracturing treatment. Both the pad stage (irid- casing, the results may be erroneous. Chapter 15 offers a
ium = red) and 2–4 ppa proppant stage (antimony = complete discussion of these topics.
blue) were contained within the zone of interest.
However, during the 4–8 ppa proppant stage (scandium
= yellow), the fracture grew down into a deeper sand.
The crack between the upper and lower zone was of suf-
ficient width to conduct the proppant into the lower

20 Well Cementing
 SC RD SC FM SC FM
0 (in.) 25 1,000 (API) 0 0 (API) 1,000

Gamma Ray SB RD SB FM SB FM Individual Tracer Logs

0 150 0 (in.) 25 1,000 (API) 0 0 (API) 1,000
Iridium Antimony Scandium

Depth (ft)
Gamma Ray Total IRA RD IRA FM IRA FM IR Total SB Total SC Total

12,800
12,900
13,000
13,100 0 150 0 (in.) 25 1,000 (API) 0 0 (API) 1,000 0 1,000 0 1,000 0 1,000

Zone of interest Zone of interest

13,200

Fig. 1-8. Tracer survey showing fracture-height growth into lower sand (courtesy of ProTechnics,
a division of Core Laboratories L.P.).

1-7 Conclusions
15 2,175
Despite recent advances in the cementing of oil and gas
wells, many of today’s wells are at risk. Failure to isolate
Shear 10 1,450 Shear sources of hydrocarbons early in the well-construction
bond bond process, or later after production has begun, has
strength 5 Standard cement 725 strength resulted in abnormally pressured casing strings. In addi-
(MPa) 9.4% BVOS† BA‡ (psi) tion, gas and produced water can contaminate other
0 18.8% BVOS BA 0 subsurface zones, such as freshwater aquifers. This
0 5 10 15 20 25 30 results in wells that are environmentally and opera-
Curing time (days) tionally hazardous. The well’s revenue stream can also
†Byvolume of solids be interrupted or reduced because of the loss of lift effi-
‡Bond-enhancing agent
ciency in the pump, excess-water or gas-handling issues,
Fig. 1-9. Cement shear bond strength development at 68°F [20°C].

Chapter 1 Implications of Cementing for Well Production and Performance 21

inefficient remediation or stimulation treatments, and 1-8 Acronym list
potential loss of the well owing to casing collapse. The BA Bond-enhancing agents
environmental impact of contaminating a single fresh-
water aquifer is extremely serious. Therefore, the initial BVOS By volume of solids
and long-term quality of the cement sheath and bond SCP Sustained casing pressure
should be of prime importance to the operator, because
it is essential for the safe and successful production of a
well.

22 Well Cementing
2-1 Introduction
Chemistry and Characterization

Ordinary Portland cement (OPC) is by far the most

important oilwell binding material in terms of quantity
of Portland Cement
Erik B. Nelson and Michel Michaux—Schlumberger

2-2 Chemical notation

2
A special chemical notation established by cement
chemists is frequently used in this textbook. The chemi-
produced; indeed, it is possibly the most abundant man- cal formulas of many cement compounds can be
ufactured material. The term “ordinary” indicates that expressed as a sum of oxides; for example, tricalcium
the cement is manufactured in a rotary kiln from a silicate, Ca3SiO5, can be written as 3CaO • SiO2.
molten matrix of suitably proportioned ingredients. OPC Abbreviations are given for the oxides most frequently
is used in nearly all well cementing operations; there- encountered, such as C for CaO (lime), S for SiO2
fore, throughout this textbook, the terms OPC and (silica), and A for Al2O3 (alumina). Thus Ca3SiO5
Portland cement shall be used interchangeably. becomes C3S. A list of abbreviations is given below.
The conditions to which Portland cements are
C = CaO F = Fe2O3 N = Na2O P = P2O5
exposed in a well differ significantly from those encoun-
A = Al2O3 M = MgO K = K2O f = FeO
tered at ambient conditions during construction opera-
S = SiO2 H = H2O L = Li2O T = TiO2
tions; as a result, special Portland cements are manu- _ –
factured for use as well cements. Certain other special Others are sometimes used, such as S = SO3 and C =
cements, used to solve particular well cementing prob- CO2. This convention of using a shortened notation was
lems, are discussed in Chapters 7 and 10. adopted as a simple method for describing compounds
Portland cement is the most common example of a whose complete molecular formulas occupy much space
hydraulic cement. Such cements set and develop com- when written.
pressive strength as a result of hydration, involving
chemical reactions between water and the compounds
present in the cement. The setting and hardening occur 2-3 Portland cement manufacture
not only if the cement/water mixture is left to stand in
Portland cement is produced by pulverizing clinker.
air, but also if it is placed underwater. The development
Clinker is the calcined (burned) material that exits the
of strength is predictable, uniform, and relatively rapid.
rotary kiln in a cement plant. Clinker consists primarily
The set cement also has low permeability and is nearly
of hydraulic calcium silicates, calcium aluminates, and
insoluble in water; therefore, exposure to water does not
calcium aluminoferrites. One
_ or more forms of calcium
destroy the hardened material. Such attributes are
sulfate (usually gypsum, CSSH2 ) are interground with the
essential to achieve and maintain zonal isolation.
clinker to make the finished product. Materials used in
This chapter presents fundamental information
the manufacture of Portland cement clinker must con-
regarding the manufacture, hydration, and classification
tain appropriate amounts of calcium, silica, alumina,
of Portland cements. In addition, the effects of various
and iron compounds. During manufacture, frequent
chemical and physical parameters upon performance
chemical analyses of all materials are made to ensure
are discussed. Several excellent textbooks were relied
uniformity and high quality.
upon heavily to produce this overview of cement tech-
nology: Taylor (1964); Ghosh (1983); Taylor (1997);
Hewlett (2001); and Bensted and Barnes (2002).

Well Cementing ■ Chapter 2 Chemistry and Characterization of Portland Cement 23

2-3.1 Raw materials Natural argillaceous materials frequently used as raw
Two types of raw materials are needed to prepare a mix- materials include clays, shales, marls, mudstones, slate,
ture that will produce Portland cement clinker: calcare- schist, volcanic ashes, and alluvial silt. Blast-furnace
ous materials, which contain lime, and argillaceous slag from steelworks and fly ash from coal-fired power
materials, which contain alumina, silica, and iron oxide. plants are the most important artificial sources.
Depending upon the location of the cement plant, a The properties of Portland cement are determined by
great variety of natural and artificial raw materials is the mineralogical composition of the clinker. The miner-
employed (Table 2-1). alogical composition of conventional Portland cement
The most important calcareous materials are sedi- clinker is shown in Table 2-2. For special cements, the
mentary and metamorphic limestones, coral, shell content of C3A and C4AF can differ significantly. The
deposits, and “cement rock,” which naturally has a com- main oxides make up about 95%:
position similar to Portland cement. Artificial calcare- CaO 60–70%
ous materials include precipitated calcium carbonate SiO2 18–22%
and other wastes from various industrial processes. Al2O3 4–6%
Fe2O3 2–4%.

Table 2-1. Various Raw Materials Used in the Manufacture of Portland Cement
Calcium Iron Silica Alumina Sulfate
Alkali waste Blast-furnace flue dust Calcium silicate Aluminum-ore refuse† Anhydrite

Calcite† Clay† Cement rock Bauxite Calcium sulfate

Cement-kiln dust Iron ore† Clay† Cement rock Gypsum†

Cement rock Mill scale† Fly ash Clay†

Chalk Ore washings Fuller’s earth Copper slag

Clay Pyrite cinders Loess Fly ash†

Fuller’s earth Shale Marl† Fuller’s earth

Limestone† Ore washings Loess

Marble Quartzite Ore washings

Marl† Rice-hull ash Shale†

Seashells Sand† Slag

Shale† Sandstone

Slag Shale†

Slag

Traprock
† Most common sources

Table 2-2. Mineralogical Composition of Classic Portland Cement Clinker

Oxide Composition Cement Notation Common Name Concentration (wt%)
3CaO • SiO2 C3S Alite 55–65

2CaO • SiO2 C2S Belite 15–25

3CaO • Al2O3 C3A Aluminate 8–14

4CaO • Al2O3 • Fe2O3 C4AF Ferrite phase 8–12

24 Well Cementing
 The total content of minor compounds such as CaO not exceed 0.6% in concrete applications, because of
(free lime), MgO, K2O, Na2O, TiO2, Mn2O3, and SO3 is nor- adverse reactions with certain types of siliceous aggre-
mally under 5%. gates.
When selecting the raw materials and kiln fuels, it is
important to consider impurities that can have signifi-
cant effects on the properties of the finished cement. 2-3.2 Raw material preparation
These include manganese, magnesia (M), chlorides, Before calcination in the kiln, the raw materials must
phosphates, lead oxide, zinc oxide, alkalis, and fuel first be pulverized to a fine powder and uniformly
residues. After clinkering in the kiln, such impurities are blended to ensure that the bulk composition corre-
often in solid solution within the principal cement sponds to that required to manufacture a particular type
phases, resulting in a change of reactivity. When present of Portland cement. Raw material homogenization is a
in quantities exceeding about 0.5 wt%, manganese can very important technological operation, because the kiln
lead to the development of excessively large alite crys- feed should have a very stable chemical composition.
tals, which may retard strength development. Excess Although each cement plant has its own specific
magnesia (more than 5%) can cause a disruptive delayed method, there are two general processes in use today:
expansion of the set cement, a condition known as the dry process and the wet process. In the dry process,
“unsoundness.” Chlorides are generally limited to grinding and blending are done with dry materials. In
0.015%, because they are volatile and can cause serious the wet process, the grinding and blending operations
difficulties in the burning zone of the kiln. The presence use a watery slurry.
of more than 0.1% fluorine in the raw materials, usually A schematic diagram of the dry process is shown in
as calcium fluoride, results in a significant decrease in Fig. 2-1. The raw materials are crushed, dried in rotary
cement strength. Phosphates can have a beneficial driers, proportioned to obtain the correct bulk composi-
effect on strength at a level of 0.20% to 0.25%; however, tion, and then ground in a roller mill that combines
they have a deleterious effect at concentrations exceed- crushing, grinding, drying, and size classification in one
ing 0.5%. Lead and zinc oxides have a deleterious effect unit. The ground material is stored in several silos. The
upon cement properties. The effect of alkalis is variable. chemical composition varies from silo to silo; therefore,
The total alkali content, expressed as sodium oxide the mixture that goes into the kiln may require reblend-
equivalent (%Na2O + [0.658 × %K2O]), generally should ing and fine-tuning.

High-pressure grinding
Feed rolls (optional, usually
used in conjunction with Raw mix
a ball mill)
Limestone

Iron ore
Sand

Clay

Dust To preheater
Discharge
collector
Proportioning
Air
of raw materials Roller To pneumatic Dry-mixing Ground raw
mill pump and blending materials
silos storage

Fig. 2-1. Schematic flow diagram of dry process (from Portland Cement Association, 2002).

Chapter 2 Chemistry and Characterization of Portland Cement 25

 The wet process is illustrated in Fig. 2-2. The raw 2-3.3 Heat treatment
materials are initially proportioned in the dry state. After achieving the appropriate degree of size reduction,
Water is added, and further size reduction occurs in a classification, and blending of the raw materials, heat
grinding mill. Size classification is performed by pump- treatment is performed in a rotary kiln that is usually
ing the resulting slurry past a vibrating screen. Coarser preceded by a preheater in more modern cement plants.
material is returned to the mill for regrinding. The slurry This step is shown in Fig. 2-3. The kiln is slightly inclined
is stored in basins equipped with rotating arms and com- and rotates at 1 to 4 rpm; as a result, the solid material
pressed air agitation to keep the mixture homogeneous. passes through the kiln as it rotates.
The chemical composition of the slurries varies slightly Production of clinker involves burning fossil fuels to
from basin to basin. Thus, final adjustments of composi- calcine large amounts of limestone. As a result, large
tion are performed by blending the slurries from various quantities of CO2 are released to the atmosphere. Today,
basins. cement producers are proactively working to reduce the
For many years, the wet process was preferred emission of greenhouse gases. Foremost among the
because more accurate control of the raw mix was possi- methods being employed is improved fuel-burning effi-
ble; however, the dry process has become predominant ciency. By using less fuel per ton of clinker produced,
because of the lower heat consumption, a typical value CO2 generated from fuel burning is reduced. In recent
being 3,000 kJ/kg of clinker for dry and 5,500 kJ/kg for years, alternative fuels such as waste oil and old rubber
wet material. In recent years, the technology has been tires have been used. Alternative raw materials that con-
developed to obtain excellent control of raw material tain calcium but have been calcined in a previous man-
composition using the dry process. ufacturing process are also used.

Water Slurry Vibrating

added screen
here Oversize
Limestone

Iron ore
Sand

Clay

To kiln

Grinding
Proportioning mill Mixing and blending Storage basins
of raw materials Slurry Fines of slurry
pumps Slurry
pump

Fig. 2-2. Schematic flow diagram of wet process (from Portland Cement Association, 2002).

Raw material feed Hot gases from preheater or clinker cooler to raw mill

Preheater (about 40% calcination)

that uses hot gases from kiln heat
raw feed before feed enters kiln
Materials stored
Flash furnace (optional; provides separately
Dust 85% to 95% calcination) Clinker
collector Tertiary air duct
Gypsum

Air

Fan Dust bin Rotating kiln Clinker cooler Clinker and gypsum to grinding mills

Fig. 2-3. Schematic flow diagram of the burning process (from Portland Cement Association, 2002).

26 Well Cementing
 Cement kilns are also used for the destruction of haz- In Zones III and IV, several important reactions occur.
ardous wastes. The neutralization and destruction of Dehydroxylation of clay minerals is completed, and the
these substances is much more effective in a cement products crystallize. Calcium carbonate decomposes to
kiln than traditional industrial incinerators. In addition, free lime, releasing large quantities of carbon dioxide.
some waste solvents can act as fuel in the kiln. The production of various calcium aluminates and fer-
A complex series of reactions takes place in the kiln rites also begins. The sintering zone, Zone V, occupies a
to convert the raw materials to clinker. There are six small portion of the kiln; however, most of the principal
temperature zones in a kiln, and the temperature ranges cement phases are produced at this stage. At this point,
and reaction profiles are shown in Table 2-3. The clinker part of the reaction mixture liquefies. At the maximum
production process from raw feed to final product is temperature in the sintering zone, also known as the
shown in Fig. 2-4. “clinkering temperature,” the formation of C2S and C3S
Evaporation of free water occurs in Zone I. Water is completed. The uncombined lime, alumina, and iron
removal occurs quickly in the dry process; however, up oxide are contained in the liquid phase. During the cool-
to one-half the length of the kiln can be devoted to ing phase (Zone VI), the liquid phase disappears, result-
drying with a wet-process system. During preheating ing in the crystallization of C3A and C4AF. Some residual
(Zone II), dehydroxylation of the clay minerals occurs. free lime is always present in the clinker.

Table 2-3. Reaction Zones in a Rotary Cement Kiln

Zone Approximate Temperature Range (°F[°C]) Reaction Profile
I Up to 390 [700] Evaporation

II 390 to 1,470 [200 to 800] Preheating

III 1,470 to 2,010 [800 to 1,100] Decarbonation

IV 2,010 to 2,370 [1,100 to 1,300] Exothermic reactions

V 2,370 to 2,820 to 2,370 [1,300 to 1,550 to 1,300] Sintering

VI 2,370 to 1,830 [1,300 to 1,000] Cooling

Chapter 2 Chemistry and Characterization of Portland Cement 27

 Cross-Section View of Kiln Nodulization Process Clinkering Reactions

To 1,292°F [700°C]: Particles are solid and do not Water is lost.

Raw materials react with each other. Dehydrated
are free-flowing clay H2O
powder. recrystallizes.
H2O

Clay particle
Limestone particle
1,292–1,652°F [700–900°C]: Particles are still solid. As calcination
Powder is still continues, free
free-flowing. lime increases. CO2 CO2
Reactive silica
combines with CaO
to begin forming CO2
C2S. Calcination
maintains feed
temperature at
1,562°F [850°C].
2,102–2,192°F [1,150–1,200°C]: Reactions begin between When calcination Free CaO
Particles start to solid particles. is complete,
become “sticky.” temperature
increases rapidly.
Small belite
crystals form
from the
combination
of silicates and CaO.
2,192–2,462°F [1,200–1,350°C]: The capillary forces of the liquid Above 2,282°F [1,250°C], Round belite crystals
As particles start to keep particles a liquid phase is
agglomerate, the liquid together. formed. Liquid
holds them together. allows reaction
The rotation of the between belite and
kiln initiates free CaO to form
coalescing of alite.
agglomerates and
layering of particles.
Angular alite crystals
2,462–2,642°F [1,350–1,450°C]: Nodules will form Belite crystals
Agglomeration and with sufficient liquid. decrease in
layering of particles Insufficient liquid amount but
continue. will result in increase in
dusty clinker. size. Alite
increases in
size and amount.

Cooling Clinker nodules remain unchanged. Upon cooling, the

C3A and C4AF
crystallize in the
liquid phase.
Lamellar structure
appears in
belite crystals.

Fig. 2-4. Process of clinker production from raw feed to the final product (from Portland Cement Association, 2002).

28 Well Cementing
2-3.4 Cooling
The quality of the clinker and the finished cement is very
dependent on the thermal profile and particularly upon
the cooling rate. The best clinker is obtained by cooling
slowly to about 2,282°F [1,250°C], followed by rapid
cooling, usually 32° to 36°F/min [18 to 20°C/min].
When the cooling rate is slow, 7° to 9°F/min [4° to
5°C/min], the C3A and C4AF develop a high degree of
crystallinity, the C3S and C2S crystals become highly
ordered, and the free MgO forms crystals (mineral name:
periclase). This results in a cement that is less hydrauli-
cally active. When hydrated at ambient temperatures,
early compressive strength is high, but longer-term Fig. 2-5. Microscope photograph of cement clinker (from Portland
strength is low. Cement Association, 2002).
When the cooling rate is fast, the liquid phase formed
at Zone V in the kiln solidifies to a glass. The C3A and
C4AF remain trapped in the glassy phase, and the crys-
tallinity of the C2S and C3S is less ordered. The free MgO
also remains in the glassy phase; as a result, it is less
active and the resulting cement is less apt to demon-
strate unsoundness. Early compressive strength is lower,
but longer-term strength is higher.
The general behavior described above is based upon
general observations of cement behavior at ambient con-
ditions. For well cements, larger C3S and MgO crystals
are usually preferred (personal communication, B.
Carruthers, 2002).
Figures 2-5 and 2-6 are microscope photographs of a
typical Portland cement clinker and finished Portland
cement, respectively. Figure 2-5 is a polished thin sec-
tion of clinker. Alite (C3S) appears as light, angular crys- Fig. 2-6. Two-dimensional image of Portland cement (from NIST,
tals. The darker, rounded crystals are belite (C2S). 2001). Red means C3S, aqua C2S, green C3A, yellow C4AF, pale green
Figure 2-6 is an example of an advanced two-dimen- gypsum, white free lime, dark blue K2SO4, and magenta MgO (peri-
sional microscopic technique that allows visualizing clase).
more of the phases. The image was obtained by combin-
ing a set of backscattered scanning electron and X-ray
images and producing a computer-generated map with
different colors assigned to the individual phases
(National Institute of Standards and Technology [NIST],
2001).

Chapter 2 Chemistry and Characterization of Portland Cement 29

2-3.5 Grinding 2-4 Hydration of the clinker phases
As shown in Fig. 2-7, the finished cement is produced by The compounds present in Portland cement are anhy-
grinding the– clinker with calcium sulfate, usually drous. When brought into contact with water, they are
gypsum (CS H2 ) which, for reasons that will be explained attacked or decomposed, forming hydrated compounds.
later, prevents a phenomenon known as flash set. Most Supersaturated and unstable solutions form, gradually
clinker is ground in tubular mills partly filled with hard depositing their excess solids. Because the solubilities of
steel balls and, depending upon the type of cement being the original anhydrous compounds are much higher than
manufactured, the clinker is ground to a given fineness those of the hydration products, complete hydration
(Appendix B). The particle size of the cement grains should ultimately occur.
varies from 1 to 100 μm. Research concerning cement hydration has largely
The ball milling process is inherently inefficient, with consisted of studying the behavior of individual clinker
97–99% of the energy input being converted to heat. phases in an aqueous environment and relating the find-
Consequently, it is necessary to continuously cool the ings to the behavior of the multicomponent system
mill. If the cement reaches an excessively high tempera- Portland cement. The principal components of Portland
ture, too much of the gypsum can dehydrate, forming cement (C3S, C2S, C3A, and C4 AF) display different
–
calcium sulfate hemihydrate (CSH1/2). While still able to hydration kinetics and form different hydration prod-
prevent flash set, dehydrated calcium sulfate com- ucts. This chapter follows the same path, first presenting
pounds can cause another phenomenon called false the contributions of the individual phases in this section
set, which will also be discussed later in this chapter. and finally discussing their combined performance in
Grinding aids are prohibited during the production of Portland cement in the following section.
well cements. Such compounds can interfere with addi-
tives used to control the performance of well cement
slurries. 2-4.1 Hydration of the silicate phases
The silicate phases in Portland cement are the most
abundant, often comprising more than 80% of the total
2-3.6 Storage material. C3S is the principal constituent, with a con-
Portland cement is a moisture-sensitive material. If kept centration as high as 68%. The quantity of C2S normally
dry, it will retain its quality indefinitely. Cement stored does not exceed 30%.
in contact with damp air or moisture sets more slowly As shown in the idealized chemical equations below,
and develops less strength. At the cement plant, bulk the hydration products for both phases are calcium sili-
cement is stored in large silos. The relative humidity in cate hydrate and calcium hydroxide (also known as port-
a warehouse used to store bagged cement should be as landite).
low as possible. Frequently, there are several silos for a
particular type of cement. In such cases, cement from 2C3S → C3S2H3 + 3 CH (2-1)
different silos can be blended to maintain a more con-
sistent product. 2C2S + 4H → C3S2H3 + CH (2-2)

High- Cement product High-

efficiency and air to dust efficiency
Exhaust
dust collector collector separator
gas
Finished
Mill product cement
Fan and air product
Gypsum
Clinker

Primary air to silo Packaging

Secondary from mill machine
air air sweep
m
gypsu
r and
Proportioning Clinke Separator rejects
(tails) to mill
of materials Air
Ambient Grinding mill
Mill discharge
air Solids

Twin high-pressure roll press to

precrush clinker entering the Bucket elevator Cement Bulk Bulk Bulk Box Truck
ball mill (optional) to separator pump storage truck car car

Fig. 2-7. Schematic flow diagram of the grinding process and storage (from Portland Cement Association, 2002).

30 Well Cementing
 The calcium silicate hydrate does not have the exact phase, the hydration of C3S is largely responsible for the
composition of C3S2H3; instead, the C:S and H:S ratios beginning of the set and early strength development.
are variable depending upon such factors as the calcium The hydration of C2S is significant only in terms of the
concentration in the aqueous phase (Barret et al., 1980a final strength of the hardened cement.
and 1980b), temperature (Odler and Skalny, 1973), the The mechanism of C2S hydration is very similar to
presence of additives (Odler and Skalny, 1971), and that of C3S; however, the process occurs more slowly. The
aging (Barnes, 1983). The material is quasi-amorphous, C-S-H phase formed by the hydration of C2S is also very
and it is therefore commonly called “C-S-H phase.” The similar to that formed by C3S. Therefore, only C3S is con-
C-S-H phase comprises roughly 65% of fully hydrated sidered in this chapter. The hydration of C3S is some-
Portland cement at ambient conditions and is consid- times considered to be an analog for the hydration
ered the principal binder of hardened cement. By con- behavior of Portland cement.
trast, the calcium hydroxide is highly crystalline and The hydration of C3S is an exothermic process; there-
occurs as hexagonal plates. Its concentration in hard- fore, the hydration rate can be followed by conduction
ened cement is usually between 15% and 20%. calorimetry. From the thermogram given in Fig. 2-9, five
After a brisk but brief initial hydration when added to hydration stages are arbitrarily defined.
water, the silicate phases experience a period of low I. Preinduction period
reactivity, called the “induction period.” During this II. Induction period
stage, they do not significantly influence the rheology of III. Acceleration period
the cement slurry. Substantial hydration eventually IV. Deceleration period
resumes and, as shown in Fig. 2-8, the hydration rate of V. Diffusion period
C3S exceeds that of C2S by a wide margin. Because of the
C3S abundance, and the massive formation of C-S-H

100
80 77°F [25°C] Ca2+
(mmoles/L)
C2S 60 Saturation
hydrated level
(%) 40 122°F [50°C]
I II III IV V
20 Rate of
heat
41°F [5°C]
0 evolution
0.01 0.03 0.1 0.5 1 3 5 10 30 100 1,000
Time (days) min hr days
Time of hydration
100
80 Fig. 2-9. Schematic representation of changes taking place in C3S-
water system.
60 77°F [25°C]
C3S
hydrated 40
(%)
20 122°F [50°C]
41°F [5°C]
0
0.01 0.03 0.1 0.5 1 3 5 10 30 100 1,000
Time (days)

Fig. 2-8. Hydration of C2S and C3S versus time.

Chapter 2 Chemistry and Characterization of Portland Cement 31

2-4.1.1 Preinduction period 2-4.1.2 Induction period
The preinduction period begins during mixing and lasts As explained earlier, relatively little hydration activity is
only a few minutes immediately following mixing. Upon observed during the induction period. The rate of heat
contact with water, a fast hydration reaction begins liberation dramatically falls. Additional C-S-H phase pre-
between C3S and the mix water. A large exotherm is cipitates slowly, and the Ca2+ and OH– concentrations
observed at this time, resulting from the wetting of the continue to rise. When critical supersaturation finally
powder and the rapidity of the initial hydration. occurs, precipitation of calcium hydroxide begins.
The first step of this reaction appears to be a Hydration resumes at a significant rate, signaling the
protonolysis of the silicate and oxygen ions at the C3S end of the induction period. At ambient temperatures,
surface, followed by a congruent dissolution of the mate- the duration of the induction period is a few hours.
rial (Barret et al., 1983; Barret, 1986; Damidot et al., The mechanism of the induction period is still a sub-
1992). The O 2– ions originally present in the C3S lattice ject of debate among cement chemists. Several theories
enter the liquid phase as OH– ions, and the SiO44– ions have been proposed; however, they are often more com-
hydrolyze and form dissociated silicic acid. The balanc- plementary than contradictory. They are summarized in
ing positive charges are from Ca2+ ions (Eq. 2-3). Table 2-4.
2Ca 3SiO5 + 8H2 O → 6Ca 2+ + 10OH − + 2H3SiO 4 − (2-3) 2-4.1.2.1 Impermeable hydrate layer theory
The impermeable hydrate layer theory postulates that
During the preinduction period, C3S dissolution is faster the “first-stage” product precipitated at the C3S surface
than the diffusion of the reaction products away from acts as a barrier that inhibits the migration of water to
the surface. As a result, the liquid phase surrounding the the unhydrated surface, the release of Ca2+, OH–, and sil-
C3S surface becomes supersaturated, and a layer of C-S- icate ions into the liquid phase, or both (Brown et al.,
H phase begins to precipitate on the surface (Menetrier, 1984; Skalny and Young, 1980). As a result, the initial
1977; Barret and Bertandrie, 1986). fast hydration slows and is followed by a period in which
the reaction barely progresses. Toward the end of this
2Ca 2+ + 2OH − + 2H3SiO 4 − → dormant period, the initial C-S-H layer becomes more
( )2
Ca 2 OH H 4Si 2 O7 + H2 O (2-4) permeable, and the hydration rate accelerates.
There are three principal theories to explain why the
C-S-H layer becomes more permeable. According to de
Equation 2-4 assumes that the initial C-S-H phase has a
Jong et al. (1967), the C-S-H layer ages and undergoes a
C:S ratio of about 1.0 (Menetrier, 1977). In addition, the
morphological change, resulting in increased permeabil-
silicate anions in the C-S-H phase are, at short hydration
ity. Consequently, water more easily penetrates the layer,
times, dimeric (Michaux et al., 1983).
and hydration continues. Another theory holds that
Addition of Eqs. 2-3 and 2-4 produces the following.
osmotic pressure is generated between the C-S-H layer
2Ca 3SiO5 + 7H2 O → and the unhydrated C3S surface (Powers, 1961; Double
et al., 1978; Birchall et al., 1980). The gel layer eventu-
( )2
Ca 2 OH H 4Si 2 O7 + 4 Ca 2+ + 8 OH − (2-5) ally bursts, exposing the unhydrated surface. A vigorous
release of hydration products then occurs, resulting in a
During the preinduction period, critical supersatura- massive formation of C-S-H phase. The third proposed
tion with respect to calcium hydroxide does not occur; mechanism involves the breakdown of the C-S-H layer
therefore, as indicated in Equation 2-5, the concentra- because of water imbibition (Dent-Glasser, 1979).
tion of lime increases as hydration continues.
According to another theory, the initial dissolution of
C3S is incongruent. A SiO2-rich layer is formed at the sur-
face that subsequently adsorbs Ca2+ ions dissolved in the
liquid phase. Thus, an electrical double layer forms at
the C3S surface (Tadros et al., 1976; Skalny and Young,
1980).

32 Well Cementing
Table 2-4. Theories on the Mechanism of C3S Hydration
Mechanism Impermeable Hydrate Electrical Double- Nucleation of Nucleation of
Layer Theory Layer Theory CH Theory C-S-H Theory
Pre-induction period First-stage C-S-H formed

Beginning of First-stage product acts as diffusion barrier Electrical double layer Supersaturation of liquid phase with
induction period forms and impedes respect to CH stops further rapid
passage of ions dissolution of C3S

Changes during Phase transformation Osmotic pressure Gradual weakening Slow nucleation Slow nucleation
induction period or aging of C-S-H phenomena across of double layer of CH of second-stage
layer the first-stage C-S-H
product layer

End of induction Increased permeability Breakdown of Breakdown of CH nuclei reach Nuclei of second-
period of C-S-H layer C-S-H layer due double layer critical size stage C-S-H
to osmotic pressure reach critical size
and/or imbibition

Acceleratory period Accelerated dissolution of C3S, growth of second-stage C-S-H and CH

† From Hewlett, 2001. Reprinted with permission from Elsevier.

2-4.1.2.2 Electrical double-layer theory 2-4.1.2.4 Nucleation of C-S-H theory

As mentioned earlier, the electrical double-layer theory According to the nucleation of C-S-H theory, the nucle-
assumes that the initial hydration of C3S is incongruent. ation and growth of a “second-stage” C-S-H controls the
An SiO2-rich layer, with adsorbed Ca2+ ions, creates an end of the induction period. The second-stage C-S-H is
electrical double layer that impedes the passage of addi- different from the initial, “first-stage” C-S-H formed
tional ions into solution. Therefore, the induction period during the preinduction period. In agreement with the
ensues (Tadros et al., 1976; Skalny and Young, 1980). The nucleation of Ca(OH)2 theory, formation of first-stage
double layer gradually breaks down, and the hydration C-S-H is controlled by the Ca(OH)2 concentration in the
rate accelerates. liquid phase. However, unlike the impermeable hydrate
layer theory, the first-stage C-S-H does not act as a bar-
2-4.1.2.3 Nucleation of Ca(OH)2 theory rier layer and does not inhibit hydration.
The liberation of Ca(OH)2 always accompanies the During the induction period, C-S-H nuclei begin to
hydration of C3S. Initially, the Ca(OH)2 dissolves in the form. When the concentration and size of the nuclei
liquid phase. As hydration continues, the Ca(OH)2 does reach a critical level, fast production of second-stage
not precipitate even when its liquid-phase concentration C-S-H begins (Odler and Dorr, 1979; Barret and
reaches or exceeds saturation. According to the nucle- Bertrandie, 1986). At the same time, solid Ca(OH)2
ation of Ca(OH)2 theory, the surfaces of the Ca(OH)2 begins to precipitate.
nuclei are poisoned by adsorbed silicate ions, preventing
precipitation (Tadros et al., 1976; Wu and Young, 1984). 2-4.1.3 Acceleration and deceleration periods
Owing to the supersaturation of Ca(OH)2 in the liquid
phase, further hydration of C3S is retarded (Odler and At the end of the induction period, only a small percent-
Dorr, 1979). age of the C3S has hydrated. The acceleration and decel-
Eventually, sufficient supersaturation (~1.5 to 2.0 eration periods, also collectively known as the “setting
times the saturation value) occurs to form stable nuclei, period,” represent the interval of most rapid hydration.
and precipitation of solid Ca(OH)2 commences. At this During the acceleration period, solid Ca(OH)2 crystal-
stage the crystalline Ca(OH)2 begins to act as a sink for lizes from solution and the C-S-H phase deposits into the
Ca2+ ions, enabling renewed intense C-S-H formation. available water-filled space. The hydrates intergrow, a
cohesive network is formed, and the system begins to
develop strength.

Chapter 2 Chemistry and Characterization of Portland Cement 33

 The porosity of the system decreases as a conse- 2-4.2 Hydration of the aluminate phases
quence of hydrate deposition. Eventually, hydrate depo- As shown in Fig. 2-11, the aluminate phases, especially
sition hinders the transportation of ionic species and C3A, are the most reactive at short hydration times.
water through the network of C-S-H phase, and the Although they are present in considerably less abun-
hydration rate decelerates. At ambient conditions, these dance than the silicates, they have a significant influ-
events occur within several days. ence on the rheological properties of the cement slurry
and early strength development of the set cement.
2-4.1.4 Diffusion period During the hydration of Portland cement, the aluminate
phases react with the added calcium
_ sulfate. Thus
_ the
Hydration continues at a slow pace owing to the ever-
reactions in the system C3A–CSS–H2O and C4AF–CSS–H2O
decreasing system porosity. The network of hydrated
are of particular importance.
products becomes more and more dense, and strength
increases. There is no evidence of major structural
changes; however, polymerization of C-S-H phase has
1.0
been observed (Dent-Glasser et al., 1978). The duration
of the diffusion period is indefinite at ambient condi-
tions. Portlandite crystals continue to grow and engulf
the hydrating C3S grains; as a result, total hydration is 0.8
C4AF
never attained (Fig. 2-10).

0.6
C3S C3A
Hydrated
fraction
0.4

C3A
+ gypsum
0.2 C2S

0
1 10 100 1,000 10,000
Hydration time (hr)

Fig. 2-11. Typical hydration kinetics of pure clinker minerals (C3A

with and without added gypsum) at ambient temperature (from
Hewlett, 2001). Reprinted with permission from Elsevier.

Fig. 2-10. Photographs of precipitated Ca(OH)2 in C-S-H

phase matrix.

34 Well Cementing
2-4.2.1 Tricalcium aluminate C3A hydration is controlled by the addition of 3 to 5%
In the absence of calcium sulfate, the first hydration gypsum to the cement clinker before grinding, as
step is analogous to that of C3S—an interfacial reaction described earlier in this chapter. Upon contact with
between the surface of the anhydrous solid and water. water, part of the gypsum dissolves. The calcium and sul-
This irreversible reaction leads to the hydroxylation of fate ions released in solution react with the aluminate
the superficial anions AlO2– and O2– into [Al(OH)4]– and and hydroxyl ions released by the C3A to form a calcium
OH– anions (Bertrandie and Barret, 1986), resulting in a trisulfoaluminate hydrate, known as the mineral ettrin-
congruent dissolution of the protonated surface. gite (Collepardi et al., 1978). In Portland cements, iron
can substitute for aluminum in the ettringite structure.
−
⎣ 4⎦
( )
Ca 3 Al 2 O6 + 6H2 O → 3Ca 2+ + 2 ⎡ Al OH ⎤ + 4OH − (2-6)
In such cases, iron substituted ettringite is frequently
called an AFt (aluminoferrite trisulfate) phase.
−
The solution quickly becomes supersaturated with
respect to some calcium aluminate hydrates, leading to ( )
6 Ca 2+ + 2 ⎡ Al OH ⎤ + 3SO 42− + 4 OH − + 26H2 O →
⎣ 4⎦
their precipitation.
− ⎣ ( ) ( )
Ca 6 ⎡ Al OH ⎤ SO 4 i26H2 O
6 ⎦2 3
(2-10)

⎣ ( )
6 Ca 2+ + 4 ⎡ Al OH ⎤ + 8 OH − + 15H2 O →
4⎦
The global reaction can be written as
⎣ ( ) 5 ⎦2 ⎣ 7
( )
Ca 2 ⎡ Al OH ⎤ i3H2 O + ⎡2Ca 2 Al OH i6H2 O ⎤
⎦ (2-7)
C3 A + 3 CSH2 + 26H → C3 A i3 CSi32H . (2-11)
By adding Eqs. 2-6 and 2-7, the following equation is
obtained using cement-chemistry notation. As shown in Fig. 2-13, ettringite occurs as needle-shaped
crystals that precipitate onto the C3A surfaces, hinder-
2C3 A + 27H → C2 AH8 + C4 AH19 (2-8) ing further rapid hydration. Thus, as shown in Fig. 2-14,
an induction period is artificially created. During this
The calcium aluminate hydrates in Eq. 2-8 are period, the gypsum is gradually consumed and ettringite
metastable and occur as hexagonal crystals. They even- continues to precipitate. When the supply of gypsum is
tually convert to the more stable cubic form, C3AH6, as exhausted, the sulfate-ion concentration sharply drops.
shown below. At ambient conditions, this reaction Ettringite becomes unstable and converts to a platy cal-
occurs within several days (Tumidajski and Thomson, cium monosulfoaluminate hydrate. If iron is present in
1994). the structure, the compound is known as an AFm (alu-
minoferrite monosulfate) phase. The protective sulfoalu-
C2 AH8 + C4 AH19 → 2C3 AH6 + 15H (2-9) minate layer breaks down and rapid hydration resumes
(Collepardi et al., 1978).
As shown in Fig. 2-12, C3A hydration does not have an
induction period. If uncontrolled hydration were allowed
C3 A + 3 CSi32H + 2C3 A+4H → 3C3 A iCSi12H (2-12)
to occur in a Portland cement slurry, severe rheological
difficulties would arise. Any remaining unhydrated C3A forms calcium aluminate
hydrate as shown in Eq. 2-8 (Bensted, 1976).

50

40
Rate of
heat 30
evolution
(cal/g/hr) 20

10
0
0 1 2 3 4 5 6 7 8
Time (hr)

Fig. 2-12. Thermogram of C3A hydration (77°F [25°C]).

Chapter 2 Chemistry and Characterization of Portland Cement 35

 2-4.2.2 Calcium aluminoferrite
Calcium aluminoferrite is classically referred to as C4AF
or brownmillerite. However, the actual composition of
calcium aluminoferrite is a solid solution with the gen-
eral formula Ca4Fe(2–x)AlxO10, where x varies from 0
to 1.4. C4AF is one point in the series in which x = 1.0.
Many cements have a ferrite phase with a composition
close to C4AF; therefore, this is the most common com-
position studied.
Under comparable conditions, the hydration products
formed in the hydration of the ferrite phases are similar
in many respects to those formed from C3A, although the
rates of reaction differ (Ramachandran and Beaudoin,
1980). As shown in Fig. 2-11, C4AF has the fastest hydra-
tion rate of all the pure clinker phases (Brown, 1993).
However, the reactivity of the ferrite phase generally
declines with increasing Fe content (Collepardi et al.,
1979).
In the absence of gypsum, the course of ferrite-phase
hydration is similar to that of C3A. The metastable iron-
substituted AFm phases C4(A,F)H19 and C2(A,F)H8 form,
which eventually convert to a hydrogarnet phase,
C3(A,F)H6. At elevated temperatures, the hydrogarnet
phase is formed directly.
In the presence of gypsum, the rates of reaction of
C4AF are higher, and an AFt phase forms that is similar
to ettringite (Bensted, 2001; Meller et al., 2004). Like
ettringite, the ferrite AFt phase eventually converts to
an AFm phase (Tong and Yang, 1994).

Fig. 2-13. Scanning electron microscope photographs of ettringite 2-5 Hydration of Portland cements—
crystals.
The multicomponent system
The hydration of Portland cement is a sequence of over-
lapping chemical reactions between clinker compo-
Stage 1 nents, calcium sulfate, and water, leading to continuous
Stage 4
20 cement-slurry thickening and hardening. Although the
Rate of hydration of C3S is often used as a model for the hydra-
heat Stages 2 and 3 tion of Portland cement, it must be kept in mind that
evolution 10 many additional parameters are involved.
(cal/g/hr) From a chemical point of view, Portland cement
hydration is a complex dissolution and precipitation
process in which, unlike the hydration of the individual
0 pure phases, the various hydration reactions proceed
0 10 20 30 40 50
simultaneously at differing rates. The phases also influ-
Time (hr) ence each other. For example, the hydration of C3A is
Fig. 2-14. Thermogram of C3A hydration with gypsum (77°F [25°C]). modified by the presence of hydrating C3S, because the
production of calcium hydroxide reinforces the retard-
ing action of gypsum. None of the clinker minerals is
pure. Depending upon the composition of the raw mate-
rials, each contains alien oxides in solid solution that
alter their reactivity.

36 Well Cementing
 The hydration products are also impure. The C-S-H the internal porosity of the system increases.
phase incorporates significant amounts of aluminum, In the confined environment of a wellbore, the
iron, and sulfur, while the ettringite and monosulfoalu- decrease in absolute volume can affect the transmission
minate phases contain silicon. The calcium hydroxide of hydrostatic pressure to the formation and can affect
also contains small quantities of foreign ions, chiefly sil- the cement's ability to prevent annular fluid migration.
icate. This subject is thoroughly discussed in Chapter 9.
A typical schematic thermogram of Portland cement
hydration is shown in Fig. 2-15. It can roughly be
described as the addition of the thermograms for C3S 2-5.2 Effect of temperature
and C3A, adjusted for relative concentration. Temperature is one of the major factors affecting the
hydration of Portland cement. The hydration rate of the
cement and the nature, stability, and morphology of the
hydration products are strongly dependent upon this
Ettringite
dissolution
Rapid formation Formation of parameter.
of C-S-H and CH monosulfate
and C-S-H gel Elevated hydration temperatures accelerate the
formation Diffusion- hydration of cement. As illustrated by the calorimetry
Rate of controlled
heat Induction period
Final reactions curves in Fig. 2-16, the duration of the induction and set-
increase
evolution in Ca2+ and OH– set ting periods is shortened, and the rate of hydration
concentration during the setting period is much higher. However, upon
Initial set extended curing, the degree of hydration and the ulti-
mate strength are often reduced. This is most probably
min hr days related to the formation of a dense layer of C-S-H phase
Time of hydration around the C3S surfaces, hindering their complete
hydration (Bentur et al., 1979).
Fig. 2-15. Schematic representation of Portland cement hydration. Up to 104°F [40°C], the hydration products are the
same as those that occur at ambient conditions. Certain
changes occur in the microstructure and morphology of
2-5.1 Volume changes during setting C-S-H phase at higher temperatures. The material
When Portland cements react with water, the system becomes more fibrous, and a higher degree of silicate
cement plus water undergoes a net volume diminution. polymerization is observed. At curing temperatures
This is an absolute volume decrease. It occurs because
the absolute density of the hydrated material is greater 200
77°F [25°C]
than that of the initial reactants. Table 2-5 shows the
113°F [45°C]
change of absolute volume with time for a number of 175
149°F [65°C]
Portland cements. 185°F [85°C]
Despite the decrease in absolute volume, the external 150
dimensions of the set cement, or the bulk volume,
remain the same or slightly increase. To accomplish this, 125

Heat flow 100

Table 2-5. Volume Diminution of Portland Cement Samples (mW)
Over Time†
75
Sample Absolute Volume Diminution
(%)
50
At 1 At 7 At 28 At 100
Day Days Days Days
25
No. 1, Portland cement 2.8 4.8 6.0 6.9

No. 2, Portland cement 1.7 4.4 –‡ 6.3 0

0 5 10 15 20
No. 3, Portland cement 2.7 8.0 8.6 8.7
without gypsum Hydration time (hr)

No. 4, Portland cement 2.6 6.3 7.5 7.6 Fig. 2-16. Effect of temperature on hydration kinetics of Class G
without gypsum Portland cement.
† Lea, 1971. Reprinted with permission from Elsevier.
‡ Not available

Chapter 2 Chemistry and Characterization of Portland Cement 37

exceeding 230°F [110°C], the C-S-H phase is no longer Clinker Sulfate Hydration Time
stable, and crystalline calcium silicate hydrates eventu- Resistivity Availability
10 min. 1 hr 3 hr
in
ally form. This subject is thoroughly discussed in Solution Ettringite Recrystallization
Chapter 10. Ettringite
The conversion of the hexagonal aluminate hydrates coating

to the cubic form (Eq. 2-9) is strongly accelerated by

Low Low
temperature. Above 176°F [80°C], C3AH6 is directly
formed.
The behavior of the calcium sulfoaluminates also Workable Workable Set
Ettringite
depends on curing temperature. Above 140°F [60°C], coating
ettringite is no longer stable and decomposes to calcium
monosulfoaluminate and gypsum (Barvinok et al., 1976). High High

C3 A i3 CSi32H → C3 A iCSi12H+2CSH2 + 16H (2-13) Workable Set Set

Ettringite coating; C4AH13 and
However, other researchers have recorded higher sta- monosulfate in pores
bility limits for ettringite, up to 230°F [110°C] (Lach and
High Low
Bures, 1976). Calcium monosulfoaluminate is reported
to be stable up to 374°F [190°C] (Satava and Veprek,
1975). Set Set Set
Ettringite coating; secondary
gypsum in pores
2-5.3 Flash set and false set
Low High
When Portland cement clinker is ground alone (i.e.,
without calcium sulfates) and mixed with water, the C3A
and C4AF rapidly react, the slurry temperature markedly Set Set Set

increases, and an irreversible stiffening occurs, followed Fig. 2-17. Schematic diagram of structure development in the set-
quickly by a pseudoset. This phenomenon is called a ting of Portland cement in relation to the reactivity of the clinker and
“flash set,” or sometimes a “quick set.” In the context of to sulfate availability (from Ghosh, 1983). Reprinted with permission
well cementing, a flash set could prevent proper place- from Elsevier.
ment of the cement slurry in the annulus.
To avoid uncontrolled C3A and C4AF hydration, cal- To relieve this condition, so-called “secondary gypsum”
cium sulfates (usually gypsum) are ground in with the is precipitated. A marked stiffening or gelation of
clinker during the manufacture of Portland cement. For the cement slurry, known as “false set,” is observed
optimal cement performance, the quantity of calcium (Fig. 2-17).
sulfates must be balanced according to the reactivity of False sets are reversible by vigorous slurry agitation;
the clinker (Fig. 2-17). however, such agitation would not be possible during
A flash set can still occur if the quantity of calcium most well cementing operations, particularly if the
sulfates in the cement is insufficient with respect to the slurry is mixed continuously. The addition of a disper-
reactivity of the clinker. Unfortunately, no simple rule sant reduces the rheological impact of false sets
exists to determine the optimal calcium-sulfates con- (Chapter 3).
tent, because this depends upon a variety of parameters,
including cement particle-size distribution, calcium-sul-
fates reactivity, alkali content, and aluminate-phase con- 2-5.4 Effects of aging
tent (Lerch, 1946; Ost, 1974). The performance of Portland cement can be affected sig-
Because of the heat generated during the grinding nificantly by exposure to the atmosphere and/or high
process at the cement mill, the gypsum added as calcium temperatures during storage in sacks or silos. The prin-
sulfate in Portland cement is dehydrated to a variable cipal effects upon neat cement slurries (no additives)
extent. In some cases, calcium sulfate hemihydrate include the following (Silk, 1986).
–
(CS H1/2) is the only form of calcium sulfate present. At ■ Increased thickening time
–
ambient temperature, the solubility of CS H1/2 is approx- ■ Decreased compressive strength
imately three times that of gypsum; therefore, upon
■ Decreased heat of hydration
hydration, the aqueous phase of the cement slurry
quickly becomes supersaturated with respect to gypsum. ■ Increased slurry viscosity

38 Well Cementing
 The effects are principally caused by carbonation of 2-5.6 Influence of surface area
the calcium silicate hydrate phases and partial hydration The surface area (sometimes called fineness) is an
of the free CaO and C3A/C4AF. The rate at which these important parameter with respect to cement reactivity
processes occur is directly related to the relative humid- and slurry rheology. The fineness of cement is usually
ity of the storage environment. The effects of limited determined by measuring the air permeability of a small
cement exposure to air during transport operations have layer of lightly compacted cement (Blaine method)
been shown to be less severe (Cobb and Pace, 1985). (Appendix B). With the assumption that the cement par-
The water liberated during this reaction can prehy- ticles are spherical, such information is used to calcu-
drate the aluminate, silicate, and alkali phases. When late a theoretical surface area; however, this method
the cement is eventually hydrated in water, an imbal- underestimates the true surface area (Vidick et al.,
ance exists between the aluminates and sulfates, often 1987), as measured by the Brunauer-Emmett-Teller
leading to a false set. This reaction also causes the (BET) gas-adsorption method (Table 2-6).
cement to clump, hindering transport between contain-
ers.
When Portland cement is stored in hot regions, the
Table 2-6. Surface Area of Anhydrous Class G Cements, as
temperature in the silo can be sufficiently high to dehy- Measured by Two Techniques†
drate the gypsum (Locher et al., 1980). Such cements
would be more apt to exhibit the false-set phenomenon. Sample Surface Area (m2/g)
Adding hot cement (more than 200°F [93°C]) to the silo Blaine BET
can have a similar effect. Thus, when designing cement
A 0.2 0.8
systems for a particular job, it is always prudent to per-
form the laboratory tests with samples of the cement to B 0.3 0.5
be used at the wellsite.
C 0.4 1.2
† From Vidick et al., 1987. Reprinted with permission from Elsevier.
2-5.5 Influence of alkalis
The principal alkaline elements found in Portland
The water-to-cement ratio required to wet the
cement are sodium and potassium. They affect setting
cement particles and prepare a pumpable slurry is
and strength development; thus, the amounts of these
directly related to the surface area (Sprung et al., 1985).
substances are usually held below 1% (expressed as
Thus, for consistency of performance, the fineness is
oxides). In well cements the maximum concentration of
controlled by the cement manufacturer.
total alkalis is 0.75% (as a convention, these are
The development of compressive strength is often
expressed as total equivalent Na2O). Total alkalis are the
correlated with the cement’s surface area (Bakchoutov
sum of insoluble alkalis (impurities in the lattices of
et al., 1980, and Frigione and Marra, 1976). Generally,
clinker phases) and soluble alkalis (in the form of alkali
the results indicate that cements with high fineness
sulfates).
tend to develop greater compressive strength. Some
The effects of alkalis upon strength development are
research shows that the rate of hydration is accelerated
unpredictable. Alkalis have been shown to improve com-
by high surface area, but that it is difficult to separate
pressive strength (Sudakas et al., 1978) and to be dele-
the effects of fineness from those of chemical composi-
terious. Jawed and Skalny (1978) demonstrated a posi-
tion. Hunt (1986) and Hunt and Elspass (1986), working
tive effect upon early strength but a negative effect upon
with a selection of well cements, found a good correla-
long-term strength.
tion between Blaine fineness and thickening time (Fig.
If sufficient potassium sulfate is present as an impu-
2-18).
rity in the cement, a reaction with gypsum can occur,
resulting in the formation of syngenite.
2CaSO 4 i2H2 O+K 2SO 4 →

( ) 2
1
CaK 2 SO 4 iH2 O+CaSO 4 i H2 O+2.5H2 O
2 (2-14)
syngenite

Chapter 2 Chemistry and Characterization of Portland Cement 39

 160 Portland cements with low C3A contents are less sus-
Supplier 1
140 Supplier 2 ceptible to sulfate attack after setting (American
Supplier 3 Petroleum Institute [API], 1955). In addition, because
120 Other suppliers
Least squares the solubility of magnesium sulfate and sodium sulfate is
API 100
Schedule 5
low above 140°F [60°C], sulfate attack is not normally a
G serious problem at that temperature or higher (Suman
thickening 80
time (min) A
and Ellis, 1977). In any event, as discussed in Chapter 3,
60
A
sulfate attack can be substantially reduced by the addi-
40 A
tion of pozzolanic materials such as fly ash to the cement
20 system.
180 200 220 240 260 280 300 320 340 360 380
Blaine fineness (m2/kg)

Fig. 2-18. Linear regression of thickening time and Blaine fineness

2-6 Advanced techniques for monitoring
from Class A and Class G cements (from Hunt, 1986). Reprinted with and modeling cement hydration
permission from Elsevier.
The previous discussion showed that Portland cement is
mineralogically complicated, and the hydration chem-
istry is not sufficiently understood. This obscurity is par-
2-5.7 Sulfate resistance tially a result of the lack of scientific techniques for
directly observing the chemical changes that occur in
Downhole brines commonly contain magnesium sulfate cement slurries over the early minutes and hours after
and sodium sulfate, and detrimental effects can result mixing.
when such solutions react with certain cement hydra- During the 1990s materials scientists began to apply
tion products. These sulfates react with precipitated cal- some advanced analytical techniques toward acquiring a
cium hydroxide to form magnesium hydroxide and better understanding of cement hydration (Luke et al.,
sodium hydroxide as well as calcium sulfate. The cal- 1995). Such techniques included
cium sulfate can in turn react with the aluminates to
■ environmental scanning electron microscopy (Mehta
form secondary ettringite.
et al., 1994)
( )2
Ca OH + MgSO 4 + 2H2 O → ■ Fourier transform infrared (FTIR) spectroscopy
(Hughes et al., 1994a and 1994b; Fletcher et al., 1995)
CaSO 4 i2H2 O+Mg OH ( )2 (2-15) ■ nuclear magnetic resonance spectroscopy (Skibsted
et al., 1993; Brough et al., 1994)
Swelling occurs because of the replacement of
■ electron energy loss spectroscopy (Byrdson et al.,
Ca(OH)2 by Mg(OH)2.
1994)
( )2
Ca OH + Na 2SO 4 + 2H2 O → ■ synchrotron energy dispersive X-ray diffraction
(XRD) (Kuzel and Pöllmann, 1991; Meller et al.,
CaSO 4 i2H2 O+2NaOH (2-16) 2004).

An increase in cement porosity occurs, because NaOH These experimental approaches are complemented
is much more soluble than Ca(OH)2. by a variety of modeling methods (Bentz et al., 1994).
This work has brought the study of Portland cement into
( )
3CaOi Al 2 O3 i6H2 O+3 CaSO 4 i2H2 O + 20H2 O → the mainstream of materials science.
For well cementing, such methods have the potential
3CaOi Al 2 O3 i3CaSO 4 i32H2 O (2-17) to open a window on the chemical changes that occur in
or cement slurries from just seconds after mixing through-
out the life of the well. As work continues to further the
C3 AH6 + 3 CSH2 + 20H → C3 A i3 CSi32H understanding of cement hydration, fine-tuning of slurry
design and, most importantly, improving the reliability of
When ettringite forms after the cement has devel- cement jobs may be possible. Some notable studies are
oped strength, an expansion occurs. As discussed in highlighted below.
Chapter 7, a limited amount of expansion can be benefi-
cial in terms of bonding; however, uncontrolled cement
expansion leads to loss of compressive strength, crack-
ing, and damage to tubulars.

40 Well Cementing
2-6.1 FTIR spectroscopy identified as α-C2SH, a common phase that can form in
FTIR spectroscopy has long been used for the qualitative well cements exposed to high temperatures (Chapter
identification of minerals. Until recently, the applica- 10). The spectra also show changes in the sulfate bands.
tions for Portland cement were limited to characterizing The strong AFt band (at 1,110 cm–1), formed during
the main clinker phases. Furthermore, sensitivity was mixing, broadens and weakens with time to form a shoul-
relatively low for the accessory minerals (such as sul- der around 111 cm–1.
fates) that strongly influence the properties of commer- The performance of Portland cements that meet the
cial cements. During the 1990s, important advances in requirements for a given API or International Organ-
spectrometer design and chemometric signal processing ization for Standardization (ISO) class (Section 2-7) is
brought radical changes to FTIR analysis. highly variable, especially when cement additives such
Hughes et al. (1994a) showed that diffuse reflectance as retarders are present in the slurry (Chapter 3).
FTIR spectroscopy allowed highly reproducible cement Therefore, prejob performance testing of a given
spectra to be obtained. Sensitivity to the accessory min- cement-slurry formulation is essential to prevent opera-
erals was also vastly improved. Using attenuated total tional difficulties. To reduce the testing requirements,
reflectance optics, Hughes et al. (1994b) extended the Fletcher et al. (1995) showed that the FTIR spectrum of
method to explore hydration chemistry. a cement yields information about the factors that con-
The spectral absorption bands that reveal the chem- trol hydration and performance variability. In essence,
istry of cement hydration lie in four distinct regions of the FTIR spectrum provides a “signature” of cement
the midinfrared spectrum. These are the Si-O stretching composition and performance (Fig. 2-20). Statistical
modes around 600–1,000 cm–1; the S-O modes around models, based on linear statistics and artificial neural
800–1,200 cm–1; the H-O-H bending modes associated networks, allow prediction of cement composition,
with free water at about 1,635 cm–1; and the broad and extent of aging, and particle-size distribution. Case stud-
complex band at 3,400–3,600 cm–1 that arises from O-H ies have demonstrated that the spectra have useful pre-
stretching vibrations both in water and OH-containing dictive value, particularly in identifying anomalous or
minerals. Figure 2-19 shows the evolution of the mid- “rogue” cements that can be expected to perform very
infrared spectrum of a neat Class G cement slurry at differently than normal. Despite the promise of this
300°F [150°C] and 2,000 psi [14 MPa]. Hydration was technique, FTIR spectrometers are not yet common in
monitored for 7 hr. oilfield laboratories in which cement-slurry design is per-
The rather amorphous C-S-H phase formed at lower formed, and the method has not been widely developed.
temperatures progressively gives way to distinct crys-
talline hydrates. This is shown very clearly in Fig. 2-19, in
Silicates
which the band at 943 cm–1 is replaced by strong and 0.80 Total sulfate
well-defined bands at 972, 928, and 855 cm–1. These are 0.70 Calcium carbonate
0.60 Calcium
Bassanite
0.50 hydroxide and gypsum
0.45 Absorbance 0.40 Bassanite Fe/Al
0.40 0.30 ratio
Gypsum
0.35 0.20
0.30 0.10 Syngenite

0.25 0.00
Absorbance 4,500 4,000 3,500 3,000 2,500 2,000 1,500 1,000 500
0.20
0.15 Wave number (cm–1)

0.10 Fig. 2-20. Typical FTIR spectrum of Portland cement (from Fletcher
0.05 et al., 1995). Reprinted with permission of SPE.
0.00
4,000 3,500 3,000 2,500 2,000 1,500 1,000 500
Wave number (cm–1)

Fig. 2-19. FTIR-attenuated total reflectance spectra of a Class G well

cement slurry at 300°F [150°C] and 2,000 psi [14 MPa]. Spectra are
shown at intervals of 38 min, extending over 7 hr from mixing.
Successive spectra are offset upwards by 0.02 absorbance units.

Chapter 2 Chemistry and Characterization of Portland Cement 41

2-6.2 XRD techniques 5,000 x 5,000 x o
30°C 100°C
Powder XRD methods have been widely used by cement o
chemists for mineralogical analysis of clinkers, anhy- 4,000 + +x k
4,000 x
xx xx
drous cements, and individual cement minerals, includ-
ing hydrates. Such techniques are routinely used to Total 3,000 3,000
counts
determine the “potential phase composition” of Portland with
cements (Section 2-7), one of the criteria used for 2,000 2,000
offset
cement classification.
1,000 1,000
Kuzel and Pöllman (1991) demonstrated that, using a
wet cell, XRD techniques could be applied toward moni- +
0 0 +
toring cement hydration. They used a time-lapse tech- 1 1.5 2 2.5 1 1.5 2 2.5
nique to track tricalcium aluminate reactions as they
occurred during cement hydration. This technique (d/nm)–1 (d/nm)–1
requires rapid data acquisition to follow the reactions 35
with sufficient detail; therefore, they used an intense
30 C4AF at 30°C
synchrotron X-ray source combined with energy-disper-
C4AF at 100°C
sive detectors. A sample result (Fig. 2-21) shows the 25
gradual development of portlandite and ettringite, cre- Corrected
peak 20
ated as C3A reacts with gypsum. When all the gypsum is area:
15 C3AH6 at
consumed after about 60 hr, it seems to trigger the trans- Total 100°C
formation of ettringite (AFt) to monosulfate (AFm). counts/103 10
C2AH8 at 100°C
5
C2AH8 at 30°C
Monosulfate Gypsum 0
Portlandite Tricalcium aluminate 0 20 40 60 80 100 120 140
Ettringite
Time (min)

83 Fig. 2-22. Hydration of C4AF at 30 and 100°C (from Meller et al., 2004).
77
72 Key: x = C4AF; + = C2AH8 /C4AH19; o = C3AH6. Reproduced by permis-
66 sion of the Royal Society of Chemistry.
Counts 60
54 Hours of
48
39 hydration
28
14
4 5 6 7 8 9 10 11 When gypsum is added to C4AF at 86°F [30°C], ettrin-
Angle (degrees) gite is observed to form within 6 min of contact with
water (Fig. 2-23). Ettringite increases throughout the
Fig. 2-21. Tracking aluminate reactions in hydrating cement using 3-hr observation time, and both the C4AF and gypsum
time-lapse XRD. decrease.
Also using synchrotron energy dispersive XRD, Meller
et al. (2004) followed the hydration of the ferrite phase
at various temperatures, with and without gypsum. The
results without gypsum are shown in Fig. 2-22. The spec-
tra on the top illustrate the mineralogical changes over
a 2-hr period. The evolution of C4AF to the metastable
AFm phases and the hydrogarnet phase is clear. Twenty-
four patterns were collected, one every 5 min.

42 Well Cementing
 18,000 x
30°C e 70°C m
16,000 g

14,000 e e

12,000
Total
counts 10,000
with 8,000
offset 6,000
4,000
2,000
0
0.8 1 1.2 1.4
(d/nm) –1

12,000
30°C
10,000 C4AF

8,000
Corrected
Gypsum
peak 6,000
intensity
4,000
Fig. 2-24. Four steps in a two-dimensional cement hydration simula-
2,000 Ettringite tion. See text for color guide.

0
0 20 40 60 80 100 120 140 160 180
Some time later in 2, the green diffusing particles
Time (min)
turn into C-S-H gel (also yellow) when they either return
Fig. 2-23. Hydration of C4AF + gypsum at 30°C (from Meller et al., to a cement grain or meet already formed gel.
2004). Key: g = gypsum; e = ettringite; x = C4AF; k = polymer cell. Production of new gel is accompanied by the production
Reproduced by permission of the Royal Society of Chemistry. of portlandite (blue), which grows in clusters in the
water (black).
In 3 (equivalent to 20% hydration), most of the
cement grains have yet to react. A layer of C-S-H gel
2-6.3 Computer modeling of Portland cement (yellow) surrounds most of the grains. There are well-
hydration formed portlandite (blue) clusters and much remaining
Using computer modeling techniques, Garboczi and water (black).
Bentz (1991) simulated all of the steps in Portland In 4 (equivalent to 75% hydration), the cement grains
cement hydration. The simulation can be either three- are smaller and the C-S-H gel (yellow) and portlandite
dimensional, in which cement particles are modeled as (blue) almost fill the void space. The simulation pro-
spheres, or two-dimensional, in which each particle can duces images remarkably similar to scanning electron
be given an individual shape based on an electron pho- microscope (SEM) photographs of partially hydrated
tomicrograph of a cement sample. Particle-size distribu- cement (Fig. 2-25).
tion can be varied at will, but generally the mean parti- More recently, the aluminate phases have been added
cle diameter is 15–30 μm. The cement particles include to the simulator (Bentz et al., 1994). Figure 2-26 shows a
ground clinker and added gypsum. A simulation limited simulation of a partially hydrated C3S/C3A/gypsum
to the silicate phases is illustrated in Fig. 2-24. system.
As the simulation begins in 1, the computer identifies
image pixels (yellow) on the surfaces of the cement
grains (orange) and lets them dissolve as particles
(green) that are free to diffuse in the water (black). The
right part of 1 shows the model before any dissolution
and the left half after dissolution.

Chapter 2 Chemistry and Characterization of Portland Cement 43

 2-7 Classification of Portland cements
Portland cements are manufactured to meet certain
chemical and physical standards that depend upon their
application. To promote consistency of performance
among cement manufacturers, classification systems
and specifications have been established by various user
groups. The best known systems are those of ASTM
International (formerly the American Society for Testing
and Materials) (ASTM C 150, Standard Specification
for Portland Cement) and API (API Spec 10A,
Specification for Cements and Materials for Well
Cementing). The API classification scheme has been
adopted by the ISO as Standard 10426-1, Petroleum and
natural gas industries - Cements and materials for
well cementing - Part 1: Specification.

2-7.1 Classification criteria

The principal chemical criterion for classifying Portland
cements is the relative distribution of the main clinker
Fig. 2-25. SEM photomicrograph of 7-day old cement paste (top) phases, known as the “potential phase composition.”
compared with simulated hydration (bottom). Limits on the amounts of alkalis, free CaO, MgO, SO3,
and insoluble residue as well as the loss on ignition
(weight loss after burning) are also specified for some
classes of Portland cements.
The most widely accepted method of expressing the
relative amounts of the principal clinker phases relies
upon a series of calculations based upon the oxide com-
position of the cement. This method, first introduced by
Bogue (1929), is based upon various phase equilibria
relationships between the cement components. Bogue’s
method suffers from various limitations, but it nonethe-
less remains a yardstick by which cements are classified.
The Bogue equations are listed in the sidebar on the
next page. These equations are based on the assumption
that chemical equilibrium is established at the clinker-
ing temperature and is maintained throughout the criti-
cal cooling period. This introduces errors that can give
seriously incorrect results (Taylor, 1964; Aldridge, 1982).
The Bogue calculation has been refined by Taylor
(1989) by using more realistic phase compositions, with
the result that the calculated mineralogical composition
is more nearly representative of the true mineralogical
Fig. 2-26. Three-dimensional cement hydration simulation involving composition, as determined by quantitative XRD analy-
both silicate and aluminate phases (red = C3S; purple = C3A; green sis. Various other corrections have been proposed, but
= gypsum) (Bentz et al., 1994). Reprinted with permission of IOP none has yet received general acceptance.
Publishing Limited.

44 Well Cementing
 Bogue equations for calculating potential phase composition 2-7.2 API and ISO classification systems
(from ASTM Method C114)† The requirements for well cements are more rigorous
than those for construction cements. Well cements must
When the ratio of percentages of aluminum oxide to ferric perform over a wide range of temperatures and pres-
oxide is 0.64 or more, the percentages of tricalcium silicate,
dicalcium silicate, tricalcium aluminate, and tetracalcium alu-
sures and are exposed to subterranean conditions that
minoferrite shall be calculated from the chemical analysis as construction cements do not encounter. Well cements
follows: require greater consistency from batch to batch to
Tricalcium silicate =(4.071 × %CaO) – (7.600 × %SiO2) ensure predictable performance when various cement
– (6.718 × %Al2O3) – (1.430 × %Fe2O3) additives are introduced (Chapter 3).
– (2.852 × %SO3) There are currently eight classes of API-ISO Portland
Dicalcium silicate = (2.867 × %SiO2) – (0.7544 × %C3S) cements, designated A through H. They are arranged
according to the depths at which they are placed and the
Tricalcium aluminate = (2.650 × %Al2O3) – (1.692 × %Fe2O3)
temperatures and pressures to which they are exposed.
Tetracalcium aluminoferrite = 3.043 × %Fe2O3 Within some classes, cements with varying degrees of
When the alumina-ferric oxide ratio is less than 0.64, a sulfate resistance (as determined by C3A content) are
calcium aluminoferrite solid solution [expressed as ss(C4AF + sanctioned: ordinary (O), moderate sulfate resistance
C2F)] is formed. Contents of this solid solution and of tricalcium (MSR), and high sulfate resistance (HSR). The chemical
silicate shall be calculated by the following formulas: and physical specifications are listed in Tables 2-7 and
ss(C4AF + C2F) = (2.100 × %Al2O3) + (1.702 × %Fe2O3) 2-8. Table 2-9 lists typical compositions and surface-area
Tricalcium silicate =(4.071 × %CaO) – (7.600 × %SiO2) ranges for certain API cements. Below is a general
– (4.479 × %Al2O3) – (2.859 × %Fe2O3) description of each API class, with its ASTM equivalent
– (2.852 × %SO3). when appropriate.
No tricalcium aluminate will be present in cements of this Classes A, B, and C are products obtained by grinding
composition. Dicalcium silicate shall be calculated as previ- Portland cement clinker, consisting essentially of
ously shown. hydraulic calcium silicates, usually containing one or
In the calculation of C3A, the values of Al2O3 and Fe2O3 more of the forms of calcium sulfate as an interground
determined to the nearest 0.01% shall be used. In the calcula- addition. At the option of the manufacturer, processing
tion of other compounds, the oxides determined to the nearest
0.1% shall be used. All values calculated as described above additions† may be used in the manufacture of the
shall be reported to the nearest 1%. cement, provided such materials in the amounts used
have been shown to meet the requirements of ASTM C
† Extractedwith permission, from C114-05 Standard Test Methods for Chemical 465, Standard specification for processing additions
Analysis of Hydraulic Cement, copyright ASTM International. for use in the manufacture of Portland cement.
Class A: Intended for use when special properties are
not required. Available only in O grade (similar
An improved technique for determining the phase to ASTM C 150, Type I).
composition of Portland cement is the Rietveld method Class B: Intended for use when conditions require mod-
for powder XRD (Young, 1995). The Rietveld method erate or high sulfate resistance. Available in
allows standardization of powder diffraction analysis both MSR and HSR grades (similar to ASTM C
through the use of calculated reference diffraction pat- 150, Type II)
terns based upon crystal structure models. The result is Class C: Intended for use when conditions require high
a set of refined crystal structure models for each phase early strength. Available in O, MSR, and HSR
in the clinker. From these data, one can obtain pattern grades (similar to ASTM C 150, Type III).
intensity information that may be related to phase abun-
dance. Today, this method is being used for quality con-
trol in cement production and to analyze the NIST refer-
ence clinkers.
Physical parameters that appear in specifications
include the fineness of the cement and the performance
of the cement according to standardized tests. The per-
formance tests include measurements of thickening
time, compressive strength, expansion, and free water.
Appendix B presents a complete description of the test
†Asuitable processing addition or set-modifying agent will not prevent a well
methods and equipment.
cement from performing its intended fuctions.

Chapter 2 Chemistry and Characterization of Portland Cement 45

Table 2-7. Chemical Requirements for API Portland Cements†
Cement Class

A B C D, E, F G H

(wt%)
Ordinary grade (O)
Max. magnesium oxide (MgO) 6.0 6.0

Max. Sulfur trioxide (SO3) 3.5 4.5

Max. loss on ignition 3.0 3.0

Max. insoluble residue 0.75 0.75

Max. tricalcium aluminate (3CaO • Al2O3) 15

Moderate sulfate-resistant grade (MSR)

Max. magnesium oxide (MgO) 6.0 6.0 6.0 6.0 6.0

Max. sulfur trioxide (SO3) 3.0 3.5 3.0 3.0 3.0

Max. loss on ignition 3.0 3.0 3.0 3.0 3.0

Max. insoluble residue 0.75 0.75 0.75 0.75 0.75

Max. tricalcium silicate (3CaO • SiO2) 58 58

Min. tricalcium silicate (3CaO • SiO2) 48 48

Max. tricalcium aluminate (3CaO • Al2O3) 8 8 8 8 8

Max. total alkali content expressed as 0.75 0.75

sodium oxide (Na2O) equivalent

High sulfate-resistant grade (HSR)

Max. magnesium oxide (MgO) 6.0 6.0 6.0 6.0 6.0

Max. sulfur trioxide (SO3) 3.0 3.5 3.0 3.0 3.0

Max. loss on ignition 3.0 3.0 3.0 3.0 3.0

Max. insoluble residue 0.75 0.75 0.75 0.75 0.75

Max. tricalcium silicate (3CaO • SiO2) 65 65

Min. tricalcium silicate (3CaO • SiO2) 48 48

Max. tricalcium aluminate (3CaO • Al2O3) 3 3 3 3 3

Max. tetracalcium aluminoferrite (4CaO • Al2O3 • Fe2O3) 24 24 24 24 24

plus twice the tricalcium aluminate (3CaO • Al2O3)

Max. total alkali content expressed as sodium 0.75 0.75

oxide (Na2O) equivalent
† From API Specification 10A. Reproduced courtesy of the American Petroleum Institute.

46 Well Cementing
Table 2-8. Physical Requirements for API Portland Cements†
A B C D E F G H
Mix water (percent BWOC‡) 46 46 56 38 38 38 44 38

Min. Blaine fineness (specific surface area), tubidimeter (m 2/kg) 150 160 220 –§ – – – –

Min. fineness (specific surface area), air permeability (m2/kg) 280 280 400 – – – – –

Max. free fluid content (mL) – – – – – – 3.5 3.5

Compressive strength test

8-hr Curing Time Schedule Curing Curing Min. Compressive Strength (psi [MPa])
Number Temperature Pressure
(°F [°C]) (psi [kPa])
– 100 [38] Atmospheric 250 [1.7] 200 [1.4] 300 [2.1] – – – 300 [2.1] 300 [2.1]

– 140 [60] Atmospheric – – – – – – 1,500 [10.3] 1,500 [10.3]

6S 230 [110] 3,000 [20,700] – – – 500 [3.5] – – – –

8S 290 [143] 3,000 [20,700] – – – – 500 [3.5] – – –

9S 320 [160] 3,000 [20,700] – – – – – 500 [3.5] – –

24-hr Curing Time Schedule Curing Curing Min. Compressive Strength (psi [MPa])
Number Temperature Pressure
(°F [°C]) (psi [kPa])
– 100 [38] Atmospheric 1,800 [12.4] 1,500 [10.3] 2,000 [13.8] – – – – –

4S 170 [77] 3,000 [20,700] – – – 1,000 [6.9] 1,000 [6.9 – – –

6S 230 [110] 3,000 [20,700] – – – 2,000 [13.8] – 1,000 [6.9] – –

8S 290 [143] 3,000 [20,700] – – – – 2,000 [13.8] – – –

9S 320 [160] 3,000 [20,700] – – – – – 1,000 [6.9] – –

Pressure temperature thickening time test

Specification Maximum Min. Thickening Time (min)
Test Consistency,
Schedule 15–30 min Stirring
Number Period (Bc)
4 30 90 90 90 90 – – – – –

5 30 – – – – – – – 90 90

5 30 – – – – – – – 120 max. 120 max.

6 30 – – – 100 100 100 – – –

8 30 – – – – 154 – – – –

9 30 – – – – – 190 – – –
† From API Spec 10A. Reproduced courtesy of the American Petroleum Institute.
‡ By weight of water
§ – = not available

Table 2-9. Typical Composition and Fineness of API Cements†

API Class ASTM Type Typical Potential Phase Composition (wt%) Typical Blaine
Fineness (cm2/g)
C3S β-C2S C3A C4AF
A I 45 27 11 8 1,600
B II 44 31 5 13 1,600
C III 53 19 11 9 2,200
D –‡ 28 49 4 12 1,500
E – 38 43 4 9 1,500
G Nominal II 50 30 5 12 1,800

H Nominal II 50 30 5 12 1,600
† From Nelson, 1983. Reprinted with permission from Oil & Gas Journal.
‡ Not applicable

Chapter 2 Chemistry and Characterization of Portland Cement 47

Classes D, E, and F: 2-8 Acronym list
The products obtained by grinding Portland AFm Aluminoferrite monosulfate
cement clinker, consisting essentially of
hydraulic calcium silicates, usually containing AFt Aluminoferrite trisulfate
one or more of the forms of calcium sulfate as API American Petroleum Institute
an interground addition. At the option of the BET Brunauer-Emmett-Teller (gas adsorption
manufacturer, processing additions may be method of determining surface area)
used in the manufacture of the cement, pro- BWOW By weight of water
vided such materials in the amounts used have
been shown to meet the requirements of ASTM C-S-H Calcium silicate hydrate
C 465. Further, at the option of the manufac- FTIR Fourier transform infrared (spectroscopy)
turer, suitable set-modifying agents may be HSR High sulfate resistance
interground or blended during manufacture. ISO International Organization for
This product is intended for use under condi- Standardization
tions of moderately high temperatures and
pressures. Available in MSR and HSR grades. MSR Moderate sulfate resistance
Classes D, E, and F are also known as NIST National Institute of Standards and Technology
“retarded cements,” intended for use in deeper O Ordinary (an indication of cement sulfate
wells. The retardation is accomplished by sig- resistance)
nificantly reducing the amount of faster-hydrat- OPC Ordinary Portland cement
ing phases (C3S and C3A) and increasing the
particle size of the cement grains. Since these SEM Scanning electron microscope
classes were first manufactured, the technology XRD X-ray diffraction
of chemical retarders has significantly
improved; consequently, these classes are
rarely found today.
Classes G and H:
The products obtained by grinding Portland
cement clinker, consisting essentially of
hydraulic calcium silicates, usually containing
one or more of the forms of calcium sulfate as
an interground addition. No additions other
than calcium sulfate or water, or both, shall be
interground or blended with clinker during
manufacture of Class G well cement. This prod-
uct is intended for use as a basic well cement.
Available in MSR and HSR grades.
Classes G and H were developed in response
to the improved technology in slurry accelera-
tion and retardation by chemical means. The
cement manufacturer is prohibited from adding
special chemicals, such as glycols or acetates,
to the clinker. Such chemicals improve the effi-
ciency of grinding but have been shown to inter-
fere with various cement additives. Classes G
and H are by far the most commonly used well
cements today.

48 Well Cementing
3-1 Introduction
Portland cement systems for well cementing are rou-
tinely designed to perform at temperatures ranging from
Cement Additives and
Mechanisms of Action
Erik B. Nelson, Michel Michaux, and Bruno Drochon—Schlumberger 3
that affect additive performance, additive examples, and
proposed mechanisms of action are discussed in detail.
A thorough review of Chapter 2 is recommended before
below freezing in permafrost zones to 700°F [350°C] in reading this chapter.
thermal-recovery and geothermal wells. Well cements
encounter the pressure range from near ambient in shal-
low wells to more than 30,000 psi [200 MPa] in deep 3-2 Variability of additive response
wells. In addition to severe temperatures and pressures, Typical performance data for many additives are pre-
well cements must often be designed to contend with sented throughout this chapter. It is important for the
weak or porous formations, corrosive fluids, and over- reader to understand that this information is presented
pressured formation fluids. Cement additives make it solely to illustrate general trends and should not be used
possible to accommodate such a wide range of condi- for design purposes. Most additives are strongly influ-
tions. Additives modify the behavior of the cement enced by the chemical and physical properties of the
system, ideally allowing successful slurry placement cement, which are highly variable even within a given
between the casing and the formation, rapid compres- American Petroleum Institute (API) or International
sive strength development, and adequate zonal isolation Organization for Standardization (ISO) classification.
during the lifetime of the well. Consequently, a wide spectrum of results can be
Today more than 100 additives for well cements are obtained with the same slurry design. The important
available, many of which can be supplied in solid or cement parameters include the following:
liquid forms. There are eight major categories of additives. ■ particle-size distribution
1. Accelerators: chemicals that reduce the setting time ■ distribution of silicate and aluminate phases
of a cement system and increase the rate of compres- ■ reactivity of hydrating phases
sive strength development
■ gypsum/hemihydrate ratio and total sulfate content
2. Retarders: chemicals that delay the setting time of a
■ free alkali content
cement system
■ chemical nature, quantity, and specific surface area
3. Extenders: materials that lower the density of a
cement system, reduce the quantity of cement per of initial hydration products.
unit volume of set product, or both Other important parameters that influence additive
4. Weighting agents: materials that increase the density performance include temperature, pressure, additive
of a cement system concentration, mixing energy, mixing order, and water-
5. Dispersants: chemicals that reduce the viscosity of a to-cement ratio.
cement slurry Figure 3-1 illustrates the variability of additive
response to cements. The figure compares the hydration
6. Fluid-loss control agents: materials that control leak-
behavior of two API/ISO Class G cements. Conduction
age of the aqueous phase of a cement system to the
calorimetry curves were generated for the neat slurries
formation
and for three additional slurries containing an accelera-
7. Lost-circulation control agents: materials that con- tor, a retarder, or a dispersant. Scrutiny of the curves
trol loss of the cement slurry to weak or vugular for- reveals significant differences in hydration behavior.
mations Because of the complexity of the cement hydration
8. Specialty additives: miscellaneous additives, such as process and the large number of parameters involved,
antifoam agents, fibers, and flexible particles. the only practical cement slurry design method (to avoid
unpleasant surprises at the wellsite) is thorough labora-
Each of the above categories is discussed individually
tory testing before the job. Computer-based databases
in this chapter. The physical and chemical phenomena

Well Cementing ■ Chapter 3 Cement Additives and Mechanisms of Action 49

 Neat
Cement with 1% BWOC CaCl2 accelerator
Cement with 0.3% BWOC PNS dispersant
Cement with 0.05% BWOC retarder

Cement A Cement B
35 35

30 30

25 25

20 20
Heat flow Heat flow
(mW) 15 (mW) 15

10 10

5 5

0 0
0 2 4 6 8 10 12 14 16 18 20 22 24 0 2 4 6 8 10 12 14 16 18 20 22 24
Time (hr) Time (hr)

Fig. 3-1. Calorimetric behavior of Cements A and B in the presence of different additives.

and expert systems can propose “first-guess” cement for- Among the chlorides, the accelerating effect becomes
mulations that are consistent with the required perfor- stronger as the valence and ionic radius of the accompa-
mance criteria (Chapter 12). This can help reduce the nying cation increases (Skalny and Maycock, 1975).
number of tests required to achieve the final design. It is Edwards and Angstadt (1966) suggested that cations
essential that the tests be performed with a representa- and anions may be ranked according to their efficiency
tive sample of the cement and mix water to be used as accelerators for Portland cement.
during the cement job.
Ca2+ > Mg2+ > Li+ > Na+ > H2O
OH– > Cl– > Br– > NO3– > SO42– = H2O
3-3 Accelerators Calcium chloride (CaCl2) is undoubtedly the most
Accelerators are added to cement slurries to shorten the efficient and economical of all accelerators. Regardless
setting time (Stages I and II of the hydration scheme of concentration, it always acts as an accelerator
described in Chapter 2), accelerate the hardening (Table 3-1). It is normally added at concentrations
process (Stages III and IV), or both. They are often used between 2% to 4% by weight of cement (BWOC)
to counteract the set delay caused by other additives, (Appendix C). Results are unpredictable at concentra-
such as dispersants and fluid-loss-control agents (Odler tions exceeding 6% BWOC, and premature setting may
et al., 1978). occur.
Sodium chloride affects the thickening time and com-
3-3.1 Examples pressive strength development of Portland cement in
different ways, depending upon its concentration and
Many inorganic salts are accelerators of Portland the curing temperature (Fig. 3-2). NaCl acts as an accel-
cement. Chloride salts are used most frequently. Other erator at concentrations up to 15% by weight of mix
salts that have an accelerating effect include carbon- water (BWOW). Between 15% and 20% BWOW, NaCl is
ates, silicates (especially sodium silicate), aluminates, essentially neutral, and thickening times are similar to
nitrates, sulfates, thiosulfates, and alkaline bases [e.g., those obtained with fresh water. The addition of NaCl at
sodium hydroxide (NaOH), potassium hydroxide (KOH),
and ammonium hydroxide (NH4OH)].

50 Well Cementing
Table 3-1. Effects of Calcium Chloride on the Performance of Portland Cement Systems
Thickening Time of Neat Cement Slurries Compressive Strength at Temperature and Time (psi)
Accelerated by Flake Calcium Chloride (hr:min)
CaCl2 91°F [33°C] 103°F [39°C] 113°F [45°C] 60°F [16°C] 80°F [27°C] 100°F [38°C]
(%BWOC)
6 hr 12 hr 24 hr 6 hr 12 hr 24 hr 6 hr 12 hr 24 hr
0 4:00 3:30 2:32 Not set 60 415 45 370 1,260 370 840 1,780

2 1:17 1:11 1:01 125 480 1,510 410 1,020 2,510 1,110 2,370 3,950

4 1:15 1:02 0:59 125 650 1,570 545 1,245 2,890 1,320 2,560 4,450

Thickening Time Compressive Strength

8 8,000
136°F [58°C] 180°F [82°C]
6 6,000 160°F [71°C]
Time to reach Compressive
100-Bc slurry 4 154°F [68°C] strength at 24 hr 4,000
consistency (hr) (psi) 140°F [60°C]
179°F [81°C]
2 210°F [99°C] 2,000

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30
NaCl in mix water (%BWOW) NaCl in mix water (%BWOW)

Fig. 3-2. Effect of sodium chloride on thickening time and compressive strength development.

concentrations above 20% BWOW causes retardation. 3-3.2 Calcium chloride—mechanisms of action
NaCl is not a very efficient accelerator and should be Calcium chloride is by far the most common accelerator
used only when CaCl2 is not available at the wellsite. of Portland cement. The mechanisms by which it oper-
Seawater is used extensively for mixing cement slur- ates are complex and still not completely understood.
ries at offshore locations. It contains up to 2.5 wt% NaCl, Several hypotheses described in the literature are sum-
resulting in acceleration. The presence of magnesium in marized below.
seawater (about 0.15 wt%) must also be taken into
account (Chapter 7).
Chloride-free accelerators, originally developed by 3-3.2.1 Effects of calcium chloride on the hydra-
the concrete industry to reduce the corrosion of rein- tion of principal Portland cement phases
forcing steel, are also used in well cements. Sodium sili- It has been proposed that calcium chloride increases
cate is normally used as a cement extender; however, it the hydration rate of the aluminate phases–gypsum
also has an accelerating effect. Sodium silicate reacts system (Bensted, 1978; Traetteberg and Grattan-Bellew,
with Ca2+ ions in the aqueous phase of the cement slurry 1975). Chloride ions enhance the formation of ettringite
to form additional calcium silicate hydrate until the gypsum is consumed (Tenoutasse, 1978).
(C-S-H) phase nuclei, thus hastening the end of the If free C3A remains, calcium monochloroaluminate
induction period. Other examples include alkaline (C3A • CaCl2 • 10H2O) forms. The more rapid setting of
earth formates, nitrates, nitrites, triethanolamine, and the cement slurry is also attributed to the crystalline
thiocyanates (Pauri et al., 1986; Ramachandran, 1973; shape of ettringite, which occurs as very fine needles
1976a). (Bensted, 1978; Young et al., 1973).
By contrast, Stein (1961a) and Edwards and Angstadt
(1966) concluded that accelerators do not promote the
hydration of C3A but instead accelerate the hydration of

Chapter 3 Cement Additives and Mechanisms of Action 51

C3S. This accelerating action of calcium chloride has 3-3.2.4 Change in aqueous phase composition
been confirmed by studying the hydration of pure C3S Michaux et al. (1989a and 1989b) showed that calcium
(Odler and Skalny, 1971a) and C2S (Collepardi and chloride strongly modifies the distribution of ionic
Massidda, 1973). species in the aqueous phase of well-cement slurries.
Chloride ions do not participate in the formation of
3-3.2.2 Change in C-S-H structure hydration products during the induction period; there-
fore, one observes a decreased concentration of hydroxyl
Many researchers have proposed that Portland cement
and sulfate ions and an increased concentration of cal-
hydration is controlled by the diffusion of water and
cium ions. Kurczyk and Schwiete (1960) proposed that
ionic species through the initial protective C-S-H phase
calcium chloride decreases the alkalinity of the aqueous
coating (Chapter 2). Therefore, the hydration rate
phase, resulting in acceleration.
should be influenced by the permeability of the coating.
Stadelmann and Wieker (1985) investigated the influ-
A morphological change of the C-S-H phase to a more
ence of a large number of inorganic salts on the hydra-
open, flocculated structure would enhance diffusion and
tion of C3S. They showed that C3S hydration is acceler-
accelerate hydration. Such a process has been con-
ated by increasing the solubility of calcium hydroxide in
firmed in studies of pure C3S (Odler and Skalny, 1971a;
the aqueous phase, e.g., by adding CaCl2. Conversely,
Ben-Dor and Perez, 1976; Traetteberg et al., 1974). The
retardation is observed when the solubility of calcium
C-S-H phase has a higher bulk lime-to-silica (C/S) ratio
hydroxide decreases, e.g., by adding high concentrations
and a “crumpled-foil” morphology rather than a spicular
of NaCl.
or needle-like appearance. In the presence of calcium
Wu and Young (1984) showed that the addition of cal-
chloride, C-S-H phase has a greater specific surface area
cium salts affects the dissolution rate of C3S. When the
(Collepardi and Marchese, 1972) and a greater degree
concentration of calcium in the aqueous phase is moni-
of silicate-anion polymerization (Hirljac et al., 1983). A
tored with time, the peak concentration always occurs
change in the pore-size distribution of hydrated C3S
earlier in the presence of chloride ions. Thus, precipita-
(Young et al., 1973; Skalny et al., 1971) and C2S (Odler
tion of calcium hydroxide (and the end of the induction
and Skalny, 1971b) has also been reported. The mor-
period) occurs earlier.
phology of calcium hydroxide (portlandite) is also
In conclusion, it is apparent that many factors inter-
affected by the presence of chloride ions (Berger and
act simultaneously in the acceleration of Portland
McGregor, 1972). The portlandite crystals become elon-
cement by calcium chloride. Both physical and chemical
gated.
phenomena are involved. The presence of chloride ions
alters the structure and increases the permeability of
3-3.2.3 Diffusion of chloride ions the C-S-H phase layer. In addition, calcium chloride sig-
Kondo et al. (1977) determined the diffusion rate of nificantly alters the distribution of ionic species in the
anions and cations from alkaline and alkaline-earth aqueous phase, resulting in a faster hydration rate.
chlorides through a set–Portland cement plate. They
concluded that the diffusion coefficient of the chloride 3-3.3 Secondary effects of calcium chloride
ion is much higher than that of the accompanying
In addition to acceleration of the initial set, several
cation. Because the chloride ions diffuse into the C-S-H
other effects are evident when calcium chloride is pre-
phase layer more quickly than the cations, a counterdif-
sent in a Portland cement system. Some effects are not
fusion of hydroxyl ions must occur to maintain the elec-
beneficial; therefore, calcium chloride should be used
trical balance. Therefore, the precipitation of port-
judiciously depending upon well conditions. A summary
landite, ending the induction period, takes place earlier.
of the more important secondary effects is given below.
Kondo et al. (1977) also established that only a small
amount of chloride is incorporated into the C-S-H lattice
but that it may be chemically bound to the C-S-H sur- 3-3.3.1 Heat of hydration
face. The presence of CaCl2 increases the rate of heat genera-
Others studied the effect of calcium chloride on C3S tion during the first hours after slurry mixing. If the well-
hydration. They concluded that the accelerating effect bore is thermally insulated to a sufficient degree, the
occurred because chloride ions have a smaller ionic size temperature of the cement, casing, and surrounding for-
than hydroxyl ions and would more easily diffuse into the mation can increase by as much as 50°–60°F [27°– 33°C]
C-S-H membrane. The resulting internal-pressure after slurry placement. An autoacceleration of hydration
increase takes place more quickly, causing the C-S-H results.
membrane to burst earlier.

52 Well Cementing
 More importantly, increased casing expansion occurs 3-3.3.6 Sulfate resistance
because of the temperature rise. Because steel casing Because the ultimate permeability of calcium chloride-
and cement do not have the same coefficient of thermal accelerated systems is greater, the resistance to aggres-
expansion, the casing may shrink away from the cement sive sulfate solutions is reduced (Shideler, 1952; Gouda
when the hydration heat eventually dissipates. This et al., 1973). However, as discussed in Chapter 2, the C3A
results in a so-called “thermal microannulus,” and zonal content of the cement is the principal controlling factor
isolation is compromised (Pilkington, 1988). Additional governing sulfate resistance.
research is needed to better quantify this effect and to
determine the most susceptible wellbore environments.
3-3.4 Chloride-free accelerators
3-3.3.2 Slurry rheology Although calcium chloride is the least expensive and
most effective accelerator, growing concern about its
According to Collepardi (1971), calcium chloride corrosion of the reinforcing steel embedded in Portland
increases the yield point of a cement slurry but initially cement concrete has led to the development of chloride-
does not affect the plastic viscosity. After a 30-min hydra- free additives. Casing corrosion from chloride accelera-
tion period at ambient conditions, the plastic viscosity tors may also be a concern during the life of a well.
begins to increase. Slurries containing calcium chloride Under normal conditions, steel reinforcement is pro-
also tend to have a greater degree of thixotropy; as a tected from corrosion (passivated) by the high pH of the
result, particle sedimentation is seldom a problem. surrounding concrete-pore solution. A thin protective
film of gamma ferric oxide (γ-Fe2O3) forms on the steel
3-3.3.3 Compressive strength development surface. The protective film prevents iron cations (Fe2+)
from entering the electrolyte and acts as a barrier to pre-
Calcium chloride significantly increases the rate of com-
vent oxygen anions (O2–) from contacting the steel sur-
pressive strength development during the first few days
face. If the passivation is compromised, corrosion of the
after slurry placement. The magnitude of this effect
reinforcement can occur at a high rate. The protective
depends upon the curing temperature and the CaCl2
film can be disrupted by a significant reduction of the
concentration (Table 3-1).
pore-solution pH because of carbonation or by the pene-
tration of aggressive ions such as chlorides to the steel-
3-3.3.4 Shrinkage concrete interface.
Calcium chloride has been shown to increase volumetric A plethora of inorganic and organic compounds,
shrinkage by 10% to 50% in concretes (Shideler, 1952). including alkali carbonates, alkali silicates, alkali alumi-
This is mainly owing to the greater degree of hydration nates, alkali sulfates, alkali hydroxides, nitrates, nitrites,
and changes in hydration products (Collepardi and thiocyanates, thiosulfates, formates, and alkanolamines,
Massidda, 1973). Such data cannot be directly translated have been evaluated as calcium-chloride replacements.
to well cements, because the service conditions are very However, very few have performed well enough to be
different. To the authors’ knowledge, a thorough investi- used on an industrial scale. Unlike calcium chloride,
gation of the dimensional stability of calcium chloride- which is generally added alone, most of the commercial
accelerated well cements has not been performed. The chloride-free accelerators are formulated and contain
magnitude of the shrinkage effect in concrete and the several components.
popularity of calcium chloride as an accelerator for well Calcium formate, Ca(HCOO)2, was patented as a
cements suggest that such a study is overdue. cement accelerator in 1965. Owing to its low solubility
in water, calcium formate is usually sold as a powder.
The normal dosage is 1–2% BWOC. Calcium formate
3-3.3.5 Permeability accelerates the hydration and setting of all types of
Initially, the permeability of set cement containing cal- Portland cement, but its effect is not significant within
cium chloride is reduced. This is caused by the greater the first 24 hr. Early strength development can be
volume of hydration products compared to an additive- improved by including sodium nitrite as an accelerator
free cement. Later, when the degree of hydration is sim- aid (Rosskopf et al., 1975).
ilar for both systems, the permeability of the set cement The use of calcium nitrite, Ca(NO2)2, as an accelera-
containing CaCl2 is greater than that of its additive-free tor was patented by Angstadt and Hurley in 1963. Its
counterpart (Gouda et al., 1973). water solubility is very high; therefore, it can be used as
a liquid additive. The effect of calcium nitrite on
strength development is comparable to that of calcium

Chapter 3 Cement Additives and Mechanisms of Action 53

chloride (Rosenburg et al., 1977). Calcium nitrite is also 3-4 Retarders
an effective corrosion inhibitor for steel embedded in Like acceleration, the mechanisms of Portland-cement
concrete (Berke, 1985). retardation are still a subject of controversy. Several the-
Calcium nitrate, Ca(NO3)2, in conjunction with tri- ories have been proposed, but none fully explains the
ethanolamine, is an effective accelerator (Tokay, 1982); retardation process by itself. Two principal factors must
however, its efficiency is highly variable depending on be considered: the chemical nature of the retarder and
the cement type. Technical calcium nitrate, a blend of the cement phase (silicate or aluminate) upon which
calcium- and ammonium-nitrate hydrates, has a dual the retarder acts. The four principal theories are sum-
function as a set accelerator and corrosion inhibitor for marized below.
reinforced concrete (Justnes and Nygaard, 1995).
Thiocyanate (SCN–) salts were introduced to the con- 1. Adsorption theory: The retarder adsorbs onto the sur-
crete market in 1983 (Rosskopf, 1983). Like the nitrates, faces of the hydration products, thereby inhibiting
the thiocyanates should be accompanied by an alka- contact with water.
nolamine to attain the desired results. The combination 2. Precipitation theory: The retarder reacts with cal-
of thiocyanate salts with nitrates and alkanolamines was cium ions, hydroxyl ions, or both in the aqueous
patented in 1984 (Gerber, 1984). phase, forming an insoluble and impermeable layer
Triethanolamine, N(C2H4OH)3, accelerates the reac- around the cement grains.
tion between C3A and gypsum and, at dosages of 0.1 and 3. Nucleation theory: The retarder adsorbs onto the
0.5% BWOC, setting can occur within a few minutes at nuclei of hydration products, arresting their future
ambient temperature (Ramachandran, 1973; 1976b). growth.
However, compressive strength development can be 4. Complexation theory: The retarder chelates the cal-
delayed because triethanolamine strongly retards C3S cium ions, preventing the formation of nuclei.
and C2S hydration. Therefore, triethanolamine is rarely
used by itself as an accelerator. It is probable that all of the above theories are
A complex mixture of calcium nitrate, calcium nitrite, involved to some extent in the retardation process.
diethylene glycol, methyldiethanolamine, and calcium Despite the uncertainty regarding the mechanisms of
bromide was recently patented as an accelerator for well retardation, the chemical technology is very well devel-
cements (Maberry et al., 2001). It is used in cold cement- oped. The major chemical classes of retarders, as well as
ing environments such as deepwater offshore wells or proposed mechanisms of action, are discussed individu-
permafrost zones. The performance of this additive ally below.
(Fig. 3-3) is superior to calcium chloride at tempera-
tures between about 40° and 70°F [5° and 20°C]. Unlike
calcium chloride, this accelerator does not affect slurry
3-4.1 Lignosulfonates
rheology at elevated concentrations. The most commonly used retarders for well cements are
the sodium and calcium salts of lignosulfonic acids
(Fig. 3-4). Lignosulfonates are polymers derived from
wood pulp; therefore, they are usually unrefined and
10:00 contain various amounts of saccharide compounds. The
9:00 average molecular weight varies from about 20,000 to
8:00 30,000. Purified lignosulfonates have much less retard-
7:00 ing power; therefore, the retarding action is often attrib-
Thickening 6:00 uted to the impurities in the bulk material. Such impuri-
time (hr) 5:00
4:00
3:00 O
2:00 HO
H
1:00 CH2 C C C CH2OH
0:00
0 0.1 0.2 0.3 0.4 0.5 SO3H O
OH
Chloride-free accelerator (gal/sk)

Figure 3-3. Effect of a chloride-free accelerator on the thickening Figure 3-4. Basic lignosulfonate chemical structure.
time of a 16.4-lbm/gal [1,970-kg/m3] Class H cement system at 65°F
[18.3°C].

54 Well Cementing
ties include low-molecular-weight carbohydrates such as (Ciach and Swenson, 1971). A waterproofing mecha-
pentoses (xylose and arabinose), hexoses (mannose, glu- nism, preventing further significant hydration, has also
cose, fructose, rhamnose, and galactose), and aldonic been proposed (Jennings et al., 1986).
acids (especially xylonic and gluconic acids) (Chatterji, Some of the lignosulfonate remains in the aqueous
1967; Milestone, 1976; 1979). phase. It may be in a free state or linked to calcium ions
Lignosulfonate retarders are effective with all through electrostatic interactions. At low lignosulfonate
Portland cements and are generally added in con- concentrations, the crystal growth (and probably the
centrations ranging from 0.1% to 1.5% BWOC (Fig. 3-5). nucleation) of calcium hydroxide is inhibited (Jawed et
Depending upon the lignosulfate retarders’ carbohy- al., 1979). Although the same experiment has not yet
drate content and chemical structure (e.g., molecular been performed with C-S-H phase, a similar result would
weight distribution and degree of sulfonation) and the be expected. A significant change in the size and mor-
nature of the cement, they are effective to about 250°F phology of the calcium hydroxide crystals is observed
[122°C] bottomhole circulating temperature (BHCT). when C3S is hydrated in the presence of lignosulfonates
When blended with sodium borate (Section 3-4.6), the (Berger and McGregor, 1972). These results suggest
effective temperature range of lignosulfonates can be that, if the nucleation and crystal growth of the hydra-
extended to as high as 600°F [315°C] BHCT. tion products are hindered by the lignosulfonate, the
Lignosulfonate retarders predominantly affect the hydration rate of C3S will be similarly affected.
kinetics of C3S hydration; however, their effects upon
C3A hydration are not insignificant (Stein, 1961b;
Angstadt and Hurley, 1963). The retardation mechanism 3-4.2 Hydroxycarboxylic acids
of the lignosulfonates is generally thought to be a com- Hydroxycarboxylic acids contain hydroxyl and carboxyl
bination of the adsorption and nucleation theories. groups in their molecular structures (Fig. 3-6). Glu-
Lignosulfonate retarders perform best with low-C3A conate and glucoheptonate salts and tartaric acid are
cements. When C3A is hydrated in the presence of the most widely used materials in this category. They
organic additives such as lignosulfonates, the solution have a powerful retarding action and can easily cause
concentration of the additives quickly falls. C3A hydra- overretardation at BHCTs less than 200°F [93°C].
tion products have a much stronger adsorptive effect As shown in Fig. 3-7, these materials are efficient to tem-
than those of C3S (Blank et al., 1963; Rossington and peratures approaching 300°F [150°C].
Runk, 1968). In a Portland cement system, C3A hydra-
tion can prevent a significant amount of lignosulfonate
from reaching the surfaces of C3S hydration products; as
OH
a result, the retarder is less efficient (Young, 1969).
Ramachandran (1972) showed that the sulfonate and HO2C CH2 C CH2 CO2H
hydroxyl groups adsorb onto the C-S-H phase layer. As a
result, the permeability of the C-S-H phase is reduced CO2H

Citric acid

5.5
5.0 103°F [40°] 113°F [45°C] CH2(OH) CH2(OH)
4.5
4.0 CH(OH) CH(OH)
3.5 125°F [51°C]
Thickening CH(OH) CH(OH)
time (hr) 3.0
2.5 144°F [62°C]
CH(OH) CH(OH)
2.0
1.5 CH(OH) CH(OH)
1.0
0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40 CH(OH) CO2H
Retarder concentration (%BWOC)
CO2H Gluconic acid
Figure 3-5. Typical effect of a lignosulfonate retarder on a 15.8-lbm/
gal [1,900 kg/m3] Class G cement. Glucoheptonic acid

Figure 3-6. Molecular structures of hydroxycarboxylic acid retarders.

Chapter 3 Cement Additives and Mechanisms of Action 55

 H OH
0.16
0.14 H OH H O CH2 O
Concentration required for H H H HOCH2 O
0.12 HO H O H
H
Retarder 4-hr thickening time OH H H OH
0.10 CH2OH HO O CH2OH
concentration 0.08
(%BWOC) H OH OH H
0.06
0.04 Raffinose
0.02
0.00 CH2OH
150 170 190 210 230 250 H H O H HOCH2 O H
BHCT (°F) HO OH H O H OH CH OH
2

Figure 3-7. Retardation performance of sodium glucoheptonate H OH HO H

with a 15.6-lbm/gal [1,870-kg/m3] Class A cement slurry. Sucrose
Figure 3-8. Molecular structures of saccharide retarders.

Another hydroxycarboxylic acid with a strong retard-

ing effect is citric acid. Citric acid is also an effective
cement dispersant (Section 3-5) and is normally used at 3-4.4 Cellulose derivatives
concentrations between 0.1% and 0.3% BWOC. Cellulose polymers are polysaccharides derived from
The retarding action of hydroxycarboxylic acids and wood or other plants. They are stable in the alkaline
their salts is generally attributed to the presence of environment of cement slurries. Set retardation occurs
alpha- or beta-hydroxycarboxylic groups (HO-C-CO2H when the polymer adsorbs onto the hydrated cement
and HO-C-C-CO2H, respectively) that are capable of surfaces. The active sites for adsorption are the ethyl-
strongly chelating a metal cation such as calcium ene-oxide links and carboxyl groups.
(Double, 1983). Highly stable five- or six-membered The most common cellulosic retarder is car-
rings form, partially adsorb onto the hydrated cement boxymethylhydroxyethylcellulose (CMHEC) (Shell
surface, and poison the nucleation sites of hydration and Wynn, 1958). Its molecular structure is shown in
products. Like the lignosulfonates, hydroxycarboxylic Fig. 3-9. CMHEC is an effective retarder at temperatures
acids are more efficient with low-C3A cements. up to about 250°F [121°C] (Rust and Wood, 1966).
Typical performance data are presented in Fig. 3-10.
A number of secondary effects are observed with
3-4.3 Saccharide compounds CMHEC. It is often used as a fluid-loss control agent
Saccharide compounds (also called sugars) are excel- (Section 3-8). In addition, CMHEC significantly
lent Portland cement retarders (Fig. 3-8). The best per- increases the viscosity of the slurry.
formers contain a five-membered ring (e.g., sucrose and
raffinose) (Bruere, 1966; Previte, 1971; Thomas and
Birchall, 1983). Such compounds are not commonly used 3-4.5 Organophosphonates
in well cementing, because the degree of retardation is In the 1980s, alkylene phosphonic acids and their salts
very sensitive to small variations in concentration. were identified as set-retarding additives for well
The retarding action of saccharide compounds cements (Nelson, 1984; Sutton et al., 1985; Childs et al.,
depends upon their susceptibility to degradation by 1986; Nelson, 1987). Phosphomethylated compounds
alkaline hydrolysis. The sugars are converted to saccha- containing quaternary ammonium groups (Crump and
rinic acids containing alpha-hydroxycarbonyl groups Wilson, 1984), as well as N-phosphonomethyl iminodi-
(HO-C-C=O), which adsorb strongly onto C-S-H phase acetic acid (Huddleston, 1995) are also efficient. Such
surfaces (Taplin, 1960). Hydration is inhibited when the materials have excellent hydrolytic stability and,
C-S-H phase nucleation sites are inactivated (poisoned) depending upon the molecular backbone, are effective
by the adsorbed sugar acid anions (Milestone, 1979). to circulating temperatures as high as 450°F [232°C].

56 Well Cementing
 CH2OR H OR
H H H
O
H OH H O
OCH2CO2Na OH H O H O
CH2 CH2OH
CH2OR m
CH2 CH2
O O DS = 2
R = alkyl group
CH2 H CH2
H O O
H OR´OR´OH
O H H OH H
OH H H CH2 H OR´OR´OH
O O H
H O OH H O
H OH OH OH H H
m
H OH CH2OR
Acetal linkage
MS = 2.5
R´ = alkylene group

Figure 3-9. CMHEC molecular structure and illustration of degree of substitution (DS) and molecular
substitution (MS) concepts.

15.6-lbm/gal [1,870 kg/m3] Class A cement (4-hr thickening time)

15.8-lbm/gal [1,900 kg/m3] Class H cement (4-hr thickening time)
15.8-lbm/gal [1,900 kg/m3] Class H cement (3-hr thickening time)
0.3

CMHEC 0.2
(%BWOC)

0.1

0
100 120 140 160 180 200 220 240
BHCT (°F)

Figure 3-10. Typical CMHEC concentrations required to obtain 3- to 4-hr thickening times
(using Class A and Class H cements).

Organophosphonates are advantageous for well circulating temperatures up to about 400°F [204°C]
cementing because of their apparent insensitivity to (Brothers, 1994; Rodrigues and Lindsey, 1995). Ultrafine
subtle variations in cement composition and their ten- cement slurries are used for squeeze cementing and
dency to lower the viscosity of high-density cement well repair (Chapters 7 and 14).
slurries. The mechanism of action involves the adsorp-
tion of phosphonate groups (Fig. 3-11) onto the nuclei of
cement hydrates, thus hindering their growth. H OH
Methylenephosphonic acid derivatives can be used to
prepare cement slurries with very long thickening times. R C P O
The set can then be activated when needed (e.g., by H OH
an aqueous solution of sodium silicate) (Childs et al.,
1987). They can also be used to retard ultrafine cements Figure 3-11. Alkylene phosphonate structure.
(Blaine fineness of greater than 6,000 cm2/g) at

Chapter 3 Cement Additives and Mechanisms of Action 57

 At temperatures above about 230°F [110°C], methyl- (K s = 1.8.10–14) and is deposited as a colloidal gel; con-
ene phosphonic acid derivatives are poor retarders of sequently, the precipitate has a low permeability. The
the aluminate (C3A) and aluminoferrite (C4AF) phases, retardation effect ends when the gelatinous zinc hydrox-
but they strongly retard the silicate phases (C3S and ide eventually transforms to crystalline calcium hydrox-
C2S) (unpublished data, M. Michaux, 1993). As a result, yzincate.
thickening times are short and compressive strength
development is slow. Adding a borate salt to the retarder 2Zn(OH)2 + 2OH– + Ca2+ + 2H2O → CaZn2(OH)6 • 2H2O
formulation solves this problem. At an appropriate phos- (3-1)
phonate/borate ratio, longer thickening times and swift
compressive strength development can be achieved at Sodium tetraborate decahydrate (borax:
circulating temperatures up to 450°F [232°C] (Nelson, Na2B4O7 • 10H2O) is commonly used as a “retarder aid.”
1984; Nelson and Casabonne, 1992; Barlet-Gouédard et It has the ability to extend the effective temperature
al., 1996; Casabonne et al., 1996) (Fig. 3-12). range of most lignosulfonate retarders to as high as 600°F
[315°C]; however, it can be detrimental to the effective-
ness of cellulosic and polyamine fluid-loss additives.
12
350°F [177°C]
3-5 Extenders
10
300°F [149°C] Cement extenders are routinely used to accomplish one
or both of the following.
8 ■ Reduce slurry density: A reduction of slurry density
Thickening reduces the hydrostatic pressure during cementing.
time (hr) This helps to prevent induced lost circulation
6
because of the breakdown of weak formations. In
addition, the number of stages required to cement a
4 well may be reduced.
■ Increase slurry yield: Extenders reduce the amount of
cement required to produce a given volume of set
2
0.6 0.7 0.8 0.9 1.0 1.1 1.2 1.3
product. This results in a greater economy.
Retarder concentration (gal/sk)
Extenders can be classified into one of three cate-
gories, depending upon the mechanism of density reduc-
Figure 3-12. Effect of phosphonate retarder on thickening times of tion and/or yield increase. Often more than one type of
16.5-lbm/gal [1,970-kg/m3) Class H cement systems. extender is used in the same slurry.
■ Water extenders: Clays and various water viscosifying
agents allow the addition of more water to achieve
3-4.6 Inorganic compounds slurry extension. Such extenders maintain a homoge-
neous slurry and prevent the development of exces-
Many inorganic compounds retard the hydration of sive free water.
Portland cement. The major classes of materials are
■ Low-density aggregates: This varied category consists
listed below.
of materials with densities lower than that of
■ Acids and salts thereof: boric, phosphoric, hydro-
Portland cement (3.20 g/cm3). The slurry density is
fluoric, and chromic reduced when significant quantities of such exten-
■ Sodium chloride: concentrations greater than ders are present.
20% BWOW (Section 3-2) ■ Gaseous extenders: Nitrogen or air can be used to
■ Oxides: zinc and lead prepare foamed cements with exceptionally low den-
In well cementing, zinc oxide (ZnO) is sometimes sities yet sufficient compressive strength. The prepa-
used for retarding thixotropic cements, because it does ration and placement of such cement systems are
not affect the slurry rheology (Chapter 7) or the hydra- described in Chapters 7 and 13.
tion of the C3A-gypsum system (Ramachandran, 1986). Table 3-2 presents a list of the common extenders, with
The retardation effect of ZnO is attributed to the precip- general information regarding their performance char-
itation of zinc hydroxide onto the cement grains (Arliguie acteristics.
and Grandet, 1985). Zn(OH)2 has a low solubility

58 Well Cementing
Table 3-2. Summary of Extenders Table 3-3. Effect of Bentonite on Cement-Slurry Properties
Extender Range of Slurry Performance % Bentonite Water Slurry Density Yield
Slurry Densities Features and (gal/sk) (lbm/gal) (ft3/sk)
Obtainable (lbm/gal) Other Benefits
0 4.97 15.8 1.14
6 11 16
2 6.17 15.0 1.31
Bentonite 11.5 15 Assists fluid-loss
control 4 7.36 14.4 1.48
Fly ashes 13.1 14.1 Resist corrosive fluids 6 8.56 13.9 1.65
Sodium 11.1 14.5 Available in solid or 8 9.76 13.5 1.82
silicates liquid form; effective
at low concentrations; 10 10.95 13.1 1.99
ideal when mixing
slurry with seawater 12 12.15 12.7 2.16

Microspheres 8.5 15 Good compressive 16 14.55 12.3 2.51

strength, low
permeability, thermal 20 16.94 11.9 2.85
stability, and insulating
properties

Foamed 6 15 Good compressive 2,400

cement strength and low
2,000
permeability 200°F [93°C]
1,600
Compressive
strength after 1,200
3-5.1 Clays 24 hr (psi)
800
100°F [38°C]
Clay minerals are hydrous aluminum silicates of the
400
phyllosilicate group (Hurlbut, 1971), in which the silica
tetrahedra are arranged in sheets. Such minerals have a 0
platy or flaky shape and one prominent cleavage. 4 6 8 10 12 14 16 18 20
Magnesium or iron can partially substitute for aluminum Bentonite (%BWOC)
in the crystal lattice, and alkalis or alkaline earths may
also be present as essential components. Figure 3-13. Effect of bentonite on compressive strength.
The most frequently used clay-base extender is ben-
tonite, also known as “gel,” which contains at least 85% of
the mineral smectite. Smectite, NaAl2(AlSi3O10)(OH)2, is increases with bentonite concentration; therefore, such
composed of two flat sheets of silica tetrahedra sand- cements are less resistant to sulfate waters and corro-
wiching one sheet of alumina octahedra. Bentonite has sive fluids. However, as explained in Section 3-3.8.1, high
the unusual property of expanding to several times its concentrations of bentonite tend to improve fluid-loss
original volume when placed in water, resulting in higher control.
fluid viscosity, gel strength, and solids-suspending abil- High concentrations of Ca2+ ion in the aqueous phase
ity. Bentonite is obtained primarily from mines in of a cement slurry inhibit the hydration of bentonite.
Wyoming and South Dakota; however, foreign sources The extension efficiency of bentonite can be greatly
are being developed (Samsuri et al., 2001). enhanced if it is prehydrated in the mix water before
Bentonite is added in concentrations up to 20% cement addition. A slurry containing 2% prehydrated
BWOC. Above 6%, the addition of a dispersant is usually bentonite BWOC is equivalent to one containing 8% dry-
necessary to reduce the slurry viscosity and gel strength. blended bentonite (Table 3-4). Complete hydration of a
The API/ISO recommends that 5.3% additional water good-quality bentonite occurs in about 30 min. The
(BWOC) be added for each 1% bentonite for all classes of thickening times of prehydrated bentonite slurries are
cement; however, testing is necessary to determine the generally the same as those of dry-blended slurries at
optimal water content with a particular cement. As the same density. It should also be noted that prehy-
shown in Table 3-3, the slurry density decreases and the drating the bentonite does not appreciably change the
yield increases with bentonite concentration; however, final compressive strength when compared to an equal-
as shown in Fig. 3-13, there is a price to be paid in terms density slurry prepared with dry-blended bentonite.
of compressive strength. Cement permeability also

Chapter 3 Cement Additives and Mechanisms of Action 59

Table 3-4. Comparison of Prehydrated and Dry-Blended Bentonite Slurry Properties
Prehydrated Dry-Blended Fresh Water Slurry Density Slurry Yield
Dry Bentonite (%) Bentonite (%) (gal/sk) (lbm/gal) (ft3/sk)
Prehydrated Dry Blend Prehydrated Dry Blend
0 0 5.2 – 15.6 – 1.18

0.5 2 6.4 14.8 14.8 1.34 1.35

1.0 4 7.6 14.1 14.2 1.50 1.52

1.5 6 8.8 13.5 13.7 1.66 1.69

2.0 8 10.0 13.1 13.3 1.83 1.86

2.5 10 11.2 12.7 12.9 1.99 2.03

3.0 12 12.4 12.4 12.6 2.16 2.20

4.0 16 14.8 11.9 12.2 2.48 2.55

5.0 20 17.2 11.5 11.8 2.81 2.89

According to API/ISO specifications, only pure, minimizes storage and handling. However, because of
untreated bentonite should be used in well cements. their tendency to accelerate, they tend to reduce the
Beneficiating agents such as polyacrylamide, which effectiveness of other additives, retarders and fluid-loss
improve the water absorption capacity of low-grade ben- agents in particular.
tonites, are prohibited. Such materials can interfere The solid sodium silicate, Na2SiO3 (sodium metasili-
with the performance of other cement additives and lead cate), is normally dry-blended with the cement. If it is
to unpredictable results (Grant et al., 1990). added to fresh mix water before slurry preparation, a gel
Bentonite can be prehydrated in seawater or light may not form unless calcium chloride is also added. If a
brine, but the salt inhibits hydration and the slurry yield gel does not form, proper slurry extension will not occur.
is reduced. Bentonite is not effective as an extender in The recommended concentration range of Na2SiO3 is
highly saline cement slurries. Under such circum- 0.2% to 3.0% BWOC. These concentrations provide a
stances, another clay mineral, attapulgite, is frequently slurry-density range of from 14.5 to 11.0 lbm/gal [1.75 to
used (Smith and Calvert, 1974). Attapulgite, 1.35 g/cm3]. The typical properties and performance of
(Mg,Al)5Si8O22(OH)4 • 4H2O, also known as “salt-gel,” sodium metasilicate-extended cement systems are
occurs as fibrous needles. When dispersed in water, the shown in Table 3-5.
attapulgite needles associate with one another and pro- The liquid sodium silicate, Na2O • (3-5)SiO2 (also
vide viscosity. Unlike bentonite, attapulgite does not called water glass), is added to the mix water before
improve fluid-loss control. Attapulgite has been banned slurry mixing. If calcium chloride is to be included in
in some countries, because the fibrous needles are simi- cement systems mixed with fresh water, it must be
lar to those of asbestos (Bensted, 2001). However, gran- added to the mix water before the sodium silicate to
ular forms of attapulgite are available that are still obtain sufficient extending properties. In the case of
permitted in many locations. slurries mixed with seawater, sodium silicate will inter-
act with the divalent cations in the seawater to achieve
slurry extension. The normal concentration range is 0.2
3-5.2 Sodium silicates to 0.6 gal/sk. Typical performance data are presented in
Silicate extenders react with lime or calcium chloride in Table 3-6.
the cement to form a calcium silicate gel. The gel struc-
ture provides sufficient viscosity to allow adding extra
mix water without excessive free-water separation.
Sodium silicates are available in solid or liquid form. A
major advantage of the silicates is their efficiency, which

60 Well Cementing
Table 3-5. Typical Performance of Class G Cement Systems Containing Sodium Metasilicate
Sodium Slurry Density Slurry Yield Water Compressive Strength Thickening Time (hr:min)
Metasilicate (lbm/gal (ft3/sk at 24 hr (psi [MPa])
(%BWOC) [kg/m3]) [m3/SI ton])
(gal/sk [L/SI ton]) % 120°F [49°C] 140°F [60°C] 103°F [39°C] 113°F [40°C] 125°F [52°C] 140°F [60°C]

0 15.8 [1,900] 1.15 [0.76] 4.97 [440] 44 4,770 [33] 5,310 [37] +4:00 3:10 2:35 –†

0.15 14.5 [1,741] 1.38 [0.91] 6.77 [600] 60 1,746 [12] 2,248 [15] 3:20 2:37 2:10 –

1.0 14.5 [1,741] 1.38 [0.91] 6.77 [600] 60 1,896 [13] 2,175 [15] 2:40 1:34 – –

0.25 14.0 [1,681] 1.51 [1.00] 7.68 [680] 68 1,420 [10] 1,510 [10] – – – –

1.0 14.0 [1,681] 1.51 [1.00] 7.68 [680] 68 1,640 [11] 1,723 [12] – – – –

0.5 13.5 [1,621] 1.66 [1.10] 8.81 [780] 78 946 [7] 1,278 [9] – 3:30 2:10 –

2.0 13.5 [1,681] 1.66 [1.10] 8.81 [780] 78 1,327 [9] 1,420 [10] 1:53 1:28 – –

0.5 13.0 [1,561] 1.84 [1.22] 10.17 [910] 90 750 [5] 927 [6] – – – –

2.0 13.0 [1,561] 1.84 [1.22] 10.17 [910] 90 120 [1] 1,080 [7] – – – –

0.75 12.5 [1,501] 2.05 [1.36] 11.75 [1,050] 104 382 [3] 625 [4] +5:00 +5:00 +5:00 +5:00

2.0 12.5 [1,501] 2.05 [1.36] 11.75 [1,050] 104 633 [4] 653 [5] +5:00 1:43 – –

1.0 12.0 [1,441] 2.32 [1.54] 13.78 [1,230] 122 265 [2] 380 [3] – – – –

2.0 12.0 [1,441] 2.32 [1.54] 13.78 [1,230] 122 420 [3] 510 [4] – – – –

1.5 11.5 [1,381] 2.69 [1.78] 16.6 [1,480] 147 147 [1] 230 [2] +5:00 +5:00 +5:00 +5:00

3.0 11.5 [1,381] 2.69 [1.78] 16.6 [1,480] 147 271 [2] 289 [2] +5:00 1:27 – –

2.0 11.0 [1,321] 3.20 [2.12] 20.34 [1,810] 180 102 [1] 175 [1] – – – –

3.0 11.0 [1,321] 3.20 [2.12] 20.34 [1,810] 180 145 [1] 205 [1] – – – –
† Not available

Table 3-6. Effect of Liquid Sodium Silicate on the Performance of Class G Cement Slurries Mixed With Fresh Water
Slurry Density Liquid Silicate Thickening Time at BHCT (hr:min) Compressive Strength at BHST† After 24 hr (psi)
(lbm/gal Concentration
[kg/m3]) (gal/sk [L/SI ton]) 103°F [39°C] 113°F [45°C] 175°F [79°C] 95°F [35°C] 110°F [43°C] 140°F [60°C] 170°F [77°C] 200°F [93°C]

14.2 [1,700] 0.20 [17.8] 2:20 1:40 –‡ 2,200 [15] 2,550 [18] 2,300 [16] 2,100 [14] 2,000 [14]

13.6 [1,630] 0.30 [26.7] 3:00 2:00 – 1,150 [7.9] – 1,450 [10] – 1,350 [9.3]

13.0 [1,560] 0.36 [32.0] 3:40 2:20 – 900 [6.2] – 1,050 [7.2] – 1,050 [7.2]

12.5 [1,500] 0.42 [37.4] +4:00 2:30 1:50 850 [5.8] 850 [5.8] 850 [5.8] 850 [5.8] 850 [5.8]

12.0 [1,440] 0.50 [44.5] +4:00 +4:00 3:10 500 [3.5] – 500 [3.5] – 500 [3.5]

11.5 [1,380] 0.60 [53.4] +4:00 +4:00 3:50 250 [1.7] 350 [2.4] 300 [2.1] 300 [2.1] 300 [2.1]
† Bottomhole static temperature
‡ Not available

3-5.3 Pozzolans ence of moisture, chemically react with calcium

Pozzolans are perhaps the most important group of hydroxide at ordinary temperatures to form com-
cement extenders. They are defined in accordance with pounds possessing cementitious properties.”
ASTM International† designation C-219-55 as follows: Thus, pozzolans not only extend Portland cement sys-
“A silicious or siliceous and aluminous material, tems, but also react with the calcium hydroxide liber-
which in itself possesses little or no cementitious ated during cement hydration and contribute to the
value, but will, in finely divided form and in the pres- compressive strength of the set product. There are two

† ASTM International was formerly known as the American Society for Testing
and Materials (ASTM).

Chapter 3 Cement Additives and Mechanisms of Action 61

types of pozzolans: (1) natural pozzolans, which include weights. The ratio for mixtures based upon equivalent
volcanic ashes and diatomaceous earth, and (2) artifi- sacks is designated as 25:75, 50:50, 75:25, or whatever
cial pozzolans, such as certain fly ashes. ratio is desired. The term 25:75 indicates 1⁄4 eq sk of poz-
When one 94-lbm sack of cement hydrates, about 20 zolan and 3⁄4 eq sk of Portland cement.
to 23 lbm of free Ca(OH)2 are liberated. By itself, The weights of other additives (except salt) are cal-
Ca(OH)2 contributes little to the strength of the set culated as a percentage by weight of the pozzolan/
cement and is fairly soluble; thus, it can be eventually cement equivalent sack. Salt is always calculated as a
dissolved and removed by water contacting the cement. percentage of the mix-water volume.
This contributes to a weakening of the set cement. When As an example, an equivalent sack of one typical fly
a pozzolan is present, the silica combines with the free ash is 74 lbm. A 50:50 blend would require 37 lbm of fly
Ca(OH)2 to form a stable cementitious compound (sec- ash and 47 lbm of Portland cement. Thus, 84 lbm of this
ondary C-S-H) that is very durable. blend would displace 3.59 gal. Additive concentrations
At typical cement densities, the water permeability would then be calculated as a percentage of an 84-lbm
of set pozzolan/cement systems is usually less than equivalent sack, not the usual 94-lbm sack of Portland
0.001 mD. The low permeability of the set cement, as cement. A more detailed discussion of pozzolanic slurry
well as the decrease of free Ca(OH)2 content, resists the calculations is presented in Appendix C.
encroachment of sulfate water and other corrosive
fluids. Should corrosive waters manage to enter the set
pozzolanic cement, damage is further prevented by 3-5.3.1 Diatomaceous earth
another mechanism. Zeolites in the pozzolan act as ion Diatomaceous earth is composed of the siliceous skele-
exchange agents and help prevent deterioration. tons of diatoms deposited from either fresh water or sea-
There are two notation systems commonly used for water. The main constituent of diatomaceous earth is
mixing pozzolan cements. The first is a bulk/volume opal, an amorphous form of hydrous silica containing up
ratio. A 1:1 ratio indicates 1 ft3 of pozzolan and 1 ft3 of to 10% water. For use as a pozzolanic extender, diatoma-
cement. The first figure always indicates the volume of ceous earth is ground to a fineness approaching that of
pozzolan, and the second indicates the volume of Portland cement; consequently, it has a large surface
cement. This system is used primarily with very light area and a high water demand.
pozzolans. Diatomaceous earth imparts slurry properties similar
The second mixing system is the most widely used. It to those of bentonite slurries; however, it does not
is based on the “equivalent sack.” A sack of Portland increase the slurry viscosity to such a high degree. In
cement has an absolute volume of approximately addition, because of its pozzolanic activity, set cements
3.59 gal. In other words, when mixed with water, one containing diatomaceous earth are stronger than their
sack of cement will increase the volume of the mix by bentonitic counterparts. The principal disadvantage of
about 3.59 gal. An equivalent sack is the weight of poz- diatomaceous earth is its cost. Typical slurry properties
zolan that also has an absolute volume of 3.59 gallons. and performance of diatomaceous earth slurries are
Different pozzolans have different equivalent sack shown in Table 3-7.

Table 3-7. Effect of Diatomaceous Earth on the Performance of Class A Cement Systems
Diatomaceous Water Slurry Weight Slurry Yield Compressive Strength After Compressive Strength
Earth (%BWOC) (gal/sk) (lbm/gal) (ft3/sk) 24 Hr of Curing (psi) After 72 Hr of Curing (psi)
80°F and 95°F and 110°F and 140°F and 80°F and 95°F and 110°F and 140°F and
Ambient 600 psi 1,600 psi 3,000 psi Ambient 600 psi 1,600 psi 3,000 psi
Pressure Pressure Pressure Pressure Pressure Pressure
0 5.2 15.6 1.18 1,360 1,560 2,005 2,620 2,890 3,565 4,275 4,325

10 10.2 13.2 1.92 110 360 520 750 440 660 945 1,125

20 13.5 12.4 2.42 70 190 270 710 240 345 645 1,000

30 18.2 11.7 3.12 na† na na na na na na na

40 25.6 11.0 4.19 15 30 50 260 70 150 220 630

† Not available

62 Well Cementing
3-5.3.2 Fly ashes result, many of them are cementitious and do not fit the
Fly ash is the residue from power plants that burn pul- strict definition of a pozzolanic material. In well cement-
verized coal (Davis et al., 1937). The ash is suspended in ing, Type F fly ash is used most frequently.
flue gases as fused particles that solidify into a roughly Normally, 2% bentonite is added to Type F fly ash/
spherical shape. The ash is very finely divided, with a Portland cement systems to improve the slurry proper-
surface area roughly approximating that of Portland ties and prevent the development of free water. In Table
cements. The major constituent of fly ash is a glass 3-9, slurry data for different ratios of Type F fly ash and
chiefly composed of silica and alumina with some iron cement are presented with various water contents.
oxide, lime, alkalis, and magnesia. Quartz, mullite, The use of Type C fly ashes as extenders for well
hematite, and magnetite, as well as some combustible cements is relatively new. Because of the significant
matter, are also found. The composition and properties amount of lime in such fly ashes, the rheological effects
of fly ash can vary widely depending upon the source of must be carefully monitored. In addition, Type C ashes
the coal and the efficiency of the power plant; accord- vary widely depending upon the source, and special
ingly, the specific gravities of fly ashes can vary from slurry preparation guidelines are required for each.
about 2.0 to 2.7 (Hewlett, 2001). Some Type C fly ashes are sufficiently cementitious to
According to ASTM International specifications, be used as the principal component of a well cement.
three types of fly ash are recognized: Types N, F, and C. Such systems have been developed for application in
As shown in Table 3-8, the distinction is made on chem- shallow wells having circulating temperatures up to
ical grounds. Types N and F are normally produced from 120°F [49°C]. Compressive strength development is
burning anthracite or bituminous coals. Type C fly ashes, often more rapid than that observed with conventional
made from lignite or subbituminous coals, are less Portland cement systems.
siliceous, and some contain more than 10% lime; as a
3-5.3.3 Commercial lightweight cements
Commercial oilwell cements, such as TXI Lightweight‡,
Table 3-8. Chemical Requirements for Fly Ashes are special formulations composed of interground
Mineral Admixture Class Portland cement clinker and lightweight siliceous aggre-
N F C
gates; consequently, some pozzolanic activity occurs.
They are convenient and time-saving for the service
Min. silicon dioxide (SiO2) + company. The particle-size distribution is finer than in
aluminum oxide (Al2O3) + 70 70 50
iron oxide (Fe2O3) (%)
Portland cements, and the normal slurry density range is
from 11.9 to 13.7 lbm/gal [1.43 to 1.64 g/cm3].
Max. sulfur trioxide (SO3) (%) 4 5 5

Max. moisture content (%) 3 3 3

‡ TXI Lightweight is a trademark of Texas Industries.
Max. loss on ignition (%) 10 12 6

Table 3-9. Properties of Fly Ash/Class H Cement Systems

Ratio† Weight of Components (lbm) Min. Water Requirement Max. Water Requirement
Fly Ash Class H Fly Ash Class H Water Slurry Slurry Water Slurry Slurry
Cement Cement (gal/sk) Density Volume (gal/sk)‡ Density Volume
(lbm/gal) (ft3/sk) (lbm/gal) (ft3/sk)‡
25 75 18.5 70.5 5.24 15.1 1.19 5.64 14.7 1.25

35 65 25.9 61.5 5.17 15.0 1.18 5.73 14.6 1.26

50 50 37.0 47.0 5.00 14.7 1.16 5.80 14.2 1.27

65 35 48.1 32.9 4.85 14.5 1.14 5.89 13.8 1.28

75 25 55.5 23.5 4.75 14.3 1.12 5.96 13.5 1.29

† Allsystems contain 2% bentonite by weight of fly ash/cement blend.
‡ Based on the weight of an equivalent sack of the specific blend.

Chapter 3 Cement Additives and Mechanisms of Action 63

3-5.3.4 Silica 3-5.4 Lightweight particles
Three forms of finely divided silica are used in well Lightweight-particle extenders reduce the slurry density
cements: α-quartz, condensed silica fume, and colloidal because they are lighter than the cement particles. Such
silica dispersions. Silica as α-quartz is used most fre- extenders include expanded perlite, powdered coal,
quently for the prevention of strength retrogression gilsonite, and either glass or ceramic microspheres.
when Portland cement systems are placed in thermal Most extenders in this category are inert with respect to
wells (Chapter 10). Two particle sizes are routinely used: the cement.
silica sand, with an average particle size of about
100 μm, and silica flour, with an average particle size of
about 15 μm. Other sizes of crystalline silica are used in 3-5.4.1 Expanded perlite
special cement systems with controlled particle-size dis- Perlite is a crushed volcanic glass that expands when
tributions (Chapter 7). Because of cost, these materials heated to the point of incipient fusion (Lea, 1971). The
are rarely used for slurry extension alone. expanded perlite product generally has a bulk density of
Condensed silica fume (also called microsilica) is a 7.75 lbm/ft3, which allows the preparation of competent
byproduct of the production of silicon, ferrosilicon, and cement slurries with densities as low as 12.0 lbm/gal
other silicon alloys. The individual particles are glassy, [1.44 g/cm3]. A small quantity of bentonite (2% to 4%
amorphous microspheres. The mean particle size is usu- BWOC) is added to prevent the segregation of the perlite
ally between 0.1 μm and 0.2 μm, about 50 to 100 times particles from the slurry.
finer than Portland cement or fly ash; consequently, the Expanded perlite contains both open and closed
surface area is extremely high (15,000 to 25,000 m2/kg). pores. Under hydrostatic pressure, the open pores fill
Condensed silica fume is highly reactive and, because with water, and some of the closed pores are crushed; as
of its fineness and purity, is a very effective pozzolanic a result, the perlite becomes more dense. To maintain
material (Parker, 1985). Its high degree of pozzolanic density control, the use of perlite is confined to shallow
activity has led to the introduction of low-density cement wells.
systems with a higher rate of compressive strength To prepare an expanded perlite slurry that will have a
development (Carathers and Crook, 1987). The high sur- given density downhole, it is necessary to mix a lower-
face area of condensed silica fume increases the amount density slurry at the surface. At 3,000 psi [20.7 MPa], the
of water required to prepare a pumpable slurry; there- specific gravity of expanded perlite is 2.40. Table 3-10
fore, slurries with densities as low as 11.0 lbm/gal [1.32 shows some typical slurry designs and illustrates how
g/cm3] have little or no free water. The normal concen- the slurry density increases between atmospheric pres-
tration of this material is about 15% BWOC; however, up sure and 3,000 psi [20.7 MPa]. Owing to the density vari-
to 28% BWOC is possible. ability, perlite is rarely used today.
The fineness of condensed silica fume also promotes
improved fluid-loss control, perhaps by reducing the per- 3-5.4.2 Gilsonite
meability of the initial cement filtercake (Mueller and
Dillenbeck, 1991). For this reason, it is also used for the Gilsonite is a naturally occurring asphaltite mineral,
prevention of annular fluid migration (Chapter 9) found primarily in northeastern Utah. The specific
(Grinrod et al., 1988; Golapudi et al., 1993). In addition, gravity of gilsonite is 1.05. The water requirement for
it is used as a source of silica in thermal cement systems gilsonite is low, about 2 gal/ft3; thus, it is possible to
(Chapter 10) (Grabowski and Gillot, 1989; Noik et al., prepare low-density cement systems with relatively high
1998). High retarder concentrations are often required compressive strength (Slagle and Carter, 1959). Up to
owing to the high specific surface area of silica fume. 50 lbm of gilsonite can be used per sack of Portland
Colloidal silica dispersions are aqueous sols of pure cement to achieve slurry densities as low as 12.0 lbm/gal
amorphous silica and traces of sodium hydroxide. Like [1.44 g/cm3]. However, mixing difficulties may occur
microsilica, colloidal silica particles are spherical; how- at such high concentrations. Bentonite is often included
ever, the particle size is about one order of magnitude in the slurry to keep the gilsonite particles evenly dis-
smaller (0.05 μm). Therefore, the surface area of col- persed.
loidal silica is about 500,000 m2/kg. The primary applica- Gilsonite is a black, angular solid, normally supplied
tions of colloidal silica are to prepare low-density as a ground material with a wide particle-size distribu-
cement systems and to prevent annular fluid migration tion (up to 0.6 cm). It is often used to prevent lost circu-
(Bjordal et al., 1993). lation (Chapter 6). Gilsonite has a melting point of
385°F [196°C]. Some softening occurs above 240°F
[116°C], and the particles may tend to fuse. As a result,
gilsonite is not recommended for use in wells with bot-
tomhole static temperatures above 300°F [149°C].

64 Well Cementing
Table 3-10. Properties of Cement Systems Containing Expanded Perlite and Bentonite†
Slurry Properties at Various Pressures
Cement Bentonite Mix Atmospheric 3,000 psi Compressive
Perlite Ratio (%BWOC) Water Strength
Slurry Density Slurry Slurry Density Slurry
(by volume) (gal/sk) (at 24 hr, 100°F,
Volume Volume
(sk:ft3) (lbm/gal) (lbm/ft3) (lbm/gal) (lbm/ft3) 3,000 psi)
(ft3/sk) (ft3/sk)
1:1⁄2 2 6.5 13.80 103.2 1.52 14.85 111.1 1.41 –‡

2 7.0 13.58 101.6 1.58 14.57 109.0 1.47 2,800

2 7.5 13.36 99.9 1.65 14.29 106.9 1.54 –

2 8.0 13.16 98.4 1.72 14.02 104.9 1.61 2,200

2 8.5 12.98 97.1 1.78 13.75 102.8 1.67 –

1:1 2 9.0 12.26 91.7 2.00 13.71 102.5 1.79 1,950

2 9.5 12.15 90.9 2.07 13.55 101.3 1.86 –

2 10.0 12.02 89.9 2.14 13.37 100.0 1.93 1,500

2 10.5 11.91 89.1 2.20 13.20 98.7 1.99 –

2 11.0 11.81 88.3 2.27 13.04 97.5 2.06 1,050

1:11⁄2 2 10.5 11.50 86.0 2.36 13.31 99.6 2.04 –

2 11.0 11.41 85.3 2.43 13.16 98.4 2.11 1,125

2 11.5 11.31 84.6 2.49 13.00 97.2 2.17 –

2 12.0 11.23 84.0 2.56 12.86 96.2 2.24 1,050

2 12.5 11.17 83.6 2.63 12.71 95.6 2.31 890

4 11.5 11.38 85.1 2.50 13.04 97.5 2.18 1,170

4 12.0 11.29 84.4 2.57 12.91 96.6 2.25 1,000

4 12.5 11.21 83.8 2.64 12.77 95.5 2.32 860

4 13.0 11.15 83.4 2.70 12.65 94.6 2.38 740

4 13.5 11.09 82.9 2.77 12.53 93.7 2.45 650

4 14.0 11.03 82.5 2.84 12.43 93.0 2.52 600

1:2 2 12.0 10.92 81.7 2.72 12.98 97.1 2.29 1,300

2 12.5 10.86 81.2 2.78 12.82 95.9 2.35 –

2 13.0 10.80 80.8 2.85 12.71 95.1 2.42 1,025

2 13.5 10.75 80.4 2.92 12.60 94.2 2.49 –

2 14.0 10.69 80.0 2.98 12.49 93.4 2.55 775

2 14.5 10.63 79.5 3.04 12.39 92.7 2.61 –

4 13.0 10.85 81.1 2.86 12.76 95.4 2.43 1,000

4 13.5 10.79 80.7 2.93 12.64 94.5 2.50 870

4 14.0 10.73 80.3 2.99 12.53 93.7 2.56 760

4 14.5 10.69 80.0 3.06 12.43 93.0 2.63 670

4 15.0 10.65 79.7 3.13 12.33 92.2 2.70 590

4 15.5 10.60 79.3 3.19 12.22 91.4 2.76 520

† Data are based on Class A cement.
‡ Not available

Chapter 3 Cement Additives and Mechanisms of Action 65

3-5.4.3 Powdered coal Between 12.5 and 25 lbm of powdered coal are nor-
The performance of powdered coal as an extender is very mally added per sack of cement to prepare slurries with
similar to that of gilsonite. Its specific gravity is slightly densities as low as 11.9 lbm/gal [1.43 g/cm3]. Bentonite
greater (1.30). Like gilsonite, it is coarsely ground (50% is often added to such slurries to prevent the coal from
of particles between 6 and 12 mesh) and is often used to segregating. Table 3-11 illustrates typical slurry designs
prevent lost circulation. Unlike gilsonite, the melting for powdered coal systems.
point of powdered coal is high—1,000°F [538°C].
Therefore, powdered coal can be used in thermal well
environments.

Table 3-11. Physical Properties of Class A Cement Slurries Containing Powdered Coal and Bentonite
Bentonite Powdered Water Slurry Slurry Bentonite Powdered Water Slurry Slurry
(%BWOC) Coal (gal/sk) Density Volume (%BWOC) Coal (gal/sk) Density Volume
(lbm/sk) (lbm/gal) (lbm/gal) (lbm/sk) (lbm/gal) (ft3/sk)
0 0 5.20 15.6 1.18 6 0 8.78 13.7 1.69

5 5.40 15.2 1.26 5 8.98 13.5 1.77

10 5.60 14.9 1.35 10 9.18 13.3 1.86

12.5 5.70 14.7 1.40 12.5 9.28 13.3 1.91

15 5.80 14.6 1.44 15 9.38 13.2 1.95

20 6.00 14.3 1.53 20 9.58 13.0 2.04

25 6.20 14.1 1.62 25 9.78 12.9 2.13

30 6.40 14.0 1.71 30 9.98 12.8 2.22

40 6.80 13.5 1.88 40 10.38 12.6 2.39

50 7.20 13.2 2.06 50 10.78 12.4 2.57

2 0 6.39 14.6 1.35 8 0 9.98 13.3 1.86

5 6.59 14.5 1.43 5 10.18 13.1 1.95

10 6.79 14.3 1.52 10 10.38 13.0 2.04

12.5 6.89 14.1 1.57 12.5 10.48 12.9 2.08

15 6.99 14.0 1.61 15 10.58 12.9 2.12

20 7.19 13.8 1.70 20 10.78 12.8 2.21

25 7.39 13.6 1.79 25 10.98 12.7 2.30

30 7.59 13.5 1.88 30 11.18 12.6 2.39

40 7.99 13.2 2.05 40 11.58 12.4 2.57

50 8.39 12.9 2.23 50 11.98 12.2 2.74

4 0 7.59 14.2 1.52 12 0 12.37 12.6 2.20

5 7.78 14.0 1.60 5 12.57 12.5 2.29

10 7.98 13.7 1.69 10 12.77 12.4 2.38

12.5 8.08 13.6 1.74 12.5 12.87 12.4 2.42

15 8.18 13.6 1.78 15 12.97 12.4 2.47

20 8.36 13.4 1.87 20 12.17 12.3 2.56

25 8.56 13.3 1.96 25 13.37 12.2 2.64

30 8.78 13.2 2.03 30 13.57 12.1 2.73

40 9.18 12.9 2.22 40 13.98 12.0 2.91

50 9.58 12.7 2.40 50 14.38 11.9 3.09

66 Well Cementing
3-5.4.4 Microspheres They have low reactivity in a Portland cement matrix;
Microspheres are small gas-filled beads with specific however, some pozzolanic behavior can occur at high
gravities normally between 0.2 and 0.9. Such low specific curing temperatures (Drochon and Maroy, 2000).
gravities allow the preparation of high-strength, low-per- The particle-size range is between 20 and 500 μm.
meability cements with densities as low as 7.5 lbm/gal The shell thickness is about 10% of the particle radius.
[1.02 g/cm3] without the need for nitrogen. Two types of The composition of the gas inside is a mixture of CO2 and
microspheres are available: glass and ceramic. N2. The microspheres are heavier than their glass coun-
The original application of microspheres was for pri- terparts, with a specific gravity of 0.6–0.9 and a bulk den-
mary cementing of conductor and surface pipes, in sity of 25 lbm/ft3; thus, a higher concentration is neces-
which washouts and low formation fracturing pressures sary to achieve low slurry densities (Harms and Sutton,
are common. However, they are used much more exten- 1981).
sively today, and in many cases microsphere cements As mentioned earlier, both glass and ceramic micros-
have eliminated the need for multistage cementing. pheres are susceptible to breakage and collapse when
Microspheres have also found use in special cement sys- exposed to high hydrostatic pressure; as a result, the
tems with controlled particle-size distributions density of the slurry increases (Messenger, 1974). The
(Chapter 7). A significant limitation of microspheres is pressure resistance of ceramic microspheres can vary
their inability to withstand high hydrostatic pressure widely depending upon the source (Fig. 3-14). This
without being crushed; thus, they cannot be used in deep increase can be predicted and, as shown in Fig. 3-15, can
wells. Microsphere cement systems require special care be taken into account in the design calculations. When
in design and mixing, and these special procedures are the microspheres are crushed, the slurry volume
briefly described below. decreases and the packing volume fraction changes.
Glass microspheres are manufactured from borosili- This can lead to significantly higher slurry viscosities.
cate glass. Several grades are commercially available, The use of ceramic microspheres is not recommended
and the specific gravity varies from 0.12 to 0.80 (Smith et when bottomhole pressures exceed 4,500 psi [31 MPa].
al., 1980). All grades have roughly the same particle-size It is important to ensure that the microspheres do not
distribution (30–40 μm), and their pressure resistance separate from the cement particles during the blending
is directly related to specific gravity. Most grades of glass process. The microspheres must be thoroughly dry-
microspheres withstand pressures up to 5,000 psi [34.5 blended with the cement and not premixed in the water.
MPa]; however, the heavier grades with thicker walls Any variation in the ratio of microspheres to cement will
will survive to 10,000 psi [68.9 MPa]. Glass microspheres result in erratic densities during mixing. The bulk
are significantly more expensive than their ceramic volume of microspheres is high relative to their specific
counterparts; thus, their use is relatively infrequent. gravity; therefore, when preparing blends, it is important
Ceramic microspheres, also called cenospheres, are to ensure that the capacity of the blending equipment is
derived from ash produced by coal-burning power adequate.
plants. Their composition is variable, but the principal
constituents are silica and alumina (Table 3-12) The
material is mainly amorphous; however, small amounts 60
of mullite are sometimes detected by X-ray diffraction. 50
Source 1
40
Volume of crushed 30
microspheres (%) 20 Source 2
Table 3-12. Typical Chemical
Composition of Cenospheres 10
0
Silica (SiO2) 50%–65% 0 1,000 2,000 3,000 4,000 5,000
Alumina (Al2O3) 20%–45% Pressure (psi)

Iron oxide (Fe2O3) 1%–10% Figure 3-14. Percent crushing of two sources of cenospheres
versus hydrostatic pressure.
Calcium (CaO) 0.2%–4%

Magnesium (MgO) 0.5%–2%

Titanium (TiO2) 0.5%–2%

Alkali (Na2O, K2O) 0.5%–4%

Chapter 3 Cement Additives and Mechanisms of Action 67

 14.5
14.0
13.5
Original density: 12.0 lbm/gal
13.0
12.5
Measured slurry 12.0
density (lbm/gal) Original density: 10.5 lbm/gal
11.5
11.0
Original density: 9.0 lbm/gal
10.5
10.0

9.5
0 500 1,000 1,500 2,000 2,500 3,000 3,500 4,000 4,500
Pressure (psi)

Figure 3-15. Density of cenosphere-extended slurries versus pressure.

Microspheres are compatible with any class of Ceramic microspheres (lbm/sk)

cement. Figure 3-16 illustrates the amount of micros- 160
0 50 100 18
pheres required to achieve slurry densities between 8.5 140 16
and 15.0 lbm/gal [1.02 and 1.80 g/cm3]. Mix water 120 14
requirements are shown in Fig. 3-17 and slurry yields in Water 12 Water
100 10
Fig. 3-18. The relationship between the density of (%BWOC) 80 (gal/sk)
8
ceramic microsphere system density and compressive 60 6
strength is illustrated in Table 3-13. 40
0 50 100 150
Ceramic microspheres (%BWOC)
Slurry density (lbm/gal)
Figure 3-17. Water requirements for cenosphere-extended cement
8 9 10 11 12 13 14 systems.
150

Ceramic 100
microspheres Ceramic microspheres (lbm/sk)
(%BWOC) 50 0 50 100
420 6
370
0 320 5
1.00 1.20 1.40 1.60 1.80 270 4
Yield Yield
Slurry specific gravity (L/100 kg) 220 3 (ft3/sk)
170
120 2
Figure 3-16. Typical cenosphere concentration requirements
70 1
versus slurry density.
0 50 100 150
Ceramic microspheres (%BWOC)

Figure 3-18. Yield of cenosphere-extended cement systems.

68 Well Cementing
Table 3-13. Compressive-Strength Data For Cenosphere- control, acceptable slurry rheology, and no solids set-
Extended Cement Systems Mixed with Class G Cement + tling. Without excellent fluid-loss control, the risk of
1% Calcium Chloride† slurry bridging is higher. If solids settling occurs, the
compressive strength and bonding will not be uniform
Curing Compressive Strength Data (psi)
Pressure
across the cemented interval. The maximum slurry den-
(psi) Slurry Mixing Densities (lbm/gal) sity attainable by reducing mix water is 18.0 lbm/gal
8.5 9 9.5 10 10.5 11 11.5 [2.16 g/cm3].
When higher slurry densities are required, materials
0 55 100 160 250 270 –‡ 420
with a high specific gravity are added. Such materials
800 115 115 125 250 250 450 470 must meet several criteria to be acceptable as weighting
agents.
2,000 – – 175 315 355 420 480
■ The particle-size distribution of the material must be
3,000 215 250 295 295 435 640 compatible with the cement. Large particles tend to
† Allslurries were cured for 24 hr at 80°F. settle out of the slurry, while small particles tend to
‡ Not available
increase slurry viscosity.
■ The mix water requirement must be low.

■ The material must be inert with respect to cement

3-5.4.5 Nitrogen hydration and compatible with other cement addi-

Foamed cement is a system in which nitrogen is incor- tives.
porated directly into the slurry to obtain a low-density The most common weighting agents for cement slur-
cement. The system requires the use of specially formu- ries are ilmenite, hematite, barite, and manganese
lated base cement slurries to prepare a homogeneous tetraoxide. A summary of their physical properties
system with high compressive strength and low perme- appears in Table 3-14. The concentrations of each mate-
ability. Incorporating nitrogen allows the preparation of rial normally required to achieve a given slurry density
competent cement systems with densities as low as are plotted in Fig. 3-19.
7.0 lbm/gal [0.84 g/cm3]. The reader is referred to
Chapter 7 for a complete discussion of this important
technology. Table 3-14. Physical Properties of Weighting Agents
for Cement Slurries
Material Specific Absolute Color Additional
3-6 Weighting agents Gravity Volume Water
High pore pressures, unstable wellbores, and (gal/lbm) Requirement
deformable or plastic formations are controlled by high (gal/lbm)
hydrostatic pressures. Under such conditions, mud den- Ilmenite 4.45 0.027 Black 0.00
sities in excess of 18.0 lbm/gal [2.16 g/cm3] are common.
Hematite 4.95 0.024 Red 0.0023
To maintain control of such wells, cement slurries of
equal or greater density are also necessary. Barite 4.33 0.028 White 0.024
One method of increasing the cement-slurry density
is simply to reduce the amount of mix water. To maintain Manganese 4.84 0.025 Reddish 0.0011
tetraoxide brown
pumpability, the addition of a dispersant is required. The
principal disadvantage of reduced-water slurries is the
difficulty of simultaneously achieving adequate fluid-loss

Chapter 3 Cement Additives and Mechanisms of Action 69

 Hematite slurry (32% water BWOC)
Ilmenite slurry (35% water and 0.5%
PNS dispersant BWOC)
Manganese tetraoxide slurry
(40% water BWOC)
Barite slurry (44% water BWOC
+ 0.024 gal of water per pound of barite)
20.0

19.5

19.0

18.5

Slurry 18.0
density
(lbm/gal) 17.5

17.0

16.5

16.0

15.5
0 20 40 60 80 100 120 140
Weighting agent concentration (%BWOC)

Figure 3-19. Densification of cement slurries with various weighting agents.

3-6.1 Ilmenite 3-6.3 Barite

Ilmenite (FeTiO3), a black granular material, has a spe- Barite (BaSO4), a white powdery material, is readily
cific gravity of 4.45. It has little effect on cement hydra- available at most oilfield locations; however, it is not an
tion. As currently supplied, the particle-size distribution efficient weighting agent compared to ilmenite,
of ilmenite is rather coarse; therefore, the slurry viscos- hematite, or manganese tetraoxide. Although it has a
ity must be carefully adjusted to prevent sedimentation. high specific gravity (4.33), additional water is required
Slurry densities in excess of 20.0 lbm/gal [2.4 g/cm3] are to wet its particles, which diminishes its effectiveness as
easily attainable with ilmenite. a densifier. The additional water also decreases the com-
pressive strength of the set cement. Nevertheless, slur-
ries with densities up to 19.0 lbm/gal [2.28 g/cm3] can be
3-6.2 Hematite prepared with barite.
With a specific gravity of 4.95, hematite (Fe2O3) is a very
efficient weighting agent. The material occurs as red
crystalline granules. Unlike ilmenite, it is currently sup- 3-6.4 Manganese tetraoxide
plied with a fine particle-size distribution. At high Manganese tetraoxide (Mn3O4), a reddish-brown powder,
hematite concentrations, addition of a dispersant is has a specific gravity of 4.84. It is available in both solid
often necessary to prevent excessive slurry viscosity. and liquid-slurry forms. The latter form is convenient for
Hematite is routinely used to prepare cement slurries offshore applications. The particles are very small (aver-
with densities up to 19.0 lbm/gal [2.28 g/cm3]; however, age size 5 μm), resulting in a significantly greater parti-
slurries with densities as high as 22 lbm/gal [2.64 g/cm3] cle surface area than other weighting agents. The ten-
can be prepared. dency to settle is less than for hematite, and manganese
tetraoxide can be added directly to the mix water
(Johnston and Senese, 1992). This allows last-minute

70 Well Cementing
slurry density changes for primary cement jobs or emer- sulfonated polymers such as polystyrene sulfonate and
gency plugback jobs. Owing to its very high fineness, polycarboxylate-based products.
magnesium tetraoxide can sometimes shorten the thick- Lignosulfonates are most frequently used as disper-
ening time of a slurry. Slurries with densities as high as sants in drilling-mud formulations (Lummus and Azar,
22 lbm/gal [2.64 g/cm3] can be prepared when man- 1986) but are also effective in cement slurries (Every
ganese tetraoxide is used in combination with hematite. and Jacob, 1978; Detroit, 1980). However, they are also
retarders and cannot be used at low temperatures. It is
important to note that the performance of some ligno-
3-7 Dispersants sulfonates, notably those containing large quantities of
sugars, is very sensitive to cement quality, and gelation
Well cement slurries are highly concentrated suspen-
problems are possible. This can be attributed to a dra-
sions of solid particles in water. Their rheological
matic acceleration of interstitial-phase (C3A and C4AF)
properties are related to those of the supporting liquid
hydration (Michaux and Nelson, 1992). A modified lig-
rheology, the solid volume fraction (volume of parti-
nosulfonate, with a molecular weight in the range of
cles/volume of slurry), and interparticle interactions.
60,000 to 120,000, has been patented as a biodegradable
The aqueous phase of a cement slurry contains ionic
dispersant for offshore applications (Chatterji et al.,
species and organic additives. Therefore, the rheological
2000). The chemical structure of a modified lignosul-
properties of the aqueous phase can differ greatly from
fonate is shown in Fig. 3-20.
those of water, notably when high-molecular-weight
water-soluble polymers (e.g., fluid-loss control agents)
are added. The viscosity of the interstitial fluid can also
vary significantly with temperature.
The solid volume fraction (SVF) can vary from about OH H OH
0.2 for extended slurries (Section 3-5) to about 0.7 for
O C C C
reduced-water slurries. High SVF values generally result
in high slurry viscosities, unless the controlled granu- H SO3Na H
OCH3
lometry process is employed (Chapter 7). Particle inter-
actions depend primarily on the surface-charge distribu- n
tion and steric hindrance effects caused by organic Figure 3-20. Chemical structure of modified lignosulfonate.
molecules adsorbed at the solid-particle surfaces.
Without modification, most cement slurries would not
have the correct rheological properties for proper place-
ment in long, narrow annuli. Cement dispersants, also Polynaphthalene sulfonate (PNS) is by far the most
known in the construction industry as plasticizers and common and cost-effective dispersant for well cements.
superplasticizers, are used to obtain the desired rheo- However, it can no longer be used in some marine envi-
logical properties. ronments owing to its toxicity to algae, its tendency to
This section discusses the various dispersant bioaccumulate, and its nonbiodegradability in seawater.
chemistries and the mechanisms by which they work. PNS is produced from naphthalene by sulfonation fol-
The most important factors affecting the response of lowed by polymerization with formaldehyde (Tucker,
cement slurries to dispersants are also discussed. 1938). Residual sulfonic acid is neutralized with sodium
hydroxide or lime. PNS is available in a wide variety of
molecular weights and degrees of branching (Rixom,
3-7.1 Chemical composition of cement 1974; Costa et al., 1982). The repeating unit has the
dispersants structure shown in Fig. 3-21. The value of n is typically
The first category of cement dispersants, known as plas- low (about 10–20), but conditions are chosen to get a
ticizers in the concrete industry, includes lignosul- proportion of higher-molecular-weight product [molecu-
fonates, modified lignosulfonates, and hydroxycarboxylic lar weight (MW) ≈ 100,000), because it is believed to be
acids such as citric acid, tartaric acid, salicylic acid, glu- more effective. The commercial material is supplied as a
conic acid, and glucoheptonic acid. Most plasticizers act powder or a 40% aqueous solution. For freshwater slur-
as powerful cement retarders (Double, 1983), and in the ries, 0.2–1.0% active PNS BWOC is normally required for
well cementing industry they are considered more in effective slurry dispersion; however, as shown in
this context than as dispersants (Messenger, 1978). Fig. 3-22, concentrations as high as 4% BWOC may be
A second category of dispersants, known as super- necessary for slurries containing NaCl (Michaux and
plasticizers in the concrete industry, includes polynaph- Oberste-Padtberg, 1986). PNS can be used at tempera-
thalene sulfonate, polymelamine sulfonate, and other tures as high as 400°F [204°C]. As shown in Fig. 3-23, the

Chapter 3 Cement Additives and Mechanisms of Action 71

 SO3Na N PNS concentration required to decrease the yield value
H OCH2 – NH C C NH CH2 O of a cement slurry below 10 lbf/100 ft2 [4.8 Pa] varies
CH2
N N depending upon the cement source (Michaux et al.,
C 1986).
n n
PNS N Polymelamine sulfonate (PMS) is frequently used in
CH2SO3Na the construction industry (Malhotra and Melanka, 1979)
but to a limited extent in well cementing, probably
PMS because it is significantly more expensive than PNS.
Figure 3-21. PNS and PMS repeating units.
Melamine reacts with formaldehyde to form trimethylol
melamine, which is in turn sulfonated with bisulfite and
condensed to form a polymer. The structure of the base
72
unit is shown in Fig. 3-21. The polymerization time influ-
ences the MW of the product. The most useful average
MW is about 30,000. The product is available commer-
60 cially in solid form or as a water solution (20% and 40%).
30% NaCl As shown in Fig. 3-24, about 0.4% PMS (BWOC) is typi-
BWOW cally required to achieve proper dispersion; however, as
48
observed with PNS (Fig. 3-23), this concentration varies
from cement to cement. This product is effective at tem-
Yield value 36 peratures up to about 250°F [121°C].
(lbf/100 ft2)

24
40 40
15% NaCl
BWOW 35 35
12 Yield value
0% NaCl 30 30
BWOW
0 25 25
0 1 2 3 4 Plastic Plastic
Yield value 20 20
viscosity viscosity
PNS dispersant (%BWOC) (lbf/100 ft2)
(cp)
15 15
Figure 3-22. Influence of NaCl concentration on the dispersing
ability of PNS (15.8-lbm/gal [1,900 kg/m3] Class G slurry, 77°F [25°C]). 10 10

5 5

48 0 0.00
A 0 0.20 0.40
B Active PMS (%BWOC)
40 C
D Figure 3-24. Yield value and plastic viscosity of Class G slurry con-
E taining PMS dispersant [120°F (49°C)].
32 F
G
Yield value 24 H
(lbf/100 ft2) Other sulfonated polymers used as dispersants
include polystyrene sulfonate (PSS) and a condensation
16 product of aldehyde and ketone that contains sulfonate
groups. PSS is an effective cement dispersant, but it is
8 rarely used because of cost (Biagini, 1982). Its chemical
structure is shown in Fig. 3-25.
0 The MW of the aldehyde/ketone condensation is
0 0.09 0.18 0.26 0.35 0.44 0.53 approximately 15,000, and the dispersant is reported to
PNS dispersant (%BWOC) be salt tolerant (Aignesberger and Plank, 1989). It is
commercially available as a powder or a 33% aqueous
Figure 3-23. Yield value versus PNS concentration for different API/ solution.
ISO Class G cements [77°F (25°C)].

72 Well Cementing
 –CH CH2 60

50 Neat

40
SO3Na Shear stress Spring factor: 1
n 30
(dial reading) Bob: 1
Figure 3-25. Chemical structure of polystyrene sulfonate (PSS). 20
Dispersed
10

Polycarboxylate-based products were developed more 0

recently than PNS and PMS. They are now extensively 0 50 100 150 200 250 300
used in the construction industry and are beginning to Shear rate (rpm)
be used for well cementing, notably at low temperatures
Figure 3-27. Rotational viscometer readings for a neat and a dis-
(Volpert, 2002). The polymer structure is generally com- persed 15.8-lbm/gal [1,900-kg/m3] Class G cement slurry at 120°F
posed of a backbone chain with carboxylic groups upon [49°C].
which polyethyleneoxide (PEO) side chains are grafted.
Sulfonate groups can also be present on the backbone
chain (Fig. 3-26). These polymers are sometimes called well-defined plastic viscosity. As shown in Fig. 3-28,
“comb polymers.” They are manufactured from the cement dispersants reduce both the yield value and plas-
monomers by a free-radical mechanism using peroxide tic viscosity. At high concentrations, the yield value
initiators and, depending on the degree of prepolymer- approaches zero and the cement slurry becomes essen-
ization, can be either block or random polymers. tially Newtonian. Unfortunately, very low yield values
often lead to slurry instability, resulting in solids sedi-
mentation or free water.
CH3 CH3 CH3
When a solid with a characteristic net surface charge
(such as a cement grain) is in contact with a solution, an
C CH2 C CH2 C CH2 electric double layer forms at the surface boundary. The
CH2 a C =O b C =O inner layer, or Stern layer, which is approximately a
SO3Na ONa O CH2 CH2 O CH3
single ion in thickness, consists of counterions that
m n remain fixed to the solid surface. This is associated with
Figure 3-26. Chemical structure of polycarboxylate cement a sharp fall in potential. The outer layer, or diffuse layer,
dispersant. extends some distance into the liquid phase. In this
region, ionic species are free to move; however, because
of the electrostatic field at the surface, there will be a
preferential distribution of positive and negative ions.
3-7.2 Effect of dispersants on the rheology of
This results in a gradual decrease in potential towards
cement slurries the bulk solution, in which the charge distribution is uni-
The rheological behavior of well cement slurries can be form. The diffuse layer in highly ionic systems such as
studied using a rotational viscometer, which enables cement slurries is usually greatly compressed and typi-
determination of the yield value and plastic viscosity cally only a few nanometers thick (Neubauer et al.,
(Chapter 4; Appendix B). This is accomplished by plot- 1998). The boundary between the Stern layer and the
ting the shear stress against the shear rate. Figure 3-27 diffuse layer is called the shear plane. The electrical
shows shear stress/shear rate plots for neat and dis- potential at the shear plane is known as the zeta poten-
persed cement slurries. The curves are approximately tial or the electrokinetic potential. Dispersant mole-
linear. The slope of the line is the plastic viscosity, and cules adsorbed at the surface of cement particles are
its ordinate at the origin is the yield value. positioned in the Stern layer and, therefore, contribute
When the cement powder and water are mixed, a to the zeta potential. Figure 3-28 also shows how the zeta
structure forms throughout the slurry owing to coagula- potential increases with PNS concentration. The zeta
tion, which prevents flow below a given shear stress potential and cement dispersion are discussed later in
threshold, the yield value. At low shear stresses, below this section.
the yield value, the slurry behaves as a solid and does not
flow. Above the yield value, it behaves as a liquid with a

Chapter 3 Cement Additives and Mechanisms of Action 73

 60 30 10 35
Zeta
potential
25
50
7.5 30
20
40 Yield Free water
value Plastic
Zeta potential 15 Yield value Free water 5 25 viscosity
(–mV) (lbf/100 ft2) (%) (cp)
30
10
2.5 20
20
5

Viscosity
10 0 0 15
0 0.25 0.5
PNS dispersant (%BWOC)

Figure 3-28. Yield value, plastic viscosity, zeta potential, and free water for a cement slurry at 185°F [85°C].

3-7.3 Mechanisms of action

The dispersive effect of superplasticizers is caused by C2SH– +Ca+ – HSC2
the adsorption of superplasticizer molecules on the sur- C2SH– +
Ca
faces of cement grains throughout the initial hydration
C2SH– +
reactions to the final set. Depending on the chemical
+Ca+ – HSC2
structure of the dispersant, the effect is attributed to
electrostatic repulsions, steric repulsions between C2SH– + Ca+
cement grains, or both.
In the absence of dispersant, Portland cements con-
tain about 80 wt% of calcium silicates, C3S and C2S Figure 3-29. Cement grain interactions.
(Chapter 2). During hydration, the surfaces of cement
particles are mainly composed of silanol groups
(–Si–OH). The silanol groups are hydrolyzed owing to 1986). Therefore, the particles tend to align with positive
the high pH (greater than 12.5) of the interstitial water. or negative patches. The resulting attractive forces
create a three-dimensional structure. The strength of
–Si–OH + OH –  –SiO – + H2 O (Eq. 3-2) this structure is measured macroscopically by the yield
value.
This should lead to electrostatic repulsion between
To properly describe the interactions between the
the particles and to complete particle dispersion.
cement grains, one must also take ion-correlation forces
However, the presence of calcium ions in the pore solu-
into account. For monovalent ions such as sodium, the
tion adds an attractive force. Calcium ions adsorb onto
effective double layer repulsion is only slightly reduced.
the negatively charged groups, conferring a global posi-
For divalent ions such as calcium, the ion-correlation
tive charge at the surface.
attraction exceeds the repulsive forces between the
– Si – O– +Ca2+  –Si–O–Ca+ (Eq. 3-3) charged surfaces (Israelachvili, 1992). The different
effects of sodium and calcium ions on interparticle
One calcium ion may bind two hydrolyzed silanol forces have been observed experimentally by atomic
groups. As shown in Fig. 3-29, the silanol groups may be force microscopy (AFM). The predominant attractive
on the same grain or bridging two grains (Thomas and force between the cement particles in the presence of
Double, 1981). It must be kept in mind that the positive calcium ions creates a structural network between the
charge on a cement particle in a neat slurry is a global or particles. This explains why a neat, nondispersed
average charge. The surface of hydrating cement is likely cement slurry has a yield stress (yield value) and does
composed of positive patches separated by neutral or not behave as a Newtonian fluid. Cement particles are
even possibly negative regions (Michaux and Défossé, aggregated because of the coagulation process induced

74 Well Cementing
by calcium ions. These aggregates contain entrapped the zeta potential of cement particles. Michaux and
interstitial water, which is not available to lubricate the Defossé (1986) performed such a study with Class G
cement grains when the cement slurry is being pumped. cement. To determine the amount of dispersant
Thus, large cement particle aggregates correspond to adsorbed onto hydrating cement particles, one measures
high slurry viscosity. the concentration of dispersant in the cement-pore solu-
In the presence of dispersant, adding an anionic poly- tion and subtracts this value from the amount of disper-
electrolyte such as polynaphthalene sulfonate or poly- sant originally added. Pore solutions are extracted from
melamine sulfonate can reduce the yield value and plas- the cement slurry by filtration, and the dispersant con-
tic viscosity of cement slurries. In theory, cationic centration is determined by appropriate analytical tech-
polyelectrolytes could also be used for this purpose; how- niques (e.g., ultraviolet spectroscopy for PNS and PMS).
ever, they could react with anionic cement additives Figure 3-30 shows an adsorption isotherm for PNS in a
(e.g., retarders and fluid-loss additives) and cause per- cement slurry. The amount of adsorbed PNS varies with
formance difficulties. In addition, it is possible that the its concentration in the interstitial water, and the
competition between the cationic polyelectrolyte and hydrating cement surfaces become saturated when a
the calcium ions at the cement surface could impair the sufficient amount of PNS is present. When the cement-
cement-hydration process. particle surfaces are saturated, any additional PNS
Cement-slurry dispersion is achieved by adsorption of remains in the pore solution. As hydration continues, the
the dispersant molecules onto the cement-particle and excess PNS can adsorb onto newly formed hydration-
hydration-product surfaces. The adsorption mechanism product surfaces. Adding extra dispersant can help
is still not clearly understood and may depend on the reduce the loss of slurry fluidity that may occur as hydra-
chemical structure of the dispersant. It is generally tion progresses; however, overdosing can result in slurry
believed that the adsorption is caused by ionic bonds sedimentation and free-water development.
between calcium ions that are chemically adsorbed at There is considerable controversy between
the cement surface and the anionic groups (carboxylate researchers concerning the validity of zeta-potential
or sulfonate) of the dispersant. The hydrophobic portion measurements on cement systems. Questions generally
of the dispersant molecule may also preferentially arise about dilution factors, sample preparation, and
adsorb onto cement-grain surfaces (Uchikawa et al., measurement methods. Common methods to measure
1992). the zeta potential, such as electrophoresis or streaming
A number of researchers have attempted to correlate potential, require a low particle concentration in the
dispersant adsorption with cement-slurry rheology and test fluid—much lower than that in a cement slurry. The

60 6
Zeta potential

50
Adsorption

40 4
Adsorbed dispersant
Zeta potential (mg of dispersant
(–mV) per g of cement)
30 2

20

10 0
0 0.50 1 1.50 2 2.25
Equilibrium dispersant concentration
(%BWOW)

Figure 3-30. Adsorption isotherm and zeta potential for a diluted cement suspension [at 77°F (25°C)].

Chapter 3 Cement Additives and Mechanisms of Action 75

zeta potential of cement particles depends on many Figure 3-28 shows that, at various PNS concentra-
parameters that can vary significantly with the water-to- tions, there is a good correlation between the zeta poten-
cement ratio. The most important parameters include tial, the rheological properties (yield value and plastic
the calcium concentration, alkali concentrations, pH, viscosity), and the free water. When the zeta potential
and nature and amount of hydration products. Hence, reaches a maximum negative value, corresponding to
such methods have only a limited value for understand- the saturation of the hydrated cement surface by disper-
ing the surface potential in a concentrated particle sus- sant molecules, the yield value drops to zero. The plastic
pension like a cement slurry. viscosity tends toward a minimum value owing to the
The only way to reliably measure the zeta potential in deflocculation of cement aggregates. Thus, the effective
a concentrated particle suspension, with a solid-to- volume of solids in the suspension is decreased, and the
liquid weight ratio as low as 0.3, is to use the electroki- cement fluidity improves. With further addition of PNS,
netic sonic amplitude method. Charged particles the free water increases dramatically because the
exposed to an alternating electric field in a solution gen- cement particles are now free to sediment according to
erate sound waves. The instrument detects the sound Stokes law. Such cement slurries are considered overdis-
waves and determines the dynamic mobility of the parti- persed.
cles, from which the zeta potential is calculated. Using When particles covered with a layer of adsorbed poly-
this technique, Uchikawa et al. (1997) measured a sur- mer approach one another, a repulsive force can be
face potential of –1.8 mV for a nondispersed cement induced when the layers overlap (Fig. 3-32). It is likely
slurry and –11 mV for a slurry containing 0.6% BWOC that, for cement dispersants with a bulky structure, this
PNS. steric hindrance effect occurs in addition to electrosta-
The time at which the zeta potential is determined is tic repulsion.
also an important parameter. Indeed, the electrical
double layer around the cement particles is not in equi-
librium (Nägele, 1985, 1986, 1987; Hodne and Saasen,
2000). This is owing to the continuous production of ions
at the interfaces between unhydrated cement surfaces
and hydration products. In addition, owing to the forma-
tion of new hydration products, the surface composition
varies continuously with time.
Figure 3-30 shows that the zeta potential increases
when PNS is added to a cement slurry. This increase x
occurs because many of the anionic groups on the PNS
molecules are not involved in the adsorption process. As
a result, as schematically shown in Fig. 3-31, an
adsorbed polyanionic molecule creates several negative
charges. Clearly, when the cement-grain surfaces
become saturated with dispersant molecules, no addi-
tional surface charge can be accommodated and the
zeta potential levels off. The cement grains are nega- Figure 3-32. Schematic representation of steric repulsion between
tively charged and repel each other because of repulsive cement grains with adsorbed polymer dispersant.
electrostatic interactions.

Polycarboxylate dispersants are often comb polymers

with grafted polyethylene oxide [(OCH2CH2)n , PEO] side
SO3–
chains. In the concrete industry, such polymers are
C SH – + Ca + – O S
2 3 among the most efficient superplasticizers. The polymer
Poly backbone consists of carboxylic groups and, sometimes,
sulfonate groups. Adsorption of these polymers onto the
C2SH – + Ca + – O3S
SO3– hydrating cement grains probably occurs by way of the
carboxylic groups. When the polymer is adsorbed onto
a cement particle, a portion of the grafted side
chains is oriented into the solution. In a good solvent
Figure 3-31. Polyanion adsorption on a cement particle surface. such as water, solvent-chain interactions are energeti-

76 Well Cementing
cally favored over chain-chain contacts. For the above- shown that interstitial-phase hydration products adsorb
mentioned comb polymers, the PEO side chains interact much greater amounts of dispersant than silicate-phase
favorably with the aqueous medium and can stretch into hydration products (Uchikawa et al., 1992; Yoshioka et
the solution. As cement particles approach each other al., 2002). This effect is more pronounced with PNS and
and the adsorbed layers of dispersant begin to overlap, a PMS than the polycarboxylate dispersants (Uchikawa et
local increase in osmotic pressure occurs. The increased al., 1995). C3A is much more reactive than C4AF, espe-
osmotic pressure induces a steric repulsive force cially at early hydration times. Thus, the amount of dis-
between the cement particles, resulting in dispersion. persant required to obtain a given level of dispersion
For polycarboxylate dispersants, steric repulsion can increases with the C3A content.
be more important than electrostatic repulsion As described in Chapter 2, sulfate compounds in
(Uchikawa et al., 1997; Yoshioka et al., 1997). Calculated Portland cement control the interstitial-phase hydra-
interparticle potentials indicate that, because of the tion. These include the alkali sulfates (Na2SO4, K2SO4,
high concentration of electrolytes in cement pore solu- and NaKSO4) and the calcium sulfates [CaSO4 • 2H2O
tion (up to 1 M), the repulsive electrostatic forces gen- (gypsum), CaSO4 • 1/2H2O (plaster), CaSO4 (anhydrite),
erated by the polycarboxylates are never capable of over- and CaK2(SO4)2 (syngenite)]. The alkali sulfates are very
coming the attractive Van der Waals forces (Flatt and soluble and readily go into solution when the cement
Bowen, 2003). For this reason, the steric hindrance powder is added to water. The solubility and dissolution
between the adsorbed layers of polymer must be the rates of the calcium sulfates are much lower (plaster >
principal contributer to the dispersion mechanism. gypsum > anhydrite) and can be altered by the presence
of organic compounds. The nature and the amount of
sulfate compounds in Portland cement strongly affect
3-7.4 Factors affecting the response of cements the behavior and efficiency of cement dispersants.
to dispersants ■ For all types of dispersants, the initial cement-slurry
The concentration of dispersant required to effectively fluidity increases with the solubility of the calcium
disperse a cement slurry varies considerably from sulfates (Moulin and Broyer, 2003). However, the
cement to cement. This is clearly illustrated in Fig. 3-23, physico-chemical parameters governing longer-term
which shows the response of several cements that cement-slurry fluidity are more complex.
conform to the API/ISO Class G specification. The PNS
■ Cement-dispersant efficiency is low when anhydrite
concentration necessary to achieve full dispersion (yield
value close to zero) varies from about 0.2% to 0.6% is the principal calcium-sulfate phase (Prince et al.,
BWOC. Many cement properties can affect the perfor- 2003). The low dissolution rate of anhydrite inhibits
mance of dispersants: ettringite formation; consequently, the interstitial-
phase hydration rate is high.
■ cement fineness
■ PNS is more effective with cements that contain
■ nature and amount of calcium sulfates
gypsum as the principal calcium-sulfate phase
■ nature and amount of soluble alkali sulfates (Basile, 1987), rather than plaster or anhydrite.
■ C3A content ■ Sodium sulfate competes with PNS for adsorption

■ distribution of aluminate and silicate phases at the sites during early hydration (Kim et al., 2000;
cement-grain surfaces (Vidick et al., 1987) Chandra and Björnström, 2002a and 2002b). This con-
■ reactivity of the cement phases (particularly C3A and
tributes to longer-term cement-slurry fluidity,
C4AF) (Michaux and Nelson, 1992) because more PNS is left in the aqueous phase to
adsorb onto future hydration products. Jiang et al.
■ cement aging (carbonation and prehydration of anhy-
(1999) determined that, for most dispersants and
drous cement). cement compositions, the optimal dispersant concen-
In general, the amount of dispersant required to tration (expressed as %Na2O) is 0.4–0.5% BWOC.
attain a given level of dispersion increases with the fine- The performance of dispersants can also be influ-
ness of the cement. The number of adsorption sites on enced by other physical and chemical factors.
the cement particles increases exponentially as the par-
■ nature and concentration of salts present or added to
ticle size decreases.
The initial aluminate and silicate hydrates form the mix water (Section 3.7.1)
around the cement grains during the preinduction ■ mixing energy and mixing method

period. The hydration of the interstitial phase (C3A and ■ mix-water temperature
C4AF) during this period is the most important parame- ■ water-to-cement ratio
ter affecting the slurry rheology. Several studies have

Chapter 3 Cement Additives and Mechanisms of Action 77

■ when the dispersant is added [introduced in the mix 60
water or added after cement is mixed with water 50
(post-added)] Mix water
■
40
chemical structure of the dispersant (e.g., MW, linear
or branched polymer and steric size) (Section 3.7.1). Yield value 30
(lbf/100 ft2)
After 5-min
The slurry mixing conditions have a strong effect on 20
delay
the behavior of dispersed slurries. High mixing energy 10
and long mixing times increase the amount of hydrates
formed during the preinduction period. As a result, a 0
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7
greater amount of dispersant is consumed. The effi-
ciency of dispersants is also inversely related to the mix- PNS concentration (%BWOC)
water temperature.
When the dispersant is present in the mix water 50
during the preinduction period, a portion is consumed 40
by the initial hydration products and is no longer avail-
Mix water
able to perform its intended function. Numerous studies Plastic 30
have demonstrated that dispersants are more efficient viscosity
(cp) 20
when they are post-added to a cement slurry (Michaux After 5-min
et al., 1986; Michaux and Nelson, 1992; Collepardi et al., 10 delay
1980; Chiocchio and Paolini, 1985). This effect is partic-
ularly strong with PMS and PNS (Uchikawa et al., 1995; 0
Hanehara and Yamada, 1999). Delayed addition ensures 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7
that all of the dispersant is available to adsorb onto PNS concentration (%BWOC)
cement particles and hydration products. Experimental Figure 3-33. Effect of post-addition of PNS on the rheological prop-
studies have shown that the optimal delay time is erties of a Class G cement slurry (Michaux and Nelson, 1992).
between 1 and 5 min after initial slurry mixing. As shown
in Fig. 3-33, delaying the addition of PNS to a Class G
cement slurry significantly improves the dispersant effi-
ciency.
Water Cement
slurry
3-7.5 Particle settling and free water
As a side effect of dispersant addition, the slurry may Cement
show sedimentation (a slurry-density gradient from the slurry
top to the bottom of a container resulting from particle Cement
settling), free water (a layer of nonparticle-laden fluid slurry
on top of the slurry), or both. Free water (also called
free fluid) can occur with a homogeneous slurry below. Cement Cement
Sedimentation can occur without forming a separate solids solids
water layer.
■ Free water: When the cement particles are not com- Free water Sedimentation Segregation
pletely dispersed, they interact through electrostatic
forces. A flocculated structure forms and supports the Figure 3-34. Three different cement-slurry settling processes.
weight of a given particle. If the annulus in the well is
sufficiently narrow, the weight of the particles is
transmitted to the walls, and the slurry is self-sup- ■ Sedimentation: As described in the previous sections,
porting. Such cases are rare; normally, the weight of dispersants suppress the interactions between
the cement particles is transmitted to the bottom by cement particles by neutralizing positively charged
the gel lattice, and structural deformation occurs. sites. When the process is complete, the particles
Water is squeezed out of the lower portion of the slurry repel each other through double-layer interactions.
and is accommodated in the higher, less-stressed The range of action of these forces is a very short dis-
layers. The ability of the upper layers to accommodate tance owing to the high ionic content of the aqueous
the additional water is limited; thus, a layer of water phase. Therefore, the repulsive forces allow smooth
may form at the top of the slurry (Fig. 3-34). particle packing. In a fully dispersed slurry, the parti-

78 Well Cementing
 cles are free to fall and collect at the container Bentonite may be used to reduce slurry settling
bottom. In reality, this ideal situation never occurs; (Morgan and Dumbauld, 1954). As discussed in
instead, a density gradient forms. Three explanations Section 3-5, bentonite has the ability to absorb large
may be proposed that incorporate the concept of par- quantities of water; as a result, slurry homogeneity is
ticle polydispersity, that is, that small and large parti- preserved.
cles do not behave identically. Various hydrosoluble polymers reduce sedimentation
1. Smaller particles have not settled yet. by increasing the interstitial-water viscosity. The most
commonly used materials are cellulosic derivatives such
2. Brownian motion prevents small-particle settling. as hydroxyethylcellulose (HEC) and welan gum (Allen et
3. A flocculated gel exists but is not sufficiently al., 1991; Skaggs et al., 2001).
strong to support the larger particles. Seawater and silicates can improve slurry stability
(Childs et al., 1984). In addition, metallic salts such as
NiCl2 and MgCl2 build weak but extensive hydroxide
3-7.6 Prevention of free water and slurry structures throughout the slurry (Defossé, 1985a; Kar,
sedimentation 1986). As shown in Fig. 3-36, such structure building sub-
Nonhomogeneous cement columns are not acceptable, stantially reduces free water.
particularly when the wellbore is highly deviated or hor-
izontal (Chapters 12 and 13). Sufficient set-cement
strength and zonal isolation are jeopardized under such 50
circumstances. Careful study of Fig. 3-28, a plot of free 0.5% PNS dispersant BWOC
water and yield value versus PNS dispersant concentra- 40
tion, reveals a narrow range (0.2– 0.3 wt% BWOC) within
30
which the slurry is sufficiently fluid and yet stable. In a Free water
field environment, it is difficult to control additive con- (mL per 250 mL
of slurry) 20
centrations within such a narrow range. Therefore, anti-
settling agents are often added to broaden the concen- 10
tration range within which low yield values and low free 0.3% PNS dispersant BWOC
water can be obtained (Fig. 3-35). Antisettling agents 0
are materials that restore some of the yield value but at 3.5 4.5 5.5 6.5 7.5
a level compatible with the pumping conditions and fric- MgCl2 concentration (%BWOC)
tion pressure the exposed formation can bear. Examples
of such materials are discussed below. Figure 3-36. Free water development of 15.8-lbm/gal [1,900 kg/m3] Class
G slurries with two PNS dispersant concentrations [185°F (85°C)].

70 70
Free water with PNS
60 and antisettling agent 60
Yield value with PNS
Free water with PNS
50 Yield value with PNS 50
and antisettling agent
40 40 Free water
Yield value
(lbf/100 ft2) (mL per 250 mL
30 30 of slurry)

20 20

10 10

0 0
0.1 0.2 0.3 0.4 0.5
PNS dispersant (%BWOC)

Figure 3-35. Yield-value and free-water behavior of 15.8-lbm/gal [1,900-kg/m3] Class G cement slurries
with and without antisettling agent at 185°F [85°C].

Chapter 3 Cement Additives and Mechanisms of Action 79

 The efficiency of antisettling additives can be evalu- solids is deposited on the formation surface. Fluid-loss
ated by measuring the density gradient in a column of agents decrease the filtration rate by reducing the filter-
set cement (Appendix B). A test slurry is placed in a cake permeability, increasing the viscosity of the aque-
cylinder and allowed to set. Wafers of the set cement ous phase, or both.
are sliced from the top, middle, and bottom of the Two principal classes of fluid-loss additives exist:
column. The weight difference between the wafers gives finely divided particulate materials and water-soluble
an indication of the extent of slurry sedimentation. polymers. The chemical and physical nature of each type
Figure 3-37 illustrates typical results for two 15.8-lbm/gal of material, as well as mechanistic hypotheses, is dis-
[1.9 g/cm3] slurries. cussed in this section.

2.4
3-8.1 Particulate materials
2.3
With PNS dispersant
The first fluid-loss control agent used for cement slurries
2.2 was bentonite. Because of the small size of its platelets
2.1 (Section 3-3), bentonite can enter the filtercake and
Slurry density 2.0 lodge between the cement particles, decreasing the per-
(g/cm3) meability of the filtercake. In addition to bentonite, par-
1.9
1.8
With PNS dispersant ticulate systems such as carbonate powder, carbon
and MgCl2 black, microsilica, asphaltenes, and thermoplastic
1.7
resins are used to control fluid loss.
1.6 As described in Chapter 7, latex cements demon-
0 40 80 120 160 200 240 strate excellent fluid-loss control. Latexes are emulsion
(top) Position (cm) (bottom) polymers, usually supplied as milky suspensions of very
small spherical polymer particles (generally between 30
Figure 3-37. Comparison of density gradients in set-cement
columns [15.8 lbm/gal [1,900 kg/m3]; 185°F (85°C)]. and 200 nm in diameter). Most latex dispersions contain
about 45% solids. Like bentonite, such small particles
physically plug small pores in the cement filtercake.
3-8 Fluid-loss control agents The most common latexes for well cements are those
of vinylidene chloride (Eberhard and Park, 1958),
When a cement slurry is placed across a permeable for- polyvinyl acetate (Woodard and Merkle, 1964), and
mation under pressure, a filtration process occurs. The styrene-butadiene (Parcevaux et al., 1985). The first two
aqueous phase of the slurry escapes into the formation, materials are limited to temperatures below 122°F
leaving the cement particles behind. Such a process is [50°C]. Styrene-butadiene latex has been applied at
commonly known as fluid loss and is described in detail temperatures up to 375°F [191°C]. Figure 3-38 is a plot
in Chapter 6. of fluid-loss rate versus styrene-butadiene-latex concen-
If fluid loss is not controlled, several serious conse- tration for various cement slurries.
quences may result that can lead to cement-job failure. A newer particulate fluid-loss additive, based on
As the volume of the aqueous phase decreases, the slurry crosslinked polyvinyl alcohol (PVA) microgels, was intro-
density increases; as a result, the slurry performance duced by Audebert et al. in 1997. It provides excellent
diverges from the original design. If sufficient fluid is fluid-loss control at temperatures up to 250°F [121°C]
lost to the formation, the slurry becomes unpumpable. (Fig. 3-39). This additive does not retard cement hydra-
The API fluid-loss rate of a neat cement slurry tion and is compatible with cement accelerators. Thus,
(Appendix B) generally exceeds 1,500 mL/30 min. As dis- it is particularly suitable for low-temperature applica-
cussed in Chapter 6, an API fluid-loss rate less than 50 tions, for which short waiting-on-cement times are diffi-
mL/30 min is often required to maintain adequate slurry cult to obtain. Crosslinked PVA can also be used in com-
performance. To accomplish such a reduction in the bination with polyvinylpyrrolidone (Moulin, 2001).
fluid-loss rate, materials known as fluid-loss control Cement slurries prepared with this additive combination
agents are included in the slurry design. show excellent gas-tight properties. Such additives can
The exact mechanisms by which fluid-loss control be used in combination with microcement and other
agents operate are not completely understood; however, chemicals for squeeze cementing during which a high
several processes are known to occur. Once fluid loss degree of fluid-loss control is often required (Barlet-
commences across a formation, a filtercake of cement Gouédard et al., 2001; Chapter 14).

80 Well Cementing
 300
Neat (15.8 lbm/gal [1,900 kg/m3])
Barite (18 lbm/gal [2,160 kg/m3])
Bentonite (13.3 lbm/gal [1,600 kg/m3])
250 Fly ash (13.6 lbm/gal [1,630 kg/m3])
Sodium silicate (13 lbm/gal [1,560 kg/m3])

200

Fluid-loss
rate 150
(mL/30 min)

100

50

0
0 0.5 1.0 1.5 2.0 2.5 3.0
Styrene-butadiene latex (gal/sk)

Figure 3-38. Fluid-loss behavior of latex-modified cement slurries at

185°F [85°C].

400

350

132°F [56°C]
300
187°F [86°C]
250

API/ISO 115°F [46°C]

fluid loss 200
(mL/30 min)

150
175°F [79°C]

100

50
97°F [36°C] 162°F [72°C]

0
0.0 0.2 0.4 0.6 0.8 1.0
Fluid-loss additive (gal/sk)

Figure 3-39. Fluid-loss behavior of Class G cement slurry containing crosslinked PVA fluid-loss additive at
various temperatures.

Chapter 3 Cement Additives and Mechanisms of Action 81

 The fluid-loss behavior of cement slurries containing Instead, an initial filtrate spurt occurs that corresponds
particulate fluid-loss additives is different from that to the formation of a thin filtercake. Then the fluid-loss
observed when employing water-soluble polymers. The rate is very low, owing to the low filtercake permeability.
leakoff is not a function of the square root of time.
3-8.2 Water-soluble polymers
Water-soluble polymers received much attention as
50,000 fluid-loss control agents in the early 1940s, when they
were first used in drilling fluids. Today, such materials
Higher are used extensively as fluid-loss control agents for well
10,000 molecular cement slurries. In general terms, they operate by simul-
weights taneously increasing the viscosity of the aqueous phase
5,000
and decreasing the filtercake permeability.

Viscosity 1,000
(cp) 0.020
500
0.016 1

100 Intrusion 0.012

Lower
molecular volume
50 increment 0.008
weights
0.004
2
0 0
0 1 2 3 4 5 6 7 8 0 1 2 3 4 5
HEC (wt%) Pore diameter (µm)

Figure 3-40. Concentration and molecular-weight effect on viscos- Figure 3-41. Pore diameters of two Class G cement filtercakes
ity of aqueous solutions of HEC at 77°F [25°C]. (15.8 lbm/gal [1,900 kg/m3], with 0.5% PNS BWOC, no fluid-loss
additive).

Table 3-15. Efficiency of Different Polymers in Decreasing Cement Filtercake

Permeability and Increasing Filtrate Viscosity at 80°F [25°C]†
Additive Filtercake Viscosity Efficiency Fluid-Loss
Permeability of the Ratio Volume
(mD) Filtrate (cp) (mL/30 min)
None 5,100 1 1 1,600

A–0.35% BWOC 924 2.24 0.280 450

A–0.60% BWOC 140 4.48 0.077 173

A–0.80% BWOC 6.1 3.70 0.018 45

A–1.00% BWOC 4.9 3.32 0.017 20

B–0.30% BWOC 770 3.10 0.217 300

B–0.80% BWOC 5.1 4.80 0.014 26

B–1.30% BWOC 1.3 2.30 0.011 12

C–0.08 gal/sk 1,825 1.01 0.596 240

C–0.20 gal/sk 21 1.05 0.058 43

C–0.40 gal/sk 15 2.05 0.038 14

† From Desbrières, 1988

82 Well Cementing
 The viscosity of a polymer solution is dependent upon Cement slurries containing water-soluble polymers
the concentration and the MW. For example, as seen in must be well dispersed to obtain optimal fluid-loss con-
Fig. 3-40, a 2 wt% solution of low-molecular-weight HEC trol. Sulfonated aromatic polymers or salts are almost
may have a viscosity of 500 cp, but the viscosity of an always added with these materials. As described in
equally concentrated solution of high-molecular-weight Section 3-7, dispersants improve the packing of cement
HEC can be as high as 50,000 cp. Such a high viscosity grains (and perhaps the polymer aggregates) in the filter-
would certainly decrease the filtration rate; however, cake. Thus, as shown in Table 3-16, dispersants reduce the
this strategy alone cannot be relied upon to provide permeability of the cement filtercake and can provide
fluid-loss control, because slurry mixing would be impos- some degree of fluid-loss control on their own (Smith,
sible. 1987). However, one must bear in mind that slurry overdis-
Reduction of filtercake permeability is the more persion and sedimentation may artificially improve the
important parameter for fluid-loss control. When a slurry results of the API/ISO fluid-loss test (Appendix B).
contains sufficient fluid-loss control agent to provide an Unlike particulate fluid-loss additives, water-soluble
API/ISO fluid-loss rate of 25 mL/30 min, the resulting polymers do not promote the formation of a thin and
filtercake is approximately 1,000 times less permeable impermeable cement filtercake. Instead, they simply
than that obtained with a neat slurry (Binkley et al., reduce the rate at which the filtercake thickens. This
1958; Desbrières, 1988). In this case, the interstitial process continues until the slurry dehydrates, leaving a
water viscosity increases, at most, five times (Table 3-15). thick filtercake.
The size of the pores in the cement filtercake can be Several classes of water-soluble polymers are used as
evaluated by mercury porosimetry. The typical size dis- fluid-loss control agents. The chemical properties and
tribution is shown in Fig. 3-41, which shows the median performance of each are discussed separately in the fol-
diameter to be 1 μm. The typical radius of gyration of a lowing sections.
polymer molecule is less than 1,000 Å [0.1 μm]; there-
fore, only clusters of molecules would be sufficiently
large to obstruct a pore in the filtercake. Water-soluble 3-8.2.1 Natural polymers
polymers can form weakly bonded colloidal aggregates 3-8.2.1.1 Cellulose derivatives
in solution that are sufficiently stable to become wedged The first fluid-loss additive based on a water-soluble
in the filtercake constrictions (Christian et al., 1976). polymer was a protein (i.e., a polypeptide) extracted
Such polymers may also adsorb onto the cement grain from soybeans (Alcorn and Bond, 1949). Shortly there-
surfaces and thus reduce the pore size. More likely, a after, ethylene diamine carboxymethylcellulose and
superposition of these two phenomena, adsorption plus other cellulose derivatives were introduced. In the late
aggregation, is the true mechanism of action of poly-
meric fluid-loss agents.
OH
HO
Table 3-16. AP/ISO Fluid-Loss Rates of Densified Slurries CH2
of Classes A and G Cement with a 325-Mesh Screen, CH2 CH2
1,000 psi Pressure, and Temperature of 80°F † O CH2
PNS Fluid Loss (mL/30 min) CH2 O
Dispersant (%) at a Water Ratio (gal/sk) of
CH2 CH2
3.78 4.24 4.75 5.2 CH2
O
0.50 490 504 580 690
CH2 O
O
0.75 310 368 476 530 O OH
1.00 174 208 222 286 O OH
O
1.25 118 130 146 224 OH CH2O OH
CH2 CH2 n
1.50 72 80 92 –‡

1.75 50 54 64 – Figure 3-42. Idealized molecular structure of HEC.

2.00 36 40 48 –
† From Smith, 1987.
‡ Not available

Chapter 3 Cement Additives and Mechanisms of Action 83

 350 tion viscosity: 180 cp), and the typical performance in
1.3% HEC (BWOC) bentonite-extended slurries is shown in Fig. 3-44.
300 1.0% HEC (BWOC)
250 0.8% HEC (BWOC) HEC, as well as hydroxypropyl cellulose (HPC), with a
API/ISO DS range of about 0.9 to 2.8 and a MS range of about
200 1.0 to 6.0, are disclosed as fluid-loss control additives
fluid-loss
rate 150 15.6-lbm/gal when used with high-molecular-weight xanthan gum
(mL/30 min) Class A cement
100 (MW 2,000,000) (Baker and Harrison, 1984).
50
Cellulosic fluid-loss additives generally share certain
disadvantages. They are effective water viscosifiers; as a
0 result, they can increase the difficulty of slurry mixing
95 100 105 110 115 120 125 130 135 140
and ultimately cause undesirable cement-slurry viscosi-
BHCT (°F) fication. At temperatures less than about 150°F [65°C],
Figure 3-43. Typical fluid-loss performance of HEC in normal-density cellulosic fluid-loss additives are efficient retarders;
cement slurries. thus, care must be taken to avoid slurry overretardation.
Also, as shown in Fig. 3-43, the efficiency of the cellulose
polymers decreases with increasing temperature.
14 [1.68] Recent changes in environmental regulations around
the world have encouraged the development of more
environmentally acceptable cement additives.
13 [1.56] Biopolymers such as celluloses are very attractive
Slurry density 2% bentonite because they pose little or no risk to the environment.
(lbm/gal [g/cm3]) Therefore, work has been performed to extend the useful
12 [1.44]
10% bentonite range of cellulosic fluid-loss additives.
Mueller and Bray (1993) found that adding ethoxy-
11 [1.32] late-coated resins in concert with HEC (and also PVA)
0 1 2 3 4 significantly improved fluid-loss performance. Greater
% HEC (% BWOW) efficiency, salt tolerance, and thermal stability were
reported.
Figure 3-44. HEC concentrations required to obtain fluid-loss rates
less than 100 mL/30 min with low-density cement slurries.
A low-molecular-weight ethoxylated HEC (MW =
Temperature range: 80 to 150°F [27 to 66°C]. All slurries are API/ISO 60,000), with 1 to 4 moles of ethylene oxide per anhy-
Class H with 0.5% PNS (BWOC) and fresh water. droglucose unit, was recently introduced (Dao and Vijn,
2002; Vijn et al., 2002). It is effective at temperatures up
to 280°F [138°C] and can be used in cement slurries pre-
1950s, CMHEC was introduced as a fluid-loss additive for pared with fresh water, seawater, or salt-saturated mix
cement slurries (Shell and Wynn, 1958; Greminger, water. A temperature-stabilizing agent, such as synthetic
1958). The basic unit structure of CMHEC is shown in hectorite [Na0.4Mg2.7Li0.3Si4O10(OH)2], sodium thiosul-
Fig. 3-9. phate, or magnesium oxide, can also be added.
More recently, the performance of CMHEC has been
improved by adjusting the DS from 0.1 to 0.7 (car- 3-8.2.1.2 Galactomannans
boxymethyl) and the mole ratio of ethylene oxide to Galactomannans are natural polysaccharides essentially
anhydroglucose (by MS) from about 0.7 to about 2.5 consisting of galactose and mannose units (Fig. 3-45).
(Fig. 3-9) (Chatterji and Brake, 1982; Chatterji et al., They are produced from the endosperm of leguminous
1984). According to Chatterji et al. (1984) the perfor- seeds such as guar or carob. Guar gum is a polymer that
mance of CMHEC in salt slurries can be improved by contains monomeric units of D-mannose linked to one
adding a hydroxycarboxy acid such as tartaric acid. another by 1–4 bonds forming the main chain on which
The most common cellulosic fluid-loss control agent units of D-galactose are branched by 1–6 bonds. In the
is HEC, with a DS range between 0.25 and 2.5 (Hook, oil field, guar and guar derivatives have been used for
1969). The basic structural unit is shown in Fig. 3-42. many years to viscosify hydraulic fracturing fluids.
Various MWs of the polymer are used, depending upon A hydrophobically modified hydroxy propylated guar
the density of the cement slurry. For normal-density slur- (HPG), with a MW of less than 2,000,000, can provide
ries, an HEC of medium MW (2% solution viscosity: fluid-loss control at temperatures up to about 230°F
40 cp) is used. The typical fluid-loss-control performance [110°C] (Audibert et al., 2001).
of this material is shown in Fig. 3-43. A higher-molecular-
weight HEC is used for lower-density slurries (2% solu-

84 Well Cementing
 CH2OH CH2OH
O O
HO H HO H
H H Galactose
H OH H H substituents
H OH
Mannose
H OH H OH backbone
O O
CH2 CH2OH CH2 CH2
O O O O
H H H H
H H H H
O O O O
OH HO H OH HO H OH HO H OH HO H

H H H OH H H H H

Figure 3-45. Molecular structure of guar.

500 poly(aryl-vinylbenzyl) ammonium chloride, i.e., polyca-

tions (Wahl and Dever, 1963), have been reported
400 as effective fluid-loss control additives. In addition,
N-vinylpyrrolidone can be copolymerized with styrene
300
sulfonate (SS) to form an effective fluid-loss control
API/ISO additive (Newlove et al., 1984; Sedillo et al., 1987).
fluid loss PVA is frequently used as a fluid-loss control additive
(mL/30 min) 200 (Harrison, 1968; Carpenter, 1986; Moran and Moran,
1998a and 1998b). This material is particularly advanta-
100 geous for low-temperature applications [less than 100°F
(38°C)] because it has no retarding effect and is com-
0 patible with accelerators such as calcium chloride. The
0 0.2 0.4 0.6 0.8 1.0 fluid-loss control behavior of PVA is shown in Fig. 3-46. It
PVA concentration (%BWOC) is important to note the sharp threshold effect associ-
ated with this additive. Within a very short concentra-
Figure 3-46. API/ISO fluid-loss rate versus PVA concentration in a tion range, the fluid-loss rate falls from 500 mL/30 min
Class A cement slurry with 46% water and 2% calcium chloride at to less than 20 mL /30 min. Therefore, to ensure ade-
100°F [38°C] temperature and at 1,000 psi [7 MPa] of pressure.
quate fluid-loss control, special care is necessary during
field blending operations to verify the correct additive
concentration. As discussed in Section 3-8.1, PVA can be
3-8.2.2 Synthetic polymers crosslinked to form a microgel that is not water soluble.
3-8.2.2.1 Nonionic synthetic polymers
Polyvinylpyrrolidone (PVP) may be used with PNS
dispersants (Boncan and Gandy, 1986). It is also known
to improve fluid-loss control when added with CMHEC
(Hale, 1981) or HEC (Chatterji and Brake, 1982;
Chatterji, et al., 1984).
Complex mixtures containing polyvinylpyrrolidone,
maleic anhydride-N-vinylpyrrolidone copolymer, and

Chapter 3 Cement Additives and Mechanisms of Action 85

 CH2=CH severe cement-slurry retardation (Crema et al., 1989).
C=O Partially hydrolyzed polyacrylamide, containing vari-
AMPS ous proportions of acrylic acid or acrylate units, is often
NH
added to drilling fluids. However, it is difficult to use in
CH3 C CH2 SO3– H+ well-cement slurries owing to the strong interaction
CH3 between the carboxylate groups and cement grain sur-
faces, often resulting in retardation or flocculation.
Nevertheless, some applications have been reported
NH CH2 CH2 using a material with a low AA/AAm ratio, about 0.1
N CH2 CH2 NH CH2 CH2 (McKenzie and McElfresh, 1982).
Poly(ethyleneimine)
CH2 CH2 NH CH2 CH2 NH2 The copolymers of AAm most often described in the
patent literature contain a sulfonate monomer: 2-acry-
lamido-2-methyl propane sulfonic acid (AMPS)
CH CH2 CH CH2 (Mueller, 1992). The structural formula is shown in
n n
CH2 CH2 Polyallylamine Fig. 3-47. AMPS has been copolymerized with the follow-
NH2 NH2 ing materials to produce fluid-loss control agents.
■ AAm (Persinski et al., 1977; Boncan and Gandy, 1986;
Figure 3-47. AMPS structure, polyethylene imine repeating struc- Oswald et al., 2001; Walker, 2002)
ture and branching, and polyallylamine structure.
■ N,N-dimethylacrylamide (NNDMA) (Rao and
Burkhalter, 1986; Brothers, 1987; George and Gerke,
1985; Fry et al., 1987; Chatterji et al., 2001).
3-8.2.2.2 Anionic synthetic polymers
The largest group of anionic polymer fluid-loss additives Terpolymers of AMPS are also used, as listed below.
is composed of co- or terpolymers derived from acry- ■ AMPS + AAm + itaconic acid (IA) (Savoly et al., 1987)
lamide (AAm). Polyacrylamides are nonionic and are ■ AMPS + AA + N-methyl-N-vinyl acetamide (NMVA)
not used by themselves as fluid-loss additives, although (Defossé, 1985)
they can provide some fluid-loss control. They are sub- ■ AMPS + vinyl sulfonate + NMVA (Hille et al., 1987)
ject to rapid hydrolysis to acrylic acid (AA), resulting in

0.9

0.8

0.7

0.6

Fluid-loss 0.5
additive
(gal/sk) 0.4

0.3

0.2
13 lbm/gal [1,560 kg/m3]
0.1 16.2 lbm/gal [1,950 kg/m3]
18.5 lbm/gal [2,220 kg/m3]
0
50 150 250 350 450
Temperature (°F)

Figure 3-48. Typical fluid-loss-control performance of AMPS/N-vinyl-2-pyrrolidone/acrylamide/acrylic acid

tetrapolymer versus temperature. The plot shows the fluid-loss additive concentration in a Class H cement
necessary to achieve an API/ISO fluid-loss rate of less than 50 mL/30 min.

86 Well Cementing
■ AA(AAm) + NMVA + AMPS (Hille et al., 1987)
AMPS may also be part of a copolymer or a terpolymer 1,000
grafted to a lignin backbone and associated with acry-
lonitrile, NNDMA, or AA. These complex polymers are 800
efficient in salt slurries (Fry et al., 1987). API/ISO
fluid loss 600
A polymer composition comprising the random poly-
(mL/30 min)
merization product of acryloylmorpholine with at least 400
one and, preferably, two other monomers within the
group consisting of AMPS, N-vinyl pyrrolidone, and 200

0
500 Low Medium High Very high
103 104 105 106
400 Molecular weight

Figure 3-50. Influence of polyamine MW on fluid-loss control.

300
API/ISO
fluid loss
(mL/30 min) 200 and Loughridge, 1994).

100

0
1.0 1.2 1.4 1.6 1.8 2.0
Fluid-loss additive (%BWOC)

Figure 3-49. Fluid-loss-control performance of a blend of sul-

fonated polyvinylaromatic polymers in a 16.7-lbm/gal [2,005-kg/m3]
salt-saturated Class H cement slurry (37% NaCl BWOW and 40%
H2O) at 200°F [93°C].

vinylphosphonic acid, is effective to reduce fluid loss of

cement slurries at temperatures up to 400°F [204°C]
(Udarbe and Hancock-Grossi, 2000).
A tetrapolymer composed of AMPS, N-vinyl-2-pyrroli-
done, AAm, and acrylic acid controls fluid loss of cement
slurries at temperatures up to 450°F [232°C] (Stephens,
1994). Typical performance of this additive is illustrated
in Fig. 3-48.
Copolymers containing AAm, 3-allyloxyhydroxypro-
panesulfonate, and other monomers are reported to
exhibit excellent fluid-loss control properties at temper-
atures ranging from 80°F [27°C] to 350°F [177°C] (Bair
et al., 2002). They are not salt sensitive and can tolerate
a salt concentration up to saturation.
A tannin grafted with AMPS and AAm provides fluid-
loss control at temperatures up to 400°F [204°C]
(Huddleston et al., 1992). This polymer is also reported
to prevent gas migration through the cement slurry (Eoff

Chapter 3 Cement Additives and Mechanisms of Action 87

Table 3-17. Typical Fluid-Loss Data from Cement Slurries Containing a Polyamine Fluid-Loss Additive†
Fluid-Loss Additive PNS Ilmenite Slurry Density Temperature Fluid-Loss
(%BWOC) (%BWOC) (lbm/sk) (lbm/gal) (°F) (mL/30 min)
0.1 0.5 –‡ 16.2 290 20

0.1 0.5 – 16.2 315 30

0.13 0.5 – 16.2 337 18

0.15 1.0 – 16.8 299 8

0.15 1.5 – 19.0 380 34

0.15 1.5 – 20.0 370 40

0.18 1.0 5 17.4 342 30

0.18 1.0 30 18.2 370 90

0.18 1.0 25 18.0 400 78

0.2 1.2 95 19.2 436 16

0.25 1.5 70 19.0 380 10

0.25 1.5 70 19.0 380 11

† Fluid-loss tests were run with a differential pressure of 500 psi (750 psi with 250-psi backpressure).
‡ Not applicable

Table 3-18. Comparison of Two MWs of Polyallylamine of amine groups (primary, secondary, and tertiary)
Polymers Added at 2% BWOC to 15.8-lbm/gal [1,897 kg/m3] should be present in the chain.
Class G Cement† The dispersant PNS must be present with PEI to
obtain significant fluid-loss control. An insoluble associa-
Molecular Weight API Fluid Loss (mL/30 min) tion is made between the two polymers to create particles
10,000 121 that provide fluid-loss control. As shown in Fig. 3-50, fluid-
loss control improves as the MW of the PEI increases.
150,000 142
† (from
The principal advantage of PEI as a fluid-loss control
Roark et al., 1987a, 1987b, and 1987c).
agent is its effectiveness at high temperatures. As shown
in Table 3-17, PEI provides excellent fluid-loss control at
circulating temperatures as high as 436°F [225°C]. A
notable disadvantage of PEI is its tendency to promote
Sulfonated polyvinyl aromatics such as sulfonated slurry sedimentation (Section 3-5). Although the sedimen-
polystyrene (SPS) (Martin, 1966; Newlove et al., 1984; tation is preventable, slurry design can be very difficult.
Sedillo et al., 1987) and sulfonated polyvinyltoluene Polyallylamine has been reported by Roark et al.
(SPVT) (Wahl et al., 1963) have been identified as useful (1986; 1987a, 1987b, and 1987c) as an effective fluid-loss
fluid-loss control agents. A blend of SPVT, PNS, and a control agent. Instead of being part of the chain back-
sulfonated copolymer of styrene and maleic anhydride is bone, the amine group is pendant (Fig. 3-47). This mate-
effective in salt cement systems (Nelson, 1986). The rial can also be slightly crosslinked to decrease slurry
fluid-loss control performance of this material in a salt- sedimentation. Table 3-18 shows the fluid-loss control
saturated cement slurry is shown in Fig. 3-49. performance of polyallylamine at two MWs.
3-8.2.2.3 Cationic synthetic polymers Various quaternary ammonium or sulfonium monomers
Polyethylene imine (PEI), shown in Fig. 3-47, is an can be copolymerized with other materials to obtain effec-
example of a polyalkylene polyamine that was widely tive fluid-loss control agents. Several are listed below.
used as fluid-loss additive (Gibson and Kucera, 1970; ■ Alkyl ammonium chloride or sulfonium chloride
Scott et al., 1970; McKenzie, 1984). The MW range within (Wahl and Dever, 1963)
which PEI is effective is 10,000 to 1,000,000. Its structure ■ Dimethyl-diallyl ammonium chloride (DM-DAAC)
is likely to be highly branched; therefore, all three types

88 Well Cementing
 (Reese et al., 1986; Oswald et al., 2002) used less often include ground walnut or pecan shells,
■ Methacrylamidopropyltrimethyl ammonium chloride Formica§ chips (available in various particle sizes),
(MAPTAC) (Peiffer et al., 1986; 1987) coarse bentonite, and even corn cobs.
Another important bridging agent is cellophane
The alkyl ammonium and sulfonium chloride are flakes. As the cement slurry encounters the lost-circula-
copolymerized with vinylbenzene to obtain poly(ar-vinyl- tion zone, the flakes form a mat at the fracture face.
benzyl) alkyl ammonium or sulfonium chlorides. DM- They can also plate out across high-permeability zones.
DAAC is copolymerized with AA or methacrylic acid. It is The thickness of the flakes is usually 0.02 to 0.06 mm,
also copolymerized with AMPS and vinylphosphonic acid and the planar dimensions are less than 1 cm on each
or vinylphosphonic salts. MAPTAC is copolymerized with side. The normal concentration of cellophane flakes is
SS or AAm. Such materials are ampholytic polymers between 0.125 and 0.500 lbm/sk.
bearing negative and positive charges at a high pH (such More recently, fibers made from glass or synthetic
as the aqueous phase of a Portland cement slurry). polymers have been added to cement slurries to prevent
lost circulation (Messier et al., 2002). As the cement
§ Formica is a registered trademark of Formica Corp.
slurry enters the lost-circulation zone, the fibers associ-
ate to form a mat that promotes cement filtercake devel-
opment. This technique has been used successfully
for both primary and remedial cementing. The concen-
tration of fibers in the slurry normally varies from 2 to
3 lbm/bbl.

3-9.2 Thixotropic cements

When vugular or cavernous zones are so large that bridg-
3-9 Lost-circulation prevention agents ing agents are ineffective, thixotropic cements are often
indicated. When such slurries enter the formation, they
The loss of circulation during a primary cementing job is
are no longer subjected to shear; as a result, they gel and
a serious problem that often necessitates remedial
become self-supporting. Eventually, the thixotropic
cementing. Circulation losses tend to occur in vuggy or
cement plugs the lost circulation. The chemical nature
cavernous formations, particularly in highly fractured,
of such systems is thoroughly presented in Chapter 7.
incompetent zones that break down at relatively low
hydrostatic pressures.
Usually, the operator will have experienced some cir-
culation difficulties during drilling; thus, measures can 3-10 Miscellaneous cement additives
be taken to prevent their occurrence during cementing. There are a number of materials added to cement slur-
A thorough discussion of the causes of and solutions for ries that do not fit into any general category. These
lost circulation is presented in Chapter 6; however, in this include antifoam agents, fibrous additives to improve
chapter, it is appropriate to briefly mention the common cement durability, radioactive tracing agents, and mud
cement additives used to prevent lost circulation. decontaminants.

3-9.1 Bridging materials 3-10.1 Antifoam agents

Many lost-circulation problems are controlled by the Many cement additives can cause the slurry to foam
addition of materials that physically bridge over frac- during mixing. Excessive slurry foaming can have several
tures and block weak zones. Such materials increase the undesirable consequences. Slurry gelation can result,
resistance of the zone to pressure parting. As a general and loss of hydraulic pressure during pumping can occur
rule, they are chemically inert with respect to Portland owing to cavitation in the mixing system. In addition, air
cement hydration. entrainment may cause higher-than-desired slurry den-
Granular materials such as gilsonite and granular sities. During slurry mixing, a densitometer is used to
coal are excellent bridging agents. As discussed in help field personnel proportion the ingredients
Section 3-5, they are also used extensively as cement (Chapter 11). If air is present in the slurry at the surface,
extenders. They are added in concentrations similar to the density of the system “cement + water + air” is mea-
those specified in Section 3-5. Other granular materials sured. Because the air becomes compressed downhole,

Chapter 3 Cement Additives and Mechanisms of Action 89

the densitometer underestimates the true downhole Another material that dramatically improves the
slurry density. Antifoam agents are usually added to the impact resistance and flexural strength of well cements
mix water or dry-blended with the cement to prevent is particulated rubber (unpublished data, F.E. Hook,
such problems. 1971). This material is usually added in concentrations
Antifoam agents produce a shift in surface tension, up to 5% BWOC. Latex-modified cements also exhibit
alter the dispersibility of solids, or both, so that the con- improved flexural strength (Chapter 7). More recently,
ditions required to produce a foam are no longer pre- the principle of adding flexible particles to cement has
sent. In general, antifoams must have the following char- been developed further. As discussed in Chapter 8, the
acteristics to be effective. well cementing industry is paying more attention to set-
■ Insoluble in the foaming system cement mechanical properties other than compressive
■ A lower surface tension than the foaming system
strength. During a well’s lifetime, the cement sheath can
be exposed to stresses that may cause conventional set
(Lichtman and Gammon, 1979)
cements to fail. Adding flexible particles improves the
The antifoam agent functions largely by spreading on set cement’s resistance to such stresses, providing
the surface of the foam or entering the foam. Because better long-term zonal isolation (Le Roy-Delage et al.,
the film formed by the spread of antifoam on the surface 2000; Bosma et al., 2000).
of a foaming liquid does not support foam, the foam sit-
uation is alleviated.
In well cementing, two classes of antifoam agents are 3-10.3 Radioactive tracing agents
commonly used: polyglycol ethers and silicones. Very Radioactive compounds are sometimes added to cement
small concentrations can achieve adequate foam pre- slurries to more easily determine their location behind
vention, usually less than 0.1% BWOW. casing. Radioactive tracers were once used to determine
Polyethylene glycol is most frequently used because the fill-up or top of the cement column; however, tem-
of its lower cost and is effective in most situations; how- perature surveys and cement bond logs have largely
ever, it must be present in the system before mixing. assumed this function. Radioactive slurries still find
Field experience has shown that post-addition of poly- occasional use in remedial cementing to locate the
ethylene glycol is inefficient, and in some cases foam sta- slurry after placement. A base radiation log is run before
bilization can result. the cement job to measure the natural formation
The silicones are highly effective antifoam agents. radioactivity. After the job is completed, another radia-
They are suspensions of finely divided particles of silica tion log is generated, and the location of the remedial
dispersed in polydimethylsiloxane or similar silicones. slurry is determined by comparison with the base log
Oil-in-water emulsions at 10% to 30% activity also exist. (Chapter 15).
Unlike the polyglycol ethers, the silicones will destroy a The most common radioactive agents for well
foam regardless of when they are added to the system. cementing are 53I131 (half-life: 8.1 days) and 77Ir192 (half-
life: 74 days). The iodine is generally available as a
liquid. Sand or glass beads tagged with iridium 192 are
3-10.2 Strengthening agents often available in areas where tracers are used with
Fibrous materials are available that, when added to well hydraulic fracturing fluids.
cements in concentrations between 0.15% and 0.5%
BWOC, increase the cement's resistance to the stresses
associated with perforating, hydraulic fracturing, and 3-10.4 Mud decontaminants
formation movement (Carter et al., 1968; Shi et al., Certain chemicals in drilling fluids, such as tannins,
1995). Such materials transmit localized stresses more lignins, starches, celluloses, and various chemically
evenly throughout the cement matrix. Nylon fibers, with treated lignosulfonates, can severely retard a Portland
fiber lengths up to 1 in., are commonly used. cement slurry. To minimize such effects in the event that
More recently, metallic microribbons were intro- the cement slurry and the mud become intermixed,
duced to improve the mechanical properties of set chemicals such as paraformaldehyde or blends of
cement (Baret et al., 2002). The microribbon concen- paraformaldehyde and sodium chromate are effective
tration is usually about 1.5% by volume of slurry. At this (Beach and Goins, 1957).
concentration, the set cement has significantly
improved impact resistance, toughness, and tensile
strength. This system is particularly effective for kickoff 3-11 Summary
plugs (Chapter 14).

90 Well Cementing
Table 3-19. Summary of Additives and Mechanisms of Action
Additive Category Benefits Chemical Composition Mechanism of Action
Accelerator Shorter thickening time CaCl2, NaCl, Na and Ca formates, Increased permeability of C-S-H phase layer†
nitrates, and nitrites
Greater early Na silicates Formation of C-S-H phase nuclei
compressive strength by reaction with Ca2+ ions

Retarder‡ Longer thickening time Lignosulfonates Adsorption onto C-S-H phase layer,
Hydroxycarboxylic acids reducing permeability
Cellulose derivatives
Organophosphonates Prevention of nucleation and
growth of hydration products

Certain inorganic compounds Precipitation of impermeable solids

on C-S-H phase layer

Extender Lower slurry density Bentonite Absorption of water

Greater slurry yield Sodium silicates Formation of C-S-H phase +
Pozzolans absorption of water
Gilsonite Lower density than cement
Powdered coal
Microspheres
Nitrogen Foamed cement

Weighting agent Higher slurry density Barite (BaSO4) Higher density than cement
Hematite (Fe2O3)
Ilmenite (FeTiO3)
Manganese tetraoxide (Mn3O4)

Dispersant Lower slurry viscosity PNS Induced electrostatic repulsion

PMS of cement grains
Lignosulfonates
Polystyrene sulfonate
Polycarboxylate Steric hindrance that prevents
Hydroxylated polysaccharides flocculation of cement solids
Hydroxycarboxylic acids

Fluid-loss additive Reduced slurry dehydration Cellulosic polymers Increased aqueous-phase viscosity
Polyamines Reduced permeability of cement filtercake
Sulfonated aromatic polymers
PVP
PVA
AMPS copolymers and terpolymers
Bentonite Particle bridging across cement filtercake
Latexes
Crosslinked PVA

Lost-circulation Prevention of loss of slurry Gilsonite Bridging effect across formation

control agent to formation Granular coal
Cellophane flake
Nut shells
Fibrous additives
Gypsum Induced thixotropic slurry behavior§
Certain soluble sulfate salts
Bentonite
Crosslinked cellulosic polymers

Antifoam agent Reduced air entrainment Polyglycol ethers Insoluble in foaming system
during slurry mixing Silicones Lower surface tension than foaming system

Strengthening agent Increased tensile strength, Glass and polymer fibers Transmission of localized stresses
flexural strength, and Metallic microribbons more evenly throughout cement matrix
toughness Ground rubber

Radioactive tracers Easier determination of 131

53I Emission of radioactivity
cement location behind casing Ir192
77
† Proposed theoretical mechanism
‡ More than one mechanism may apply for certain classes of retarders. See text for clarification.
§ Discussed in Chapter 7

Chapter 3 Cement Additives and Mechanisms of Action 91

4-1 Introduction
Rheology is the study of the deformation and flow of
materials. This chapter considers materials that are
Rheology and Flow
of Well Cement Slurries
Dominique Guillot—Schlumberger 4
gestions for possible improvements. Finally, the equa-
tions governing the flow of cement slurries in pipes and
annuli (both concentric and eccentric) are presented.
essentially in a liquid state. In this particular case, rhe- All of the principal equations are also gathered in
ology can be described as the science that attempts to Appendix A.
determine the intrinsic fluid properties—mainly viscos-
ity—necessary to determine the relationships between
the flow rate (shear rate) element and the pressure gra- 4-2 Basic rheological principles and
dient (shear stress) element that causes fluid move- calculations
ment. To properly design, execute, and evaluate a pri- This section presents basic information about fluid rhe-
mary cement job, one must first understand the ology and fluid flow. It illustrates how the basic rheolog-
rheological properties of the cement slurries. Proper ical calculations are usually performed in most field lab-
rheological characterization is essential to oratories. After reading this section, the reader will have
■ evaluate the slurry’s mixability and pumpability a foundation for the more theoretical discussion pre-
■ optimize mud removal and slurry placement sented later in this chapter.
■ determine the friction pressure when the slurry flows
in pipes and annuli 4-2.1 Types of flow
■ evaluate the slurry’s ability to transport large parti- Under steady-state and isothermal conditions, fluids
cles (e.g., some lost circulation materials and fibers) flow in either laminar flow or turbulent flow. These flow
■ predict how the wellbore-temperature profile affects conditions are separated by a transition zone. Pipe flow
slurry placement will be used as an example to illustrate these concepts.
■ predict the annular pressure after slurry placement.

Despite a large body of research performed during Laminar flow

the past 50 years, a complete rheological characteriza- When a fluid is flowing in a pipe in laminar flow, the indi-
tion of a cement slurry has yet to be achieved. This is vidual particles in the fluid move forward in straight
because the rheological behavior of cement slurries lines parallel to the pipe’s axis (Fig. 4-1). In most cases
depends on many different factors: the particles in contact with the pipe wall are stationary.
■ water-to-cement ratio The velocity of the other particles across the pipe varies
■ specific surface area of the cement powder, and more
according to their proximity to pipe walls, with the par-
ticles at the center of the channel moving at the greatest
specifically the size and shape of the grains
speed.
■ chemical composition of the cement and the relative
The shape of the velocity profile varies from fluid to
distribution of the components at the grain surfaces fluid, depending on the rheological behavior. The veloc-
■ presence of additives ity profile for water, for example, displays a parabolic
■ mixing and testing procedures. shape in which the particles are stationary at the wall
and flow at twice the average velocity at the center. In
This chapter concentrates on the rheological charac- some other fluids, the central portion moves as a solid
terization and flow behavior of cement slurries in a well- plug with similar particle velocities, while the velocity
bore; however, a large portion also applies to spacers. An decreases sharply to 0 close to the pipe wall (Fig. 4-2).
introduction to the basic rheological principles precedes This is not a true flow regime—though it is sometimes
a discussion of common equipment and procedures to called the plug flow regime—but simply a pattern that
characterize cement-slurry rheology. Next, the limita- is specific to some fluids in laminar flow.
tions of current practices are discussed, along with sug-

Well Cementing ■ Chapter 4 Rheology and Flow of Well Cement Slurries 93

Fig. 4-1. Velocity profile of a Newtonian fluid in laminar flow
in a pipe.
Fig. 4-3. Illustration of turbulent flow.

At low flow rates, fluids always flow in the laminar

regime, and their behavior can be characterized by a
property called viscosity that will be discussed below. As
the flow rate increases, fluids enter a transition regime
before turbulent flow is established, and the flow
becomes less dependent on viscosity and more depen-
dent on inertial forces.

4-2.2 Basic definitions

In the laminar flow regime, fluid movement may be com-
pared to a large number of platelets moving parallel to
one another at different velocities (Fig. 4-4).

v1
Fig. 4-2. Velocity profile of a fluid exhibiting plug flow, a type of
laminar flow.
F

A
Turbulent flow
When fluids are in turbulent flow, particles swirl within L
the pipe in a rolling motion that is quite distinct from
the sliding motion of laminar flow (Fig. 4-3). Fluid parti-
cles now have a velocity with components that are not
parallel to the pipe wall. The velocity is also time depen-
dent. In such a flow regime, fluids witness a constant A
transfer of momentum from one region to another. With
turbulent flow, the speed of flow increases rapidly away v2
from the pipe wall and becomes fairly constant through-
out the main part of the fluid. Fig. 4-4. Flow between parallel plates.

94 Well Cementing
 In this simple flow geometry, the velocity of the fluid now, the shear rate can be considered as being propor-
particles varies linearly from one plate to the other. tional to the flow rate; however, this is not correct in the
From Fig. 4-4, the shear rate (or velocity gradient) is strictest sense. Viscosity is the fluid property that gov-
constant and described mathematically in Eq. 4-1. erns the friction pressure gradient/flow rate relation-
ship. This property usually depends on the temperature
the velocity difference betweenn 2 platelets and pressure. For most fluids used in well construction,
Shear rate = ,
the distance between 2 plateleets viscosity also depends on shear rate. This will be
explained in the following section.
or In industrial applications such as cementing opera-
tions, fluids are not always exposed to the simple-shear
d v v1 − v2
= , (4-1) situation described in Fig. 4-4. The velocities of fluid par-
dx L ticles may vary in more than one direction (e.g., axial
flow in an eccentric annulus), or they may exhibit more
where x is an axis perpendicular to the plates.
than one component (e.g., axial flow in an annulus with
The dimensions of Eq. 4-1 are
the inner pipe rotating). The shear-rate and shear-stress
length × time −1 fields are thus described by a tensor, but the viscosity
= time −1 . remains a scalar that depends on some invariants of
length these tensors. Understanding the relationship between
Therefore, the unit of shear rate is sec–1. The symbol the shear-rate and shear-stress tensors involves multiple
. fluid properties; however, this is beyond the scope of the
for shear rate is γ .
Shear stress, denoted by the symbol τ, is the force per discussion. In this chapter, the discussion will be limited
unit of surface area that causes the shearing, or, from to fluid viscosity.
Fig. 4-4:
F 4-2.3 Rheological models
τ= . (4-2) The relationship between shear stress and shear rate in
A
steady laminar flow defines Newtonian and non-
The dimensions of Eq. 4-2 are Newtonian fluids.

force
2
= force × length −2 . Newtonian fluids
length
Newtonian fluids comply with the Newtonian model, in
In common oilfield units, the unit of shear stress is which the shear stress, τ, is directly proportional to the
.
lbf/100 ft2. In the SI system, the unit is the pascal (Pa). shear rate, γ . The equation is
The viscosity of a fluid is the ratio of the shear stress,
. τ = μγ (4-4)
τ, to the shear rate, γ. The symbol for viscosity is μ.
τ This relationship is illustrated in Fig. 4-5. The slope of
μ= (4-3) the line represents the viscosity, μ, of the fluid. This is a
γ
constant that does not depend on the flow conditions
The dimensions of Eq. 4-3 are but simply on temperature and pressure. Common
Newtonian fluids include water, gasoline, and light oil.
force × length −2 In Fig. 4-6 one can see that Newtonian fluids begin to
−1
= force × length −2 × time. flow immediately when a pressure gradient is applied. As
time long as the fluid is in laminar flow, the friction-pressure
In common oilfield units, the unit of viscosity is the gradient/flow rate relationship is linear, as is the shear-
centipoise (cp). In the SI system, the unit is the pascal- stress/shear-rate relationship. However, as the flow rate
second (Pa-s). increases and the flow starts to become turbulent, the
Returning to the simple case of laminar flow in a pipe, relationship is no longer linear, and friction pressure
the shear stress can be considered to be proportional to increases faster than in laminar flow.
the friction pressure gradient (or friction losses). For

Chapter 4 Rheology and Flow of Well Cement Slurries 95

 Newtonian Fluid Power-law fluids
Power-law fluids are part of a class known as pseudo-
plastic fluids. Like Newtonian fluids, power-law fluids
flow immediately when a pressure gradient is applied.
However, unlike Newtonian fluids, the relationship
between shear rate and shear stress is not linear
Shear
(Fig. 4-7).
stress
Viscosity = Slope
Power-Law Fluid

0
0 Shear rate

Fig. 4-5. Shear-stress/shear-rate relationship for a Newtonian fluid. Shear

stress

Newtonian Fluid

0
0 Shear rate
Transition
Friction Turbulent flow Fig. 4-7. Shear-stress/shear-rate relationship for a power-law fluid
pressure with a power-law index, n, of < 1. The power-law curve passes
gradient through the origin.

Laminar flow
Power-law fluids are described by the following equa-
0 tions (k stands for the consistency index).
0 Flow rate τ = kγ n (4-5a)
Fig. 4-6. Friction pressure gradient/flow rate relationship for a
Newtonian fluid flowing in pipe. μ = kγ n−1 . (4-5b)
When n is greater than 1, power-law fluids are shear
thickening. When n is less than 1, they are shear thin-
Non-Newtonian fluids ning. Of course, when n is equal to 1, the power-law
model reduces to the Newtonian model. The viscosity of
The term non-Newtonian covers any fluid whose behav-
shear thinning power-law fluids (the most common)
ior deviates from the classic Newtonian model (i.e., the
varies from infinity at 0 shear rate to 0 at infinite shear
shear-stress/shear-rate relationship differs from a
rate. This lower limit is not physically sound; therefore,
straight line that goes through the origin). In addition to
one should exercise caution when applying the power-
being temperature and pressure dependent, these fluids’
law model to situations outside of the shear-rate range in
viscosities can either decrease with shear rate (in which
which the rheological characterization of a fluid was
case they are called shear thinning) or increase with
performed.
shear rate (in which case they are called shear thicken-
As long as a power-law fluid is in laminar flow, the
ing). Most drilling muds, cement slurries, and heavy oils
friction pressure gradient/flow rate relationship follows
are shear thinning. There are three mathematical
the power law (Fig. 4-8). However, as the flow rate
models commonly used in the well cementing industry to
increases and the flow starts to become turbulent, the
describe the behavior of such fluids.
relationship changes and friction pressures increase
■ Power-law model
more quickly than predicted by the laminar model.
■ Bingham model

■ Herschel-Bulkley model

96 Well Cementing
 Power-Law Fluid or
τy
μ=μp+ . (4-6c)
Transition γ

Therefore, Bingham plastic fluids are shear thinning

Turbulent flow fluids whose viscosities vary from infinity at 0 shear rate
Friction
pressure to their plastic viscosities at infinite shear rates.
gradient Bingham plastic fluids require a minimum pressure
Laminar flow gradient to initiate flow (Fig. 4-10). As long as the fluid
is in laminar flow, the friction pressure gradient/flow
rate relationship increases nonlinearly and then tends
0 towards linear behavior. Thus, it differs from the shear-
0 Flow rate stress/shear-rate behavior because, strictly speaking,
the flow rate cannot be considered to be proportional to
Fig. 4-8. Friction pressure gradient/flow rate relationship for a the shear rate. As in other fluid models, as the flow rate
power-law fluid flowing in a pipe n < 1. The curve passes through
the origin.
increases and the flow starts to become turbulent, the
relationship changes and friction pressures increase
much faster than predicted by the laminar relationship.
Bingham plastic fluids
As shown in Fig. 4-9, the distinguishing characteristic of
Bingham Plastic Fluid
a Bingham plastic fluid is that it will remain unsheared
until the applied stress reaches a minimum value.
Two parameters define the Bingham plastic model: Transition
.
■ the value of τ for γ = 0, τy

■ the slope of the straight line, μp.

Friction
μp is constant and is called the plastic viscosity. τy is pressure Turbulent flow
called the Bingham yield stress. gradient
Bingham fluids behave in a manner described by the
following equations: Laminar flow
τ = τ y + μ p γ when τ > τ y , (4-6a) 0
0 Flow rate
γ = 0 when τ ≤ τ y , (4-6b)
Fig. 4-10. Friction pressure gradient/flow rate relationship for a
Bingham plastic fluid flowing in pipe. The curve does not pass
through the origin.
Bingham Plastic Fluid

Herschel-Bulkley fluids
Herschel-Bulkley fluids combine power-law and
Bingham plastic behaviors. Like Bingham plastic fluids,
there is a yield stress that must be exceeded before flow
Shear
stress
commences (Herschel and Bulkley, 1926). Above the
Slope = Plastic viscosity yield stress, as with power-law fluids, the shear-
rate/shear-stress relationship follows the power law.
Herschel-Bulkley fluids are described by Eqs. 4-6b,
Offset at 0 s–1 = Yield stress 4-7a, and 4-7b, and their behavior is illustrated in
0 Fig. 4-11.
0 Shear rate
τ = τ y + kγ n when τ > τ y (4-7a)
Fig. 4-9. Shear-stress/shear-rate relationship for a Bingham plastic
fluid. The relationship is linear and does not pass through the origin. or
τ y + kγ n
μ= (4-7b)
γ

Chapter 4 Rheology and Flow of Well Cement Slurries 97

 Herschel-Bulkley Fluid Other rheological models
Two other rheological models are often quoted in the lit-
erature concerning cement-slurry rheology: the Casson
model (1959) and the Robertson and Stiff model (1976).
They are described by Eqs. 4-8 and 4-9, respectively.

( )
Shear 2
stress τ= τ y + μ p × γ (4-8)
n
⎡ 1 1 ⎤

( )
Yield stress = Offset at 0 s–1
τ=⎢ τ n + k γ ⎥
n (4-9)
⎣ y ⎦
0
0 Shear rate Like the Herschel-Bulkley model, these two models
combine the use of a yield stress with shear-thinning
Fig. 4-11. Shear-stress/shear-rate relationship for a Herschel- behavior for n < 1.
Bulkley fluid with n < 1. The curve does not pass through the origin.

4-2.4 Shear-rate range encountered in a wellbore

In the most common case, in which n is smaller than Before presenting the equipment and procedures
1, Herschel-Bulkley fluids are shear thinning. Their vis- employed to characterize the rheological properties of
cosities vary from infinity at 0 shear rate to 0 at infinite cement slurries, it is important to discuss the shear-rate
shear rate. As with power-law fluids, this lower limit is range to which these fluids are exposed during place-
not physically sound, and caution should be exercised ment in the wellbore. This discussion is limited to flow in
when applying this three-parameter model to situations pipes and in narrow concentric annuli. Also, the follow-
outside the shear-rate range in which the rheological ing assumptions are made.
characterization was performed. ■ The fluid is incompressible and inelastic.
As long as the fluid is in laminar flow, the friction
■ The flow is laminar, steady, and isothermal.
pressure gradient/flow rate relationship begins with an
offset at 0 flow rate and then increases nonlinearly ■ The velocity at the wall(s) is nil.
(Fig. 4-12). As the flow rate increases and the flow
For a fluid flowing in a pipe or in a concentric annu-
begins to become turbulent, the relationship changes
lus, the shear rate varies from 0 at the pipe axis
. or some-
and friction pressures increase much faster than pre-
where in the annulus to a maximum value, γw, at one of
dicted by the laminar relationship.
the walls.
With Newtonian fluids, the shear rate at the walls
depends only on the size of the flow path and the mean
Herschel-Bulkley Fluid velocity of the fluid, v– (or, by another name, the flow
rate, q), and can be expressed as (Whorlow, 1980)
Transition
8v 32 q
γ NW = = , (4-10)
dw π d 3
w ( )
Friction Turbulent flow
pressure for a pipe of internal diameter dw, and
gradient
12 v 48 q
Laminar flow
γ NW = = (4-11)
o (
do − dw π d + d d − d
w o w )( )2

0
0
for a narrow concentric annulus of inner diameter dw
Flow rate
and of outer diameter do. For non-Newtonian fluids,
Fig. 4-12. Friction pressure gradient/flow rate relationship for a Rabinowitsch (1929) and Mooney (1931) derived the fol-
Herschel-Bulkley fluid with n < 1 flowing in pipe. lowing equations.

98 Well Cementing
 Pipe flow is expressed as for narrow concentric annular flow. In Eq. 4-14 and 4-17,

3 n′ + 1 8 v 3 n′ + 1 ⎛ dp⎞
γ w = × = γ , (4-12) ⎜⎝ d z ⎟⎠
4 n′ dw 4 n ′ NW f

with is the friction pressure per unit length along the axis of
the flow.
n′ =
( ).
d log τ w
(4-13)
.
Thus, the shear rate at the wall, γ w, for non-
⎛ 8v ⎞ Newtonian fluids cannot be defined in a simple way
d log ⎜ ⎟ unless the precise rheology of the fluid is known. For
⎝ dw ⎠ pipes or narrow concentric annuli, Eqs. 4-10 and 4-11
represent only a lower limit for the shear rate at the wall
τw is the shear stress at the wall for pipe flow, and for non-Newtonian fluids, provided they are shear thin-
ning (i.e., n < 1, which is the case for most cement slur-
dw ⎛ d p ⎞ ries and spacers). But it is always worthwhile to calcu-
τw = . (4-14)
4 ⎜⎝ d z ⎟⎠ f late the value that a Newtonian fluid would experience
in a given application because it gives a rough idea of the
Narrow concentric annular flow is expressed as order of magnitude . of the shear rate. Table 4-1 shows
typical figures for γ NW, varying from a few reciprocal sec-
2 n ′ + 1 12 v 2 n′ + 1 onds up to more than 1,000 sec–1.
γ w = = γ , (4-15) As can be expected from Eqs. 4-10 and 4-11, the
3 n ′ do − dw 3 n ′ NW Newtonian shear rate at the wall is extremely sensitive
to the pipe diameter or annular size. Generally speaking,
with the variations in the true shear rate at the wall owing to

n′ =
( )
d log τ w
. (4-16)
variations in hole geometry may be greater than those
caused by variations in n′ (i.e., in the non-Newtonian
⎡ 12 v ⎤ behavior of the fluids).
d log ⎢ ⎥
(
⎢⎣ do − dw ) ⎥⎦
As stated earlier, the shear rate is not uniform across
the gap in either of these geometries. Therefore, theoret-
ically speaking, solving flow equations for time-indepen-
The shear stress at the wall is dent† non-Newtonian fluids in pipes or concentric annuli
do − dw ⎛ d p ⎞ requires a knowledge of the shear-stress/shear-rate rela-
τw = , (4-17) tionship in the range from the shear rate at the wall
4 ⎜⎝ d z ⎟⎠ f down to 0 shear rate. In fact, for a given friction pressure
gradient and therefore a given shear stress at the

Table 4-1a. Newtonian Shear Rates for Various Pipe Diameters and Flow Rates
Pipe flow rate (bbl/min) 1 2 5 10

Pipe flow rate (L/min) 159 318 795 1,590

Pipe flow rate (m3/s) 2.65 × 10–3 5.30 × 10–3 1.32 × 10–2 2.65 × 10–2

Shear rate in 2-in. [50.8-mm] 206 412 1,030 2,060

pipe (sec–1)

Shear rate in 5-in. [127-mm] 13.2 26.4 65.9 132

pipe (sec–1)

Shear rate in 10-in. [254-mm] 1.65 3.29 8.24 16.5

pipe (sec–1)

†Fortime-dependent fluids, the issue can be very different

(see Section 4–4.3).

Chapter 4 Rheology and Flow of Well Cement Slurries 99

Table 4-1b. Newtonian Shear Rates For Various Annular Geometries and Flow Rates
Annular flow rate (bbl/min) 1 2 5 10

Annular flow rate (L/min) 159 318 795 1,590

Annular flow rate (m3/s) 2.65 × 10–3 5.30 × 10–3 1.32 × 10–2 2.65 × 10–2

Shear rate in 4- to 51⁄2-in. 116 231 578 1,160

[101.6- to 139.7-mm]
annulus (sec–1)

Shear rate in 7- to 81⁄2-in. 70.8 142 354 708

[177.8- to 215.9-mm]
annulus (sec–1)

Shear rate in 95⁄8- to 121⁄4-in. 16.4 32.8 81.9 164

[244.5- to 311.2-mm]
annulus (sec–1)

Shear rate in 133⁄8- to 171⁄2-in. 4.60 9.19 23.0 46.0

[339.8- to 444.5-mm]
annulus (sec–1)

wall, the flow rate depends mainly on the local shear- Torsional
stress/shear-rate relationship in a region, from shear spring
.
stress at the wall, γ w, down to typically one-tenth of that
value. For eccentric annuli, the shear range that must be
covered is much wider because of the uneven distribu-
tion of both the gap and the velocity around the annulus
(Section 4-6.3).

4-2.5 Laboratory determination of rheological Inner cylinder

shaft bearing
properties
The standard equipment used to characterize the rheo-
logical properties of cement slurries and other oilfield
Rotor
fluids (drilling muds, spacers, fracturing fluids, etc.) is
a coaxial cylinder viscometer, described by Savins
and Roper in 1954 (Fig. 4-13) (Appendix B). The test
Bob
fluid, contained in a large cup, is sheared between
an outer sleeve (the rotor) and an inner cylinder (the
bob) (Fig. 4-14). The rotor spins at various preselected Cup
rotational speeds. The bob is attached to a torsion spring
that deflects as torque is exerted by the fluid. The tor-
sion-spring deflection is indicated on a dial that reads
from 0 to 300. The dial readings and rotational speeds Fig. 4-13. Schematic drawing of a coaxial cylinder viscometer
are converted to shear stress in lbf/100 ft2 (or Pa) and (Savins and Roper, 1954). Reproduced courtesy of the American
shear rate in sec–1. Petroleum Institute.
The characteristics of the rotor-and-bob geometry are
r2 = 0.725 in. [1.8415 cm]
r1 = 0.679 in. [1.7247 cm]
10, 6, 3, 2, and 1 rpm) rotational speeds. This covers a
L = 1.5 in. [3.8 cm]. Newtonian shear-rate range from at least 5 sec–1 to
Depending upon the model, the outer sleeve can 1,022 sec–1 (at the inner cylinder surface). The 6-speed
rotate at two (600 and 300 rpm), six (600, 300, 200, 100, models are the most common, but 12-speed models are
6, and 3 rpm), or more (previous values plus 60, 30, 20, preferred for reasons that will be explained later.

100 Well Cementing

 The basic procedure is to place the test fluid in the
viscometer and measure the instrument-dial readings at
r2 Ω the various rotational speeds. The dial readings are con-
r verted to shear stress and the rotational speeds are con-
r1 verted to shear rate. Then, a shear-stress/shear-rate plot
Bob L is generated. Depending on the nature of the resulting
curve, one can determine the most appropriate rheolog-
Rotor
ical model (Newtonian, Bingham plastic, or power-law).
Fluid Then, the fluid parameters (e.g., plastic viscosity, yield
value, flow-behavior index, and consistency index) can
be calculated.
(a) (b)
The slurry is first mixed according to a procedure
Fig. 4-14. Schematic diagrams of a couette-type coaxial cylinder defined in API RP 10B and ISO 10426-2 (Appendix B).
viscometer (from Whorlow, 1980). Reprinted with permission from Ideally, the slurry is then brought to temperature and
Horwood Publishing, Ltd. pressure in a consistometer according to the appropri-
ate thickening-time-test schedule.‡ Once the precondi-
The standard torsion spring has a nominal range from tioning temperature and pressure are reached, the
0 to 116,000 dynes-cm [0.0116 N-m], which corresponds slurry is stirred for an additional 20 min. If the test tem-
to a shear-stress range from 0 to 320 lbf/100 ft2 [0 to perature exceeds 185°F [85°C], the preconditioning
153 Pa] (calculated at the inner cylinder surface, and period takes place in a pressurized consistometer. After
assuming a length correction factor of 1.064% to account the preconditioning period, the slurry is cooled as
for end effects). Most manufacturers provide other quickly as possible to 185°F [85°C] and the pressure is
springs with stiffnesses of one-fifth, one-half, two, or five released. Once the preconditioning period is over, the
times that of the standard spring. slurry is vigorously stirred to redisperse solids that may
This equipment satisfies the principal requirements have settled and is then transferred to the preheated vis-
described in Recommended Practice for Testing Well cometer.
Cements (American Petroleum Institute [API] The test consists of shearing the fluid at the lowest
Recommended Practice [RP] 10B (1999), and rotational speed for 10 sec before recording the corre-
International Organization for Standardization [ISO] sponding torque reading. All the remaining readings are
10426-2, (2003). The radial ratio of the cylinders is larger then taken in ascending order, and then in descending
than 0.9 to ensure homogeneous shear. The instrument order, after 10 sec of rotation at each speed. The objec-
provides at least 5 readings within a shear-rate range tive of this sequence, known as a hysteresis loop, is to
from near 0 to at least 511 sec–1. The size of the annular detect true time-dependent effects such as thixotropy or
gap (1.17 mm) is such that slurries with particles as other artifacts that will be discussed later.
large as 120 μm can be considered homogeneous. During the test, the slurry should not be exposed to
shear rates higher than 511 sec–-1 (i.e., 300 rpm with the
standard spring). This restriction greatly improves the
Pressurized equipment reproducibility of results among laboratories (Beirute,
Several coaxial cylinder viscometers can measure fluid 1986). This point is illustrated in Figs. 4-15 and 4-16. It is
rheology at elevated temperatures and pressures important to point out that these data were generated
(Appendix B). Pressure is applied to prevent water evap- using an obsolete version of the API procedure.
oration or boiling at temperatures above 185°F [85°C].
One may also investigate the effect of pressure on a
fluid’s rheological properties. The couette geometry con- 4-2.5.2 Gel strength
forms to the requirements described in API RP 10B and The yield value (expressed in lbf/100 ft2 or Pa) is a mea-
ISO 10426-2. surement of the attractive forces that exist between the
particles in a fluid while under flowing conditions. The
gel strength is a measure of the attractive forces
4-2.5.1 Experimental procedure between the particles in a fluid under static conditions.
The key steps of the test procedure and common devia- The gel strength may be measured after the hystere-
tions from it are summarized below. More details are sis loop or as an independent measurement. The vis-
given in Appendix B. The complete procedure is cometer is turned off for 10 sec, after which the rota-
described in API RP 10B (ISO 10426-2). tional speed is set to be equivalent to 5.1 sec–1 (3 rpm).
‡ The preconditioning period may be omitted if the objective is to measure the
rheology of a cement slurry at ambient temperature just after mixing.

Chapter 4 Rheology and Flow of Well Cement Slurries 101

 Class H Cement + 38% Water (BWOC) at 100°F γ = 16.28 × Ω, (4-18)
200
where Ω is in rad/s, or
160
γ = 1.705 × Ω,
Instrument 120
dial Lab A
Lab B
where Ω is in rpm.
reading 80
Lab C
Lab D τ = 0.5109 × θ, (4-19)
40 Lab E
Lab F
0 Lab G where θ is in Pa, or
0 100 200 300 400 500 600
Rotational speed (rpm) τ = 1.067 × θ, (4-20)
Fig. 4-15. Poor repeatability of rheology data measurements where τ is in lbf/100 ft2.
involving shear rates above 511 sec–1. Several laboratories The peak readings are converted to 10-sec and 10-min
used the same cement materials, mixing method, and test gel strength values using Eq. 4-19 or 4-20.
procedure (after Beirute, 1986). Reprinted with permission from Oil
& Gas Journal.
Strictly speaking, the shear-rate formula is only truly
valid for Newtonian fluids; nevertheless, it leads to rea-
sonably accurate values for other fluids (Section 4-3.4).
These shear-rate/shear-stress values are fitted to various
Class H Cement + 38% Water (BWOC) at 100°F
rheological models. The model that best fits the data is
160
selected.
140
120 Example
100 The data analysis will be illustrated using the specific
Dial 80 example of a slurry containing a polymeric fluid-loss
reading additive that has a strong effect on slurry rheology. Using
60
the standard API/ISO procedure (Appendix B), the fol-
40 lowing data were recorded.
20
0
Table 4-2. Viscometer Readings for a Cement Slurry†
0 50 100 150 200 250 300 350
Rotational speed (rpm) Rotational Ramp-Up Ramp-Down Average
Speed Readings Readings Readings
Fig. 4-16. Improvement of repeatability of rheology data as a result (rpm)
of limiting shear rate to 511 sec–1 (compare to Fig. 4-15) (after
300 97 97 97
Beirute, 1986). Reprinted with permission from Oil & Gas Journal.
200 69 70 69.5

100 39 39 39
The maximum reading is the 10-sec gel strength. The dial
60 27 26 26.5
reading after 1 min of rotation at 5.1 sec–1 is sometimes
also recorded and compared to the 10-sec gel to quantify 30 18 16 17
the gel’s shear sensitivity. To determine the 10-min gel
6 9 7 8
strength, the above operation is repeated 10 min after
turning off the viscometer. 3 8 5 6.5
† Measurement is performed with the standard couette-viscometer geometry
and standard spring F1.
4-2.5.3 Data analysis
At any given rotational speed, Ω, the ramp-up/ramp-
down dial readings, θ, are averaged§ and then converted The data were then plotted and analyzed (Fig. 4-17).
to shear rates and shear stresses at the inner cylinder The analysis of the data using the procedure just
(bob) using the following equations: described gives the following results for the Bingham
plastic and power-law models.
§ Ofcourse, this averaging is not meaningful if there are large differences
between the ramp-up and ramp-down readings.

102 Well Cementing

 60 ries. The Bingham plastic model tends to overestimate
Measured
50 Bingham plastic
shear stresses at low shear rates, while the power-law
Power law model can lead to significant errors across the entire
40 shear-rate range. The power-law model†† usually underes-
Shear timates shear stresses at both ends of the shear-rate
30
stress
(Pa)
spectrum and overestimates them at intermediate values.
20
10
Comments
0
0 100 200 300 400 500 600
Because API/ISO standards discourage using the 600-rpm
speed, the 2-speed equipment should no longer be used.
Shear rate (s–1)
The six-speed models also suffer from a severe limitation.
Fig. 4-17. Rheological plot corresponding to measured data shown Because the 6- and 3-rpm readings are not very accurate,
in Table 4-2. Fitting to Bingham plastic and power-law models is or may be affected by wall slip (Section 4-3.7), the user
also shown. The shear stresses and shear rates were determined may be left with only 3 useful readings—100, 200,
from Eqs. 4-18 and 4-19. Model parameters are given in Table 4-3. and 300 rpm. These rotational speeds correspond to a
fairly narrow shear-rate range (170 sec–1 to 511 sec–1).
Therefore, when the maximum shear rate experienced by
Table 4-3. Rheological Model Parameters† a cement slurry during placement in the wellbore is likely
Correlation n or τy k or µp to be lower than 170 sec–1, using equipment allowing mea-
Coefficient surements between 6 and 100 rpm [10 sec–1 and 170 sec–1]
Bingham plastic 0.9992 3.70 Pa 91.3 mPa-s
is strongly recommended. Section 4-2.4 describes how
one determines the anticipated wellbore shear rates.
Power law 0.9878 0.5859 1.063 Pa-sn When standard oilfield equipment is well maintained,
† Determined using API-recommended equations for the data presented in Table 4-2. the accuracy of a built-in torque-measuring device is rea-
sonable. Once calibrated, a typical expected shear-stress
error at low dial readings (≈10) is ±15%. The error can be
Table 4-4. Relative Shear-Stress Errors Between the much higher if the bearing spring is damaged. It is not
Bingham Plastic and Power-Law Models and the unusual to encounter equipment for which the relative
Measured Data† error is on the order of ±50% at such low shear stresses
Rotational Speed Bingham Plastic Power Law
(Fig. 4-18).
(rpm) (%) (%)
300 –1.6 17.2
160
Damaged bearing
200 2.0 8.8 140 New bearing
100 3.4 –8.2 120
100
60 3.7 –18.1
Relative 80
30 3.7 –22.6 error 60
(%)
6 –13.4 –1.5 40
3 –25.5 16.8 20
y −y 0
† Relative error is defined as: calculated measured .
y measured
–20
1 5 10 50 100
Shear stress (Pa)
Even though the fit is not perfect, the Bingham plas- Fig. 4-18. Relative error of shear-stress measurements using stan-
tic model better describes the rheological behavior of dard oilfield equipment. Tests were performed with a Newtonian oil
the cement slurry. using the standard API/ISO procedure.
This example highlights the typical weaknesses of
these models in describing the rheology of cement slur-
†† Thepower-law model parameters are determined from a linear regression
performed on the logarithm of the shear stresses and the shear rates as rec-
ommended in API RP 10B (1999) and ISO 10426-2 (2003). Better prediction
of shear stresses in the intermediate and high-shear-rate range could be
obtained by performing a nonlinear regression directly on the shear stresses
and shear rates.

Chapter 4 Rheology and Flow of Well Cement Slurries 103

 This creates problems when trying to characterize Because both limits of the integral are functions of the
the rheology of low-viscosity fluids such as dispersed torque, there is no general analytical expression for the
cement slurries. shear rate and the viscosity of a non-Newtonian fluid in
By the same token, it is equally important to make such a geometry. Therefore, the shear-rate profile cannot
sure that the bob and rotor are concentric and that they be determined a priori, because it depends on the non-
end up on the same plane at the bottom of the coaxial Newtonian behavior of the fluid, as well as on the rota-
geometry, because this design minimizes end effects. tional speed and the viscometer dimensions. To use such
equipment to measure the flow curve of a non-Newtonian
fluid, it is necessary to either assume a specific rheologi-
4-3 Limitations of field practices cal model to use in conjunction with Eq. 4-22 or to make
r2 /r1 sufficiently close to unity that the variations of
4-3.1 Principle and flow equations shear stress across the gap are negligible.
As discussed earlier, in a coaxial cylinder viscometer the For fluids exhibiting a yield stress, Eq. 4-22 is valid
test material is confined between two concentric cylin- only when τ2 ≥ τy. When τ2 ≤ τ y ≤ τ1 the shear stress from
ders of radii r2 and r1 (r2 > r1), one of them being rotated r2 to ry is smaller than the yield stress of the fluid; there-
at a velocity Ω. Assume for the time being that the fluid fore, the effective annular gap is reduced. Because the
elements are moving in concentric circles around the shear rate is 0 from r2 to ry, this parameter is defined by
common axis (Fig. 4-14).
In steady state, a momentum balance shows that the χ
shear stress τ at any radius r is given by Whorlow (1980). ry = , (4-23)
2πτ y
χ
τ= , (4-21) and Eq. 4-22 then becomes
2 πr 2
1 τ1 γ
2 ∫τ y τ
where χ is the torque acting per unit length on a cylin- Ω= dτ. (4-24)
drical surface of radius r. In practice, χ is measured
from the torque acting on one of the cylinders of length
When the condition to apply Eq. 4-24 is satisfied, i.e.,
L. This expression shows that the shear stress decreases
the fluid is not sheared from r2 to ry, fluid particles in
from a maximum value of
this part of the annular gap are moving at the same rota-
χ tional velocity as the outer cylinder as shown in Fig. 4-19.
τ1 =
( )2
2 π r1
Consequently, this situation is sometimes referred to as
plug flow. Of course, when τ1 ≤ τy there is no flow at all.
at the inner cylinder surface to
χ
τ2 = 12
( ) 2 Sheared Unsheared
2π r2 10 fluid fluid

at the outer cylinder surface. Shear stress (and there- 8

Angular
fore shear rate) will be uniform only if the radius ratio, velocity 6
r2 (rpm)
4
Rr = ,
r1 2

is close to unity. The radius ratio is especially important 0

Inner ry Outer
for shear thinning fluids. As the radius ratio increases, so cylinder cylinder
does the range of shear rates the fluid will experience in
Radius (cm)
the viscometer.
The governing flow equation in a coaxial cylinder vis- Fig. 4-19. Velocity profile of a Bingham plastic fluid in a standard oil-
cometer is (Whorlow, 1980) field viscometer. r 1 = 1.7247 and r 2 = 1.8415 cm, corresponding to the
inner and outer radii of the cylinders. In this particular case, the
1 τ1 γ fluid is not sheared between r = 1.8181 and r 2, because the local
2 ∫τ2 τ
Ω= dτ. (4-22) shear stress is smaller than the fluid yield stress.

104 Well Cementing

4-3.2 Flow of model fluids in coaxial cylinder
viscometers γ 1 =
( )2 × Ω + ⎡⎢ 2 × ( Rr )2 × ln ( Rr ) − 1⎤⎥ × τ y ,
2 × Rr
When a rheological model is assumed for the fluid to be ( Rr )2 − 1 ⎢⎣ ( Rr )2 − 1 ⎥ μp
⎦
characterized, a simple analytical expression can some-
times be determined for the torque as a function of the (4-32)
rotational speed.
For a Newtonian fluid, the flow equation is and

χ ( )2 × Ω ,
2 × Rr γ 2 =
2Ω
+
⎡ 2 ln R
⎢ ( )
r
⎤ τ
− 1⎥ ×
y
, (4-33)
= μ× ⎢ ⎥ μp
2π r1( ) 2
( Rr )2 − 1
(4-25)
( )
2
( )
2
Rr − 1 ⎣ Rr − 1 ⎦

and the shear rates at the inner and outer surfaces are, If 2π(r1)2τy ≤ χ ≤ 2π(r2).2τy, part
. of the fluid is not
respectively, sheared, the expressions for γ 1 and γ 2 are implicit, and:

( )2 × Ω
2 × Rr
χ τy ⎢
⎡ ⎛
χ
⎞⎤
γ 1 = (4-26) Ω= − ⎜ ⎟ ⎥ . (4-34)
( Rr )2 − 1 ⎢1 + ln ⎜ ⎥
2
( )
4 π r1 μ p 2 μ p⎢
⎣ ⎝
2
( )
2π r1 τ y ⎟⎠ ⎥
⎦
and
If χ ≤ 2π(r1)2τy, then none of the fluid can flow and
2Ω
γ 2 = , (4-27) Ω = 0. (4-35)
( Rr )2 − 1 Thus, for Newtonian and power-law fluids flowing in a
where Rr = r2 /r1. coaxial-cylinder geometry, the relationship between the
For a power-law fluid, the corresponding equations are torque and the rotational speed is similar to the relation-
ship between shear stress and shear rate. It is linear for
n Newtonian fluids and follows a power law for power-law
⎧ 2 ⎫
χ ⎪⎪ 2 × Rr
= k× ⎨
( ) n × Ω ⎪⎪ fluids.
⎬ , (4-28) For Bingham plastic fluids, as for all fluids exhibiting
( )
2π r1
2
⎪ n⎡ R
⎪⎩ ⎢⎣ r ( )
2 ⎤⎪
n − 1⎥
⎦ ⎪⎭
a yield stress, the equations are more complex. In the
absence of a plug-flow region, there is a linear relation-
ship between the torque and the rotational speed, with
2 an apparent intercept equal to
γ 1 =
2 × Rr( )n × Ω ,
⎡ 2 ⎤
(4-29)
(
4 π r1)2
( r2 )
2
r
( )
n⎢ R
⎣ r
n − 1 ⎥⎦ χ = τy
( r2 ) − ( r1 ) r1
2 2
ln 2 . (4-36)

and
2Ω Below a given torque value, χ = 2π(r2)2τy, the rela-
γ 2 = . (4-30) tionship is independent of the outer radius, r2, and
⎡ 2 ⎤
n⎢ R
⎣ r ( ) n − 1⎥⎦ nonlinear with an intercept of χ = 2π(r1)2τy for Ω = 0
(Fig. 4-20).
For a Bingham plastic fluid, different equations apply Therefore, deriving the rheological parameters for
depending on the torque value. If χ ≥ 2π(r2)2τy, then the Newtonian and the power-law models from a series
all of the fluid in the gap is sheared in laminar flow, and of torque/rotational speed measurements is straightfor-
the governing equations are ward. However, this simple approach cannot be applied
to Bingham plastic fluids (and to fluids exhibiting a yield
χ
=
( )2
2 Rr
× ⎡⎣ μ pΩ + τ y ln ( Rr ) ⎤⎦ , (4-31)
stress in general). Indeed, the flow behavior is described
by Eqs. 4-31 and 4-34, whose limits of validity depend on
2 π ( r1 ) ( Rr ) − 1
2 2 the yield stress. This problem is usually overlooked, and
all data are fitted according to the linear equation
(Eq. 4-31).

Chapter 4 Rheology and Flow of Well Cement Slurries 105

 fτ is the shear-stress factor and is given by
K
fτ = , (4-39)
( ) 2
2 π r1 ( L + Δ L)
Torque per
unit length where K is the standard spring calibration constant and
χ = 2π(r2)2τy
(χ)
Δ L is the length correction accounting for end effects.
When stresses are expressed in Pa, fτ is equal to 0.5109.
When stresses are expressed in lbf /100 ft2, fτ is equal to
χ = 2π(r1)2τy 1.067 (see Eq. 4-19 and 4-20). The Newtonian shear rate
at the inner cylinder can be written as
τy r2 Angular velocity (Ω)
μp
ln
r1
γ NW =
( )2 × Ω = f × Ω,
2 × Rr
(4-40)
γ
Fig. 4-20. Torque/angular velocity graph for a Bingham plastic fluid ( Rr )2 − 1
in a coaxial-cylinder viscometer. χ = torque per unit length.
where fγ. is the shear-rate factor. When rotational speeds
are expressed in rpm, fγ. is equal to 1.705. When rota-
Similar equations for the Casson model and the tional speeds are expressed in rad/sec, fγ. is equal to
Robertson and Stiff model can be found in Whittaker 16.28 (see Eq. 4-18). Using these two new parameters for
et al. (1985). For Herschel-Bulkley fluids the governing Newtonian fluids, Eq. 4-25 can be reformulated as
flow equation is
fτ × fspring × θ = μ × fγ × Ω, (4-41)
1

1 τ1 (τ − τ )y
n and therefore the fluid viscosity can be determined by
Ω= 1 ∫X τ
dτ, (4-37) performing a linear regression between the rotation
speeds and the torque readings, forced through the
2k n origin. If m is the slope of this regression, the fluid vis-
where, as explained above, X is either equal to τ2 or τy. cosity is given by
There is no simple analytical solution to this expression, fτ × fspring
and determining the rheological parameters from a μ= × m. (4-42)
series of rotational speeds and torque readings is more fγ
complicated than using rheological models involving
only two parameters such as the power-law model or Using the standard spring, μ = 0.3 × m when rota-
Bingham plastic model. tional speeds are expressed in rpm and viscosities
in Pa-s.
For power-law fluids, Eq. 4-28 can be reformulated as
4-3.3 Application to standard oilfield equipment
n
When using the standard oilfield equipment, the shear ⎧ 1 − R –2 ⎫
( )
⎪ ⎪
( )
r n
stress at the inner cylinder can be written as fτ × fspring × θ = k × ⎨ ⎡ –2 ⎤ ⎬ × fγ × Ω .
χ
= fτ × fspring × θ, (4-38) ⎩ ⎣ r ( )
⎪ n ⎢1 − R n ⎥ ⎪
⎦⎭
( )2
2 π r1 (4-43)
where θ is the torque reading (dial reading). fspring is a Therefore, the power-law and consistency-fluid
normalized spring calibration factor equal to 1 for the indices can be determined from a linear regression on
standard spring. Springs are also available with the fol- the logarithm of the readings and the logarithm of the
lowing factors: 0.2, 0.5, 1.0, 2, and 5, which allow the rotational velocities. If m is the slope of this regression
same equipment to perform measurements throughout a
wide range of fluid viscosities.

106 Well Cementing

and b its intercept, the power-law parameters of the When using the standard spring,
fluid are given by
μ p = 0.3 × m (4-51)
n= m (4-44)
and When the rotational speeds are expressed in rpm and
the plastic viscosities are expressed in Pa-s,
n
⎧ ⎡ −2 ⎤ ⎫
τ y = 0.5109 × 0.9372 × b = 0.4789 × b
( )
⎪ n ⎢1 − Rr n ⎥ ⎪
⎪
k = ⎨ ⎢⎣ ⎥⎦ ⎪⎬ f × f b
τ spring × 10 . (4-45) when τ y is in Pa and
⎪⎡ ⎪
( ) –2 ⎤
⎪⎩ ⎢⎣1 − Rr ⎥⎦ × fγ ⎪⎭ τy = b (4-52)

Using the standard spring and expressing the rota- when τy is in lbf/100 ft2.
tional speeds in rpm, For the power-law and Bingham plastic models, the
two rheological parameters can in principle be derived

( )
n
⎡ −2 ⎤ from readings taken at only two rotational speeds
k = ⎢ 15 n 1 − 1.068 n ⎥ 0.5109 × f b
spring × 10 ,
(Whittaker et al., 1985). But this so-called two-point
⎢ ⎥ method should always be applied with caution, particu-
⎣ π ⎦ larly outside the shear-rate range limited by the two
(4-46) rotational speeds used.
Applying Eqs. 4-43 and 4-48 to the example given in
where k is in Pa-sn, and Table 4-2 leads to the following values of the rheological

( )
n parameters.
⎡ −2 ⎤ One can see that the yield stress and the consistency
k = ⎢ 15 n 1 − 1.068 n ⎥ 1.067 × f b
spring × 10 , index are different from those presented in Table 4-3. An
⎢ ⎥
⎣ π ⎦ explanation for the discrepancy will be given in the next
section.
(4-47)
where k is in lbf-sn/100 ft2.
For Bingham plastic fluids, and assuming the fluid is Table 4-5. Rheological Model Parameters†
fully sheared in the annular gap, Eq. 4-31 can be rewrit- n or τy k or µp
ten as
Bingham plastic 3.47 Pa 91.3 mPa-s

fτ × fspring × θ = μ p × fγ × Ω + τ y
( ) ( ).
2
2 × Rr ln Rr Power law 0.5859 1.036 Pa-sn

( Rr )2 − 1
† Determined according to Eqs. 4-43 and 4-48 from the data presented in Table 4-2.

(4-48)

Therefore, the plastic viscosity and the yield stress of 4-3.4 Newtonian and narrow-gap approximations
the fluid can be determined from a linear regression. If Earlier it was stressed that the formulas giving the shear
m is the slope of this regression and b its intercept, the rate at the inner cylinder surface (Eqs. 4-18 and 4-26)
Bingham plastic parameters of the fluid are given by are valid only for a Newtonian fluid. Therefore, the rec-
ommended API/ISO procedure, which consists of con-
fτ × fspring
μp = × m (4-49) verting rotational speeds to Newtonian shear rates at the
fγ inner cylindrical surface, is not completely correct. It
leads to an overestimation of the consistency index for
and
( Rr ) − 1
2

τ y = fτ × fspring × × b. (4-50)
2( Rr ) ln ( Rr )
2

Chapter 4 Rheology and Flow of Well Cement Slurries 107

power-law fluids and of the yield stress for Bingham plas- For Bingham plastic fluids,
tic fluids. The expressions for the rheological parameter
ratios are given by Eqs. 4-53 and 4-54, respectively. (μ p ) ng
=
(
16 Rr )2

(4-58)
⎧ –2 ⎫
( )
n μp
( Rr + 1)4
kNW ⎪ 1 − Rr ⎪
=⎨ ⎡ −2 ⎤ ⎬ (4-53) and
k ⎪ n ⎢1 − R n ⎥ ⎪ ( )
⎩ ⎣ r ⎦⎭ (τ y ) ng
=
( )2 × ln ( Rr ) .
8 × Rr
(4-59)
(τ y ) NW ( )
2 ln Rr
τy
( Rr − 1)( Rr + 1)3
= (4-54)
τy 1 − ( Rr )
–2
With the standard oilfield geometry, this error leads
to an overestimation (Eqs. 4-58 and 4-59) of 0.2% for
Using Eqs. 4-53 and 4-54, the corresponding errors for the plastic viscosity and an underestimation of 0.8% for
the standard geometry used in the oil industry the yield stress. For power-law fluids, the errors are of
(Rr = 1.068) range from 0.0% to 5.4% for the consistency the same order of magnitude (i.e., negligible).
index of power-law fluids when the power-law index It can be shown that we can use torques and rota-
varies from 0 to 1. For Bingham plastic fluids, the error tional speeds for other rheological models that are used
is 0 for the plastic viscosity and 6.7% for the yield stress. to describe the behavior of cement slurries (e.g., Casson,
One may consider these errors to be negligible for Robertson and Stiff, Herschel-Bulkley). With the narrow-
practical purposes when using the standard oilfield gap approximation, the shear-rate and shear-stress
geometry. However, when a larger annular gap is used— values could then be calculated from the following
for example when testing cement slurries containing a rather than Eqs. 4-18 to 4-20. Ω is rotational velocity,
significant volume of particles with a diameter larger and θ is the viscometer reading.
than 120 μm—these errors can become significant. In
such cases, one can use the true flow equations (Eqs. γ ng = 15.26 × Ω (4-60)
4-28, 4-31, and 4-34) or the narrow-gap approximation where Ω is in rad/s, or
described below (Mannheimer, 1982).
When the radial ratio of the cylinders is close to 1, the γ ng = 1.598 × Ω
shear stress and the shear rate can be considered to be
uniform in the annular gap and given by where Ω is in rpm.

2χ τ ng = 0.4780 × θ (4-61)
τ ng = (4-55)
( )
2
π r2 + r1 where τng is in Pa and
τ ng = 0.9984 × θ (4-62)
and

γ ng =
( r + r )Ω .
2 1
where τng is in lbm/100 ft2. Eqs. 4-61 and 4-62 are writ-
ten to correspond with standard spring F1.
2( r − r )
(4-56)
2 1
4-3.5 What about the rheological model?
Therefore, the values for the shear stress and the It is now well recognized throughout the industry that
shear rate can be derived directly from the torques and both the power-law model and the Bingham plastic
the rotational speeds. The errors resulting from using model suffer from serious limitations when describing
this approximation can easily be determined. the rheology of drilling fluids, spacers, and cement slur-
For power-law fluids, ries over a wide shear-rate range. The most commonly
n used alternative to these two parameter models is the
⎡ 2 ⎤
kng
=
4 ⎢ 4 Rr − 1
× ×
Rr n ⎥ ( ) Herschel-Bulkley model. It does not suffer from the
same limitations; however, at this writing the Herschel-
2⎢ ⎥ . (4-57)
k
( ⎣
)
Rr + 1 ⎢ n Rr + 1 R n − 1 ⎥
r
2

⎦ ( ) Bulkley model is not officially recognized in the API/ISO

standards.

108 Well Cementing

 When the narrow gap approximation applies, deter- 1.8
Measured
mining the Herschel-Bulkley model parameters becomes 1.6 Bingham plastic
easier. After converting the rotational speeds and torque 1.4
Power law
readings to shear rates and shear stresses using Herschel-Bulkley
1.2
Eqs. 4-18 to 4-20, or preferably Eqs. 4-60 to 4-62, an esti- Log
mated fluid yield stress, (τy)est, is used. A good starting [shear rate] 1
point is 80% of the measured shear stress at 3 rpm. The (Pa) 0.8
shear-stress reading at 3 rpm is the lowest that is mea- 0.6
sured; therefore, it is closest to the fluid yield stress. The
0.4
corresponding values of the power-law index, n, and 0.5 1 1.5 2 2.5 3
consistency index, k, are determined by performing a
Log [shear rate] (s–1)
linear regression analysis on the logarithm of the shear
rates and the logarithm of the difference between the Fig. 4-21. Rheological plot illustrating improvement of Herschel-
measured shear stresses and the assumed yield stress: Bulkley model compared to the Bingham plastic and power-law
models. Measured shear stresses and shear rates were deter-

{ ( ) ( ) ( )
log γ ng i ;log ⎡ τ ng i − τ y
⎣⎢ est ⎥ }
⎤ .
⎦
mined from Eqs. 4-18 and 4-19.

Then the sum of the squares of the deviations is deter- Table 4-6. Relative Shear-Stress Errors Between Herschel-
mined. Bulkley Model and the Measured Data†
Velocity (rpm) Herschel-Bulkley Error (%)
( ) ( )
2
Δ = ∑ ⎡ τ ng idata − τ ng
⎤ (4-63)
⎣⎢ icalc ⎥
⎦ 300 0.2

200 –0.7
The procedure then consists of choosing the yield-
stress value that minimizes Δ, which is usually easy to do 100 0.9
(Klotz, 1998). For the example given in Table 4-2, the
60 0.9
rheological parameters calculated according to Eqs. 4-18
to 4-20 are 30 –1.7
τy = 2.65 Pa 6 –1.4
n = 0.902
3 2.0
k = 0.169 Pa-sn. Calculated shear stresss
† Relative error = 1 −
The same parameters calculated according to Eqs. Measured shear stress

4–60 to 4-62 are

τy = 2.48 Pa
n = 0.902 annular gap where torque is measured is not homoge-
neous. The flow pattern is significantly modified close to
k = 0.168 Pa-sn.
the top and the bottom of the gap. In addition, the fluid
The improved performance of the Herschel-Bulkley that is sheared above and below the inner cylinder con-
model versus the power law and Bingham plastic models tributes to the measured torque. Such end effects are
at low shear can be clearly seen in Fig. 4-21. Comparing often assumed to be proportional to the undisturbed
the relative errors between the calculated and measured stress, and an extra cylinder length or a torque correc-
shear stresses (presented in Table 4-6) and those pre- tion factor allows them to be taken into account. This
sented in Table 4-4 confirms the improved performance factor is usually measured for Newtonian fluids and
of the Herschel-Bulkley model over the entire shear-rate applied to all fluids without regard to which rheological
range. model is most appropriate. A more reliable procedure
consists of performing the measurements with different
4-3.6 End effects levels of fluid in the gap. For each rotational speed, the
measured torque is a linear function of the fluid height
In the equations developed in Section 4-3.1, the torque in the gap, and the slope is the torque per unit length.
per unit length of any cylindrical surface with radius r is With standard oilfield equipment, the end correction
assumed to be known. However, because coaxial cylinder factor recommended by manufacturers is 1.064. That
viscometers have a finite length, the shear flow in the factor is in fact hidden in the spring calibration con-

Chapter 4 Rheology and Flow of Well Cement Slurries 109

stant,‡‡ which is 1.064 times lower than the nominal con- a quantity which is much more difficult to determine.
stant. This value is in agreement with measurements In the absence of quantitative information, rheological
performed on Newtonian oils by Mannheimer (1988) and measurements should be performed with different gap
by the author. However, in 1987 Denis and Guillot found sizes. If the experimental data are dependent upon the
that end effects can account for up to 16% of the mea- gap size, the homogeneity of the fluid is questionable.§§
sured torque when testing cement slurries (Fig. 4-22), Several authors (Tattersall, 1973; Mannheimer, 1983
indicating that the current standard procedure can over- and 1988; Lapasin et al., 1983; Denis and Guillot, 1987;
estimate shear stresses by up to 10%. Haimoni, 1987) have shown that cement slurries are
Unfortunately, there is no clear understanding today indeed not homogeneous, particularly at low shear rates
of how end effects vary with the non-Newtonian behavior (Fig. 4-23). The correct interpretation of this effect is a
of fluids; therefore, no simple systematic procedure can very involved procedure.
be proposed. Nevertheless, when trying to compare
results obtained with different instruments, one must be
aware that end effects can account for differences in 102
measured shear stresses.
Flow is driven Flow is shear
by slip at driven.
180 the wall.
Neat Class G: Linear fit
160 Newtonian oil: Linear fit Shear
stress 101
140 (Pa)
120
ΔL = 0.60 cm
100 Gap
Torque
reading 80 ΔL = 0.24 cm 2.25 mm
Standard 100 0.75 mm
60
length 100 101 102 103
40 Newtonian shear rate (s ) –1

20
Fig. 4-23. Flow curves from a neat Class G cement slurry in a coax-
0 ial cylinder viscometer with two different annular gaps (after Denis
–1 0 1 2 3 4 5 et al., 1987). The curves show the dependence of shear rate and
Annular length (cm) shear stress on the annular gap. Reprinted with permission of SPE.

Fig. 4-22. Graphical determination of end effects with a modified

coaxial cylinder viscometer (ΔL is the length that would be added to Mannheimer (1983; 1988) and others have attempted
the inner-cylinder length L to account for end effects). to analyze this phenomenon in terms of a slip velocity,
(i.e., the velocity of the test fluid at the wall is assumed
to be nonzero). Such an assumption implies that
4-3.7 Homogeneity of the cement slurry Eq. 4-22 is no longer valid and should be replaced by
and wall slip
The flow equations in Section 4-3.1 also assume the fluid 1 τ1 γ ( )
vslip r1 vslip r 2 ( )
to be homogeneous in the annular gap. Because cement Ω= ∫
2 τ2 τ
dτ +
r1
+
r2
. (4-64)
slurries are concentrated suspensions, they can only be
considered homogeneous if the annular-gap size is at
Assuming that the slip velocity depends only on the
least 10 times that of the largest particles. In view of the
shear stress at the wall for a constant shear stress at the
particle-size distribution of cement powder, the gap size
outer cylinder surface (Mooney, 1931),
should be approximately 1 mm, which is the case with
the standard oilfield geometry. Strictly speaking, what
should be considered is the size of particle aggregates,
§§ When testing a suspension exhibiting a yield stress, like a spacer or a
‡‡ Becausethe end correction factor is embedded in the spring calibration cement slurry in a coaxial cylinder viscometer, as the imposed shear rate
constant, the same factor is applied regardless of the coaxial cylinder geom- decreases, part of the fluid becomes unsheared and the size of the sheared
etry used, which is obviously not correct. So when changing the rotor and layer becomes smaller and smaller (Section 4–3.1). Therefore there is
bob combination on the standard oilfield viscometer, it is recommended to always a critical shear rate below which the suspension can no longer be
experimentally determine the associated end correction factor. considered a continuous medium.

110 Well Cementing

 lim Ω ( when r 1→ r2 ) =
( ).
2 vslip τ2
(4-65)
Using data affected by slippage at the wall, if the
influence is not detected, can lead to erroneous conclu-
r2 sions concerning the behavior of the test fluid at low
shear rates. For example, if one fits the data measured
Therefore, the effect of wall slip could be accounted with the cement formulation presented in Table 4-7 to a
for by performing experiments with different inner- power-law or Herschel-Bulkley model, relatively good
cylinder radii. This analysis, which has been simplified results are obtained over the entire shear-rate range.
by Mannheimer (1982) for narrow annular gaps and by This is shown graphically in Fig. 4-25. One could con-
Yoshimura and Prud’homme (1988), has not been con- clude that the fluid exhibits a very low yield stress or
clusively validated for cement slurries. As can be seen in even no measurable yield stress. However, rerunning the
Fig. 4-24, the percentage of the flow caused by slip does test with a wider gap would show that the data measured
not vary consistently with shear stress. at 5 and 10 sec–1 (3 and 6 rpm) with the standard annu-
lar gap are affected by slippage at the wall and, there-
fore, should not be used to characterize the rheological
120 properties of the fluid.
100
0.3% bentonite
80 Table 4-7. Viscometer Readings for a Cement Slurry†
4 different
slurries Rotational Speed (rpm) Average Dial Readings
Slip 60
(%) 300 68
40
200 58
20 100 47
No slip for average
0 shear stress > 50 lbf/100 ft2 60 41
0 25 50 75 100 125 150 175 200 30 36
Average shear stress (lbf/100 ft ) 2

6 19
Fig. 4-24. Effect of shear stress on percent slip measured with a
concentric cylinder viscometer (slurry contains 38% water BWOC) 3 12
(after Mannheimer, 1988). † Measurement was performed with the standard R1B1 geometry
and standard spring F1.

In a first series of tests, Mannheimer (1982) found

40
the effect of slip velocity to be negligible above a given
shear stress. Later, using different cements, conflicting 35
results were obtained. The coaxial cylinder viscometer 30
data, corrected for wall slip, did not agree with laminar 25
friction-pressure data in large-diameter pipes Shear
stress 20
(Mannheimer, 1988). (Pa) 15
Another approach to minimize wall slip is to use
grooved cylindrical surfaces. However, the reliability of 10 Measured
Bingham plastic
the procedure with cement slurries is questionable, 5 Power law
because the measured shear stresses depend on the 0 Herschel-Bulkley
depth of the serrations (Haimoni, 1987). 0 100 200 300 400 500 600
Haimoni (1987), tried to combine these two Shear rate (s–1)
approaches (i.e., varying the gap size and the surface
roughness of the cylinders) while making measurements Fig. 4-25. Graph of measured data displayed in Table 4-7. Data are
fitted to the Herschel-Bulkley, Bingham plastic, and power-law
on the same material. Although he was not able to pro- models.
pose a method to account for apparent slippage at the
wall, he concluded that, once plug flow is eliminated,
this phenomenon seems to have negligible conse-
quences upon measurements performed in a coaxial
cylinder viscometer.

Chapter 4 Rheology and Flow of Well Cement Slurries 111

 Thus, without a proven method for preventing or Using a transparent rotor and a transparent beaker,
accounting for wall slip, coaxial-cylinder-viscometer and adding small black tracer particles to a Newtonian
data that are affected by this phenomenon should not be fluid or polymer solution, Meeten and Pafitis (1993)
used when trying to determine rheological parameters. were able to show that the tracer particles migrated in
It is the author’s experience that these data points can the annular gap of a coaxial-cylinder viscometer because
often be detected on a log-log plot of the torque and rota- of gravitational and/or centrifugal forces. Meeten and
tional speed, which usually shows a drastic change in Pafitis also noticed that, when sheared at relatively high
curvature (e.g., between 6 and 30 rpm for data presented rates, suspensions such as drilling fluids, spacers, or
in Table 4-7 (Fig. 4-26). cement slurries sometimes required several minutes to
Frequently, experimental data falling below this reach a steady state: An unusually long shearing period
breaking point are affected by wall slip. One can check was required before the viscometer dial reading was
this assumption by rerunning the test with a different constant. Meeten and Pafitis attributed this time-depen-
gap size. The consequences of using data that are dent behavior to particle migration.
affected by wall slip can be quite important. In the par- Standard oilfield equipment is designed so that these
ticular example of the data presented in Table 4-7, the effects are not severe. But when using slurries that have
calculated yield stress of the cement slurry using the a low viscosity at high shear rates, or when shearing
Herschel-Bulkley model varies from less than 1 Pa when slurries for a long period of time (e.g., during heat-up
the 3- and 6-rpm data are used to nearly 13 Pa when they periods in a high-pressure, high-temperature [HPHT]
are not. rheometer), this problem may become significant.
Different solutions have been proposed.
Meeten and Pafitis (1993) suggested using a pump to
1.6
ensure better fluid homogeneity not only in the cup but
1.5
also in the annular gap (Fig. 4-27).
1.4
1.3
Log
[shear 1.2
Torque head
stress] 1.1
(Pa) 1 Rotor boss
Measured
0.9 Bingham plastic
0.8 Power law Sample cup
0.7 Herschel-Bulkley
0.5 1 1.5 2 2.5 3 Rotor
Log [shear rate] (s–1)
Stator
Fig. 4-26. Rheological plot corresponding to measured data dis-
played in Table 4-7. The power-law model is fitted to the entire
dataset, while the Bingham plastic and the Herschel-Bulkley
models are fitted to the five highest readings. Pump

4-3.8 Particle migration

Fig. 4-27. Rotational viscometer equipped with pump to ensure
Particle migration caused by gravitational or centrifugal better fluid homogeneity (from Meeten and Pafitis, 1993). Reprinted
forces may also affect the rheological measurements. with permission from Elsevier.
For the results to be meaningful, the test fluid should
not segregate during the measurement. Before measur-
ing the rheological properties of a cement slurry, it is
Ravi and Sutton (1990) included an independent
essential to ensure that particle segregation does not
impeller at the bottom of the cup of a HPHT viscometer
occur under static conditions (leading to free water and
to better homogenize the slurry in the gap. The impeller
sedimentation). Unfortunately, this does not necessarily
rotates during the heat-up period and stops while rheo-
mean that particle segregation will not occur under
logical measurements are being taken.
dynamic conditions, because
An interesting approach for limiting particle migra-
■ the viscosity of the fluid usually decreases with shear
tion effects, as well as the wall slip effect, was investi-
■ under dynamic conditions, the centrifugal forces can gated by Vlachou and Piau (2000). The geometry of the
be greater than the gravitational forces. rotational viscometer (Fig. 4-28) was modified to incor-

112 Well Cementing

porate a series of grid tools combined in a multiple par- than 3 to 4 mm with the standard oilfield viscometer
allel plate configuration. The geometry seems particu- does not work well for several reasons. As the gap size
larly interesting for cement slurries exhibiting some set- increases, the following effects occur.
tling under shear. The grids are placed in the bulk of the ■ The shear-rate range of the test is shifted towards
sample, and the particles can easily settle through the lower values; consequently, the measured torques
grids. This design minimizes the effect of particle set- become smaller and less accurate.
tling on the rheological measurement. But this geometry ■ The shear rate and shear stress become less homoge-
has not yet gained industry recognition.
neous across the gap, and the likelihood of obtaining
plug flow at low velocities increases. This requires
some precautions when analyzing and fitting the data
Ω
to a rheological model.
Rotating ■ Particle migration becomes more and more influential.
part Sample-containing cup
■ Instabilities like Taylor vortices (deviations from lam-
Rotating plates inar flow) may be present at high rotational speeds
External axe when the inner cylinder rotates (e.g., with some
(one out of three) HPHT viscometers).
Fixed plates
a Fixed part
4-3.10 Foamed cement
Cement slurries can be foamed with air or nitrogen, as
explained in Chapter 7. Foamed cement is usually charac-
terized by its quality, defined as the volume of gas per
unit volume of slurry. The quality of a foamed cement
during placement in the wellbore rarely exceeds 70%.
b
This is the lower limit for a fluid to be technically consid-
ered to be foam. Nevertheless, in the well cementing
industry, the term “foam” is used regardless of the quality.
Very little is known about foamed cement rheology
other than that the apparent viscosity of foamed cement
increases with quality. Al-Mashat (1976) used the
Bingham plastic model to describe pipe flow of nitrogen-
foamed cement up to Newtonian shear rates of 5,000
sec–-1. He found that the effect of foam quality, Qfoam, on
c the rheological parameters could be described by the
equations below.

Fig. 4-28. Schematic diagram of a new rotational geometry (from

Vlachou and Piau, 2000).
(
μ p = 13.00021 − 71.336573 × Q foam )
⎡
( ) ( )
2⎤ ⎡ 3⎤
+ ⎢367.12431 × Q foam ⎥ − ⎢257.9818 × Q foam ⎥
⎣ ⎦ ⎣ ⎦
4-3.9 Testing slurries containing large particles (4-66a)
When a cement slurry contains a significant volume frac- for 0 ≤ Qfoam ≤ 0.65
tion of particles larger than 120 μm, the size of the stan-
dard annular gap must be increased for the slurry to
be considered homogeneous. If this is not done, the
(
τ y = 16.2297 − 83.9824 × Q foam )
⎡
( )
2⎤
measurements are likely to underestimate the true fluid + ⎢146.36276 × Q foam ⎥ (4-66b)
viscosity. ⎣ ⎦
One way to determine if this is indeed a problem is to
perform the same rheological measurements with differ- for 0.35 ≤ Qfoam ≤ 0.65
ent gap sizes. If the measured shear stresses are gap-size where the plastic viscosities are in mPa-s and the yield
dependent, the gap should be widened until this is no stresses are in Pa.
longer the case. In practice, using a gap much larger

Chapter 4 Rheology and Flow of Well Cement Slurries 113

 Unfortunately, the rheology of the base slurry itself is 4-3.11 Cement rheology measured with different
not reported; therefore, these equations are of little equipment
practical interest other than for giving an idea of how When attempting to characterize the rheological behav-
much the apparent viscosity of the foam varies as a func- ior of fluids as complex as cement slurries, it is essential
tion of foam quality. If the plastic viscosity equation is to ensure that the measurements are not equipment
valid at 0 foam quality, it shows that plastic viscosity dependent. Thus, several authors have compared rheo-
varies by a factor of 3.6 when the foam quality increases logical measurements performed with different types of
to 60% from 0%. equipment, usually a coaxial cylinder viscometer and a
Because of the lack of data on foamed cement, place- pipe viscometer. For such a comparison to be significant,
ment simulators (Purvis and Smith, 1994; Garcia et al., it must be performed within a shear-rate range common
1993) capable of handling foamed cements or spacers to both apparatuses.
usually rely on empirical correlations that were devel- Denis and Guillot (1987) showed that, with some
oped by Reidenbach et al. (1986) with linear (i.e., cement-slurry formulations, reasonable agreement can
uncrosslinked) polymer-base fracturing fluids. These be obtained between a pipe viscometer and a specific
authors assumed that the rheological behavior of foamed coaxial cylinder viscometer, provided the rheological
fracturing fluids can be described by the Herschel- data are not affected by wall slip (Fig. 4-29).
Bulkley model. The effects of foam quality and tempera- However, when cement slurries are characterized
ture on the model rheological parameters are then with the standard oilfield viscometer, the results often
described as follows. differ from those obtained with pipe viscometers, even
τ y = 3.35 × Q foam for Q foam ≤ 0.6 when using large-diameter pipes to minimize the effects
of apparent wall slip (Bannister, 1980; Mannheimer,
τ y = 0.00958 × exp(9 × Q foam ) for Q foam > 0.6 1983; Denis and Guillot, 1987).

(⎡
) ( )
2⎤
k = kbf × exp ⎢ C1 × Q foam + 0.75 Q foam ⎥ (4-67)
⎣ ⎦ 102
Specific cut coaxial gap 2.25 mm
Specific cut coaxial gap 0.75 mm
The consistency indices are expressed in Pa-sn and Pipe, r = 10 mm
the yield stresses in Pa. Pipe, r = 16 mm
The power-law index is assumed to be independent of Pipe, r = 20 mm
foam quality. In these equations, kbf is the consistency
Shear
index of the base fracturing fluid used to prepare the stress
foam; C1 is a constant that depends on the power-law (Pa)
index of the base fluid. Harris and Reidenbach (1987)
later refined this model to account for the effects of
101
foam quality and temperature on the rheological para-
meters. Such equations cannot be used directly for
cement slurries, but they provide an order of magnitude
of the effect of foam quality on the apparent viscosity.
For example, the constants kbf, C1, and n are such that 102 103
the apparent viscosities of foamed fracturing fluids Shear rate (s )–1

(linear gels) in a 100 to 1,000 sec–1 range increase by a Fig. 4-29. Pipe- and coaxial-flow results for a neat Class G cement
factor between 2.8 and 7.8 when the foam quality slurry. Shear rates are corrected for non-Newtonian effects. Above
increases to 60% from 0%. 200 sec–1, there is good agreement between the datasets. Reprinted
This range of relative viscosities is consistent with the with permission of SPE.
values calculated by Al-Mashat (1976) on foamed
cement. It is also clear from Harris and Reidenbach
(1987) that base fluids with higher polymer concentra- In an attempt to solve this problem, Shah and Sutton
tions are more shear thinning. As a result, the relative (1989) tried to obtain a statistical correlation between
viscosity of the foam does not increase as quickly with the measurements performed with a standard oilfield
quality. From these data, it is difficult to determine how viscometer and a pipe viscometer. They used a modified
the yield stress of a foamed cement slurry is affected by coaxial-cylinder viscometer that allowed vertical circula-
foam quality. Unlike cement slurries, linear fracturing- tion of the slurry in the annular gap. Circulation was
fluid gels do not exhibit a true yield stress. stopped while taking a measurement at a given rota-

114 Well Cementing

tional speed. For a wide variety of cement-slurry formu- the pipe viscometer were between 0% and 27% greater
lations, they compared the rheological parameters by fit- than those obtained with the coaxial-cylinder viscome-
ting experimental data from their modified viscometer ter. This empirical procedure is quite useful, but it suf-
fers from one limitation: the cement slurries are
⎣( )( )
⎡ μ p c, τy c ⎤
⎦
assumed to follow the Bingham plastic model, which is
not necessarily the case.
and a pipe-flow loop

( )p ,( τ y )p ⎤⎥⎦
⎡ μp
⎢⎣
100

80
to the Bingham plastic model. They found the following
correlation for the plastic viscosities when expressed in Viscometer 60
centipoise (Fig. 4-30). yield stress
(lbf/100 ft2) 40

( μ p ) p = 0.962 × ⎡⎣( μ p )c ⎤⎦
0.9815
, (4-68) 20

indicating that the plastic viscosities obtained with the 0

pipe viscometer were about 10% lower than those 0 20 40 60 80 100 120 140
obtained with the coaxial-cylinder viscometer. Pipe yield stress (lbf/100 ft )
2

Fig. 4-31. Uncorrected yield-stress relationship between standard

coaxial-cylinder and pipe viscometers (after Shah and Sutton,
103 1989). Reprinted with permission of SPE.

102
Viscometer
plastic
4-4 Pressure, temperature, and time
viscosity dependency
(cp) 101 The pressure and temperature dependence of the rheo-
logical properties of cement slurries is not always prop-
erly investigated, because the standard oilfield equip-
100 ment allows measurements to be performed only at
100 101 102 103 atmospheric pressure and at temperatures below about
Pipe plastic viscosity (cp) 185°F [85°C]. Limited studies at higher temperatures
suggest that cement-slurry stability can be problematic.
Fig. 4-30. Plastic-viscosity relationship between standard coaxial-
cylinder and pipe viscometers (after Shah and Sutton, 1989).
Reprinted with permission of SPE.
4-4.1 Temperature dependency
Temperature can substantially affect the rheological
properties of cement slurries, but the extent is highly
The pipe- and coaxial-cylinder viscometers overesti- dependent on the cement slurry rheology and the addi-
mated the yield stresses by factors of 1.333 and 1.067, tives in the formulation. The differences in temperature
respectively (Fig. 4-31). In both cases, the shear rate at dependence are shown in Figs. 4-32 and 4-33.
the wall was assumed to be Newtonian, which is not the The first formulation contains a water-soluble poly-
case for a Bingham plastic fluid. Therefore, once the mer (hydroxyethylcellulose) that viscosifies the intersti-
yield stresses are corrected, the correlation of Shah and tial water and contributes significantly to the slurry vis-
Sutton (1989) becomes cosity. Because the solution viscosity of the polymer
( τ y ) p = 1.273 × ( τ y )c − 1.611 (4-69)
itself is temperature sensitive, the plastic viscosity of the
slurry follows the same continuous downward trend,
while the yield stress remains almost constant.
where the yield stresses are expressed in lbf/100 ft2. This The behavior of the second system (containing a dis-
equation indicates that the yield stresses obtained with persant and latex) is more complicated. The plastic vis-

Chapter 4 Rheology and Flow of Well Cement Slurries 115

 70 specific additives, a general data trend was observed by
300 Plastic viscosity Ravi and Sutton (1990). Using the Bingham plastic
Yield stress
60 model to fit the rheological data, they were able to show
250 that the variation in the rheological parameters with
50 temperature could be described by the following equa-
tions, in which the plastic viscosity and yield value were
200
40
measured at 80°F [27°C].
Plastic Yield
viscosity
(cp)
150
30
stress
(Pa) ) ) (
μ p ( T = a + ( b × T + 0.00325 × T 2 , ) (4-70)

100 where μp is in mPa-s and T is in °F, and

20
a = 65.0729 + 1.3054 × μp at 80°F
50 10 b = 1.0734 – 0.00381 × μp at 80°F.

0 0 ) ( )
τ y = a′ + ( b′ × T + 0.002 × T 2 . (4-71)
10 20 30 40 50 60 70 80 90
where τy is in lbf/100 ft2 and T is in °F, and
Temperature (°C)
a′ = 36.651 + 1.4047 × τy at 80°F
Fig. 4-32. Temperature dependence of the Bingham plastic parame-
ters of a cement slurry containing a cellulose derivative.
b′ = −0.61813 − 0.00505 × τy at 80°F.
Each of these equations are limited to a maximum
temperature, Tmax. Above this limit, the plastic viscosity
30 70 or the yield stress is considered to be constant (other-
Plastic viscosity wise these factors would increase with temperature).
Yield stress For spacers, the temperature dependency is usually
25 60
somewhat simpler because spacers are usually com-
posed of water, salts, water-soluble polymers, clays, and
50 weighting agents. Figure 4-34 is a typical example show-
20
ing how the viscosity of a water-base spacer varies with
40 temperature (Théron et al., 2002).
Plastic Yield
viscosity 15 stress In this particular case, the data were fitted to the
(cp) 30 (Pa) Herschel-Bulkley model; however, the parameters of this
10 model do not show a simple trend with temperature.
20 Therefore, interpolating data at various temperatures
from this type of measurement is usually performed in a
5 different way. For each shear rate used to build the dif-
10
ferent rheograms, a virtual shear stress is determined at
0 0 a given temperature using an interpolation of the mea-
10 20 30 40 50 60 70 80 90 sured shear stresses as a function of temperature. These
Temperature (°C) virtual shear stresses are then fitted to the Herschel-
Bulkley model. A more sophisticated approach, using a
Fig. 4-33. Temperature dependence of the Bingham plastic parame- neural network model, has been used by Théron et al.
ters of a cement slurry containing a dispersant and a latex. (2002) to describe how spacer viscosity is affected by
temperature, chemical composition, and density.
In the absence of a specific model to describe how the
cosity of the slurry first decreases by a factor of 2 rheological properties of a given spacer or cement slurry
between 77 and 113°F [25 and 45°C], then increases are affected by temperature, one should perform the
more slowly from 113 to 185°F [45 to 85°C]. Meanwhile, measurements at three different temperatures at least,
the yield stress increases slowly but continuously covering the range the fluid will encounter during the
throughout the temperature range investigated. operation. As explained above, for a given shear rate,
These two examples illustrate that there is currently shear stresses can then be interpolated as a function of
little hope of finding a general model to describe the temperature and a shear-stress/shear-rate diagram can
temperature dependence of the cement-slurry rheology. be rebuilt for each temperature before fitting the data to
Limiting the investigation to cement systems containing a rheological model.

116 Well Cementing

 35 The viscosity of the second slurry increased by 25% as
the pressure increased to 11,800 psi [81.1 MPa]. With
30
another cement system, Cartalos et al. (1994) did not
25 see a detectable effect of pressure from 100 to 11,600 psi
[0.7 to 80 MPa] at 248°F [120°C]. These results seem
Fann 20
inconsistent because the cement systems have nearly
deviation the same density (15.8 lbm/gal [1,900 kg/m3]), and they
(°) 15
have approximately the same solid volume fraction. The
10 only differences were the choice of retarder, choice of
Temperatures (°C)
5 25 100 fluid-loss additive, and the type of instrument used to
50 125 measure the rheological properties. Further investiga-
0 85 150
tion is necessary to quantify the effect of pressure on the
0 50 100 150 200 250 300 rheological behavior of cement systems.
Fann speed (rpm)

Fig. 4-34. Raw rheological measurements (markers) fitted to the 4-4.3 Time dependency under shear
Herschel-Bulkley model (solid lines) at different temperatures for a Spacers and cement slurries sheared in a coaxial cylin-
spacer fluid with ρ = 1,680 kg/m3. Reprinted with permission of SPE.
der viscometer may exhibit time-dependent behavior. As
discussed earlier, this apparent time dependency can be
an artifact caused by particle migration. However, the
4-4.2 Pressure dependency material may be exhibiting a truly time-dependent
behavior. This is more common for cement slurries than
The effect of pressure on the rheological properties of for spacers. There are two main reasons for the time
spacers or cement slurries is often ignored because, dependency of cement slurries. First, these slurries con-
under dynamic conditions, this effect seems to be small tain colloidal charged particles that may cause a
when compared to the effect of temperature. Pressure thixotropic (reversible) behavior. Second, they are
effects can be accounted for by an apparent increase in chemically reactive suspensions that are irreversibly
the solid volume fraction (SVF) owing to the higher com- affected by time.
pressibility of the liquid phase versus the solid phase. The API/ISO procedure is not very well adapted to
For a Class G cement slurry mixed at 15.8 lbm/gal [1,893 detect thixotropic behavior, but, when the time scales
kg/m3], assuming the densities of the mix water and the associated with its structural buildup and breakdown
cement powder are 8.33 lbm/gal [1,000 kg/m3] and 26.6 are not too short, an open hysteresis cycle may indicate
lbm/gal [3,200 kg/m3], the SVF is 40.6%. Assuming the true time-dependent behavior. Usually, when cement
isothermal compressibility of the liquid phase is 4 × slurries that are not designed to be thixotropic are
10–10 Pa–1, and neglecting that of the solid phase, the sheared for periods longer than 20 sec, and particle
SVF increases to only 40.7%, and the slurry density migration is avoided, the stress response depends on the
increases to 15.84 lbm/gal [1,900 kg/m3] at 10 MPa. At amount of applied shear. At relatively high shear rates,
100 MPa, the SVF is 41.6% and the slurry density is 16.18 the measured shear rate stabilizes relatively quickly
lbm/gal [1,939 kg/m3]. within less than 20 sec (Fig. 4-35). This behavior sup-
During a cementing operation, the slurry-density con- ports the time step adopted as an industry standard pro-
trol is ±0.2 lbm/gal [0.024 kg/m3] at best. Therefore, cedure. At lower shear rates, shear stress tends to
unless the slurry contains compressible or pressure-sen- increase with time towards an asymptotic value, indicat-
sitive additives (e.g., ceramic microspheres) the effect of ing that the suspension undergoes a structural buildup.
pressure should be negligible. It is possible to show that this behavior is usually
Laboratory measurements show ambiguous results, reversible within several minutes. It can therefore be
even when they are not performed close to the end of the qualified as thixotropic. When the applied shear rate is
thickening time. Pressure affects the thickening time extremely low, a steady flow does not necessarily occur
and compressive strength development. Kellingray et al. (Vlachou, 1996). The measured stress may then continu-
(1990) determined the apparent viscosities of two spe- ously increase over long time periods, as will be dis-
cific cement slurries measured at 511 sec–1 and about cussed below.†††
230°F [110°C]. The viscosity of the first slurry increased
by 8% as the pressure increased to 5,870 psi [40.5 MPa].
††† Other
phenomana such as wall slip and fracturing of the material can be
observed (Vlachou, 1996).

Chapter 4 Rheology and Flow of Well Cement Slurries 117

 102 28
24 Dispersant A
Dispersant B
100 s–1 20 Dispersant C
Plastic 16
viscosity 12
Shear (cp)
101 10 s–1
stress 8
(Pa)
4
1 s–1 0
0 20 40 60 80 100 120
Time (min)
100
0 10 20 30 40 50 60
45
Time (s) Dispersant A
40 Dispersant B
Fig. 4-35. Shear stress versus time at three different shear rates. Dispersant C
Steady flow is established within a few seconds. However, at 35
Yield
1 sec–1, the shear stress increases over a longer time period (from
stress
Vlachou, 1996). Reprinted with permission from Elsevier. 30
(lbf/100 ft2)
25
When observing the behavior of cement slurries over 20
longer time scales—a significant fraction of the thicken- 0 20 40 60 80 100 120
ing time—cement slurries may exhibit an irreversible
Time (min)
behavior caused by the ongoing chemical reactions that
modify the characteristics of the suspension. For exam- Fig. 4-36. Rheological data from 15.8-lbm/gal [1,900-kg/m3] cement
ple, the rheological behavior of a cement slurry at a spe- slurries containing various dispersants as a function of shearing
cific temperature (Fig. 4-36) can vary significantly as a time in an atmospheric consistometer. Conditioning and measure-
function of the conditioning time (i.e., how long the ments were performed at 77°F [25°C]. Dispersant A is polynaphtha-
lene sulfonate, B is polymelamine sulfonate, C is an acrylate.
slurry is sheared at temperature before performing the
rheological measurement).
This time dependency raises questions concerning
the procedures used to measure cement-slurry proper- main reasons, the results obtained with this experimen-
ties, because the measurements are merely a “snapshot” tal method are questionable.
of the behavior as it evolves with time. When performing ■ It has already been mentioned that the low-shear
simulations of the placement process, and when rheol- behavior of cement slurries is very often affected by
ogy has a significant impact on the results, it is impor- wall slip. This is especially true for thixotropic sys-
tant to ensure that the rheology data are representative tems, because most experimental results show that
of the operation. The time and temperature history may the higher the yield stress of the fluid, the larger the
need to be considered for the simulations to be meaning- shear-rate range affected by wall slip.
ful. ■ The results obtained may vary from one piece of
equipment to another, depending on the inertia of the
4-4.4 Gel strength measurements fixture and on the spring stiffness of the measuring
device.
The rheological behavior of cement slurries at rest is one
of the factors that govern formation-fluid migration into With standard oilfield equipment, little can be done
the annulus after a primary cementing operation to address the second point. Even in the absence of wall
(Chapter 9). Therefore, the industry has developed sev- slip (e.g., with drilling muds), gel strength values can be
eral techniques to better characterize the yield stress of underestimated (Speers et al., 1987). Therefore, other
cement slurries at rest. This is called the gel strength. techniques have been considered.
Following the standard procedure defined by the Sabins et al. (1980) developed a device to measure
API/ISO for drilling muds, gel strengths of cement slur- the gel strength of cement slurries. Derived from a pres-
ries are often evaluated by measuring the peak value of surized consistometer (Appendix B), their device mea-
the torque upon sudden application of a shear rate of sures the torque necessary to rotate a specific paddle at
5.11 sec–1 after a given rest period. Unfortunately, for two a low rate. However, the stress distribution in this device

118 Well Cementing

is not known, and it is not entirely clear whether the placed vertically on top of the fluid to be analyzed. The
instrument is measuring a fluid “consistency” that would cylinder is allowed to sink into the fluid until its position
correlate with the true gel strength or whether it is mea- stabilizes. The gel strength is determined from the cylin-
suring a stress at the interface between the cement der geometry and weight, the depth to which the cylin-
sample and the sample cup. Nevertheless, Sabins et al. der sank into the material, and the fluid density. The
(1980) were able to show some correlation between advantage of these methods is that the sample is submit-
their gel strength measurements and friction pressure ted to extremely small deformations. The drawback is
measurements performed at low shear rates. that it relies on a correlation that may not apply to a
To ensure that the measured rheological characteris- wide range of cement systems.
tics are valid and not an artifact of friction between the The structural buildup of a cement slurry can also be
instrument and slurry, other authors have used the followed through oscillatory dynamic tests, measuring
shear-vane method. The standard coaxial cylinder is the evolution of the storage (elastic) and loss (viscous)
replaced by a vane (Fig. 4-37). moduli over time (Hannant and Keating, 1985; Chow
If the vane is rotated at a sufficiently low speed, the et al., 1988), but these techniques do not give direct
sheared surface is cylindrical. The maximum torque access to the gel strength.
recorded can be used to calculate the gel strength of the These different techniques do not produce the same
material (Banfill and Kitching, 1990). The advantage of results on the same cement slurry. Therefore, more work
this method, which is commonly used in soil mechanics, will be required to understand these differences, in
is that results are not affected by wall slip because the hopes that the industry will eventually be able to stan-
shear surface is within the material itself. dardize equipment and procedures to measure gel
More recently, Moon and Wang (1999) derived strength. In the meantime, when reporting gel strength
cement-slurry gel strength by measuring the attenuation data, it is essential to specify the equipment and proce-
of acoustic waves propagating through the tested mater- dure used to perform the measurements.
ial. They correlated the measured attenuation and
cement-slurry gel strength measurements performed at
low shear in a pipe (Section 4-3.11) or a shearometer 4-5 Flow of spacers and cement slurries
tube. A shearometer tube is a hollow cylinder that is
in the wellbore
The previous sections showed how the rheological prop-
erties of spacers and cement slurries are characterized.
This section examines how this information is used to
determine velocity profiles, circulation efficiency, and
friction pressures of spacers and cement slurries. The
following also applies to drilling fluids.

4-5.1 Generic laminar flow equations

The axial flow of a fluid in a concentric annulus between
two cylinders of radii ro and rw < ro is considered. It is
assumed that the fluid is incompressible and inelastic.
Provided the flow is laminar,‡‡‡ steady, and isothermal,
the z component of the equation of motion along the axis
of symmetry reduces to (Bird et al., 1960)

( )
1 d dp
rτ rz = − mod , (4-72)
r dr dz
where
gz = z component of gravity
p = total pressure

Fig. 4-37. Schematic diagram of a six-blade vane. ‡‡‡ Flowregimes are discussed in Section 4–6.2. For the time being, the fluid
particles are assumed to flow along streamlines that are parallel to the
main direction of flow.

Chapter 4 Rheology and Flow of Well Cement Slurries 119

pmod = modified pressure, given by pmod = p + ρgz z τw τ τw
)
r
v( r = −
2
r = radial distance from the symmetry axis such that ⎛ d pmod ⎞ ∫τ( r ) ⎛ dv ⎞
dτ = − w
τw ∫τ( r ) γ dτ.
rw < r < ro ⎜ dz ⎟ μ⎜ ⎟
⎝ ⎠ ⎝ dr ⎠
ρ = fluid density (4-78)
τrz = rz component of the stress tensor (shear stress).
The volumetric flow rate, q, or the mean velocity, v–, of
Eq. 4-72 can be integrated for any kind of fluid. the fluid (i.e., the volumetric flow rate per unit cross-sec-
tional area) can be derived from the velocity profile
τ rz =
1 d pmod ⎡
( )2 ⎤⎥,
⎢ r − λro (4-73) through an integration by parts, and rearranged to give
2 dz ⎢⎣ r ⎥⎦
4q 4v 4 τw τ3
= = ∫0 dτ
where λro is the radial position at which τ rz = 0.
Because
( )3
π rw rw τw ( )3 ⎛ dv ⎞
μ⎜ ⎟
⎝ dr ⎠
⎛ dv⎞ dv τw 2
( )
τ rz = μ γ rz × γ rz = μ ⎜ ⎟ × , (4-74) =
4
∫0 τ γ ( τ dτ.
 ) (4-79)
⎝ dr ⎠ dr (τ w ) 3

then If the fluid exhibits a yield stress, τy, then the profile
is flat around the pipe axis and the constant velocity is
⎛ dv ⎞ dv 1 dp ⎡
μ ⎜ ⎟ × = − × mod × ⎢ r − λro ( )2 ⎤⎥, (4-75) given by
⎝ dr ⎠ dr 2 dz ⎢⎣ ⎥⎦
r τw r τy
)
rw
v( r = − ∫τ γ dτ when: ≤ . (4-80)
where τw y rw τ w
μ = viscosity of the fluid
. The expression for the volumetric flow rate is slightly
γrz = rz component of the rate-of-strain tensor (shear modified.
rate).
τw 2
This general expression is used for various flow situa-
4q
=
4
∫τ  τ γ ( τ dτ) (4-81)
tions relevant to the wellbore geometry. π rw ( ) (τ w )
3 3
y

4-5.1.1 Pipe flow A particularly useful form of Eq. 4-79

. gives the expres-
sion for the shear rate at the wall, γ w.
For the particular case of a pipe of radius rw (or diame-
ter dw), λ = 0. Using Eq. 4-73, the shear-stress profile 3 n ′ + 1 4 v 3 n ′ + 1 8 v 3 n ′ + 1 32 q
varies linearly from 0 along the symmetry axis to a max- γ w = = = ,
imum value at the wall τw.
4 n ′ rw 4 n ′ dw 4 n′ π d 3
( w)
(4-82)
r dp r
τ = τ rz = − × mod = × τ w = xτ w (4-76) where
2 dz rw
n′ =
d log τ w( )= d log τ w( ). (4-83)
Equation 4-75 reduces to ⎛ 4v ⎞ ⎛ 8v ⎞
d log ⎜ ⎟ d log ⎜ ⎟
⎛ dv ⎞ dv r dp ⎝ rw ⎠ ⎝ dw ⎠
μ ⎜ ⎟ × = – × mod . (4-77)
⎝ dr ⎠ dr 2 dz
From now on, the term
Integrating from the radius r to the wall (r = rw), and d pmod
assuming the velocity at the wall is 0, one obtains a gen-
dz
eral expression for the velocity at a distance r from the
pipe axis. will be replaced by
⎛ dp⎞ d pmod
⎜⎝ d z ⎟⎠ = − d z
f

120 Well Cementing

to recognize the fact that this pressure loss is caused by are extremely sensitive to pipe diameter when compared
friction and to avoid using a negative sign in the follow- to fluid viscosity and flow rate.
ing equations. For a power-law fluid, the corresponding equations are
For a Newtonian fluid, the pipe-flow equations are 1 1
4v r r 3n + 1 4v ⎛ r ⎞ n 3n + 1 ⎛ r ⎞ n
γ = = γ NW γ = = γ NW , (4-87)
4 n rw ⎜⎝ rw ⎟⎠ 4 n ⎜⎝ rw ⎟⎠
(4-84)
rw rw rw

⎡ 2⎤ ⎡ 1⎤
v( r ) ⎢ ⎛ r ⎞ ⎥ v( r ) 3 n + 1 ⎢ ⎛ r ⎞ 1+ n ⎥
= 2 1− ⎜ ⎟ (4-85) = , (4-88)
v ⎢ ⎝ rw ⎠ ⎥ v n + 1 ⎢1 − ⎜ r ⎟ ⎥
⎣ ⎦ ⎢⎣ ⎝ w ⎠ ⎥⎦

⎛ dp⎞ 32μv 128μq and

⎜⎝ d z ⎟⎠ = = . (4-86)
f ( )
dw
2
π dw
4
( ) ⎛ dp ⎞ 4k ⎛ 3n + 1 8v ⎞
n

⎜⎝ d z ⎟⎠ = d ⎜ 4 n × d ⎟
Notice that, for Newtonian fluids, the shear-rate pro- f w⎝ w⎠
file varies linearly from zero at the pipe axis to a maxi- 23 n+2 kpipe v n 25 n+2 kpipe q n
mum value at the pipe wall (Fig. 4-38). The velocity pro- = = , (4-89)
file is independent of the fluid viscosity and flow rate.
Friction pressure is proportional to the fluid viscosity
( dw )n+1 π n dw( )3 n+1
and flow rate. It is inversely proportional to the fourth
power of the pipe diameter. Therefore, friction pressures

Pipe Pipe Pipe

wall axis wall
2.00 3
Normalized (1) (1) n = 1.00
velocity (2) n = 0.50
profiles
1.75 (3) n = 0.20
(4) n = 0.10
(2)
1.50

2
1.25 (3)

(4)
Velocity/ Shear rate/
average 1.00 Newtonian shear
velocity rate at the wall
0.75
Normalized 1
shear-rate
profiles
0.50

0.25
(1) (2) (3)
(4)
0 0
–1 –0.50 0 0.50 1
Reduced abscissa
Fig. 4-38. Normalized velocity- and shear-rate profiles for Newtonian and power-law fluids flowing in a pipe
(n = power-law index) (n = 1.00 for Newtonian fluids). The reduced abcissa represents the distance to the
axis of the pipe divided by the pipe radius (i.e., r /rw in Eq. 4-85). Particle velocity is normalized by the fluid
mean velocity. Shear-rate values are normalized by the Newtonian shear rate at the wall.

Chapter 4 Rheology and Flow of Well Cement Slurries 121

 ( )
where kpipe is often referred to as the pipe consistency 1/ 4 ⎛ θ⎞
index, defined by ψ = M + P − 2 M 2 + P 2 − MP cos ⎜ ⎟
⎝ 2⎠
n
⎛ 3 n + 1⎞
kpipe = ⎜ k. (4-90) ⎡ −( M + P ) ⎤
⎝ 4 n ⎟⎠ θ = arccos ⎢ ⎥
⎢⎣ 2 M 2 + P 2 − MP ⎥⎦
Notice that, for power-law fluids, the shear-rate pro-
file follows a power-law curve. The shear rate is 0 at the 1/ 3
1⎡ 2 4 ⎤
M = ⎢⎛ 1 + 3ξ ⎞ + ⎛ 1 + 3ξ ⎞ − 1 ⎥
pipe axis and reaches a maximum at the pipe wall
(Fig. 4-38). The velocity profile is independent of the 2 ⎢⎜⎝ 4 ⎟⎠ ⎜⎝ 4 ⎟⎠ ⎥
fluid consistency index and the flow rate. It depends only ⎣ ⎦
on the power-law index. The lower the power-law index
1/ 3
(the less Newtonian the fluid is), the flatter the velocity 1⎡ 2 4 ⎤
profile (Fig. 4-38). The friction pressure is proportional P = ⎢⎛ 1 + 3ξ ⎞ − ⎛ 1 + 3ξ ⎞ − 1 ⎥ (4-93)
to the consistency index and the flow rate to the power 2 ⎢⎜⎝ 4 ⎟⎠ ⎜⎝ 4 ⎟⎠ ⎥
⎣ ⎦
n. It is inversely proportional to the (3n + 1) power of
the pipe diameter. Therefore, friction pressures are Eq. 4-91 shows that friction pressures depend in a
much more sensitive to the pipe diameter than the flow nonlinear fashion on flow rates (Fig. 4-39). But when the
rate, regardless of the power-law index value. Newtonian shear rate is high compared to the ratio of
For Bingham plastic fluids, the flow rate/friction pres- the plastic viscosity to the yield stress, the relationship
sure relationship can be described by what is known as becomes nearly linear. The term to the fourth power in
the Buckingham-Reiner equation. Eq. 4-91 can be neglected, leading to the following
approximation.
8v μ p 32 q μp
ξ= × = ×
( )
dw τ y π d 3 τy 1
w
ψ= (4-94)
4
ξ+
1 ⎛ 4 1 ⎞ 3
= × ⎜1− ψ + ψ4⎟ , (4-91)
ψ ⎝ 3 3 ⎠
⎛ dp⎞ 16 τ y 128μ p q 16 τ y 32μ p v
⎜⎝ d z ⎟⎠ = 3 d + = + (4-95)
where
f w π dw
4
( )
3 dw dw
2
( )
τy τy
ψ= = . (4-92)
τw dw ⎛ dp ⎞ This approximation can be made to determine fric-
tion pressures, provided the dimensionless shear rate, ξ,
4 ⎜⎝ dz ⎟⎠ f is sufficiently large. For example, if
ψ is a reciprocal dimensionless shear stress and is 8v μ p
inversely proportional to the friction pressure. ξ, which × > 1.10,
dw τ y
Eq. 4-84 shows to be
μp Eq. 4-95 will overestimate the friction pressures by less
γ NW , than 1%.
τy The shear rate and velocity profile†††† are given by two
equations. Part of the velocity profile around the pipe
is a dimensionless shear rate.§§§ Therefore, when the axis is flat, while the rest of it is a parabola.
friction pressure is known, ψ can be determined, and
the fluid mean velocity can be directly calculated. To
determine the friction pressure knowing the fluid mean
velocity, an analytical solution for Eq. 4-91 exists
(Whittaker et al., 1985).

†††† Thevelocity-profile equations show that the reciprocal dimensionless

many textbooks, ξ ⁄ 8 is called the Bingham number or the plasticity
§§§ In
shear stress, ψ, also represents the fraction of the pipe diameter where
number for pipe flow. It is equal to the ratio of the Bingham-Reynolds
the velocity profile is totally flat. This is why this parameter is sometimes
number (NRe)BG to the Hedström number (NHe). These dimensionless
called the plug-to-pipe ratio.
numbers are defined in Section 4-6.2. The reciprocal of the Bingham
number, 8⁄ ξ, is also called the yield number.

122 Well Cementing

 This parameter embeds the roles of pipe diameter,
fluid velocity (or flow rate) through the Newtonian
shear-rate term, and the rheological parameters of the
fluid through the plastic viscosity/yield stress ratio.
Friction Linear asymptotic When compared to the yield stress/plastic viscosity ratio,
pressure behavior the lower the Newtonian shear rate is, the less
16τy Bingham plastic Newtonian the fluid will be. The viscosity profile will also
flow curve be flatter (Fig. 4-40). Thus, unlike that of power-law
3dw
fluids, the velocity profile of Bingham plastic fluids is
4τy
flow rate dependent.
dw The complete set of Herschel-Bulkley equations is
given in Appendix A. The equation relating the flow rate
Flow Rate (q) to the friction pressure,
Fig. 4-39. Flow curve of a Bingham plastic fluid in a pipe. 1 1
8v ⎛ k ⎞ n 32 q ⎛ k ⎞n
ξ= ×⎜ ⎟ = ×⎜ ⎟
τy 1 ⎛ r ⎞ 1 ⎛ r ⎞
dw ⎝ τ y ⎠ π dw ( )3 ⎝ τy ⎠
γ = ⎜ − ψ⎟ = ⎜ − ψ ⎟ γ NW 1
μ p ψ ⎝ rw ⎠ ξψ ⎝ rw ⎠
=
4 n (1 − ψ )1+ n × ⎡⎢ (1 − ψ )2 + 2ψ (1 − ψ ) + ⎤
ψ2 ⎥
,
⎡
1 ⎢ 3n +1 2n + 1 n+1⎥
⎞ ⎤
2
⎛ ψ ⎣ ⎦
)
⎢ (1 − ψ 2 − r − ψ ⎥
n
⎜r ⎟ ⎥
= 2⎢
v( r ) ⎝ w ⎠ (4-98)
and
v ⎢ ⎥
⎢ 1− 4 ψ + 1 ψ4 ⎥ must be solved numerically to determine the friction
⎢⎣ 3 3 ⎥⎦
pressure for a given flow rate. As for Bingham plastic
τy fluids, this equation relates a dimensionless shear rate,
r ξ,‡‡‡‡ to a reciprocal dimensionless shear stress, ψ. An
when ≥ψ= . (4-96)
rw τw approximation to Eq. 4-98 can be used for large values of
ξ, or when ψ is small enough. This leads to the following
explicit expression:
v( r ) 2
γ ( r ) = 0 and =
16 τ y ⎛ 3 n + 1 ⎞ 2 n+2 kv n
n
v 2 1 ⎛ dp ⎞
1+ ψ + ψ2 = + . (4-99)
⎜⎝ d z ⎟⎠ ⎜⎝ n ⎟⎠
τy
3 3
f
3 dw d ( w)
n+1

r
when ≤ψ= . (4-97)
rw τw Reviews of pipe-flow equations for other model fluids
(Casson; Robertson and Stiff) were presented by
The shear rate is 0 in the center of the pipe, but then Whittaker (1985) and Fordham et al. (1991).
it varies linearly until it reaches a maximum value at the
pipe wall (Fig. 4-40).
The velocity profile depends on a single parameter—
either the dimensionless shear rate,
μp
ξ = γ NW ,
τy

or the reciprocal dimensionless shear stress, ψ—

because these two parameters are interrelated by
Eq. 4-91.
Herschel-Bulkley fluids, ξ ⁄ 8 is sometimes called the Herschel-Bulkley
‡‡‡‡ For
number for pipe flow.

Chapter 4 Rheology and Flow of Well Cement Slurries 123

 Pipe Pipe Pipe
wall axis wall
2.00 3
Normalized (1) ψ = 0.20 ξ = 3.67
velocity (2) ψ = 0.40 ξ = 1.19
profiles (1) (3) ψ = 0.60 ξ = 0.405
1.75
(4) ψ = 0.80 ξ = 0.0873

(2)
1.50

(3) 2
1.25
(4)
Velocity/ Shear rate/
average 1.00 Newtonian shear
velocity rate at the wall

0.75
1

0.50
Normalized
shear rate
profiles
0.25

(1) (2) (3) (4)

0 0
–1 –0.50 0 0.50 1
Reduced abscissa

Fig. 4-40. Normalized velocity- and shear-rate profiles for a Bingham plastic fluid flowing in a pipe
(ψ = reciprocal dimensionless shear stress; ξ = dimensionless shear rate).

4-5.2 Narrow concentric annular flow 6q 6v 3 τw τ2

= = ∫0 dτ
( )2
As will be discussed later, when the radius ratio in an ⎛ dv⎞
wL2 L τw
annulus is relatively close to unity, the flow equations μ⎜ ⎟
can be reasonably well approximated by neglecting the ⎝ dr ⎠
curvature effect and using a parallel-plate approxima- τw
tion. Annular flow is then assumed to be equivalent to =
3
∫0 )
τγ ( τ dτ (4-102)
flow between parallel plates—a rectangular slot of infi- (τ w ) 2

nite width—separated by a distance L = (ro – rw).The

actual width of the slot, w = π (ro + rw), is only taken 2 n ′ + 1 6 v 2 n ′ + 1 12 v
γ w = = (4-103)
into account to convert the mean fluid velocity to flow 3 n′ L 3 n ′ do − dw
rate, or vice versa. Expressions for the shear-stress pro-
file, velocity profile, and volume flux can be easily
derived in the same way as for a pipe, with r now being n′ =
d log τ w ( )= ( )
d log τ w
(4-104)
the distance from the plane of symmetry of the slot. ⎡ 6v ⎤ ⎡ 12 v ⎤
d log ⎢ ⎥ d log ⎢ ⎥
r ⎛ dp⎞ r
⎣ L⎦ (
⎢⎣ do − dw ) ⎥⎦
τ = τ rz = ⎜ ⎟ = τ w = xτ w (4-100)
2 ⎝ dz ⎠ f L For fluids exhibiting a yield stress, τy, the velocity
profile is flat around the pipe axis, and the constant
−1
⎛ dp⎞ τw τ τw velocity is given by
)
v( r = ⎜ ⎟
⎝ dz ⎠ ∫τ( r ) ⎛ dv⎞
dτ =
L
2τ w ∫τ( r ) γ dτ τw τy
f μ⎜ ⎟
⎝ dr ⎠
v( r = ) L
2τ w ∫τ y
γ dτ when
r
L
≤ψ=
τw
. (4-105)

(4-101)

124 Well Cementing

 The expression giving the volumetric flow rate is ⎛ d p⎞ 6τ y 48 μ p v
slightly modified. ⎜⎝ d z ⎟⎠ = d − d +
τw
f o w do − dw ( )2
6q
=
3
∫τ )
τγ ( τ dτ (4-106)
wL2 (τ w ) 2
y

=
6τ y
+
192μ p q
,
Because the flow equations for different rheological
do − dw
(
π do − dw )(
3
do + dw )
models look very similar to the ones described earlier for (4-113)
flow through a pipe, only a few of them are given here.
The other ones will be found in Appendix A. For example, if
For power-law fluids, the main flow equation is
12 v μp
2 n+ 2 n n × > 2.23,
⎛ d p⎞ 2 × 3 × kann × v do − dw τ y
⎜⎝ d z ⎟⎠ =
f ( do − dw )n+1 calculating the friction pressures from Eq. 4-113 will
4 n+ 2 n n overestimate them by less than 1%.
2 × 3 × kann × q
= , (4-107) Examples of velocity profiles and shear-rate profiles
(
π n do − dw )2 n+1 ( do + dw ) for the power-law and Bingham plastic models are pre-
sented in Figs. 4-41 and 4-42.
with For Herschel-Bulkley fluids, the flow equation needs
n to be solved numerically. Again, one relates a dimension-
⎛ 2n + 1⎞
kann = ⎜ k. (4-108) less shear rate, ξ,§§§§ to a reciprocal dimensionless shear
⎝ 3 n ⎟⎠ stress:
For Bingham plastic fluids, the main flow equation is 1

μp μp 12 v ⎛ k ⎞n
12 v 48 q ξ= ×⎜ ⎟
ξ= × = × do − dw ⎝ τy ⎠
do − dw τ y π d − d
o w
2
(
do + dw τy
)( ) 1
1 ⎛ 3 1 ⎞ ⎛ k ⎞n
= × ⎜1− ψ + ψ3 ⎟ , (4-109) =
48 q
×⎜ ⎟
ψ ⎝ 2 2 ⎠
(
π do − dw )2 ( do + dw ) ⎝ τy ⎠
where 1

τy τy =
3 n (1 − ψ )1+ n (1 + n + nψ ) . (4-114)
ψ= = . (4-110) 1
τw do − dw ⎛ d p ⎞
4 ⎜⎝ d z ⎟⎠
( n + 1)( 2 n + 1) ψ n

f
An approximation of Eq. 4-114 can be used for large
A solution to Eq. 4-109 is given by (Whittaker, 1985) values of ξ, or when ψ is sufficiently small, and it leads
to the following explicit expression:
⎡ −3 ⎤
2ξ ⎢ ⎛ ξ ⎞ 2 ⎥
ψ=2
3
+ 1 × sin arcsin
⎢3
1 2
⎜⎝ 3 + 1⎟⎠ ⎥
. (4-111) ⎛ dp⎞ 6τ y ⎛ 2n + 1⎞ 2
2( n+1)
k( v
n
)n
⎣ ⎦ ⎜⎝ d z ⎟⎠ = d − d + ⎜⎝ n ⎟⎠ . (4-115)
f o w do − dw
n+1
( )
Again, when the dimensionless shear rate, ξ, is suffi-
ciently large compared to the yield stress/plastic viscos- Reviews of infinite plane slot flow equations for other
ity ratio, the term to the third power in Eq. 4-109 can be model fluids (Casson, Robertson and Stiff) are pre-
neglected. This leads to the following approximations. sented by Whittaker (1985) and Fordham et al. (1991).
1
ψ= (4-112)
3
ξ+
2
§§§§ ForHerschel-Bulkley fluids, ξ ⁄12 is sometimes referred to as Herschel-
Bulkley number for annular flow.

Chapter 4 Rheology and Flow of Well Cement Slurries 125

 Pipe Pipe Pipe
wall axis wall
2.00 3
(1) n = 1.00
(2) n = 0.50
1.75 (3) n = 0.20
(4) n = 0.10

1.50 (1)
Normalized
velocity (2)
2
1.25 profiles
(3)
Velocity/ Shear rate/
1.00 (4)
average Newtonian shear
velocity rate at the wall
0.75
Normalized 1
shear rate
0.50 profiles

0.25
(3) (4)
(1)
(2)
0 0
–1 –0.50 0 0.50 1
Reduced abscissa
Fig. 4-41. Normalized velocity- and shear-rate profiles for a power-law fluid flowing in a slot
or narrow annulus.

Pipe Pipe Pipe

wall axis wall
2.00 3
(1) ψ = 0.20 ξ = 3.52
(2) ψ = 0.40 ξ = 1.08
1.75 (3) ψ = 0.60 ξ = 0.347
(4) ψ = 0.80 ξ = 0.0700

1.50
Normalized
velocity (1)
profiles
2
1.25 (2)
(3)
Velocity/ (4) Shear rate/
average 1.00 Newtonian shear
velocity rate at the wall
0.75
1

0.50
Normalized
shear rate
0.25 profiles

(1)
(4)
(2) (3)
0 0
–1 –0.50 0 0.50 1
Reduced abscissa
Fig. 4-42. Normalized velocity- and shear-rate profiles for a Bingham plastic fluid flowing in a slot
or narrow annulus (ψ = reciprocal dimensionless shear stress; ξ = dimensionless shear rate).

126 Well Cementing

4-5.3 Concentric annular flow ⎛ d p⎞ 23 n + 2 k ( v )n
Gn
For flow in concentric annuli with a large radius ratio, ⎜⎝ d z ⎟⎠ = n+1
there is no general expression for the velocity profile and f ⎡ do (1 − α ⎤
⎣ ⎦)
the volume flux. The equations relating friction pres-
sures to flow rate will first be described. The effect of the 25 n+2 kq n
= Gn,
radius ratio on velocity profiles will be briefly discussed.
For Newtonian fluids, the principal equation is
π n
( do ) (1 − α ) (1 + α )
3 n+1 2 n+1

(4-119)
⎛ dp⎞ 32μv 1
⎜⎝ d z ⎟⎠ =
)( )
which is the pipe-flow equation with a correction factor
( ) 2 ⎡
1 − α2 ⎤
(
f do that depends on the power-law index and the annulus
⎢ 1 + α2 + ⎥
⎢
⎣ ln ( α ⎥
⎦ ) diameter ratio.
⎛ S⎞
=
128 μq 1
, )
⎜⎝ 1 + 2 ⎟⎠ ⎡⎣( 3 − S n + 1 ⎤⎦
( )
)( )
4 ⎡
2 ⎤
2
π do G=
⎢ 4
⎢ 1− α +(
1− α ⎥
⎥
(4 − S n )
ln ( α ⎥⎦ )
⎢⎣
( )
1/ Y
S = 1 − 1 − αY
(4-116)
where α is the ratio of the inner diameter to the outer Y = 0.37 × n−0.14 (4-120)
diameter (dw ⁄ do). For Bingham plastic fluids, the relevant equations are
Fredrickson and Bird (1958) first developed the flow given below.
equations for power-law and Bingham plastic fluids. An
improved formula for power-law fluids has since been
developed by Hanks and Larsen (1979). The same year, 32 q μp 1 ⎢ ( )
⎡ 1 − α 4 − 2λ ( λ − β 1 − α 2 ⎤
⎥
)( )
× = × ⎢ ⎥,
Hanks (1979) presented the flow equation for Herschel-
Bulkley fluids. ( )
π do
3

⎣ 3
(
3
( )
τ y β ⎢ − 4 1 + α 3 β + 1 2λ − β 3 β ⎥
⎦
)
For power-law fluids, the flow is described by
(4-121)
1 where
32 q 4 n ⎡ do ⎛ d p ⎞ ⎤n τy
= ⎢ ⎥ 4
3 n + 1 ⎢ 4 k ⎜⎝ d z ⎟⎠
( )
π do
3
⎣ f⎥
⎦ β=
do
. (4-122)
⎛ d p⎞
⎡ 1 1 1⎤
( ) ( ) ⎜⎝ d z ⎟⎠
1+ 1− 1+
⎢ 1 − λ2 n −α n 2
λ −α 2 n⎥, f
⎣ ⎦
(4-117) Here, λ is the largest normalized distance from the
pipe axis where the shear stress of the fluid becomes
where λ is the normalized distance from the pipe axis equal to the yield stress. Its value is defined by the fol-
where the shear stress is 0, or where the velocity reaches lowing implicit equation.
its maximum. Its value is given by the solution of
⎡ ⎛ λ −β⎞ ⎤ ⎡
λ⎛ λ 2 ⎞
1
n 1⎛ λ
1
2⎞ n )
⎢2λ ( λ − β ln ⎜ ⎟
⎝ αλ ⎠ ⎦ ⎣
2⎤
⎥ − ⎢1 + ( α + β ⎥
⎦
)
⎣
∫α ⎜⎝ x − x⎟⎠ d x − ∫λ ⎜⎝ x − x ⎟⎠ d x = 0. (4-118)
)
+ ⎡⎣2(1 − λ β ⎤⎦ = 0 (4-123)

λ depends only on the radius or diameter ratio, α, and For these last two rheological models, the flow equa-
the power-law index of the fluid. Hanks and Larsen tions are implicit, and they can only be solved numeri-
(1979) published tabulated values for λ for a series of α cally. This is also the case for Herschel-Bulkley fluids, for
and n values. Later, an approximation to the above equa- which flow equations are described and solved by Hanks
tions was developed by Whittaker et al. (1985). (1979). Notably, Fordham et al. (1991) presented a gen-
eral method for the practical calculation of laminar flow
in centered annuli of arbitrary radius that is very useful.

Chapter 4 Rheology and Flow of Well Cement Slurries 127

 Because the narrow-gap equations are much simpler 1.30
n = 1.0
to solve, the question that must be addressed is: “What n = 0.5
are the errors associated with this approximation when 1.25 n = 0.2
compared to the exact solution?” The answer really
Ratio of 1.20
depends on the application. For determining the flow rate
corresponding to a given friction pressure, this approxi- actual
flow to 1.15
mation remains reasonable for diameter ratios larger calculated
than about 0.5, but, as shown in Fig. 4-43, it deteriorates slot flow 1.10
quickly for small-diameter ratios. The lower the power
index is (i.e., the less Newtonian the fluid behavior), the 1.05
larger the error will be. Similar errors are obtained with
Bingham plastic and Herschel-Bulkley fluids. 1
On the other hand, when trying to do the reverse cal- 0 0.2 0.4 0.6 0.8 1
culation (i.e., determine the friction pressure correspond- Ratio of inner to outer diameter
ing to a given flow rate), even for an annulus diameter Fig. 4-43. Comparison of flow rates at the same friction pressures,
ratio as low as 0.3, the corresponding error is lower than calculated using Eqs. 4-107, 4-117, and 4-118.
2.5%, both for power-law and Bingham plastic fluids. This
is likely to be true for any generalized non-Newtonian
model, provided that the fluid is shear thinning. where r now represents the distance to the axis of sym-
Velocity profiles are affected by the annular radius metry of the two cylinders. The distortion of the velocity
ratio. As this ratio decreases, the shear rate at the wall profile compared to a narrow gap approximation is illus-
of the inner cylinder increases faster than that at the trated in Fig. 4-44. The same trend is observed for non-
outer cylinder. Also, the point of maximum velocity (or Newtonian fluids. Again, unless the radius ratio is very
the position of the plug zone) is shifted towards the small and the fluid is highly non-Newtonian, the distor-
inner cylinder, and the maximum velocity itself is modi- tion of the velocity profile is not very significant.
fied. This is illustrated by the following equations, which
are given only for Newtonian fluids.
2 ⎡
⎛ r ⎞ ⎢ 1− α
1− ⎜ ⎟ −
(2
)
⎛ r ⎞⎥
× ln ⎜ ⎟
⎤ 1.6
1.4

v( r )
⎝ ro ⎠ ⎢ ln ( α
⎢⎣
)⎝ ro ⎠ ⎥
⎥⎦
1.2
1.0
=2 , (4-124)

)( )
Normalized
v ⎡ 1− α ⎤ 2 0.8

(
velocity True solution
⎢ 1 + α2 + ⎥ 0.6
⎢
⎢⎣
ln ( α ⎥
⎥⎦
) 0.4
0.2
Narrow slot

2
α +
(1 − α ) 2 0.2 0.4 0.6 0.8
Normalized distance to the axis of the cylinders
1

( )
γ rw =
2v ln ( α
,
) (4-125)
rw ⎡
(
⎢ 1 + α2 +
)(
1 − α2 ⎤
⎥ ) 1.6
1.4
⎢
⎢⎣
ln ( α ⎥
⎥⎦
) 1.2
1.0 True solution
Normalized 0.8 Narrow slot
and shear rate 0.6

1+
(1 − α ) 2 0.4
0.2

( )
γ ro =
2v ln ( α
,
) 0
0.2 0.4 0.6 0.8 1
)
(4-126)

)(
ro ⎡ 1 − α2 ⎤
(
Normalized distance to the axis of the cylinders
⎢ 1 + α2 + ⎥
⎢
⎢⎣
ln ( α ⎥
⎥⎦
) Fig. 4-44. Velocity and shear rate profiles for a Newtonian fluid flow-
ing in an annulus with a radius ratio of 0.2. Shear rates are normal-
ized by 12v–/(do – dw); local velocities are normalized by the mean
fluid velocity.

128 Well Cementing

 Therefore, it is reasonable to conclude that the The dimensionless breakthrough time, t*break, is
narrow-gap approximation is often a good engineering defined by
approximation to determine laminar friction pressures
of spacers and cement slurries in annuli because * vt zt *
tbreak = = , (4-128)
■ in most circumstances, annuli are relatively narrow vmax vmax
during cementing operations
■ for the diameter ratio in question, this approximation
where vmax is the maximum velocity of the fluid particles
(i.e., the velocity at the pipe axis). It is simply the dimen-
provides an upper limit for the friction pressures
sionless time at which fluid particles that entered the
■ in practice, friction pressures are often negligible for
pipe first at t = 0 exit the pipe section.
large-diameter ratios. The breakthrough radius, rbreak, can be calculated
from
4-5.4 Circulation efficiency
Some of the flow equations described above can be
(
v rbreak = ) z v
= .
t t*
(4-129)
used to determine the process efficiency while circulat-
ing a fluid in a pipe or an annulus. Consider a section of It is the distance from the pipe axis at which particles
pipe or annular section of length z. The circulation effi- that entered the annulus at t = 0 come out of the pipe
ciency after a circulation time t is defined as 1 minus the section at time t.
ratio of two volumes. The first is the volume of the pipe With these variables, the efficiency of the circulation
or annular section that still contains the original fluid process is expressed as

( )
before starting circulation (i.e., the red-colored fluid in
Fig. 4-45). The second is the volume of the entire pipe or η circ t * = t * when t * ≤ tbreak
*
(4-130)
annular section.
and
We will first examine circulation efficiency in a pipe
and then in a concentric annulus. To simplify the equa- 2

( )
⎛r ⎞ rw r v( r ) r
tions, it is useful to introduce a few dimensionless para- η circ t = ⎜ break ⎟ + 2 t * ∫
*
d
meters. The dimensionless time, t*, is the number of ⎝ rw ⎠ rbreak r
w v rw
pipe-section volumes pumped at time t. It is given by
when t * ≥ tbreak
*
. (4-131)
vt
t* = . (4-127)
z

Unwrapping an Eccentric Annulus Single Fluid Displacing Itself

Lmin Lmax Lmax Fluid A Fluid B

Lmin

z z

Fluid A is fluid originally present in the

annulus before starting circulation.
Fluid B is fluid injected at the bottom
of annulus after starting circulation.
0.2 0.4 0.6 0.8 1.0
Annular volumes circulated
Fig. 4-45. Schematic of the circulation-efficiency concept.

Chapter 4 Rheology and Flow of Well Cement Slurries 129

 For a narrow annulus, the equation that is equivalent ρvdw
to the previous one is N Re = . (4-133)
μ

( ) ⎡ r
η circ t * = 2 ⎢ break + t * ∫
⎣ ro − rw
ro or
rbreak
rw v( r )
v
d
r ⎤
⎥
ro − rw ⎦
Departure from laminar flow occurs as the Reynolds
number increases beyond a value of 2,100. A transition
regime that is not very well characterized exists until
when t * ≥ tbreak
*
, NRe = 3,000. Above this value, flow becomes turbulent.
(4-132) The resistance to flow at the pipe wall can then be
expressed as

( )
with r now being the distance from the plane of symme- 1
try of the narrow annulus. Similar equations can be = H × log N Re × f fr + J , (4-134)
developed for large annuli, but the calculations are more f fr
complex because the velocity profile is no longer sym-
metrical. where ffr, the Fanning friction factor, is defined by
The equations giving the dimensionless breakthrough
time and the circulation efficiency for flow in pipes and 2τ W
f fr =
( )
. (4-135)
in narrow annuli are presented in Appendix A. Most of ρ v2
the conclusions drawn about velocity profiles can be
applied to circulation efficiency.
In Eq. 4-134, which was first proposed by Von Karman
■ For power-law fluids, circulation efficiency depends
in 1930 (Schlichting, 1979), parameters H and J depend
only on the power-law index. The smaller the power- on the roughness of the pipe. For turbulent flow in
law index, the larger the dimensionless breakthrough smooth pipes, H = 4.0 and J = –0.4. With these defini-
time and the more efficient the circulation process. tions it should be noticed that, in laminar flow
■ For Bingham plastic fluids, circulation efficiency
depends on the dimensionless shear rate. The smaller 16
f fr = . (4-136)
the dimensionless shear rate, the larger the dimen- N Re
sionless breakthrough time and the more efficient
the circulation process. In the transition regime, the friction-factor/Reynolds
■ For Herschel-Bulkley fluids, circulation efficiency number relationship is not uniquely defined. However,
depends on both the power-law index and the dimen- for most engineering applications, a linear interpolation
sionless shear rate. The smaller these two parameters is performed on a log-log scale between the laminar
are, the larger the dimensionless breakthrough time values of ffr at a Reynolds numbers of 2,100 and its tur-
and the more efficient the circulation process. bulent value at a Reynolds number of 3,000 (Fig. 4-46).
These points are discussed in more detail in
Chapter 5.
Fanning friction factor = 16/(N Re)MR

4-6 Friction pressure calculations for all 0.01 n′

flow regimes Fanning 2.0
1.4
friction
1.0
4-6.1 Pipe flow factor 0.8
0.6
Regardless of the type of fluid, once a critical flow rate 0.4
in a given pipe is exceeded, streamlines are no longer 0.3
Experimental regions
parallel to the main direction of flow. Fluid particles Extrapolated regions 0.0
0.2
0.001
become subject to random velocity fluctuations in both
1,000 10,000 100,000
amplitude and direction. In fact, velocity fluctuations
are not completely random. Near the wall, fluctuations Reynolds number
in the axial direction are greater than in the radial direc- Fig. 4-46. Relationship between Fanning friction factor and the gen-
tion, and both approach 0 at the wall. Such flow instabil- eralized Reynolds number. Note that, for a given Reynolds number,
ity starts at a given value of a dimensionless parameter, ffr is strongly dependent on the value of n ′ (from Dodge and
the Reynolds number (NRe), which, for Newtonian Metzner, 1959). Reprinted with permission of Wiley-Liss Inc., a sub-
sidiary of John Wiley & Sons, Inc.
fluids, is defined by

130 Well Cementing

 Similar equations have been developed for purely vis- 30
cous non-Newtonian fluids.††††† The main problem is to
determine which viscosity value should be used in the 20
expression for the Reynolds number, because viscosity is
shear-rate dependent. For power-law fluids, Metzner
and Reed (1955) proposed a Reynolds number definition
such that Eq. 4-136 still holds true. This leads to Fanning
friction 10
9
ρ( v ) ( dw )
2− n n factor 8
( NRe ) MR = 8 n−1
kpipe
, (4-137)
× 10–3 7
6
5
where kpipe is the pipe consistency index previously
4
defined in Eq. 4-90. Using a dimensional analysis, Dodge
and Metzner (1959) found that turbulent flow of power- 3
law fluids in smooth pipes could be described by a rela- 1 2 3 4 5 6 7 8 9 10
tionship similar to Eq. 4-134: Bingham plastic Reynolds number × 103

1 ⎡ n ⎤
) MR × ( f fr )
1− Fig. 4-47. Fanning friction factor/Reynolds number graph for a given
f fr
= H n × log ⎢ N
(
⎣ Re
2 ⎥ + Jn,
⎦
cement formulation. Circles and triangles are experimental data for
a 0.6- and 0.8-in. [16- and 20-mm pipe]. The continuous (16 mm) and
(4-138) the dotted (20 mm) lines were calculated following API/ISO proce-
dures for Bingham plastic fluids (i.e., in turbulent flow, fluids are
where Hn and Jn are a function of the power-law index assumed to behave like Newtonian fluids with a viscosity μp (after
n. Guillot and Denis, 1988). Reprinted with permission of SPE.
For Bingham plastic fluids, the simplest method
(Hedström, 1952) is to assume that, once turbulent flow
is reached, the fluid is Newtonian, with a viscosity equal ⎡ ⎛ n′ ⎞ ⎤
) MR × ( f fr )
1−⎜ ⎟
to its plastic viscosity. This indicates that the relevant
Reynolds number in turbulent flow, often referred to as
1
= H n′ × log ⎢ N Re
⎢
( ⎝ 2⎠ ⎥+ J ,
⎥ n′
f fr
the Bingham-Reynolds number, is ⎣ ⎦
(4-140)
ρvdw
( NRe ) BG = μp
. (4-139) where Hn´ and Jn´ are function of n′ only. The general-
ized Reynolds number,(NRe)MR, is defined by Metzner
and Reed (1955) as
Equation 4-134 is then used to calculate friction pres-
sures for a given flow rate (Fig. 4-46). This assumption ρ( v )2− n′ ( dw )n′ .
has been established empirically for smooth pipes by
several authors working with different types of fluids
( NRe ) MR = 8 n′−1 k′pipe
(4-141)

(Govier and Aziz, 1977). Unfortunately, it does not seem

to hold for all cement slurries. Guillot and Denis (1988)
showed that this procedure can lead to a considerable The local power-law parameters, n′ and k′pipe, are
overestimation of friction factors (Fig. 4-47). defined by

( )
Other methods for calculating turbulent friction pres-
sures of Bingham plastic and Herschel-Bulkley fluids in d log τ w
n′ = (4-142)
pipes have been developed (Govier and Aziz, 1977), but ⎛ 8v ⎞
their validity has not been fully established for cement d log ⎜ lam ⎟
slurries or they are limited to a specific rheological ⎝ dw ⎠
model. A more general approach that does not suffer
from this limitation was proposed by Dodge and Metzner and
(1959). They generalized Eq. 4-138 to describe the turbu- τW
lent flow of nonelastic non-Newtonian fluids in smooth k′pipe = n′
. (4-143)
pipes (Fig. 4-46). ⎛ 8 vlam ⎞
⎜ d ⎟
⎝ w ⎠

††††† This
section does not consider elastic fluids that can exhibit drag
reduction.

Chapter 4 Rheology and Flow of Well Cement Slurries 131

 v–lam is the average velocity for the same shear stress 0.009
Solid points for suspensions
at the wall, τw, assuming the flow regime is laminar. Open points for polymeric gels
Notice that, for power-law fluids, 0.008
+2.5%
n′ = n (4-144) –2.5%
Experimental 0.007
and Fanning
n friction
⎛ 3 n + 1⎞ factor 0.006
k′pipe = kpipe = ⎜ k. (4-145)
⎝ 4 n ⎟⎠
0.005 Power-law range
Extending the application of Eq. 4-136, and using the 0.30 to 0.45
0.45 to 0.60
generalized equations for power-law fluids defined by 0.60 to 0.75
0.004
Metzner and Reed (Eqs. 4-141 to 4-143), is straightfor-
0.004 0.005 0.006 0.007 0.008 0.009
ward. However, for Eq. 4-140, the extension has been
questioned by Hanks (1986) as not being theoretically Predicted Fanning friction factor
correct. Dodge and Metzner (1959) justified this gener- Fig. 4-48. Comparison of experimental and predicted Fanning fric-
alization by pointing out that, in turbulent flow, only the tion factors (turbulent non-Newtonian points only) (from Dodge and
shear in very close proximity to the wall contributes sig- Metzner, 1959). Reprinted with permission of Wiley-Liss Inc., a sub-
nificantly to the friction. Dodge and Metzner (1959) gave sidiary of John Wiley & Sons, Inc.
experimental evidence that this is correct. For the non-
Newtonian fluids they tested, with n′ values from 0.36 to
1.0, and (NRe)MR values from 2,900 to 35,000, they empir- 30
ically found that, for smooth pipes:
4.0 0.4 20
H n′ = and J n′ = − . (4-146)
( n′ )0.75 ( n′ ) 1.2
Fanning
friction 10
9
Dodge and Metzner (1959) found that their method factor 8
gave remarkable friction-pressure predictions for the × 10–3 7
fluids they tested (Fig. 4-48). Very good results were also 6
5
obtained by Guillot and Denis (1988) with cement slur-
ries whose rheological properties were described by a 4
three-parameter model (Fig. 4-49). 3
Notice that, even for power-law fluids, Eqs. 4-138 and 5 10 50 100
4-140 are implicit with regard to the friction factor. For Generalized Reynolds number × 102
most engineering applications, it can be replaced by an
explicit expression such as that given in Appendix A Fig. 4-49. Fanning friction factor versus generalized Reynolds
(Tables A-4 and A-5). The same equation can also be number for a given cement formulation. Symbols correspond to raw
modified to account for pipe-wall roughness (Reed and data. Lines correspond to values calculated by the Dodge and
Metzner equation using a three-parameter model.
Pilehvari, 1993):
1
= −4.0 × log End of laminar flow regime
f fr
Now that the friction factor/Reynolds number relation-
⎧⎛ 0.27 h ⎞ 1.259 n−1.2 ⎫ ships are defined in laminar flow and turbulent flow, the
⎪⎪⎜ ⎟+ −0.75
⎪⎪, situation in the transition regime needs to be addressed.
⎨⎝ dw ⎠ ⎡ ⎛ n⎞ ⎤
n
⎬ There are numerous empirical or semiempirical models
) MR × ( f fr )
1−⎜ ⎟
⎪
⎪⎩
⎢ N
⎣ Re ( ⎝ 2⎠ ⎥
⎦ ⎪⎭
⎪ to predict the Reynolds number, NRe1, at which the lam-
inar flow regime of purely viscous non-Newtonian fluids
no longer applies. The criteria are either global (Mishra
(4-147) and Tripathi, 1971) or local (Ryan and Johnson, 1959;
– Hanks, 1963a and 1969). These semiempirical criteria
where h is the mean height of protuberances at the pipe
wall. are defined in such a way that they correspond to a

132 Well Cementing

Reynolds number value of 2,100 for a Newtonian fluid. Present chapter
For power-law fluids, NRe1 increases as n decreases. Schuh (1964)
Ryan and Johnson (1959)
Despite the differences between the theories, the dis- Mishra and Tripathi (1971)
crepancies in critical flow rates or velocities are usually Reed and Pilehvari (1993)
reasonable down to n = 0.3, as shown in Fig. 4-50. This is Desouky (1991)
because the Reynolds number is proportional to the flow 1.3
rate to the power (2 – n). For some spacer fluids, the 1.2
Mishra and Tripathi (1971) criterion seems to give a rea-
sonable prediction of the laminar flow stability (Brand Normalized 1.1
et al., 2001). If a good experimental dataset does not critical
exist, the author proposes Eq. 4-148. This is reasonable flow rate 1.0
because fluids with a power-law index of less than 0.3 are 0.9
rarely found at shear rates where laminar flow ends.
0.8
N Re 1 = 3, 250 − 1,150 n (4-148) 0 0.2 0.4 0.6 0.8 1
For Bingham plastic fluids, the theory proposed by Power-law index
Hanks (1967) is the most popular. The upper limit for the Fig. 4-50. Normalized critical flow rates corresponding to the end of
laminar flow regime is often expressed as a relationship the laminar flow regime in a pipe, determined according to various
between a critical value of the reciprocal dimensionless models. The flow rates are calculated according to a (NRe)MR of
shear stress (ψ1) and the Hedström number (NHe). 2,100.

N He =
( )2
ρ dw τ y
(4-149) 30,000
(μ )
2 Present chapter
p 25,000 Hanks and Pratt (1967)
Mishra and Tripathi (1971)
20,000 Slatter (1999)
After the Hanks theory, Bingham
Reynolds 15,000
ψ1 N number 10,000
= He . (4-150)
(1 − ψ 1 ) 3 16, 800 5,000
0
The critical value of the reciprocal dimensionless 100 1,000 10,000 100,000 1,000,000
shear stress is then inserted into expressions giving the Hedström number
Bingham plastic Reynolds number or the Metzner and
Reed Reynolds number as a function of the Hedström Fig. 4-51. Bingham Reynolds number corresponding to the end of
the laminar flow regime in a pipe determined according to different
number and of the reciprocal dimensionless shear models. In this case, the corresponding flow rates are proportional
stress. Figure 4-51 gives the critical value of the to the Bingham Reynolds number.
Bingham plastic Reynolds number for different models.
Again, differences between the different theories are
not significant up to Hedström numbers of the order
of 5 × 104. Above these values, it is relatively difficult End of the transition regime
to place Bingham plastic fluids in turbulent flow in a Defining the end of the transition regime is even more
wellbore. debatable than for NRe1. As shown above, a simple equa-
Most of the theoretical models predicting the end of tion can be proposed that has a few clear limitations.
the laminar flow regime are not specific to a given rheo-
logical model. They can be used to predict the critical N Re 2 = 4,150 − 1,150 n (4-151)
flow rates at which Herschel-Bulkley fluids significantly
depart from laminar flow. One of the limitations of Eq. 4-151 is that it is not
Because of the relatively small differences between compatible with the Fanning friction factor/Reynolds
the different theories, the author has chosen to use Eq. number relationship when the local power index is too
4-148, where the Reynolds number is the Metzner and low (typically lower than 0.3). The reason is that, at NRe2,
Reed Reynolds number (NRe)MR (Eq. 4-141), and n is the turbulent friction factor given by Eq. 4-138 or 4-140
replaced by the local power-law index. This way, the would be lower than the laminar one. This is not physi-
same equation can be applied regardless of the rheolog- cally sound. This difficulty can easily be avoided by basing
ical model used to describe the behavior of the fluid.

Chapter 4 Rheology and Flow of Well Cement Slurries 133

the definition of NRe2 on the Reynolds number value at number. The wall shear stress and the friction pressure
the intersection of the Fanning friction factor/Reynolds gradient are determined as above, which gives the fol-
number curves in laminar and turbulent flow. lowing.
Other constraints are imposed by the local power-law Flow rate (m3/s) 0.024
approach with some rheological models (e.g., Bingham
Fanning friction factor at NRe1 0.00598
plastic) because the calculated friction pressures may
not be a monotonic function of the flow rate at low Fanning friction factor at NRe2 0.0686
power-law index values (typically lower than 0.3). When Fanning friction factor 0.0637
such cases are encountered, the values of NRe2 (n′) Wall shear stress (Pa) 14.9
should be higher than those calculated from Eq. 4-151.
Friction pressure gradient (Pa/m) 468
For the Fanning friction factor in the transitional
regime (e.g., for Newtonian fluids), a linear interpola- Instead of performing the calculations, one can use
tion is made on a log-log scale between the laminar value charts such as that presented in Fig. 4-46; however, the
of the friction factor at NRe1 and its value at NRe2. results are not likely to be as accurate.

Example of application to a power-law fluid Example of application to nonpower-law fluids

A power-law fluid with a density of 1,300 kg/m3, a power- Application of the Dodge and Metzner method to non-
law index of 0.5, and a consistency index of 1.0 Pa-sn is power-law fluids is not straightforward, because it leads
flowing in a tube with an inner diameter of 127 mm. to a system of implicit equations that must be solved
Determine the friction pressure gradient for a series of numerically (see Appendix A in which equations are
three different flow rates: 0.01, 0.03, and 0.024 m3/s. developed for Bingham plastic fluids). Results can be
The first step in the calculation is to determine the presented in the form of master curves for specific rhe-
flow regime. One must calculate the mean velocity of the ological models. Figure 4-52 shows the relationship
fluid and its Reynolds number using Eq. 4-137. between the friction factor and the Bingham plastic
Flow rate (m3/s) 0.01 0.03 Reynolds number for a series of Hedström number
values (Eq. 4-149).
Mean velocity (m/s) 0.789 2.37
Reynolds number, (NRe)MR 822 4,271
The Reynolds number values are then compared to Bingham Plastic Fluid
0.1
the critical values given by Eqs. 4-148 and 4-151 (i.e.,
2,675 and 3,575). Therefore, at the lowest flow rate, the
flow regime is laminar. At the highest flow rate it is tur-
Fanning
bulent. One can now calculate the Fanning friction friction 0.01
factor for the appropriate flow regime (i.e., Eq. 4-136 for factor
0
laminar flow and Eqs. 4-138 and 4-146 [or Eq. 4-147, in
– 10,000
which we will assume that h = 0] for turbulent flow). 100,000
Once the friction factor is known, the shear stress at the 0.001 1,000,000

wall is derived from Eq. 4-135 and the friction pressure 100 1,000 10,000 100,000
gradient from Eq. 4-14. This yields the following results. Bingham Reynolds number
Flow rate (m3/s) 0.01 0.03 Fig. 4-52. Fanning friction factor/Bingham Reynolds number relation-
Friction factor 0.0195 0.00645 ship as a function of the Hedström number for pipe flow.
Wall shear stress (Pa) 7.88 23.5
Friction pressure gradient (Pa/m) 248 740
Figure 4-53 shows the same graph for a Herschel-
For a flow rate of 0.024 m3/s, the Reynolds number is Bulkley fluid, where the Reynolds number is now defined
equal to 3,056 and the flow regime is transitional. as
Therefore, one must determine the friction factor at NRe1
(using Eq. 4-136) and at NRe2 (using Eqs. 4-138 and ρ( v )2− n ( dw )n
4-146). Then a linear interpolation is made on a log-log ( NRe ) HB = n
, (4-152)
scale between these two values to determine the value ⎛ 3 n + 1⎞
8 n−1 × ⎜ ×k
of the Fanning friction factor at the calculated Reynolds ⎝ 4 n ⎟⎠

134 Well Cementing

 Herschel-Bulkley Fluid: n = 0.5 Note that using the hydraulic diameter, do – dw , as
0.1
the characteristic length in the Reynolds number is arbi-
trary. In laminar flow these definitions lead to
Fanning 24
f fr = . (4-157)
( NRe ) AN
friction 0.01
factor
0
10,000
100,000 In turbulent flow, the friction factor is slightly higher
0.001 1,000,000 than what Eq. 4-138 or 4-140 would give. These equations
100 1,000 10,000 100,000 can be replaced by‡‡‡‡‡
Herschel-Bulkley Reynolds number
⎡ ⎛ n′ ⎞ ⎤
) AN ( )
1−⎜ ⎟
Fig. 4-53. Friction factor/Herschel-Bulkley Reynolds number rela-
tionship as a function of the generalized Hedström number for pipe
1
⎢3
2
= H n′ × log ⎢ N Re( × f fr ⎝ 2⎠ ⎥+ J ,
⎥ n′
f fr
flow for a power-law index of 0.5. ⎣ ⎦
(4-158)
where Hn´ and Jn´ are defined by Eq. 4-146.
and the generalized Hedström number is For the transition zone, common industry practice is
2 to use the same criteria as those for pipe flow. This is the
( )
ρ dw
2
⎛ τy ⎞ n option taken in this chapter and in Appendix A. In fact,
( N He ) HB = τy
⎜ ⎟ (4-153) there is experimental evidence (Patel and Head, 1969)
⎝ k⎠ and theoretical reasons (Hanks, 1963a) to believe that
this practice leads to an underestimation of the critical
for a specific value of the power-law index (e.g., 0.5). Reynolds number. For Newtonian fluids flowing between
As with power-law fluids, it is possible to determine parallel plates, Patel and Head (1969) observed a devia-
the flow regime and the Fanning friction factor by enter- tion from laminar flow for Reynolds number values
ing the pipe diameter, fluid density, rheological proper- around 2,700 (as defined by Eq. 4-154), while Hanks
ties and the flow rate or velocity. One uses Eqs. 4-139 and (1963a) predicted a theoretical value of 2,800.
4-149, or 4-152 and 4-153, plus Figs. 4-52 or 4-53 (for n = Several theories exist for purely viscous non-
0.5). Once the friction factor is known, the correspond- Newtonian fluids, including some that take into account
ing wall stress friction-pressure gradient can be deter- the effect of the annular radius ratio. Comparison of these
mined as described above for the power-law fluid. theories leads to discrepancies similar to those described
for a pipe-flow geometry in the previous section.
4-6.2 Concentric annular flow
The concentric annular flow discussion will be limited to Example of application to a power-law fluid
the case of narrow annuli. The same local power-law A power-law fluid with a density of 1,300 kg/m3, a power-
approach described above can be used, with the friction law index of 0.5, and a consistency index of 0.4 Pa-sn is
factor being defined by Eq. 4-135 and the Reynolds flowing in an annulus of 215.9 × 177.8 mm. Determine
number by the friction pressure gradient for a series of three differ-
ent flow rates: 0.008 m3/s, 0.02 m3/s, and 0.016 m3/s.
ρ( v )2− n′ ( do − dw )n′ ,
( NRe ) AN = 12 n′−1 kann
(4-154)
The first step in the calculation is to determine the
flow regime. One must calculate the mean velocity of the
′
fluid and then its Reynolds number using Eq. 4-154.

n′ =
( )
d log τ w
,
Flow rate (m3/s) 0.008 0.02
(4-155)
⎛ 12 vlam ⎞ Mean velocity (m/s) 0.679 1.77
d log ⎜ ⎟ Reynolds number 1,065 4,210
⎝ do − dw ⎠

and
τw ‡‡‡‡‡ Because the Reynolds number is multiplied by 2⁄3 in the turbulent flow
′ =
kann n′
. (4-156) equation, it would be tempting to use 2(NRe)AN/3 as the definition of the
⎛ 12 vlam ⎞ Reynolds number. Equations 4–136, 4–138, and 4–146 (i.e., the pipe-flow
⎜ d −d ⎟ equations) could then be used to describe laminar and turbulent flow in
⎝ o w⎠ a narrow concentric annulus. But this simple analogy does not work in the
transition regime.

Chapter 4 Rheology and Flow of Well Cement Slurries 135

 The Reynolds numbers are then compared to the crit- and
ical values given by Eqs. 4-148 and 4-151 (i.e., 2,675 and
3,575). At the lowest flow rate, the flow regime is lami- N He =
(
ρ do − dw τ y )2 . (4-160)
nar. At the highest flow rate, it is turbulent. One can now
( )
2
calculate the friction factor for the appropriate flow μp
regime (i.e., Eq. 4-157 for laminar flow and Eqs. 4-158
and 4-146 for turbulent flow). Once the friction factor is Figure 4-55 shows a similar graph for a Herschel-
known, the shear stress at the wall is derived from Bulkley fluid in which the Reynolds number is now
Eq. 4-135 and the friction pressure gradient defined as
from Eq. 4-17. This gives the following.
ρ( v )2− n ( do − dw )n
Flow rate (m3/s) 0.008 0.02 ( NRe ) HB = n
, (4-161)
Friction factor 0.0225 0.00749 ⎛ 2n + 1⎞
12 n−1 × ⎜ ×k
Wall shear stress (Pa) 6.75 14.0 ⎝ 3 n ⎟⎠
Friction pressure gradient (Pa/m) 709 1,473
and the generalized Hedström number as
For a flow rate of 0.016 m3/s, the Reynolds number is
equal to 3,012, and the flow regime is transitional. One 2
must therefore determine the friction factor at NRe1
( N He ) HB =
(
ρ do − dw )2 ⎛ τ y ⎞ n
⎜ ⎟ (4-162)
using Eq. 4-157 and at NRe2 using Eqs. 4-158 and 4-146. τy ⎝ k⎠
Then a linear interpolation is made on a log-log scale
between these two values to determine the value of the for a power-law index of 0.5.
friction factor at the calculated Reynolds number. The
wall shear stress and the friction pressure gradient are
determined as above, giving the following. Bingham Plastic Fluid
0.1
Flow rate (m3/s) 0.016
Friction factor at NRe1 0.00897
Friction factor at NRe2 0.0796 Fanning
friction 0.01
Friction factor 0.0854
factor
Wall shear stress (Pa) 10.2 0
10,000
Friction pressure gradient (Pa/m) 1,075 100,000
0.001 1,000,000
An alternative to these calculations is to use charts 100 1,000 10,000 100,000
like those presented in Fig. 4-46 for pipe flow. Of course, Bingham Reynolds number
the results are not likely to be as accurate.
Fig. 4-54. Friction factor/Bingham Reynolds number relationship as a
function of the Hedström number for a narrow concentric annulus.
Example of application to nonpower-law fluids
As stated earlier for pipe flow, the application of the
local-power-law method to nonpower-law fluids is not Herschel-Bulkley Fluid: n = 0.5
0.1
straightforward. To make the results more useful, they
can be presented in the form of master curves for spe-
cific rheological models. Figure 4-54 shows the relation- Fanning
ship between the friction factor and the Bingham plastic friction 0.01
Reynolds number for a series of Hedström-number factor
0
values where these dimensionless numbers are defined 10,000
as 100,000
0.001 1,000,000

( NRe ) BG =
(
ρv do − dw ) (4-159)
100 1,000 10,000 100,000
μp Herschel-Bulkley Reynolds number

Fig. 4-55. Friction factor/Herschel-Bulkley Reynolds number rela-

tionship as a function of the Hedström number for a narrow con-
centric annulus.

136 Well Cementing

 As with power-law fluids, it is now possible to deter- Pipe eccentricity plays a predominant role in the
mine the flow regime and the friction factor using Eqs. mud-circulation and mud-displacement processes
4-159 and 4-160, or 4-161 and 4-162 (for the specific case (Chapter 5). When the flow regime is laminar, the effects
of n = 0.5). One knows the annular diameters, the fluid of eccentricity on the flow of wellbore fluids in annuli
density, and the fluid’s rheological properties. Once the can be considered using numerical models and software
friction factor is known, the corresponding wall shear that quickly provide accurate answers. This section will
stress and friction pressure can be determined as describe these effects, starting from the laminar flow
described above for a power-law fluid. regime with Newtonian and non-Newtonian fluids, and
then attempting to extend the analysis to nonlaminar
flows.
4-6.3 Effect of pipe eccentricity Flow in eccentric annuli has been the subject of many
The previous discussion assumed that flow occurred in a experimental, theoretical, and modeling studies. Early
perfectly centered annulus. Of course, this is not realis- publications on Newtonian fluids clearly identified that
tic because gravity pushes the pipe downward. The pipe eccentricity in laminar flow leads to the following
resulting effect on fluid flow is significant. Consider a effects.
cross section of an eccentric annulus, shown in Fig. 4-56. ■ The velocity profile is distorted. The fluid flows faster
The eccentricity, ε, is defined as the ratio of the distance on the wide side of the annulus than on its narrow
between the centers of the cylinders to their radial dif- side (Fig. 4-57).
ference. The well completion industry also uses the pipe
standoff ratio, defined in Specification for Bow-Spring
Casing Centralizers (API Spec 10D, 2002) as
Standoff + 100% Standoff = 90% Standoff = 50%

)
RSTO = ⎡⎣(1 − ε × 100 ⎤⎦ . (4-163) 2.5 2.5 2.5

2.0 2.0 2.0

θ = 0, 2π 1.5 1.5 1.5

Power
Lmax = (ro – rw)(1 + ε) law 1.0 1.0 1.0

θ 0.5 0.5 0.5

ro
L 0 0 0

θ=
π rw B (
e = ε ro − rw) θ=
3π
2.5 2.5 2.5
2 2 2.0 2.0 2.0

L min = (ro – rw)(1 – ε) 1.5 1.5 1.5

Newtonian
Line of symmetry 1.0 1.0 1.0

θ=π 0.5 0.5 0.5

L max 0 0 0
L 2.5 2.5 2.5
L min
θ=0 θ = 2π
2.0 2.0 2.0
Fig. 4-56. Profile of the slot equivalent of an eccentric annulus (from
Iyoho and Azar, 1981). Reprinted with permission of SPE. 1.5 1.5 1.5
Bingham
plastic 1.0 1.0 1.0

When an annulus is concentric, its eccentricity is 0 0.5 0.5 0.5

and the pipe standoff ratio is 100%. Conversely, when the
inner cylinder is touching the wall of the outer cylinder, 0 0 0
the eccentricity is 1 while the pipe standoff ratio, RSTO, Fig. 4-57. Typical example of velocity profile on the narrow and wide
is 0%. sides of eccentric annuli for model fluids. Reprinted with permission
of SPE.

Chapter 4 Rheology and Flow of Well Cement Slurries 137

■ As eccentricity increases, friction pressure decreases
(Fig. 4-58).
■ The laminar flow regime ends earlier as eccentricity
increases (based on a bulk Reynolds number).
The effects of pipe eccentricity on the laminar flow of
non-Newtonian fluids are similar to those observed with
Newtonian fluids; however, they are more difficult to pre-
dict for fluids exhibiting a yield stress. Complete analyt-
ical and numerical solutions were presented for the flow (a) ε = 0 (b) ε = 0.25
of a Bingham plastic fluid in a narrow eccentric annulus
(Walton and Bittleston, 1991). For Herschel-Bulkley
fluids, most authors developed numerical solutions
based on a finite difference method. Figure 4-59 shows
examples of velocity profiles of a Herschel-Bulkley fluid
flowing in annuli of different eccentricities. It illustrates
that, for fluids exhibiting a yield stress, the distortion of
the velocity profile can be so severe that the fluid is
static on the narrow side of the annulus. This is because
the shear stress in the fluid is lower than the fluid yield
stress. (c) ε = 0.5 (d) ε = 0.75

Fig. 4-59. Velocity profile for a Herschel-Bulkley fluid as a function

of eccentricity. Outer diameter = 10 in. [254 mm]. Inner diameter = 5
ε = 0.00 ε = 0.6 in. [127 mm]. Flow rate = 200 gal/min [12.6 L/s]. Fluid rheological
ε = 0.1 ε = 0.7 parameters: n = 0.7; k = 0.00522 lbf-secn/ft2 [0.250 Pa-sn]; τy = 0.05
ε = 0.2 ε = 0.8 lbf/ft2 [2.4 Pa] (from Azouz et al., 1992).
ε = 0.3 ε = 0.9
ε = 0.4 ε = 1.0
ε = 0.5
1 A general conclusion can be drawn concerning veloc-
ity profiles. For equivalent flow geometries, distortion of
the velocity profile increases as a fluid becomes more
0.9 shear thinning. This leads to an interesting conse-
quence. The reader will recall that shear thinning fluids
have better circulation efficiency than Newtonian fluids
0.8 in a concentric annulus. The reverse is true for an eccen-
tric annulus. This effect is discussed in Chapter 5.
NRe The numerical models described above are used for
ffr × 0.7
24 predicting friction pressures. Every other condition
being the same, laminar flow friction pressures decrease
as pipe eccentricity increases. The largest reduction—
0.6
60%—is observed for Newtonian fluids when the pipe is
totally eccentered. In the case of power-law fluids,
0.5
Haciislamoglu and Langlinais (1990) solved the laminar
flow equations expressed in a bipolar coordinate system,
also using a finite difference method. They were able to
0.4 propose a simple method to account for eccentricity
0 0.2 0.4 0.6 0.8 1 effects. For power-law indices (n) between 0.4 and 1.0,
Radius ratio they found that the laminar flow friction pressures in
annuli with eccentricities (ε) from 0 to 0.95, standoff
Fig. 4-58. Friction factor × Reynolds number for a Newtonian fluid (RSTO) from 5 to 100%, and diameter ratio (α) from 0.3
flowing in an eccentric annulus as a function of the radial ratio and to 0.9 could be determined from concentric-annulus fric-
the eccentricity. The product is normalized by its value in a narrow tion pressures using a correction factor, Blam.
concentric annulus.

138 Well Cementing

 ⎛ d p⎞ ⎛ d p⎞ is assumed to slowly vary with azimuthal position; there-
⎜⎝ d z ⎟⎠ = Blam × ⎜ ⎟ (4-164) fore, results will be presented only for high-diameter
f − ecc
⎝ dz ⎠ f − conc ratios (i.e., dw/do ≤ 0.8).
The basic-slot model is not theoretically correct
⎛ ε ⎞
(
Blam = ⎜ 1 − 0.072 α 0.8454 ⎟ − 1.5 ε 2 n α 0.1852
⎝ n ⎠
) because it neglects shear in the azimuthal direction.
However, this model presents two main advantages.

( )
■ It is qualitatively correct (i.e., it clearly shows the
+ 0.96 ε 3 n α 0.2527 (4-165) principal effects of pipe eccentricity).
■ It is quantitatively correct for friction-pressure pre-
Blam takes a minimum value of 0.388 (theoretical dictions and to a lesser degree for velocity profiles,
minimum is about 0.4) for Newtonian fluids. provided the fluid is not highly non-Newtonian, the
Extending this analysis to nonlaminar flow regimes is annulus is relatively narrow, and the eccentricity is
a difficult task, and there is no consensus on the solu- not too high.
tions even for Newtonian fluids for the following reasons.
■ Velocity-profile distortion and friction-pressure
As mentioned earlier, the major effect of eccentricity
reduction in turbulent flow are less pronounced than is to distort the velocity distribution around the annulus.
in laminar flow. The flow favors the widest part of the annulus
(Fig. 4-58). Because both the velocity and the annular
■ As a consequence of the skewed velocity distribution,
gap vary azimuthally around the annulus, some local
the transition regime is extended. Compared to a con- parameters must be defined. For example, the local
centric annulus, laminar flow ends earlier, and turbu- Reynolds number for a given annular gap, L, is defined
lent flow begins later. by
Measurements of turbulent velocity profiles (Nouri
et al., 1993) have also highlighted the presence of
N Re ( L) =
ρ v( L )2− n′ (2 L)n′ , (4-167)
azimuthal secondary flows. Needless to say, the situation 12n ′–1 kann
′
is even less clear for non-Newtonian fluids. Using a
simple model, Haciislamoglu and Cartalos (1994) pro- where v(L) is the mean axial velocity along the local
posed a turbulent flow equation for power-law fluids annular gap L.
similar to Eq. 4-161, but that was based on the assump- First, situations are considered in which the fluid is in
tion that Bturb takes a minimum value of 0.6 for laminar flow all around the annulus (i.e., all local
Newtonian fluids. Reynolds numbers are smaller than the critical Reynolds
number NRe1). For power-law fluids, one can show that
⎛ ε ⎞
(
Bturb = ⎜ 1 − 0.048 α 0.8454 ⎟ − 1.5 ε 2 n α 0.1852
⎝ n ⎠
) the velocity distribution is a function of the annular
diameter ratio, the standoff, and the power-law index.

(
+ 0.285 ε 3 n α 0.2527 ) (4-166)
Because only narrow annuli are considered, the velocity
profiles depend on only two parameters, RSTO and n. To
assess the distortion of the velocity distribution caused
The transitional regime variation of B with the by eccentricity, it is worthwhile to calculate the ratio of
Reynolds number is even more complex. the average velocity along the widest and the narrowest
Because of this complexity, many authors use the annular gap (v–max and v–min) to the average velocity
basic-slot model (McLean et al., 1967; Iyoho and Azar, through the total section area (v– ). These two parame-
1981). In the basic-slot model, the eccentric annular ters are plotted in Fig. 4-60 as a function of API standoff
geometry is considered equivalent to a series of indepen- for three different power-law indices. As standoff
dent rectangular slots of varying heights (Fig. 4-56).§§§§§ decreases, the average velocity on the widest side first
For a fixed pressure gradient, the contribution of each increases and then levels off. On the narrowest side, it
angular sector to the flow rate is determined using the decreases quickly toward negligible values. In addition,
equations given in Appendix A. The reverse problem of when the power-law index is low, the distortion of the
calculating the friction pressure for a given flow rate is velocity distribution is worse.
then solved numerically. Thus, this model is based on a For fluids exhibiting a yield stress, the velocity reduc-
narrow-annulus approximation for which the annular gap tion on the narrow side can be even more pronounced
because the shear stress may be lower than the yield
stress of the fluid, implying that the local velocity of the
§§§§§ Forfluids exhibiting a yield stress, this approximation introduces errors
fluid is 0. For Bingham plastic fluids in particular, the
that lead to an incorrect description of plug flow on the wide side of the dimensionless parameters relevant to the velocity distri-
annulus (Walton and Bittleston, 1990).

Chapter 4 Rheology and Flow of Well Cement Slurries 139

 2.6 1.0
2.4 n = 1.0
0.9 n = 0.5
2.2
n = 0.2
2.0 0.8
1.8
1.6 Pressure 0.7
Velocity 1.4 gradient
1.2 ratio 0.6
ratio v max /v n = 1.0
1 v max /v n = 0.5 0.5
0.8 v max /v n = 0.2
0.6 v min /v n = 1.0 0.4
0.4 v min /v n = 0.5
0.2 v min /v n = 0.2 0.3
0 0 10 20 30 40 50 60 70 80 90 100
0 10 20 30 40 50 60 70 80 90 100 API standoff ratio
API standoff ratio Fig. 4-61. Eccentric annulus/concentric annulus friction-pressure
ratios (dw /do = 0.8) for different power-law indices calculated using
Fig. 4-60. Ratio of average velocities along the wide and narrow
the basic-slot model—laminar flow. Reprinted with permission of
side of the annulus to the average velocity. Calculations are per-
SPE.
formed for different power-law indices using the basic-slot model in
laminar flow. Reprinted with permission of SPE.

bution are the same as for a power-law fluid, except ■ An average critical Reynolds number (NRe1)ecc, at
that the power-law index is replaced by the dimension- which laminar flow ends on the wide side of the annu-
less shear rate, ξ. The effect of ξ on the velocity distri- lus, represents the maximum average Reynolds
bution is similar to that of the power-law index (i.e., the number for the fluid to be in purely laminar flow all
lower the value of ξ, the higher the velocity distribution around the annulus.
distortion). A critical dimensionless shear rate (ξ crit) ■ An average critical Reynolds number (NRe2)ecc, at
for the fluid to start flowing on the narrow side of the which the fluid begins to be in full turbulent flow on
annulus can be defined (Chapter 5). the narrow side of the annulus, represents the mini-
Pipe eccentricity also affects the friction- mum average Reynolds number for the fluid to be in
pressure/flow rate relationship. A typical example of full turbulent flow all around the annulus.
friction-pressure reduction caused by eccentricity is
These two parameters, normalized to the correspond-
shown in Fig. 4-61. The ratios of eccentric-annuli friction
ing values for a concentric annulus (NRe1 and NRe2) are
pressures to the corresponding concentric-annulus fric-
plotted in Figs. 4-62 and 4-63 for power-law fluids with
tion pressures are plotted against standoff for a power-
three different power-law indices. These curves are typ-
law fluid with n = 1.0, 0.5, and 0.2. For narrow annuli,
ical of any nonelastic, shear-thinning fluid. They show
the relative friction-pressure reduction depends on the
that, as standoff decreases, the average Reynolds
power-law index for power-law fluids and on the dimen-
number range in which both flow regimes coexist
sionless shear rate for Bingham plastic fluids.
becomes increasingly wider. The flow regime begins to
Theoretically, the normalized friction pressure can vary
be turbulent on the wide side of the annulus earlier than
between 1.0 and slightly less than 0.4 for shear thinning
would be expected from concentric flow calculations.
fluids.
The fluid remains in laminar flow on the narrow side of
As mentioned earlier, the uneven velocity distribution
the annulus later than expected from concentric flow
in eccentric annuli affects the transition to turbulent
calculations. Notice also that (NRe1)ecc is much more
flow. Because the Reynolds number depends on the local
dependent on the power-law index than (NRe2)ecc.
average velocity and the annular gap, parameters that
When a given fluid is in turbulent flow all around the
both vary around the annulus, turbulence is likely to
annulus, the velocity distribution is less distorted. In
appear first at the widest point in the annulus and
addition, the friction-pressure reduction caused by
extend all around the annulus as the flow rate increases.
eccentricity is less in turbulent flow than in laminar flow
Consequently, laminar flow and turbulent flow regimes
(Figs. 4-64 and 4-65).
can coexist in a given eccentric annulus. Using the basic
slot model, one can define the following.

140 Well Cementing

 1.0 1.4
n = 1.0
0.9 n = 0.5 1.2
0.8 n = 0.2
Reynolds 1
0.7
number 0.8
ratio 0.6 Velocity

⎢
( )
⎡ NRe1 ecc ⎤
⎥
0.5 ratio 0.6
v max /v n = 1.0
⎢ NRe1 ⎥ 0.4 0.4 v max /v n = 0.5
⎣ ⎦
v max /v n = 0.2
0.3 v min /v n = 1.0
0.2
0.2 v min /v n = 0.5
0 v min /v n = 0.2
0.1
0 10 20 30 40 50 60 70 80 90 100
0 10 20 30 40 50 60 70 80 90 100
API standoff ratio
API standoff ratio

Fig. 4-62. Maximum normalized average Reynolds numbers for dif- Fig. 4-64. Ratio of average velocities on the wide and narrow sides
ferent power-law fluids to be in laminar flow around an eccentric of the annulus to the total average velocity for different power-law
annulus (dw /do = 0.8). Reprinted with permission of SPE. indices. Values are calculated using the basic-slot model in turbu-
lent flow.

1.0 1.0
n = 1.0 n = 1.0
0.9 n = 0.5 0.9 n = 0.5
n = 0.2 n = 0.2
0.8
0.8
0.7
Reynolds
number 0.6 Pressure 0.7
ratio gradient
0.5 0.6
( )
⎡ NRe 2 ecc ⎤
⎢ ⎥ 0.4
ratio

⎢ NRe 2 ⎥ 0.5
⎣ ⎦ 0.3
0.2 0.4

0.1 0.3
0 10 20 30 40 50 60 70 80 90 100 0 10 20 30 40 50 60 70 80 90 100
API standoff ratio API standoff ratio

Fig. 4-63. Minimum normalized average Reynolds numbers for dif- Fig. 4-65. Eccentric annulus/concentric annulus friction-pressure
ferent power-law fluids to be in turbulent flow around an eccentric ratios (dw ⁄ do = 0.8) for different power-law indices calculated using
annulus (dw /do = 0.8). the basic-slot model in turbulent flow.

As stated earlier, although the results presented ■ friction pressures play a significant role, for example
above have not been fully validated quantitatively, their U-tubing prediction in relatively small annuli
trends provide qualitative information concerning the (Chapters 5 and 12) or friction pressures in slim holes
effects of pipe standoff on the flow of wellbore fluids in ■ the velocity distribution plays a significant role (e.g.,
the annulus. One can conclude that the effect of eccen- mud circulation).
tricity should be taken into account when

Chapter 4 Rheology and Flow of Well Cement Slurries 141

4-7 Conclusions 4-8 Acronym list
Accurate and reliable rheological characterization of oil- API American Petroleum Institute
well cement slurries still presents a problem for the ISO International Organization for Standardization
industry. These fluids exhibit complex rheological behav-
SVF Solid volume fraction
ior that depends not only on their composition and the
mixing procedure (as explained in other chapters), but
also on shear history, temperature, and the testing pro-
cedures.
Unless extreme care is taken, the coaxial-cylinder vis-
cometer, most commonly used to measure the rheologi-
cal properties of cement slurries, can suffer from a
number of limitations resulting from particle migration,
end effects, or slippage at the rheometer wall. Similar
problems have been encountered with other types of
equipment (e.g., pipe-flow rheometers). Therefore, coax-
ial-cylinder and vane rheometers remain the most useful
instruments for characterizing cement slurries today. It
is clear that research is needed to improve testing equip-
ment and procedures.
Equations describing the flow of cement slurries in
pipes and annuli have focused on three widely used rhe-
ological models—the power-law, Bingham plastic, and
Herschel-Bulkley models. These models’ limitations
have also been discussed, and some more realistic alter-
natives presented. Because of the lack of experimental
data, the validity of many of these equations (at least in
turbulent flow) is not yet fully established. Further
progress in this domain depends on acquiring better
characterizations of the rheological properties of
cement slurries.

142 Well Cementing

5-1 Introduction
Mud removal has been a subject of intense interest in
the well-cementing community for many years because
Mud Removal
Gérard Daccord, Dominique Guillot, and
Frederik Nilsson—Schlumberger

■
5
Mud displacement by completion fluids has yet
another objective: to remove the maximum amount of
suspended solids in the wellbore fluid. The result is
of its effect on cement quality and zonal isolation. The evaluated by monitoring of the clarity of the returned
main objective of a primary cement job is to provide fluid.
complete and permanent isolation of the formations In all the above three domains—replacement of a
behind the casing. To meet this objective, the drilling drilling fluid by a cement slurry, another drilling fluid, or
mud and the preflushes (if any) must be fully removed a completion fluid—the outlined technical objective is
from the annulus, and the annular space must be com- balanced by economic considerations, which include the
pletely filled with cement slurry. Once in place, the rig time required for the displacement operations, the
cement must harden and develop the necessary cost of the cleaning products, and the eventual loss of
mechanical properties to maintain a hydraulic seal recovered fluid because of intermixing.
throughout the life of the well. Therefore, good mud Mud removal before cementing is closely related to
removal and proper slurry placement are essential to the borehole quality resulting from the drilling phase.
obtain well isolation. From a chronological point of view, mud removal begins
Incomplete mud displacement can leave mud chan- when drilling operations are completed.
nels or mud layers on the walls across the zones of inter- 1. The well is circulated and cleaned before and while
est, thereby favoring interzonal communication. the drillstring is pulled out. This process is known as
Therefore, bonding of the cement to the pipe and for- mud conditioning.
mation, as well as cement-seal durability, are affected by
the efficiency of the displacement process. 2. The well is logged. During this process, the mud is
Mud removal in cementing operations is not funda- mainly static.
mentally different from mud-to-mud displacement or 3. Casing is run in and the mud is conditioned again.
mud-to-completion fluid displacement, although the 4. Mud displacement and cement placement begin.
means and objectives are slightly different.
Krause (1986) calls the operations before mud dis-
■ For a cementing engineer, the heart of the mud-
placement and cement placement “predisplacement”
removal process consists of optimizing casing cen- steps. In this text, these steps are discussed in
tralization, selecting the sequence of fluids, deter- Section 5-2 (Well preparation).
mining the volume and properties of each of the Research concerning the cement-placement process
fluids, and to some extent determining the pumping began in the 1930s. Some key factors influencing pri-
rate. Often, these are the only variables the engineer mary cement job failures were identified, and solutions
can control. Ideally, after determining the optimal were proposed as early as 1940. Using a large-scale sim-
fluid properties, one can determine the correct fluid ulator, Jones and Berdine (1940) showed that poor zonal
compositions. In the end, the final objective is to isolation could be attributed to channeling of the
achieve the most efficient zonal isolation. Success is cement slurry through the mud, a phenomenon they
most often evaluated using cement logs (Chapter 15). found to be promoted by casing eccentricity. They also
■ Mud displacement while drilling has a different identified that a residual mud filtercake at the
objective: to replace one drilling fluid with another cement/formation interface was caused by poor mud dis-
with a minimum amount of intermixing. Also, the placement. To minimize cement channeling, Jones and
conditions are different (e.g., the pipe- to hole-diam- Berdine (1940) proposed casing centralization. They
eter ratio is much smaller), and the evaluation also found effective ways to remove the mud filtercake,
method is direct: One monitors the amount of mixed including fluid jets, scrapers or scratchers, casing recip-
fluid recovered on the surface. rocation, and pumping acid ahead of the cement slurry.

Well Cementing ■ Chapter 5 Mud Removal 143

 Since those early days, many theoretical, experimen- filtercake deposition and erosion). Usually, the models
tal, and field studies have been performed to better assume that the fluids are separated by a sharp interface
understand mud displacement. This is necessary partly with no interfacial mixing.
because of the increasing complexity of the problem It is worthwhile to mention that attempts to model
(e.g., deeper wells and deviated wells), together with the interfacial instabilities caused by density or viscosity dif-
development of new knowledge. Unfortunately, the ferences are still at an early stage. They are often limited
experimental and theoretical approaches suffer from to two dimensions and are valid only for Newtonian
severe limitations. At first glance, a theoretical approach fluids. Incidentally, such instabilities have been
seems attractive, because there are major drawbacks observed experimentally by Tehrani et al. (1993), and
associated with experimental devices. they can lead to enhanced mud removal (Section 5-3.6).
■ One of the key parameters in the mud-removal In the last 10 years, progress has been made by combin-
process, the ratio of casing length to the annular gap, ing experimental and theoretical studies (Vefring et al.,
is difficult to reproduce. In the laboratory, one is typ- 1997; Biezen et al., 2000; Frigaard et al., 2001) with
ically limited to a ratio of about 500, while in the field attempts to apply new models developed in other indus-
the order of magnitude of this parameter is 104. This tries. For example, hole-cleaning models have been
prevents observation of, for example, axial deforma- adapted to mud-removal situations (Nguyen, 1997).
tion of fluid interfaces owing to eccentricity on a long Despite these technical advances, the accuracy of pre-
length scale. Therefore, the effect of gravity in eccen- dictions is still limited. Predictions are expressed in
tric annuli cannot be fully investigated. One may terms of chance of success, which is a clear acknowledge-
argue that, in theory, the casing-length/annular-gap ment that all of the phenomena are not under control.
ratio could be maintained with an experimental Field data supporting successful cementing practices
device (reducing the annular gap to a very small are cited in this chapter. However, it is difficult to quan-
value); unfortunately, this is not practical because tify the success of a job for the following reasons.
dynamic similarity requires that at least six dimen- ■ Controlling and monitoring an actual cement job is
sionless parameters be matched to the corresponding not as easy as in a laboratory experiment.
field values. ■ The properties of fluids mixed in the field are often
■ Second, in view of the number of parameters not measured, and they can be quite different from
involved, an experimental displacement-efficiency laboratory-mixed systems. Similarly, the wellbore
investigation over the complete dimensionless para- conditions are often not accurately known.
meter space would represent an enormous amount of ■ There is no direct method to measure mud removal,
work. so indirect evaluation techniques are used.
Therefore, one must be careful when attempting to Interpretation of the measurements is often not
extrapolate experimental results outside of the condi- straightforward (Chapter 15).
tions within which they were obtained. It is also impor- Remaining technical challenges include developing a
tant to mention that some of the key parameters (e.g., better quantification of the effects of casing movement
rheology) are difficult to measure (Chapter 4). In addi- during placement, chemical interactions between fluids,
tion, in most published experimental studies in which and unstable flow situations. Although it is possible to
cement was used as the displacing fluid, little information predict the fluid properties that will provide optimal
is available regarding its compatibility with the displaced mud displacement, it remains a tedious process to for-
mud. Compatibility may strongly affect the results. mulate such fluids.
The theoretical approach also has its limitations. The The link between fluid performance in the laboratory
complete modeling of the displacement process pre- and the field also requires further study to develop labo-
sents a formidable task, even for the most sophisticated ratory procedures that are representative of downhole
computers. For example, one must contend with conditions. Such procedures would assess the effective-
unsteady mass and momentum transfer between differ- ness of a proposed solution. Of course, any new proce-
ent non-Newtonian fluids in an asymmetric geometry. At dures must be accepted by industry standardization
present, the best models for field use are based on a two- organizations such as the American Petroleum Institute
dimensional (2D) representation of annular displace- (API) and the International Organization for
ment (Bittleston et al., 2002). More complex three- Standardization (ISO) (Appendix B).
dimensional (3D) models (King et al., 2000) remain Despite all of the problems cited above, a few areas of
limited by computational power. In addition, some of the industry consensus have emerged regarding the optimal
key parameters cannot yet be modeled well (e.g., chem- design of a primary cement job for successful mud
ical interactions between fluids and static and dynamic

144 Well Cementing

removal and cement placement. This chapter distin- washed-out pockets tend to hold gelled or dehydrated
guishes areas of consensus from those of controversy. mud that may be dragged out by the cement slurry, con-
The organization of the chapter is outlined below. taminating the cement column above. While good
1. Well preparation. This section focuses on mud condi- drilling practices do not guarantee a successful cement
tioning and circulation. This is a key point, because job, they may prevent a failure.
the success of a primary cement job largely depends Although it is understandable that the objective for
on the suitability of the wellbore for cementing. the drilling engineer is to drill the well safely and as
quickly and economically as possible, this should be
2. Mud displacement. This section discusses various
accomplished bearing in mind that one of the ultimate
methods and important considerations in removing
goals of drilling is to provide an optimal wellbore for sub-
mud.
sequent operations. In summary, well operations and
a. The ideal case of mud displacement in a concen- especially cementing require “a team effort that must
tric annulus between two impermeable walls (the include drilling management, drilling operations,
so-called mobile mud) drilling engineering, rig contractor, and service person-
b. Effects of eccentricity nel. (It is not) a process that starts when the pipe is on
c. The difficult problem of removing the immobile bottom. It must start during the drilling of the hole to be
mud, which can easily be bypassed by the displac- cemented” (Sauer, 1987).
ing fluid
d. Casing movement strategies to overcome some of 5-2.1 Borehole quality
the previously described problems and contribute A borehole in good condition for cementing operations
to the success of critical primary cementing jobs has the following attributes:
3. Drilling fluids, spacers, and washes. This section ■ controlled subsurface pressures
outlines the basic properties and chemistries of
■ a smooth wall with doglegs of low severity
drilling fluids, spacers, and washes, and why they are
required ■ in-gauge (i.e., the expected diameter according to the

4. Other cement placement problems. This section dis- bit size)

cusses the mixing of fluids as they travel down the ■ stable (i.e., free of formation encroachment or

casing when no mechanical plugs are used and con- spalling)

tamination of the shoe once cement is placed ■ clean (i.e., free of cuttings)

5. Qualitative recommendations. The chapter con- ■ a correctly treated and mobile mud that will deposit
cludes with qualitative recommendations for achiev- thin filtercakes in front of permeable zones.
ing successful mud removal and cement placement.
Unfortunately, such an ideal situation cannot always
be achieved. Therefore, cement-placement techniques
often must be designed to minimize the influence of
5-2 Well preparation poor well preparation.
C.W. Sauer (1987), in his review on the state of the art of Failure to achieve these requirements may have the
mud displacement, noted that events occurring during following consequences:
the drilling and casing phases will affect the cement job. ■ differential sticking problems while logging or
“It should not be believed that the cement job should go running the casing
all right regardless of what else is or has taken place ■ high drag forces, preventing the running of casing
during the drilling to casing point.” to the planned depth
A poorly drilled hole may have large dogleg sections,
■ influxes or losses while running the casing
preventing engineers from running casing with the
proper number of centralizers. Crooked holes make ■ poor mud removal resulting in poor cement

casing centralization difficult; consequently, the removal placement.

of the mud from the narrow side of the annulus is prob- Some of these requirements are highly dependent
lematic. Poorly treated mud could induce washouts, upon the drilling method. For example, traditional direc-
thick filtercakes, or settled solids beds, which would be tional drilling in sliding or rotating mode has a tendency
difficult or impossible to remove. The hole may also have to lead to helicoidal or undulating trajectories. The more
several washed-out zones that are difficult to clean out, recent techniques of rotary steerable drilling (Schaaf et
regardless of the displacement rate. Furthermore, these

Chapter 5 Mud Removal 145

al., 2000), together with optimized drilling bits—which that the annulus still contains undisplaced Fluid No. 1
drill with less instability (Mensa-Wilmot and Stacey, (Fig. 5-2). There is no restriction in using this definition.
2001)—provide much smoother and easy-to-clean well- It applies to any situation, including, for example, when
bores (Downton et al., 2000; Gaynor et al., 2001). the casing is moved.
The most common method to clean cuttings out of the
wellbore is to periodically pump viscous fluids while
Fluid No. 1
rotating the pipe at 150 to 200 rpm, if possible (Sewell
and Billingsley, 2002). It is also important to remove vis- (d)
cous pills before cementing. This method has emerged
from field practices rather than from physical or phe-
nomenological studies. Even if the borehole is clean z=1
(c)
after the drillstring has been pulled out, special care
must be taken to maintain a good-quality hole until the
beginning of cementing operations. Long static periods
favor barite settling in deviated sections and the devel- (b)
opment of thick mud filtercakes that are difficult to
remove and that impair running the casing.
Assessing the quality of the hole before running the
casing requires caliper and survey logs. A thick mud fil-
tercake can be misidentified as an undergauge hole (Al
Khayyat et al., 1999).

5-2.2 Circulation and displacement efficiency

The most common parameter to define the ability of a
fluid to displace another is the displacement efficiency. (a)
z=0
When the same fluid is used to displace itself, this para- Fluid No. 2
meter is called the circulation efficiency.
Consider an annulus of volume Vann and length L that
is filled with Fluid No. 1 (the fluid to be displaced) flow- (a) t* = 0
ing at a given volumetric rate q (Fig. 5-1). At time t = 0, (b) t* < t*break (before breakthrough)
(c) t* = t*break (at breakthrough)
Fluid No. 2 (the displacing fluid, which may be fresh (d) t* > t*break (after breakthrough)
mud) suddenly replaces Fluid No. 1 at the inlet of the
annulus (Z = 0). At any time t > 0, the displacement effi- Fig. 5-1. Schematic diagram of interface profiles at different times
ciency ηdisp is the fraction of annular volume occupied during the displacement of Fluid No. 1 by Fluid No. 2.
by Fluid No. 2. In other words, for case (d) in Fig. 5-1, the
displacement efficiency would be 1 – the hatched area
divided by the area between z = 0 and z = 1. The natural
time scale that allows one to define a nondimensional
1
time, t*, is the ratio of the annular volume, Vann, to the
flow rate, q.
Displace-
q ment
t* = t × (5-1) efficiency (d)
Vann
(c)
(b)
This dimensionless time is equal to the number of
annular volumes pumped. Notice that, with these defin-
itions, ηdisp is equal to t* [case (b)] until Fluid No. 2 first 1 2
appears at the outlet of the annulus [case (c)]. This is - (a) Number of annular volumes
defined as the breakthrough time, t*break. After break- Fig. 5-2. Schematic of a displacement-efficiency curve, with an
through [case (d)], ηdisp approaches but never reaches asymptotic value equal to unity.
a constant value that may be less than 1, indicating

146 Well Cementing

 The circulation efficiency may easily be monitored in small proportion of the total flow area. This situation is
the field. For example, one can use carbide pills (Section illustrated in Fig. 5-3. Around most of the casing the dis-
5-2.4.7). Or, for simple situations, numerical models may placement is very efficient; however, there is a continu-
be used to predict the evolution of the efficiency as a ous mud channel. In addition, a thin layer of mud may
function of the pumped volume. A difference between remain immobile on the formation and casing walls.
the measured and predicted values is that the measured Again, the corresponding volume of fluid will be small,
efficiency considers the internal casing or drillstring but this can affect cement logs significantly if the mud is
volume plus the annular volume, while the models are in contact with the casing.
limited to the annular volume. This difference is usually In Fig. 5-3, each rectangle at the upper right repre-
negligible, as mixing and fingering inside of tubulars is sents an annulus at a given time. It is produced by math-
insignificant compared with that observed in the annu- ematically “unwrapping” the annulus. Each set of five
lus. During cementing, when wiper plugs are used, no rectangles represents a progression in the circulation
mixing is possible inside the tubulars unless their pre- from 0.2 to 1 annular volume. The set on the left shows a
mature wear leaves a layer of mud on the casing walls single non-Newtonian fluid displacing itself. The set on
(Chapters 11 and 13). the right shows an 8.33-lbm/gal [1,000-kg/m3] fluid dis-
The circulation or displacement efficiency is a clear placed by an 11.7-lbm/gal [1,400-kg/m3] fluid. The den-
and simple concept; however, it can sometimes be mis- sity contrast induces azimuthal flows, and the displace-
leading, especially for high eccentricities. Under such ment is significantly improved, although some fluid is
circumstances, a large angular channel of bypassed mud still left behind on the narrow side.
in the narrow part of the annulus would correspond to a

Single fluid displacing itself Two fluids of different densities

Unwrapping an eccentric annulus Mud Cement

Lmin Lmax Lmax

Lmin

z z

0.2 0.4 0.6 0.8 1.0 0.2 0.4 0.6 0.8 1.0
Annular volumes circulated Annular volumes circulated

Bypassed
mud

Fig. 5-3. Schematic diagrams of bypassed mud in an eccentric annulus and a photograph of an actual mud
channel between two casings (Bittleston and Guillot, 1991; Piot, unpublished diagrams and photograph).

Chapter 5 Mud Removal 147

5-2.3 Mud conditioning run continuously, the velocity is not constant, and iner-
Drilling muds are designed to facilitate drilling opera- tial forces also contribute to the annular pressure.
tions and provide proper cuttings transport. They are not Mathematical models for calculating the associated
necessarily conducive to efficient mud displacement, pressure surges can be found in the literature (Mitchell,
logging, or completion operations. Therefore, it is often 1988; Zamora et al., 2000).
necessary to condition the mud (i.e., to modify its prop- The mud should be circulated after the casing is in
erties). Before placing cement in the wellbore, two mud place, because the well may have been static for a long
characteristics can be changed—density and rheology. period, allowing the mud to gel or build a filtercake. The
The required adjustments vary according to the particu- minimum circulation volume should be at least “bottoms
lar situation. This issue will be discussed in detail later. up” (pumping one annular volume; the internal volume
It is generally desirable to reduce the mud density of the tubulars is ignored), and preferably greater. Mud
without compromising well control. Reducing the mud’s circulation is also beneficial because it
gel strength, yield stress, and plastic viscosity is also ■ helps clean cuttings out of the hole

beneficial. Doing so reduces the driving forces necessary ■ ensures that gas flow is not occurring
to displace the mud and increases mud mobility. Of ■ helps detect any gas flow into the well
course, these steps require prior removal of the cuttings
■ homogenizes the mud after treatment on the surface
from the borehole and the drilling fluid. When the cut-
tings are removed from the drilling fluid, its yield point ■ reduces the yield stress and plastic viscosity because

usually decreases. One must be careful to prevent set- most drilling muds are thixotropic
tling of weighting agents (e.g., barite); otherwise, con- ■ erodes the gelled or dehydrated mud that is trapped
trol of mud density is lost. This may represent a major in washouts, on the narrow side of an eccentric annu-
constraint for highly deviated holes (Crook et al., 1987) lus, and at the walls of permeable formations
in terms of minimum values for the low-shear-rate rhe- (because the shear stress exerted by the flowing mud
ology. is usually less than the shear strength of the mud
Once the mud is clean, its rheology can be further film, this erosion is only possible if the hydraulic
modified by adding dispersants, water, or base oil (which action of the fluid is supplemented by casing move-
also reduces the density). It is necessary to circulate the ment, especially when scratchers are installed on the
mud until its rheological properties fall within the casing [Chapter 11]).
desired range.
This requires circulation for at least one hole volume, Unfortunately, at this stage, operators commonly per-
which ideally should be performed before removing the form only mud conditioning.
drillpipe. Otherwise, unconditioned mud may have suffi- If cuttings, or gelled or dehydrated mud, are scraped
cient time to gel while the crew is removing drillpipe, into the mud while running the casing, an excessive
logging, and running casing. pressure buildup can occur when circulation is resumed.
Moving the drillstring during conditioning aids the Therefore, it is often desirable to circulate the annulus
displacement of gelled mud and helps keep the cuttings at intermediate depths before the casing reaches the
suspended. High-speed pipe rotation is particularly ben- bottom of the hole.
eficial in horizontal sections because of the orbital These qualitative recommendations are only margin-
movement of the string, while in deviated sections, reci- ally helpful for the completion engineer, who must
procation allows the string to move up and down the design the mud-circulation phases before removing the
hole (Al Khayyat et al., 1999; Section 5-2.4.4). drillpipe and after the casing is in place. For a better
The running of casing should be performed carefully understanding of the relative importance of each para-
to avoid fracturing the formations. The equivalent flow meter, it is possible to use mathematical models that
rate in the annulus (qann) as a function of the speed at predict mud circulation. However, continuous monitor-
which the casing is run (vrun) is given by the following ing of mud circulation remains the best method to verify
equation: adequate mud conditioning.

qann = vrun × Apipe , (5-2)

5-2.4 Modeling mud circulation
where Apipe = cross sectional area of pipe. Modeling mud circulation is a fluid-mechanics problem
A quick calculation shows that these rates are not that can be solved without making too many assump-
negligible. For example, a 7-in. outside diameter (OD) tions; therefore, the predictions have practical value.
pipe run at 3 ft/s [1 m/s] gives rise to an annular flow The model describes the isothermal flow of an incom-
rate of 8.6 bbl/min [23 L/s]. Because the casing is not pressible and inelastic fluid in the annulus between two

148 Well Cementing

pipes. The principal part of the solution is the fluid- Bingham plastic fluids, it also depends upon a single
velocity profile calculation, which accounts for the cor- parameter, the dimensionless shear rate, ξ, given by
rect rheological model and boundary conditions. Walton
and Bittleston (1991) calculated the flow of a fluid with 12v ⎛ μ p ⎞
a yield stress in a narrow eccentric annulus, allowing the ξ= ⎜ ⎟, (5-3)
identification of various flow regimes as the flow rate do − dw ⎝ τ y ⎠
increases (Fig. 5-4).
Bittleston and Hassenger (1992) considered the flow where
of viscoplastic fluids in a concentric annulus during do = outer diameter of annulus
casing rotation. Barton et al. (1994) presented a flow dw = inner diameter of annulus
simulation of a Herschel-Bulkley fluid in an eccentered v– = average fluid velocity
annulus during casing rotation or reciprocation. These μp = plastic viscosity
models are briefly introduced below, along with the τy = yield stress of the Bingham plastic fluid.
effects of flow regime, eccentricity, and casing move- For Herschel-Bulkley fluids, two parameters are
ment. involved—the power-law index and a dimensionless
The effects of gelled mud and mud filtercake are dis- shear rate now defined as
cussed; however, the models do not consider them
1/ n
12 v ⎛ k ⎞
directly. Nevertheless, the models are valid in a qualita-
tive sense, and they allow one to consider the worst-case ξ= ⎜ ⎟ (5-4)
scenario. The notations used in this chapter are the do − dw ⎝ τ y ⎠
same as those defined in Chapter 4, where the basic flow
equations are presented. where
k = consistency index
τy = yield stress of the Herschel-Bulkley fluid.
5-2.4.1 Laminar flow in a concentric annulus
Notice that the breakthrough time corresponds to the
In laminar flow, the circulation efficiency can be calcu-
ratio of the average velocity to the maximum velocity. As
lated by tracking marked particles. This is done by com-
explained in Chapter 4, this value is equal to 2⁄3 for
puting the velocity profile and solving the streamline
Newtonian fluids in a slot and is higher for shear-thin-
equations. For a concentric annulus, the circulating effi-
ning fluids. After breakthrough, the efficiency
ciency is directly derived from the nonzero velocity com-
approaches 100%, a value that theoretically can be
ponent described in Chapter 4. The data given in this
reached at infinite time, because the velocity of the fluid
section were calculated using either the rectangular-slot
particles at the annular walls is assumed to be zero (no
approximation (Chapter 4 and Appendix A, Table A-8) or
wall slip).
the finite-element solution of the true annular flow. In
Figure 5-5 also shows that, as a fluid becomes more
Fig. 5-5, the circulation efficiency is plotted versus the
shear-thinning (i.e., as the power-law index, the dimen-
number of annular volumes. For power-law fluids, the
sionless shear rate, or both become smaller), circulation
curve depends only upon the power-law index, n. For

Increasing flow rate

No flow Laminar flow Turbulent flow

Fig. 5-4. Flow regimes for fluids with yield stress flowing in an eccentered annulus as the flow rate
is increased (Walton and Bittleston, 1991). Copyright Cambridge University Press.

Chapter 5 Mud Removal 149

 1.00
0.95
0.90
0.85
0.80
Circulation 0.75
efficiency
0.70
0.65
0.60 n = 0.3
n = 0.5
0.55 n = 0.8
0.50 n = 1.0
0.5 1 1.5 2.0 2.5 3.0
Number of annular volumes

1.00
0.95
0.90
0.85
0.80
Circulation 0.75
efficiency
0.70
0.65
0.60
0.55 ξ = 0.18
0.50 ξ = 4.00
0.5 1 1.5 2.0 2.5 3.0
Number of annular volumes

1.00
0.95
0.90
0.85
0.80
Circulation 0.75
efficiency
0.70
0.65
0.60 q = 1.9 bbl/min
0.55 q = 5.7 bbl/min
0.50 q = 9.4 bbl/min
0.5 1 1.5 2.0 2.5 3.0
Number of annular volumes

Fig. 5-5. Circulation-efficiency plots for various fluids in a narrow concentric annulus. Top: Circulation efficiency for
Newtonian and power-law fluids. Middle: Circulation efficiency for Bingham plastic fluids for various dimensionless
shear-rate values. Bottom: Circulation efficiency for a Herschel-Bulkley fluid at different flow rates (in bbl/min).

150 Well Cementing

will become more efficient. A circulation efficiency of 0.94
100% at the breakthrough time would occur with an Fluid 1
Fluid 2
entirely flat velocity profile (which is equivalent to a 0.93
power-law index or a dimensionless shear rate of zero).
As shown in Table 5-1, the dimensionless shear-rate Circulation 0.92
values in Fig. 5-5 correspond to typical field conditions. efficiency
High dimensionless shear-rate values correspond to a 0.91
low yield-stress/wall-stress ratio, while low values are
close to plug-flow conditions. In this table, ψ is the 0.90
dimensionless shear stress that is equal to the ratio of 0 2 4 6 8 10
the fluid yield stress to the wall shear stress. Flow rate (bbl/min)
It is also important to point out that, for power-law
fluids, circulation efficiency does not depend on the flow 14
rate. However, for fluids with a yield stress, the flow rate
12
is important. Thus, everything else being equal, the cir-
culation efficiency in a concentric annulus improves 10
under the following conditions. 8
Shear
■ Decreasing average velocity stress 6
■ Increasing annular-gap size
4
■ Decreasing μp/τy or k/τy ratio (Fig. 5-6)
2 Fluid 1
0 Fluid 2
0 200 400 600 800 1,000
5-2.4.2 Turbulent flow in a concentric annulus
Shear rate (s–1)
The flatter average velocity profiles that result from tur-
bulent flow (Chapter 4) generally give much higher cir- Fig. 5-6. Variation of circulation efficiency for an annular volume
culation efficiencies than those for laminar flow. of 1 versus flow rate for two Herschel-Bulkley fluids in a centered,
However, calculating the circulation efficiencies is much narrow concentric annulus. Rheograms of the fluids are shown on
the bottom plot. Fluid 1: k = 0.02 Pa-s–n, n = 0.9, τy = 2 Pa. Fluid 2: k
more complicated. A detailed discussion is beyond the = 0.03 Pa-s–n, n = 0.8, τy = 3 Pa.
scope of this chapter. The interested reader is invited to
review textbooks by Schlichting (1979) and Nauman and
Buffham (1983).
annulus is not centered, the velocity distribution around
the annulus is distorted, and the flow favors the wider
5-2.4.3 Influence of string eccentricity side. This may lead to unusual situations in which the
Although this section discusses the effect of eccentricity flow regime can be laminar on the narrow side of the
on circulation efficiency, it also addresses the effects of annulus and turbulent on the wide side (Fig. 5-4),
asymmetric flow geometry caused by oval holes. The because the local Reynolds number varies azimuthally
qualitative effect of casing eccentricity on velocity pro- around the annulus (see Eq. 4-131 for the definition of
files and pressure gradients was presented in Chapter 4, the local Reynolds number).
using the basic-slot model. When the inner pipe of an

Table 5-1. Parameters of the Simulations Shown in Fig. 5-5, Middle Plot
Hole Size, Pipe Outside Fluid Plastic Fluid Yield Flow Rate, q Velocity, v– Dimensionless Dimensionless
do Diameter, dw Viscosity, μp Stress, τy (bbl/min (ft/s [m/s]) Shear Stress, Shear Rate,
(in. [mm]) (in. [mm]) (cp [mPa-s]) (lbf/100 ft2 [Pa]) [L/s]) ψ ξ
8 ⁄2 [216]
1
7 [178] 50 [50] 15 [7.2] 8 [21] 5.9 [1.8] 0.18 4.07

20 [508] 16 [406] 30 [30] 15 [7.2] 10 [26.5] 1.2 [0.36] 0.7 0.174

Chapter 5 Mud Removal 151

 When the flow around the annulus is laminar, the ning the fluid, the worse the circulation efficiency.
effect of eccentricity on circulation efficiency can be Therefore, in this standoff range, the circulation effi-
derived from the calculated velocity profiles. The results ciency of power-law fluids will increase with the power-
for a Newtonian fluid are plotted in Fig. 5-7 (assuming law index. For Bingham plastic fluids, the higher the
the diameter ratio dw / do = 0.8). In this simple case, the dimensionless shear rate, the better the circulation effi-
circulation efficiency only depends on the pipe standoff ciency will be (Table 5-2). Thus, for such standoffs, the
and the volume pumped. circulation efficiency can be improved by increasing the
The situation is more complex for shear thinning flow rate or increasing the μp /τy ratio. As mentioned ear-
fluids. When standoff decreases, the distortion of the lier, the opposite is true for concentric annuli. Because a
velocity profile is such that the breakthrough time perfectly concentric annulus never exists in the field,
occurs earlier and the circulation efficiency deteriorates recommendations for improving circulation efficiency in
(Chapter 4). In eccentric annuli, such fluids have a more eccentric annuli rather than concentric annuli should
uneven velocity distribution, and the effect of eccentric- be adopted.
ity on the circulation efficiency is even more pro- It is important to note that the circulation efficiency
nounced. The breakthrough time, tbreak, and the rate at is more sensitive to the standoff than to the dimension-
which the circulation efficiency increases after break- less shear rate (Fig. 5-10). If the dimensionless shear
through are reduced to a greater extent with decreasing rate increases by a factor of 50, the circulation efficiency
standoff. varies by no more than 10%. As the standoff increases to
For power-law fluids, with the casing-to-hole diame-
ter ratio close to 1, the circulation efficiency at a given
number of annular volumes pumped depends upon the Table 5-2. Minimum Flow Rates Required to Achieve
pipe standoff and the power-law index. Typical examples Complete Flow Around the Annulus†
of circulation efficiency curves for a power-law index of
0.5 are shown in Fig. 5-8. Standoff Minimum Flow Rate (bbl/min [L/s])
(%)
For Bingham plastic fluids, the circulation efficiency Laminar Flow Around Mixed-flow Regime
depends upon the pipe standoff and the dimensionless the Annulus (Laminar and Turbulent)
shear rate. For different standoffs, Fig. 5-9 shows the cir- Around the Annulus
culation efficiency of a Bingham plastic fluid flowing at 80 2.0 [5.3] 2.0 [5.3]
a rate such that the dimensionless shear rate = 0.178.
60 11 [29.3] 11 [29.3]
Comparing Figs. 5-8 and 5-9 to Fig. 5-7 confirms that
shear thinning fluids are much more affected by pipe 40 38 [101] 19 [50.5]
eccentricity than Newtonian fluids. The velocity distrib-
ution is sensitive to the power-law index or the dimen- 20 >100 [>266] 33 [87.8]
sionless shear rate. Typically, for standoffs less than 80% † Ratesare calculated for a Bingham plastic fluid under two conditions: single-flow regime
(laminar flow) and mixed-flow regime (laminar and turbulent flow).
to 90% (compare Fig. 5-5 to Fig. 5-8 and Fig. 5-9) and a
given number of annular volumes, the more shear thin-

1.0
0.9
0.8
0.7
0.6
Circulation 0.5
efficiency
0.4
0.3 Standoff: 100%
0.2 Standoff: 75%
0.1 Standoff: 50%
Standoff: 30%
0
0 1 2 3 4
Number of annular volumes

Fig. 5-7. Circulation efficiency for Newtonian fluid in an eccentric annulus, calculated using a finite-
element model (dw /do = 0.8), for various standoff values.

152 Well Cementing

 1.0
0.9
0.8
0.7
0.6
Circulation 0.5
efficiency
0.4
0.3
0.2 Standoff: 100%
0.1 Standoff: 80%
Standoff: 50%
0
0 0.5 1 1.5 2 2.5 3 3.5 4
Number of annular volumes pumped

Fig. 5-8. Circulation efficiency for a power-law fluid in an eccentric annulus, calculated using a finite-
element model (dw /do = 0.8, n = 0.5), for various standoffs.

1.0
0.9
0.8
0.7
0.6
Circulation 0.5
efficiency
0.4
0.3
0.2 Standoff: 100%
0.1 Standoff: 80%
Standoff: 50%
0
0 0.5 1 1.5 2 2.5 3 3.5 4
Number of annular volumes pumped

Fig. 5-9. Circulation efficiency for a Bingham plastic fluid in an eccentric annulus, calculated
using the finite-element model (dw /do = 0.8, ξ = 0.178).

1.00
ξ = 0.178
0.95
ξ = 1.0
0.90 ξ = 5.0
0.85
0.80
Circulation 0.75
efficiency
0.70
0.65
0.60
0.55
0.50
0 20 40 60 80 100
API standoff (%)

Fig. 5-10. Circulation efficiency for a Bingham plastic fluid after one hole volume.
Sensitivity to dimensionless shear rate and standoff.

Chapter 5 Mud Removal 153

90% from 50%, the effect on the circulation efficiency is basic-slot model. For various standoffs, Table 5-2 shows
more dramatic. the minimum flow rates necessary to achieve laminar
Among shear thinning fluids, those exhibiting a yield flow all around the annulus. However, this is often not the
stress present a specific case for the following reason. case, especially at low standoff, because the fluid might
When the flow rate is sufficiently low, such fluids are sta- not be in laminar flow on the wide side (Chapter 4).
tionary in the narrowest part of the annulus. The shear For example, consider a mud with a plastic viscosity
stress is less than the yield stress. The basic-slot model of 20 cp [20 MPa-s] and a yield stress of 10 lbf/100 ft2
shows that this will occur if the shear stress at the wall, [4.79 Pa], flowing in a 121⁄2-in. OD, 95⁄8-in. inside diame-
τw(L), calculated for a local annular gap, L, is such that ter (ID) annulus. The data presented in Table 5-2 show
the minimum flow rates necessary to achieve flow
L ⎛ dp ⎞ around the annulus, calculated for two conditions:
τ w ( L) = × < τ y. (5-5)
2 ⎜⎝ dz ⎟⎠ f ■ the fluid is flowing in laminar flow all around the
annulus
This situation is not desirable during mud circulation, ■ the fluid on the wide side may be flowing in turbulent
because the circulation efficiency would asymptote flow.
toward a value smaller than 1 (Fig. 5-9). To avoid this, it
is necessary to have all of the fluid moving around the One can see that, at low standoffs, the minimum flow
annulus. This can be achieved if the minimum shear rate can be overestimated if flow is assumed to be truly
stress at the wall (i.e., the shear stress at the wall at the laminar everywhere.
minimum annular gap Lmin = RSTO[do – dw]/2) is greater The effect of pipe standoff on the transition from lam-
than the yield stress. Figure 5-11 shows the variation of inar to turbulent flow is discussed in Chapter 4. Once a
wall shear stress versus angular position. fluid is in turbulent flow around the annulus, there is
much less distortion of the velocity distribution when
⎛ d p⎞ 4τ y compared to a laminar flow situation. The velocity field
⎜⎝ d z ⎟⎠ > R ( d − d ) or ψ < RSTO (5-6) is also not sensitive to the shear-thinning behavior of the
f STO o w
fluid. To the best of the authors’ knowledge, there is no
where published model for predicting circulation efficiency
under such conditions. However, the efficiency is bound
RSTO = standoff ratio (%). to be much better than in laminar flow.
For a Bingham plastic fluid, the corresponding mini-
mum dimensionless shear rate to have fluid movement
around the annulus can also be determined using the

100 16
90 14
80
12
70
60 10
Gap Wall
thickness 50 8 stress
(mm) 40 (Pa)
6
30
4
20 Gap width
10 Wall stress 2
Fluid yield stress
0 0
0 1.57 3.14 4.71 6.28
Angle (rad)

Fig. 5-11. Variation of wall shear stress around an eccentered annulus (standoff of 66%), at the minimum
condition for having all the fluid in movement according to Eq. 5-5.

154 Well Cementing

5-2.4.4 Effect of casing movement ■ When reciprocating in inclined sections, the pipe has
Rotating the drillstring aids the cleaning of cuttings an alternating up and down or lateral movement in a
from deviated boreholes (Sanchez et al., 1999). One can section of the well in addition to axial movement.
also expect a similar positive effect on mud circulation ■ Pipe rotation also induces Taylor vortices that
and displacement. Indeed, this practice is part of the increase the wall shear stress. Taylor vortices are
“good cementing practices” detailed in Table 5-4. toroidal or helicoidal flows that result from a combi-
The two basic ways to move tubulars are rotation and nation of mean axial flow and azimuthal flow (Fig.
reciprocation along their axes. Secondary movements 5-13). The induction of Taylor vortices is believed to
may also occur that have an even greater effect on the be one of the main reasons why one sees increased
mud-circulation efficiency. annular friction pressure during pipe rotation (Philip
■ When rotating at high speed, the tubular starts to et al., 1998).
have an orbital or whirling movement in addition to
rotation around its axis (Fig. 5-12).

Pull out Run in

Stable motion Snaking motion

Irregular motion Whirling motion

Upward
movement

Downward
movement

Fig. 5-12. Schematic diagram showing orbital or whirling pipe motion during rotation
and lateral pipe motion during reciprocation.

Circular couette flow

Annular Poiseuille flow

Taylor vortices

Fig. 5-13. Annular flow patterns.

Chapter 5 Mud Removal 155

 Sanchez et al. (1999) mention that “it is the orbital effects of these mechanical devices on mud circulation
motion and not the rotation that improves hole cleaning.” are poorly quantified; however, there is no doubt that
Using reasonable assumptions, the effects of tubular they contribute significantly to the efficiency of the
movement on the fluid-velocity map in a cross section process (Smith and Ravi, 1991). With some muds the fil-
can be calculated using finite-element methods. tercake redeposits quickly after scraping and becomes
However, for the secondary movements discussed above, tougher with age.
only experimental and field observations are available. Other techniques or effects are available to help
Therefore, the following discussion is restricted to the remove the immobile mud or deposits. These include
effects of the primary movements. ■ jetting using a special pipe shoe while running casing
Speers et al. (1987) developed numerical models to in the hole (Way et al., 2000)
study the influence of casing movement on the flow pat- ■ rotating the tubular to allow Taylor vortices to
tern of simple non-Newtonian fluids (power-law or
develop, which may significantly increase mud circu-
Bingham plastic fluids) flowing in laminar flow. Similar
lation (Erbil, 1995)
results were presented by Bittleston and Guillot (1991).
■ vibrating the casing during circulation to break adhe-
Circulation efficiencies derived from these models show
that casing movement can indeed partially counteract sion because of static gel strength (Sutton and Ravi,
the detrimental effect of pipe eccentricity (Figs. 5-14 1991) (while this technique is hardly applicable in
and 5-15). However, it must again be emphasized that the field, vibration certainly occurs when rotating the
these models do not account for lateral motion of the casing at high rotary speed)
casing—a likely occurrence during reciprocation and ■ reverse circulating the fluid—down the annulus and
rotation at high rotary speed (McLean et al., 1967). In up the tubular (commonly done in completion opera-
Fig. 5-15, note that rotation is found to be more effective tions but is rarely possible while running the casing)
than reciprocation for displacing mud at intermediate ■ using specially designed hardware elements placed
standoff values. around the casing for inducing spiral flow or turbu-
When performed in combination with scratchers, lence to help displace immobile mud or deposits in
scrapers, or cable wipers (Chapter 11), casing move- the vicinity of the tool (these tools can be specially
ment mechanically erodes the filtercake and consider- designed centralizers [Kinzel and Martens, 1998]).
ably improves the displacement process (Table 5-4). The

Fig. 5-14. Fluid velocity maps comparing casing rotation (right) versus no rotation (left) while pumping a
Bingham plastic fluid (from Bittleston and Guillot, 1991).

156 Well Cementing

 Newtonian, no movement; Herschel-Bulkley; no movement;
standoff of 20% n = 0.9; τy = 2Pa; k = 0.05 Pa-s–n
200 200 2.0
180 180 1.8
160 160 1.6
140 140 1.4
120 120 1.2
100 100 1.0
80 80 0.8
60 60 0.6
40 40 0.4
20 20 0.2

20 40 60 80 100 120 140 160 180 200 20 40 60 80 100 120 140 160 180 200

Herschel-Bulkley; Herschel-Bulkley; reciprocation

rotation (modulus of velocity)
200 200 2.2
180 180 2.0
160 160 1.8
140 140 1.6
1.4
120 120
1.2
100 100
1.0
80 80
0.8
60 60 0.6
40 40 0.4
20 20 0.2

20 40 60 80 100 120 140 160 180 200 20 40 60 80 100 120 140 160 180 200

1.00
No movement
0.95 Reciprocation at 9 m/min
Rotation at 30 rpm
0.90

0.85

0.80
Circulation
efficiency 0.75

0.70

0.65

0.60

0.55
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
API standoff
Fig. 5-15. Top row, left: Velocity map for 20% standoff for a Newtonian fluid.
Top row, right: Same map for a Herschel-Bulkley mud at 800 L/min.
Center row, left: Same Herschel-Bulkley fluid with casing rotation at 6 rpm.
Center right: Same Herschel-Bulkley fluid with casing reciprocation at 9 m/min.
Bottom: Circulation efficiency at one annular volume versus standoff, with and
without casing movement, for a Herschel-Bulkley fluid.

Chapter 5 Mud Removal 157

5-2.4.5 Discussion of theoretical models ■ Drilling fluids are thixotropic. Upon rest or low-shear
The theoretical models mentioned above are sufficiently conditions, their yield stress increases with time. For
sophisticated to allow reasonable simulations of the instance, the mud filling a large washout may be par-
principal steady-state hydrodynamic factors. However, tially immobile and develop a high gel strength.
several important parameters are still not taken into ■ In front of permeable formations, filtercakes are

account because of limited knowledge of their physics deposited (Chapter 6). Filtercakes may have widely
and the overlapping nature of their effects. varying properties depending on the fluid chemistry
■ Mud thixotropy and on whether filtration takes place in a dynamic or
static regime.
■ Mud filtercake and solids-bed formation and erosion
■ In deviated sections, the solids that should be carried
■ Lateral movement of tubulars
by the drilling mud may settle on the low side, build-
■ Effects of casing hardware ing a bed. These solids could be cuttings or weighting
■ Temperature profile material.
In recent years, cuttings-transport models have been In these situations, part of the annulus will be locally
extended to mud circulation. Prototype models exist, but filled with a material with different properties than
they cannot be used routinely to solve field problems those of the flowing mud (Fig. 5-16). This material may
(Nguyen, 1997; King et al., 2000). be removed in various ways during mud circulation.
Channels of gelled mud, mud filtercake, or solids beds
have different effects on zonal isolation. In addition, dif-
5-2.4.6 Removing gelled mud and mud filtercake ferent methods are required to remove them.
during the circulation process ■ They do not form at the same places in the well.
The theoretical results presented thus far are based on ■ Their mechanical and chemical properties differ
the assumption that the annulus is filled with a fluid of widely.
uniform properties. In practice, however, three effects ■ Their response upon contact with cement will be
complicate the analysis.
different.

Drill-in fluid Dehydrated gel

Cuttings bed Filtercake

Formation

Gelled mud channel or

Mudcake
pocket; layer of gelled mud
immobile on impermeable Solids layers
surfaces.
Permeable
layer

Fig. 5-16. Schematic diagram of the potential occurrences of gelled mud and solids layers (Mathis et al., 2000).
Reprinted with permission of SPE.

158 Well Cementing

5-2.4.6.1 Gelled mud For practical use, the flow rate corresponding to this
When allowed to remain static, most drilling muds friction pressure is calculated as if no static gelled mud
develop a structure that is characterized by gel strength. exists in the annulus. This provides a conservative esti-
Gel strength represents the minimum shear-stress value, mate of the minimum required flow rate. In this way, one
τgel, necessary to induce flow. Drilling muds are designed can only determine whether mud is flowing on the
to exhibit thixotropic properties, because they must be narrow side of the annulus, not the velocity at which it is
able to suspend cuttings and the weighting agent when flowing.
circulation is stopped. Thus, gelled mud can be found
everywhere in the hole when circulation is stopped. 5-2.4.6.2 Mud filtercake
While the mud is being circulated, wellbore ovality, The presence of a mud filtercake at the wall of perme-
irregularities (washouts), and casing eccentricity can able formations is another factor that affects the circu-
create zones in which the local fluid velocity is zero. lation process. When mud is flowing across a permeable
Such mud is commonly called the immobile mud. zone, it is subjected to dynamic filtration and then to
Immobile mud is partially responsible for the well- static filtration when the flow is stopped (Chapter 6).
head pressure peak when circulation is resumed. If Without sufficient fluid-loss control, an excessively thick
channels or layers of immobile mud are left in the hole filtercake can grow and reduce the size of the annulus.
after cement placement, they may dehydrate and shrink There is a possible relationship between mud filtra-
while the cement sets, leaving an empty space. This tion and pipe eccentricity that would be detrimental to
empty space compromises zonal isolation, allowing for- the circulation process. Because the erosion of the
mation fluids to flow along the well. deposited filtercake is a function of the shear stress at
Gelled mud can return to its original state if a suffi- the formation wall, the mud filtercake thickness during
cient amount of shear is applied. The minimum shear circulation is likely to be larger at the narrow side of the
required to restore movement is the yield stress or gel annulus. The resultant nonuniform thickness of the fil-
strength. Much has been published concerning the gel tercake would favor an uneven distribution of the flow
strength development of muds as a function of time. path around the annulus and would further reduce the
Garrison (1939) found that, depending on the fluid velocity of the fluid on the narrow side. In extreme cases,
chemistry, bentonite muds can require a few minutes to the fluid could stop flowing, be subjected to static filtra-
a few hours to develop a stable gel strength. The gel tion, and become very difficult to mobilize later.
strength value is also dependent on the way it is mea- When in contact with cement, a mud filtercake can
sured, in particular the applied shear rate. This observa- further dehydrate and, in the same way as gelled mud,
tion is a clear indication of the complex kinetics and leave an empty space. However, a major difference is
equilibria between gel-structure formation and rupture that mud filtercakes do not tend to form in front of
(Darley and Gray, 1988). impermeable caprock layers, where zonal isolation is
Thus, the interpretation of the experimental results especially important. Thus, the potential zonal isolation
obtained by the standard oil industry procedure problems associated with mud filtercakes are often less
(Chapter 4; Appendix B) is questionable. For day-to-day critical than other types of problems such as differential
applications, the situation is even worse, because the sticking.
standard practice for measuring the gel strength con- Mud filtercakes are very compressible; as a result,
sists of readings after 10 sec and 10 min. Ten minutes is their mechanical properties vary widely. Close to the for-
not representative of the long static periods that muds mation wall, yield stresses are typically a few hundreds
can experience before being circulated (several hours or of pascals for oil-base muds (OBMs) and 10 to 1,000
even days). Once the gel strength is exceeded, it is usu- times more for water-base muds (WBMs) (Cerasi et al.,
ally assumed that the mud instantaneously regains its 2001). The outer part of the filtercake is more like gelled
previous rheological properties. mud (Fig. 5-17). In other words, mud filtercakes can be
Based on the above simplifying assumptions, gelled consolidated. Ravi et al. (1992) called this layer “par-
mud can be restored to flowing mud if the shear stress is tially dehydrated gelled drilling fluid.”
equal to the gel strength, τgel. The corresponding mini- The thickness of the mud filtercake is strongly depen-
mum friction pressure to achieve flow on the narrow side dent on the shear history. When circulation stops, a
of an eccentered annulus is similar to Eq. 5-6. static filtercake grows. The filtercake is partly eroded
when circulation and pipe movement resume. With time,
⎛ d p⎞ 4τ gel τy the filtercake’s resistance to flow and the ratio of filter-
⎜⎝ d z⎟⎠ > or ψ < (5-7)
cake thickness to its mean permeability (known as
f
RSTO ( do − dw ) τ gel
hydraulic resistance) tend to increase; therefore, the
rate of filtercake growth decreases during subsequent

Chapter 5 Mud Removal 159

static periods. Mud-filtercake thicknesses measured in The shear stress at the formation wall that corre-
the laboratory vary from a few millimeters to more than sponds to the high yield stresses mentioned above is
1 cm (Haut and Crook, 1979). unattainable under normal flow conditions. For exam-
The mud composition and chemistry strongly influ- ple, achieving a wall shear stress of 250 Pa in a 1-cm-
ence the cohesion of the mud filtercake and its adhesion wide rectangular slot requires a pressure gradient of
to the rock surface (Amanullah, 2002). Cohesion and 1.1 psi/ft [25 kPa/m]. Thus, except for the outer layers,
adhesion are key properties for mud-filtercake removal pure shear stress is not an efficient way to remove mud
by backflow during completion operations (Ali et al., filtercake. This is well supported by experimental obser-
1999). In this context, cohesion is the resistance of the vations (Zuiderwijk, 1974; Crook et al., 1987).
filtercake in tension, while adhesion is the filtercake’s
ability to stick to formation wall. During backflow, high 5-2.4.6.3 Solids beds
adhesion and low cohesion will favor the creation of pin- Solids beds are the result of incomplete cuttings clean-
holes, while high cohesion and low adhesion will lead to ing or weighting-agent settling. Thus, solids beds are
peeling of the cake (Table 5-3). often found in deviated and horizontal sections. Their

Scraping Strength Profile of Filtercake Scraping Moisture Profile of Filtercake

on Birchover Rock on Birchover Rock
10,000 3.0
8,000 2.5
2.0
Yield 6,000
Void 1.5
stress
4,000 ratio
(Pa) 1.0
2,000 0.5
0 0.0
0.0 0.3 0.6 0.9 1.2 1.5 1.8 2.1 2.4 0.0 0.3 0.6 0.9 1.2 1.5 1.8 2.1 2.4
Slice height above rock (mm) Slice height above rock (mm)

Fig. 5-17. Yield-stress and void-ratio measurements made from a WBM filtercake (from Cerasi et al., 2001).
Reprinted with permission of SPE.

Table 5-3. Summary of Cleaning Mechanisms for Mudcakes and Solids Beds
Mechanism Relevant Physical Parameters Domain of Application
Flow Solid Layer
Erosion by shear Wall shear stress Shear strength Mud circulation and
instability of interface displacement

Erosion by pressure Reynolds number Cohesion Mud circulation and

fluctuations, rupture displacement
in tension

Pinholing or peeling Differential pressure Cohesion and adhesion Backflow during

completion operations

Weakening of layer Chemical composition, Cohesion, shear strength, Completion operations,

mechanical properties, duration of treatment chemical nature mud displacement
dissolution

Fluidization, increase Pressure gradient Permeability, buoyant Hole cleaning

in lift force density of particles

Abrasion by Inertia of particles Cohesion Gravel-pack placement

suspended particles Cutting transport and
corresponding hole cleaning

160 Well Cementing

occurrence is independent of formation properties and tion of polymeric components (Ryan et al., 1995).
tubulars. Solids-bed removal is critical to the successful Chemicals that can cause such effects include acids,
placement of casing on the bottom of the well. Its effect chelating agents, surfactants, dispersants, and oxidizers.
on local zonal isolation is not well understood, although The various filtercake and solids-removal mechanisms
one may expect it to be similar to that of a mud filter- are summarized in Table 5-3.
cake. The fact that solids beds occur over long intervals During mud circulation, erosion is the only relevant
makes their presence potentially more detrimental than solids-removal mechanism. Eventually, detached solids
mud filtercakes. may induce some limited abrasion. Chemical effects
The properties of solids beds have not been described later come into play when spacers and washes are
quantitatively. Barite beds are known to be very cohesive pumped into the annulus. Such effects are discussed
and adhesive. Following laboratory tests, Bern et al. later in this chapter.
(1998) found barite beds to be very different from cut- From the preceding discussion, it is clear that a com-
tings beds. Barite beds can be readily dispersed with a plete understanding of the effects of mud gelation, mud
combination of high flow rates and drillpipe rotation. filtration, and solids settling has not yet been achieved.
The presence of polymers in the drilling fluid increases Nevertheless, a qualitative analysis shows that all three
the consolidation of solids beds (Saasen, 1998). may have a detrimental effect on circulation efficiency,
in particular when the pipe is not centered. The global
5-2.4.6.4 Removal of mud filtercakes and solids beds cleaning process is usually a series of steps. Optimiza-
Removal methods for mud filtercakes or solids beds are tion consists of selecting the correct order. This issue
analogous to other engineering domains such as “clean- will be discussed in Section 5-5.
ing-in-place” in the food industry, “back-pulse cleaning”
in chemical engineering, or beach erosion in marine geo-
physics. Fluid velocities and contact time are critical 5-2.4.7 Measuring mud-circulation efficiency
parameters. in the field
It is still unknown whether the shear stress at the Traditions such as “circulating bottoms up” are often
wall or the Reynolds number is the key parameter (Plett, relied upon when designing the mud-circulation period
1985). If the wall shear stress is the relevant physical before cementing. Results show that this approach is
parameter, the removal mechanisms are more likely to insufficient for complete mud removal in most circum-
involve shear instabilities at the fluid/solids-bed inter- stances. A better approach is to measure the mud-circu-
face. If the Reynolds number predominates, then pres- lation efficiency (Smith, 1984). This is performed by
sure and stress fluctuations would be expected to play monitoring the volume of mud that is actually circulating
the major role. All of these effects are grouped under the (the circulatable mud) with a fluid caliper or tracer. A
general term “erosion.” This lack of basic knowledge is a small volume of mud is tagged with a tracer and injected
root cause of the debate about whether viscous or dense at the wellhead. The time necessary for the tracer to
fluids are more effective at cleaning wells (Sewell and return to the surface indicates the volume of mud being
Billingsley, 2002). circulated. This volume is then compared to the bore-
Abrasion is another process that is based on the iner- hole volume determined from caliper measurements. An
tial energy of solids particles carried by the flowing fluid. illustration of this concept is shown in Fig. 5-18.
Upon impact with the solid layer, the particles can break Tracers have included inert particles (oats, rice
some part of it. Johnson et al. (1992) and Becker and hulls), dyes (hematite, fluorescein), radioactive materi-
Gardiner (2000) observed this effect during flow loop als (scandium), chemical tracers (bromide, iodine, thio-
tests of gravel placement. This is similar to a sand-blast- cyanate, nitrate) and reactive materials (carbide pills).
ing effect. Abrasion can also be achieved by scratchers Carbide pills are composed of calcium carbide. When
fastened around the casing or by fluid jets directed per- added to the drilling mud, they dissolve in water and
pendicular to the solid layer. generate acetylene.
Solids beds are permeable; therefore, some flow may Detection of the tracer can be visual (dyes or parti-
take place through them. Such flow may weaken the bed cles retrieved on the shaker screen) or chemical (e.g., by
through a fluidization mechanism and increase the lift using gas detectors, titration kits, ion-selective elec-
force on individual particles (Cho et al., 2001). trodes, and ion chromatographs) (Hall and Hughes,
Finally, chemical or physico-chemical effects can lead 1993). Most of these techniques only provide qualitative
to solids removal. The various phenomena include inter- results, because it is often not clear what is being mea-
facial-tension reduction, weakening (cohesion or shear sured—time of first appearance or time of maximum
strength), dissolution of solid components, and degrada- tracer concentration. A quantitative analysis would

Chapter 5 Mud Removal 161

require continuous monitoring of the tracer concentra- Indirect methods based on surface or downhole pres-
tion at the return line and an interpretation scheme to sure measurement provide indications of hole cleaning
infer an average circulation velocity. and cutting suspension. Ravi et al. (1993) and Griffith
Nevertheless, these techniques are quite useful. For and Ravi (1995) performed surface pressure measure-
example, using carbide pills, Smith (1990) advocated ments that, when combined with continuous measure-
designing the flow rate and the circulation time on the ment of fluid rheology, density, and flow rate, allowed
assumption that 95% of the calipered hole volume was them to determine a pseudodisplacement efficiency.
circulating. His measurements (Fig. 5-19) led him to rec- Hutchinson and Rezmer-Cooper (1998) showed that
ommend circulation velocities in excess of 250 ft/min annular pressure-while-drilling data can be interpreted
[76.2 m/min]. Such a velocity is quite high by oilfield in terms of hole cleaning and mud-solids carrying capac-
standards. ity. They also demonstrated the effectiveness of pipe
rotation (Fig. 5-20).

Flowmeter 60
55
50
45
Mud pump 40

Shale shaker 35
Lithium
concentration 30
[(mol/L) × 10–4] 25
20
Tagged mud
15
Multiarm caliper
hole volume 10
5
0
0 20 40 60 80 100 120 140
Fluid caliper
Elapsed time (min)

Fig. 5-18. Left: Schematic of a well showing the fluid-caliper concept used to determine mud-circulation efficiency
(from Smith, 1990). Right: Lithium-ion chemical log for tracer test: lithium bromide concentration in mud filtrate
(mol/L mud) plotted against elapsed time (from Hall and Hughes, 1993). Reprinted with permission of SPE.

100
95
90
85
80
Circulation
efficiency 75
(%) 70
65
60
55
50
30 50 70 90 110 130
Annular average velocity (m/min)

Fig. 5-19. Effect of annular velocity on circulation efficiency (from

Smith, 1990). Reprinted with permission of SPE.

162 Well Cementing

 Standpipe Pressure
0 (psi) 5,000
Geometry Weight Time Rotation Flow ECD/ESD
–10 (ft/s) 10 0 (klbf) 500 0 (rpm) 200 0 (gpm) 1,000 13.2 (ppg) 14.2

15:00 Rotation
stops
ECD
decreases
Sliding
interval
Lower standpipe
pressure when
pipe rotation is
16:00 stopped
Rotation
starts ECD
increases

Fig. 5-20. Field data demonstrate the effectiveness of pipe rotation to suspend cuttings and achieve good
hole cleaning (from Hutchinson and Rezmer-Cooper, 1998). Pressure varies as equivalent circulating
density (ECD) increases and decreases. Reprinted with permission of SPE.

5-2.5 Mud circulation—conclusions ■ Whenever predictions are doubtful, fluid calipers

Ensuring that a large percentage of the mud is actually should be used to measure the efficiency of the circu-
in circulation is a key to the success of a primary cement lation process. Circulation should be maintained
job. In view of the complexity of the problem, there is no until the caliper indicates that 90% of the borehole
doubt that sufficient time should be devoted to the volume is in circulation.
design, execution, and evaluation of the mud condition- ■ The borehole pressure should be maintained between
ing and circulation stages before cementing. The follow- the pore and fracture pressures.
ing qualitative guidelines can be distilled from the pre- ■ To prevent the erosion of unconsolidated formation, it
ceding discussion. is a current practice to keep the annular velocity of
■ The rheological properties of the mud (namely the the mud below the maximum value encountered
μ p / τy ratio), mud gel strength, and pipe standoff during drilling.
should be such that the mud is moving completely ■ It is preferable to drill the hole with high-quality mud
around the annulus at an achievable flow rate. This from the start. If poor-quality mud is used, it may be
can be done by improving pipe standoff, increasing impossible to fully remove it before the cement job.
the μ p / τy ratio, decreasing the mud gel strength, or
increasing the flow rate.
■ If the above criteria for rheology, gel strength, and
5-3 Mud displacement
pipe standoff cannot be met, reciprocation or rotation
of the pipe should be performed during mud circula- Determining mud-displacement efficiency is more com-
tion. plex than determining mud-circulation efficiency. The
interface profile between two dissimilar fluids is very dif-
■ When available, circulation models should be used to
ferent from the velocity profile of a single fluid flowing in
better optimize the above parameters in view of the same geometry at the same flow rate (Fig. 5-1).
improving the circulation efficiency. However, the tendency to use the velocity profile to
■ The volume of circulated mud should represent at explain displacement efficiency still persists. In addition
least one full hole volume; however, circulation to the parameters mentioned earlier, mud displacement
models can be used to obtain a better estimate of the depends upon the relative properties of the fluids
required mud circulation time. involved (density and rheology), their relative flow
regimes, and their eventual interactions when mixed.

Chapter 5 Mud Removal 163

 To introduce the subject, the origin of current field ity hierarchy, and minimum pressure gradient are now
practices is summarized. Next, the physical description part of the design criteria to optimize mud displace-
of the problem is presented. Phenomena that are under- ment. The quantitative requirements attached to these
stood best are described first, followed by those that are rules were derived from simple physical reasoning, pilot
more complex and qualitative. This is followed by a dis- experiments, or field observations and were confirmed
cussion of laminar flow and turbulent-flow displacement by field studies (Kelessidis et al., 1995). Increased mod-
of one fluid by another in concentric and eccentric eling capabilities and understanding have allowed the
annuli. Then, complicating situations such as gelled industry to test designs that more closely resemble field
mud, solids layers, and casing movement are covered. reality (Tehrani et al., 1992; Bittleston et al., 2002).

5-3.1 Field practices and observations 5-3.2 Problem description

One of the first parameters recognized to influence mud- Efficient displacement of one fluid by another results
displacement efficiency was the flow regime of the dis- from the interplay between various forces. Some forces
placing fluid. From pilot-scale studies, Howard and Clark aid displacement, while others inhibit it. These forces
(1948) concluded that, when the Reynolds number of include buoyancy, viscosity, inertia, and other rheologi-
the cement slurry was low, only 60% of the “circulatable” cal properties such as yield stress and gel strength.
mud was displaced, whereas 90% to 95% could be dis- Physico-chemical effects must also be considered; how-
placed when the cement slurry was in the upper laminar ever, these are more difficult to include in global models.
flow or turbulent-flow regime. In 1964, Brice and Holmes These forces are closely related to various parame-
surveyed 46 cement jobs performed in southwest ters. For example, buoyancy forces are a direct function
Louisiana, an area that was experiencing primary of density differences and hole inclination, while viscous
cementing failures. They reiterated the need for turbu- forces are functions of fluid rheology, wellbore geometry,
lent flow, and suggested that the annular space should and flow conditions. In general, mud displacement is a
be in contact with a turbulent displacing fluid for a suf- competition between driving and resisting forces. The
ficient time. Subsequent pilot-scale studies lead to the predominant driving or resisting force will depend upon
establishment of rules, called good cementing practices, the local conditions.
to maximize mud removal (Table 5-4). Today, these prac- ■ In a large vertical annulus, with a heavy fluid displac-
tices are generally accepted in the industry. Numerous ing a lighter one at slow velocity, buoyancy forces will
case studies have demonstrated their effectiveness in be the main driving force.
many situations. However, these practices remain quali- ■ Conversely, in a deviated and slim annulus, viscous
tative because no accepted thresholds or recommended
forces will generally govern the flow.
values are provided.
More recently, quantitative rules were developed Because each type of force has a different depen-
from improved knowledge or more experiments (Brady dence on length, the mechanisms driving the displace-
et al., 1992; Ryan et al. 1992). Density hierarchy, viscos- ment are a function of the length scale.

Table 5-4. Good Cementing Practices†

Wellbore geometry The ideal annulus is straight, free of washouts, and at least 1.5 in. wide.

Centralization The casing is kept as near the middle of the wellbore as possible. A standoff of about 75%
is generally considered to be acceptable.

Mud conditioning The mud is circulated before the cementation to remove gas and cuttings, break the mud’s
gel strength, and lower the mud’s viscosity.

Casing movement Rotation is preferred to reciprocation. Movement should start during mud conditioning.

Chemical washes Washers and spacers are pumped ahead of the slurry, acting as a buffer between
and spacers possibly incompatible mud and cement. A 10-min contact time is recommended.

Cementing plugs Top and bottom plugs are pumped ahead and behind the slurry to separate it from the mud.

Other casing hardware Scratchers are sometimes attached to the outside of the casing to remove
the mud filtercake.
† Bittleston and Guillot (1991).

164 Well Cementing

■ On a large length scale, typically the length of the A basic displacement-efficiency model for these con-
interval being treated, various displacement patterns ditions is built as follows. Consider a thin rectangular
occur, e.g., plug flow. This is the domain of fluid slot of width L, with static layers of displaced fluid along
mechanics, both theoretical and experimental. the walls and the displacing fluid flowing in the middle
■ On a small length scale, typically the thickness of the at a mean velocity, v–. The displacing fluid creates a shear
annular gap, mud filtercakes, immiscible fluids, and stress. As long as this shear stress is less than the yield
other physico-chemical phenomena predominate. stress of the displaced fluid, the layers along the wall are
stable and their maximum thicknesses can be deter-
A typical order of magnitude for the ratio of these two mined. Although this seems simple, it is only recently
length scales is 10 4:1. As previously mentioned, no labo- that Allouche et al. (2000) provided an exact solution for
ratory apparatus can be scaled to simultaneously repro- two yield-stress fluids. They showed that this maximum
duce field conditions and consider all of the indepen- thickness depends on only three parameters: the ratio of
dent variables. the yield stress of the two fluids, Ry; the Bingham
The present understanding of the influence of these number of the displacing fluid, NBG; and the ratio of
forces on mud displacement is incomplete. On a small buoyancy forces to the yield stress, Rb (Fig. 5-21). Four
scale, the general term erosion is frequently used to situations are possible, labeled (a) to (d) on Fig. 5-21,
cover poorly known phenomena. Large-scale phenom- depending on the relative values of the shear strength of
ena are somewhat better understood. the two fluids with respect to the shear stress at the fluid
Everything else being equal and at low flow rates, the interface.
upward displacement of a dense fluid by a lighter one in
a vertical conduit leads to an unstable phenomenon
known as a buoyant plume. Conversely, when the dis- Ry =
( τ y )2 NBG =
( τ y )2 L Rb =
(ρ2 − ρ1 ) gL
placing fluid is heavier than the displaced fluid, buoy- ( τ y )1 2( μ p ) v ( )1
2 τy
2
ancy forces tend to flatten the interface and promote
efficient displacement. Similarly, the displacement of a (5-8)
viscous fluid by a less viscous fluid of identical density This model provides two predictions.
leads to viscous fingering, in which a stream of the less
■ The maximum possible displacement efficiency, i.e.,
viscous fluid penetrates the interface and enters the
more viscous fluid. Conversely, displacing a thin fluid by the displacement efficiency when steady-state condi-
a more viscous one leads to better displacement. More tions are reached
complex combinations of density, rheology, and borehole ■ The conditions necessary to remove the static layer
geometry remain a subject of academic research. (as shown in Eq. 5-9, this condition is independent of
Modeling large-scale displacement relies on the use the buoyancy ratio)
of fluid-mechanics laws (mass and momentum balance
for each fluid) and consistent boundary conditions. ( τ y )2 = ( τ y )cem = 1 ,
( τ y )1 ( τ y ) mud ξ
Further, considering the large borehole-height/gap- (5-9)
thickness ratio, the problem is usually reduced to two
dimensions by averaging physical quantities across the
gap width. Alternatively, true 3D codes have been devel- with dimensionless shear rate, ξ, being the solution of
oped; however, the running time is on the order of days 2ξ3 – 2(3 + 6/NBG) ξ2 + 1 = 0.
rather than minutes. Two different methods are The main limitation of this work is that it neglects the
employed to track the interface between two fluids: a displacement front.
mesh moving with the interface and a fixed virtual mesh Flumerfelt (1973) and Beirute and Flumerfelt (1977)
with the concentration calculated for each fluid in each presented the first complete 2D model that provides the
mesh cell. time variation of the displacement efficiency. The main
assumptions of these analyses are given below.
■ Displacement takes place in a narrow rectangular
5-3.3 Laminar displacement of mobile mud and vertical slot with upward flow.
in concentric vertical annuli ■ For both fluids, the flow regime is laminar, and the
In a vertical and centered annulus with favorable den- Reynolds numbers of each fluid, (NRe)i, are small com-
sity and viscosity ratios, the displacement is stable in the pared to the length-to-gap ratio of the annulus, i.e.,
azimuthal direction. Any variation in displacement effi-
ciency is caused by the shape of the interface profile in
the radial direction. Therefore, the problem is reduced ( NRe )i << d Z
.
o − dw
to two dimensions, axial and radial coordinates.

Chapter 5 Mud Removal 165

 (a) (b) (c) (d)

1.0

v 0.8

0.6
1
NBG
Yi Y = 1 Yi Y = 1 Yi Y = 1 Yi Y = 1 0.4

0.2
–(τy)1,Y
0
–(τy)1,Y 0 1 2 3 4 5
–τ, xy –(τy)1,Y
1
–(τy)1,Y
–(τy)2,Y NBG
–(τy)2,Y –(τy)2,Y
–(τy)2,Y

Yi Y = 1 Yi Y = 1 Yi Y = 1 Yi Y = 1

Fig. 5-21. Left: Schematic of the four different possible characteristic axial velocity and stress profiles
for positive buoyant displacements (a to d); Y is the dimensionless radial coordinate. Right: a condition
sufficient to obtain a zero-thickness wall layer. NBG is the Bingham number of the displacing fluid
(from Allouche et al., 2000). Copyright Cambridge University Press.

■ Displacement occurs under stable conditions (i.e., the power-law index of the displacing fluid is lower
the interface is smooth, a condition that is not quan- than that of the displaced fluid.
titatively defined in the papers). ■ In the case of Herschel-Bulkley fluids, yield stresses
■ The fluids are miscible (i.e., interfacial tension is are critical. High displacement efficiencies result
neglected). when the dimensionless yield stress values of the dis-
■ Molecular diffusion at the interface is negligible. placing fluid are higher than those of the displaced
■
fluid.
The horizontal velocities are negligible.
■ For fluids with a nonzero yield stress, displacement
The authors recognized that the mass balance was not efficiencies decrease with increasing dimensionless
correct. In addition to the dimensionless time, the flow rates.
approximate solution depends on five dimensionless
When looking carefully at the data presented in both
parameters for power-law fluids (two more for fluids with
papers, it appears that the last conclusion is somehow too
a three-parameter rheological model such as Herschel-
drastic. For example, in the case of power-law fluids,
Bulkley): density ratio, effective viscosity ratio, dimen-
Flumerfelt (1973) varied the power-law indices while
sionless flow rate, and the two power-law indices. The
keeping the other dimensionless parameters constant
principal conclusions of these works are given below.
(Table 5-5). The ratio of apparent viscosities of the two
■ The density ratio plays a predominant role, provided
fluids (calculated at the wall for both fluids as if they were
the dimensionless flow rate is not too high. A density pumped independently) changed greatly. Also, depending
ratio greater than 1 flattens the profile of the inter- on the relative value of the power-law indices, this ratio
face and enhances the displacement efficiency could increase or decrease with increasing flow rate, lead-
(Fig. 5-22). ing to an improvement or a deterioration of the displace-
■ Although the displacement efficiencies increase with ment efficiency. More specifically, the displacement effi-
increasing effective viscosity ratios, their effect is less ciency decreases with flow rate when the ratio of apparent
important than that of the density ratio (Fig. 5-22). viscosities is less than unity (stable situation), and it
■ Power-law indices do not seem to be important in the increases in the unstable situation (apparent viscosity
case of Herschel-Bulkley fluids. For power-law fluids, ratios higher than unity). The same argument applies to
better displacement efficiencies are obtained when the effect of yield stresses for Herschel-Bulkley fluids.

166 Well Cementing

 1.00
100
2 30
0.95 20
1.5 15
0.95
1.2 10
0.90 8
1.1
0.90 6
0.85 5
Displace- 1.02 Displace- 4
ment ment 0.85
R2 =1 3
efficiency 0.80 efficiency
0.80 2
0.75 1
0.75 R 3 = 0.1
0.70
0.70
0.7 0.8 0.9 1.0 1.1 1.2 1.3 1.4 1.5 1.6 0.7 0.8 0.9 1.0 1.1 1.2 1.3 1.4 1.5 1.6
Dimensionless time Dimensionless time
Fig. 5-22. Effect of density ratio (noted as R2) and effective viscosity ratio (noted as R3) on displacement
efficiency versus time (after Beirute and Flumerfelt, 1977). Reprinted with permission of SPE.

Table 5-5. Effect of Dimensionless Flow Rate (R4) on Displacement Efficiency†

Circulation Efficiencies for n 1 = 1, n 2 = 1 (Newtonian Fluids) Effective Viscosity Ratio (R 3)
R4 t* = 1 t * = 1.2 t * = 1.5
10–4 0.808 0.844 0.879 1

10–3 0.808 0.844 0.879 1

10–2 0.808 0.844 0.879 1

Circulation Efficiencies for n 1 = 1, n 2 = 0.6 (Stable Conditions) Effective Viscosity Ratio (R 3)

R4 t* = 1 t * = 1.2 t * = 1.5
10–4 0.937 0.982 0.991 28.9

10–3 0.926 0.966 0.980 11.5

10–2 0.901 0.937 0.958 4.6

Circulation Efficiencies for n 1 = 0.6, n 2 = 1 (Unstable Conditions) Effective Viscosity Ratio (R 3)
R4 t* = 1 t * = 1.2 t * = 1.5
10–4 0.467 0.487 0.511 0.035

10–3 0.568 0.592 0.621 0.087

10–2 0.672 0.702 0.734 0.22

† In
this table, density ratio = 1. The column on the right gives the ratio of apparent viscosities of the fluids
when pumped individually. All dimensionless ratios represent displacing to displaced fluid.

Chapter 5 Mud Removal 167

 These theoretical developments regarding the dis- tions when displaced by a very thin cement slurry at high
placement process in a concentric annulus have empha- velocities (in which case ncem/nmud > 1), and the effi-
sized the role played by gravitational and viscous forces. ciency of the process was attributed to the turbulent
With the density ratio and the low shear-viscosity ratio eddies in the displacing fluid (this point is discussed
higher than 1 (i.e., power-law index ratio lower than 1 and later). For velocities ranging from 0.5 to 1.5 ft/sec [0.15
a yield-stress ratio higher than one for fluids exhibiting a to 0.46 m/s] and values of ncem/nmud on the order of 1,
yield stress), low flow rates contribute to a flattening of the displacement efficiency was almost constant.
the interface profile and lead to efficient displacement. The slow-flow technique, which was developed in the
With increasing flow rate, the influence of the density 1960s to overcome some of the practical limitations of
ratio logically decreases, and the viscous properties con- turbulent flow displacement, also relies on experimental
tinue to have an impact. Everything else being equal, the observations that are partially in agreement with the
more viscous the displacing fluid, the better the results. Beirute and Flumerfelt (1977) studies. Parker et al.
However, care should be taken to compare apparent fluid (1965) showed that good mud displacement could be
viscosities in a shear-rate range representative of the flow obtained at low flow rates, provided the displacing fluid
conditions to be encountered. (a cement slurry in this case) is at least 2 lbm/gal
Similar conclusions on the effect of viscosity can be [0.24 g/cm3] heavier than the mud and the initial gel
deduced from the work of Nguyen et al. (1992), who strength and viscosity of the mud are less than those of
studied the displacement of a power-law fluid by another the cement slurry. They also observed that, under these
power-law fluid in a concentric and horizontal annulus, conditions, the displacement efficiency deteriorated
assuming uniform azimuthal pressures. This assumption with increasing flow rate. For good mud displacement,
allowed them to use a 2D model. They presented their the shear stress applied by the displacing fluid should be
result in terms of the thickness of the displaced-fluid higher than the gel strength of the mud.
layer remaining on the wall. At a given time, varying only Excellent results were obtained in the presence of
the ratio of fluid consistencies has a significant effect on washouts when the annular velocity was less than 90
the thickness of the wall layer. ft/min [27 m/min]. The efficiency of the process was
attributed to the action of a coagulated mass at the
cement/mud interface, which provided a piston-like dis-
5-3.3.1 Experimental validation placement even in large washouts and irregularities. At
Most of the experimental studies performed on laminar higher flow rates, the cement slurry broke through this
flow displacement in concentric annuli (Childers, 1968; gel mass; consequently, poor displacement efficiencies
Zuiderwijk, 1974) are in qualitative agreement with the resulted. However, the results available did not allow the
theoretical study presented above, as far as the relative clear definition of an optimal displacement velocity
importance of fluid properties (density and rheology) range (Fig. 5-23).
and flow rates is concerned. One of the most extensive The slow-flow technique was later refined. The com-
studies is that of Zuiderwijk (1974), who performed bined effect of density and gel strength differential on
more than 200 tests. Muds were displaced by five annu- mud-displacement efficiency was evaluated and, in the
lar volumes of cement, and displacement efficiencies most common case in which the mud density was lower
greater than 80% were observed. All fluids were assumed than that of the cement, the minimum gel strength
to follow the power-law model. Zuiderwijk (1974) con- required for 100% mud displacement could be calculated
cluded that annular velocity is a key parameter in the empirically (Fig. 5-24). Parket et al. made a qualitative
displacement process. The results showed that, at low conclusion that “a sufficiently high-density ratio can
velocities, a density ratio greater than 1 enhanced compensate for an unfavorable yield-stress ratio.”
the displacement efficiency, and gravitational forces These experimental results provide only a qualitative
appeared to be less important for velocities higher indication concerning the efficiency of the displacement
than 1 ft/s [0.3 m/s]. Depending on the prevailing condi- process. The measurements do not allow one to quanti-
tions, efficient displacement was obtained at both low tatively validate the efficiencies predicted by the numer-
and high displacement rates. At low velocities, good ical displacement models. A common conclusion of both
results were obtained with cement slurries having a the numerical models and the experiments is that a
higher viscosity than the mud (power-law index ratio favorable density ratio can compensate for an unfavor-
ncem/nmud < 1). Well-treated muds (i.e., muds with a low able effective viscosity ratio, bearing in mind that the
yield point) were also easily removed from washout sec- effective viscosity ratio is flow rate-dependent.

168 Well Cementing

 The most common way to model displacement in an
90 eccentric annulus is to use the parallel-slot analogy or
the sectored concentric-annulus analogy, described in
Displace- 80 Chapter 4 (Section 4-6.3). The three principal limita-
ment tions to this approach are as follows.
efficiency OD 14-in. OD 6.75-in. OD 4.75-in.
section section section ■ The validity of the results strongly degrades when the
(%)
70 radius ratio is far from unity.
■ The drag stress between consecutive sectors or slots

60 is ignored.
0 100 200 ■ The pressure is assumed to be uniform in any cross-
Annular average velocity (ft/min) section.
Fig. 5-23. Displacement efficiency as a function of annular velocity in This model is intrinsically one-dimensional (1D), but
various sections of an annulus (ID = 23⁄8 in.) (from Parker et al., 1965). it leads to some semiquantitative criteria that are useful
OD means outside diameter. Reprinted with permission of SPE. for job design. The job-design restrictions imposed by
the above limitations will be described later.
Jones and Berdine (1940) and Howard and Clark
80
(1948) described the channeling of the displacing fluid
on the wide side of the annulus. In 1967, McLean et al.
60 described a more fundamental effort to understand the
role of casing eccentricity on mud removal. First, they
developed a 1D model describing the flow of a single
40 Bingham plastic fluid in an eccentric annulus, based on
the sectored concentric-annulus analogy. Next, the
model was extended to the displacement problem in the
Gel strength absence of gravitational forces. The results suggested
differential 20
(lbf/100 ft2) that, in laminar flow, displacements in eccentric annuli
are more effectively optimized by increasing the yield-
0
100% stress ratio rather than the plastic-viscosity ratio. The
reasons for this are twofold.
90%
1. Once the yield-stress ratio is higher than a critical
–20 value equal to
80%
60%
2 − RSTO
,
–40 RSTO
–1 0 1 2 3 4 5
Density differential (lbm/gal) the shear stresses generated by the cement are suffi-
ciently high for the mud to flow in the narrowest part
Fig. 5-24. Mud displacement efficiency as a function of density and of the annulus. Under creeping flow conditions (i.e.,
gel strength differential between cement slurry and mud. Reprinted
with permission of SPE. near-zero flow rate), the velocity of the mud in the
narrowest part of the annulus is equal to the average
velocity of the displacing fluid. In effect, there is 100%
displacement efficiency. As the flow rate increases,
5-3.4 Modeling of mud removal in eccentric the displacement efficiency decreases (Fig. 5-25).
annuli
2. Increasing the plastic viscosity ratio improves the dis-
Until now, the mud-removal discussion has been con- placement efficiency. This effect is stronger as the
fined to concentric annuli. As discussed earlier, eccen- flow rate is higher (Fig. 5-26).
tric annuli are encountered in most cement jobs, and
many cementing difficulties are exacerbated by casing Both points can be qualitatively understood by con-
eccentricity and geometrical asymmetry (e.g., in dual sidering the effective
. viscosity of a Bingham fluid,
completions). The main difference with respect to a cen- μeff = μp + τy /γ. Increasing the flow rate, hence the
tered annulus is that the problem is now truly 3D. The shear rate, decreases the importance of the yield stress
displacement fluid may channel in both the radial and with respect to the plastic viscosity. Point 1, however,
azimuthal directions. seems paradoxical. Under creeping-flow conditions, the

Chapter 5 Mud Removal 169

 1.0 were obtained with a high-viscosity displacing fluid in
Mud
τy = 10 lbf/100 ft2 laminar flow than a low-viscosity fluid in partial turbu-

0 bbl/min

min

n
in
lent flow. These experiments were performed at the

mi
0.8 μp = 10 cp

bl/m

bl/
3 bbl/

9b
ρ = 10 lbm/gal

6b
Velocity of Cement
same flow rate and, in both cases, the displacement effi-
bypassed mud 0.6 τy = 10 lbf/100 ft2 ciency increased with the flow rate. On the basis of their
divided by μp = as shown theoretical work and limited experimental studies (they
average ρ = 10 lbm/gal
velocity 0.4 intentionally did not allow the muds to gel), McLean et
9 in.
of cement
Cement al. (1967) suggested that viscous displacing fluids are
(dimensionless) 0.2 6 in. better at achieving high displacement efficiency than
thin fluids. While thin displacing fluids extend the area
0
3⁄4 in. Critical value of turbulent flow, the drag and pressure gradient are sig-
15 20 25 30 35 40 45 nificantly reduced.
Yield stress of cement (lbf/100 ft ) 2 Other authors have attempted to model the effect of
eccentricity on the displacement process. On the basis
Fig. 5-25. Effect of cement yield stress on displacement of a mud of the same annular geometry used by McLean et al.
from an eccentric annulus with a standoff of 50% (from McLean (1967), Graham (1972) modeled the fluids as a 1D
et al., 1967). Reprinted with permission of SPE. bundle of parallel cylindrical pipes. He reemphasized
the relative-viscosity concept in the absence of gravita-
tional forces, and used a mobility ratio concept defined
1.6 as the ratio of the flow rate to the friction pressure. A
Mud Cement
1.4 τ = 10 lbf/100 ft2 τ = 10 lbf/100 ft2
y y
mobility ratio greater than 1 was shown to be desirable
Velocity of 1.2 μp = 10 cp μp = as shown for optimal displacement. Because different fluids
ρ = 10 lbm/gal ρ = 10 lbm/gal
bypassed mud 1.0 exhibit different changes in mobility with changing flow
divided by conditions, optimum results could be obtained at either
average 0.8
9 bbl/min high or low flow rates. Even under the best conditions,
velocity 0.6
of cement Cement the velocity of the interface was always greater in the
0.4 wide part of the annulus than in the narrow side
(dimensionless) 6 in.
9 in.
0.2 (Fig. 5-27). To overcome the resulting difference in the
6 bbl/min
0 3⁄4 in. level of the interface and to ensure that cement would
0 20 40 60 80 100 reach the target level on the narrow side, Graham
Plastic viscosity of cement (cp) (1972) applied the knowledge of the interfacial velocity
around the annulus and proposed pumping an excess
Fig. 5-26. Effect of cement plastic viscosity on displacement of a volume of cement slurry.
mud from an eccentric annulus with a standoff of 50% (from Graham’s theoretical developments led him to draw
McLean et al., 1967). Reprinted with permission of SPE.
completely different conclusions from those of McLean et
al. (1967). Muds with low yield points and low plastic vis-
cosities, displaced at the highest possible rate by viscous
cement-slurry velocity in the narrow side of the annulus cement slurries, were recommended. Specific conditions
should be zero, because the yield stress of the cement were imposed on the mud rheology: (τ y)mud < 5 lbf/100 ft2
slurry is much higher than that of the mud. This is effec- and (μp)mud < 12 cp. Although the effects of density dif-
tively what McLean et al. (1967) observed; however, this ferences were not included in this analysis, Graham rec-
did not prevent the mud from being driven by the ommended relying more on viscous forces—hence high
cement from the narrow side to the wide side. Although rates and low yield points—than density differences. In
Point 2 is logical, the authors presented little data to most practical situations, the density difference is favor-
support it. The limited number of tests they performed able; thus, Graham’s conclusions are conservative.
with density differences tend to show that gravitational Jamot (1974) extended Graham’s model by introduc-
forces do play a role on mud displacement in an eccen- ing the effect of gravitational forces. He found that the
tric annulus. deformation of the fluid interface caused by eccentricity
McLean et al. (1967) also investigated displacement was minimized at low displacement rates. The best
at high flow rates in the extended transition from lami- results were obtained when the density of the mud was
nar to turbulent flow but using extremely severe condi- significantly lower than that of the displacing fluid (typ-
tions (with the inner pipe lying against the outer pipe). ically >4.2 lbm/gal [500 kg/m3]). Care was taken to min-
Under these circumstances, when displacing a mud or a imize the gel strength of the mud and to use viscous dis-
cement slurry exhibiting a yield stress, better results placing fluids. On the other hand, turbulent flow was

170 Well Cementing

more effective when the density differences were small Ryan et al. (1992) describe the use of a similar model
(typically <1.7 lbm/gal [0.2 g/cm3]). In between, both implemented into a simulator for field use.
flow regimes showed equivalent efficiencies, and lami- An advance in the modeling was presented by Martin
nar flow gave the poorest results in all cases (Fig. 5-28). et al. (1978). They considered the flow as 2D, the values
of all parameters being averaged along directions per-
pendicular to the cylindrical surfaces. The flow equa-
3.0 tions were solved by making an analogy with those gov-
Mud Cement erning the flow of two immiscible fluids in porous media.
2.8 Plastic viscosity (cp) 10 30
Yield point (lbf/100 ft2) 10 50 Jamot’s (1974) recommendations regarding the opti-
2.6 Density (lbm/gal) 9.5 13.8 mum flow regime and density ratio were largely con-
Hole diameter (in.) 8.0
2.4 Pipe diameter (in.) 5.5
firmed. Martin et al. claimed that the density and vis-
Standoff (in.) 1.0 cosity ratios had a similar effect, but this statement was
2.2 Standoff (%) 80 purely qualitative.
2.0 Cement top, wide side
Unlike previous studies, Martin et al. (1978) investi-
of the annulus gated the displacement of a given fluid by two others
1.8 (e.g., a spacer fluid and a cement slurry). Their model
Fractional 1.6 demonstrated that, at low displacement rates, a spacer
volume Total volume could have a negligible or even a negative effect on mud
1.4
displaced displacement. They assumed the fluids were separated
1.2 Cement top, narrow side by sharp interfaces, with no mixing zones or possible
1.0
of the annulus chemical interactions. The results showed that, to be
effective, the density and rheology of the spacer fluid
0.8 must be between those of the mud and cement, espe-
0.6 3 5 10 15 cially for smaller standoffs. When outside this range, the
Pump rate (bbl/min) spacer would tend to flow preferentially on the wide
0.4
side. In extreme cases, the flow was confined to either
0.2 the wide side or the narrow side of the annulus; conse-
0
quently, the cement slurry would directly contact the
0 10 20 30 40 50 mud.
Friction-pressure gradient for mud (psi/1,000 ft)
Tehrani et al. (1992) presented results from a numer-
ical model close to that of Martin et al. (1978), together
Fig. 5-27. Effect of flow rate on displacement of a mud by a cement with experiments that validated the model. This model
slurry from an eccentric annulus (from Graham, 1972). Reprinted was further refined by Bittleston et al. (2002): The
with permission from Oil & Gas Journal.
azimuthal-flow simulation was improved, and a lubrica-
tion assumption was used in the radial direction. This
allowed more realistic simulations of displacement in
100 horizontal sections. The focus was to understand the
Cement slurry specific gravity 1.95 fluid behavior close to the displacement front. In partic-
90
Flow regime
80 Slow flow Laminar Turbulent
ular, significant azimuthal flows that aid displacement
Mud specific gravity
were observed in the vicinity of the front. This feature is
70 1.95 reminiscent of similar observation by Szabo and
Badly 60 Hassager (1997).
cemented 50
Allouche et al. (2000) developed a model that simu-
length lates displacement in the axial-radial direction to com-
(%) 40 plement the Bittleston et al. model. The model shows
1.85 that complex flow takes place at the displacement front,
30
1.70 with recirculation loops that correlate with the presence
20
1.50 of a static layer of undisplaced fluid (Fig. 5-29).
10 1.35
1.20
This flow makes the basic axial-displacement model
0 described by McLean et al. (1967) largely conservative;
80 100 1,000 10,000 in fact, precise numerical simulations showed that the
Flow rate (L /min) actual static-layer thickness is less than the maximum
value (Fig. 5-30). Allouche et al. (2000) argued that the
Fig. 5-28. Effect of flow rate and flow regime on the displacement layer thickness is better predicted by using a calculation
of muds of various densities by a cement in an eccentric annulus
(standoff = 80%) (Jamot, 1974).

Chapter 5 Mud Removal 171

(a) 1. Stable or piston-like displacements are obtained
when the combination of buoyant and viscous forces
is sufficient to overcome the destabilizing effects of
standoff and viscous fingering.
2. When viscous fingering occurs, a mud channel is
formed. The mud velocity in the channel may be such
that the channel eventually disappears once enough
X=0 X
(b)
displacing fluid has been pumped.
3. In cases in which the yield stress of the mud is large,
buoyant and viscous forces are insufficient to break
the gelled mud, and the mud channel is truly static.
4. Finally, gravity instabilities (reminiscent of Rayleigh-
Taylor instabilities) may appear when the buoyant
forces are large and unfavorable. Such instabilities
X–0 X
occur when a dense fluid is placed on top of a light
Recirculation fluid, with viscosity dampening the exchange flow.
Fig. 5-29. Schematic illustration of the two types of streamline This last pattern is often observed when low-viscosity
behavior at the displacement front: a) no recirculation in displaced and low-density washes are pumped in turbulent flow.
fluid; b) with recirculation in displaced fluid (from Allouche et al.
2000). Copyright Cambridge University Press. These results show that good modeling of azimuthal
flow is necessary to capture the effect of azimuthal
velocity components, especially in the displacement
front region between each pair of fluids. Note how the
0.5 velocity field is complex around the fluid interfaces,
even for piston displacement. This means that, locally,
0.4 there are large azimuthal velocities. Only a correct
Comparison of account of these azimuthal velocities can produce mean-
shear stress ingful simulations.
with yield stress
0.3 Nguyen et al. (1992) and Nguyen and Rahman (2002)
Layer
thickness described a model specifically designed to simulate dis-
0.2
placement in horizontal sections, first in concentric
annuli, then in eccentric situations. However, the rheo-
logical model did not include a yield stress component,
0.1 which restricts the use of this model.
Energy As computing power increased, computational fluid
dissipation dynamics (CFD) codes were used to model fluid dis-
placement. Vefring et al. (1997) and Biezen et al. (2000)
0.2 0.4 0.6 0.8
proposed using software that simulates the 3D flow of
Yield stress (Pa) power-law fluids in an eccentric annulus. However, no
Fig. 5-30. Variation of layer thickness versus yield stress of dis- details were provided on the computations, and few
placed fluid. Points marked by a circle are experimental transient results were presented in terms of displacement effi-
layers, while those marked by a cross are experimental static ciency. Such simulations are presently used to comple-
layers. The two lines correspond to the two different theoretical cri- ment physical experiments. Various boundary conditions
teria (from Allouche et al., 2000). Copyright Cambridge University can be simulated, including casing rotation, but
Press.
presently available computer power prevents general
application of this approach.
Szabo and Hassager (1997) simulated the 3D dis-
based on energy dissipation than by comparing the shear placement of two immiscible Newtonian fluids in a verti-
stress of the flowing fluid with the yield stress of the cal annulus and provided detailed analyses of several
immobile fluid. interesting situations. For concentric annuli, starting
Ladva et al. (2001) showed typical displacement pat- from a circulation situation—for which the interface
terns that arise depending on the principal forces has a parabolic velocity profile, known as a Poiseuille
described in Section 5-3.2 (Fig. 5-31). shape—the interface shape becomes increasingly flat as

172 Well Cementing

 Velocity (ft/sec) Fluids Concentrations Velocity (ft/sec) Fluids Concentrations
4,600 4,600 4,600 4,600
4,800 4,800 4,800 4,800
5,000 5,000 5,000 5,000
5,200 5,200 5,200 5,200
Depth 5,400 Depth 5,400 Depth 5,400 Depth 5,400
(ft) 5,600 (ft) 5,600 (ft) 5,600 (ft) 5,600
5,800 5,800 5,800 5,800
6,000 6,000 6,000 6,000
6,200 6,200 6,200 6,200
6,400 6,400 6,400 6,400

Wide Narrow Wide Wide Narrow Wide Wide Narrow Wide Wide Narrow Wide

5.27–5.82 4.71–5.27 Fluid 1 5.59–6.23 4.94–5.59 Fluid 1

4.16–4.71 3.61–4.16 Fluid 3 4.30–4.94 3.65–4.30 Fluid 3
3.06–3.61 2.51–3.06 Fluid 2 3.00–3.65 2.36–3.00 Fluid 2
1.96–2.51 1.41–1.96 1.71–2.36 1.07–1.71
0.86–1.41 0.42–1.07

Velocity (ft/sec) Fluids Concentrations Velocity (ft/sec) Fluids Concentrations

4,600 4,600 4,600 4,600
4,800 4,800 4,800 4,800
5,000 5,000 5,000 5,000
5,200 5,200 5,200 5,200
Depth 5,400 Depth 5,400 Depth 5,400 Depth 5,400
(ft) 5,600 (ft) 5,600 (ft) 5,600 (ft) 5,600
5,800 5,800 5,800 5,800
6,000 6,000 6,000 6,000
6,200 6,200 6,200 6,200
6,400 6,400 6,400 6,400

Wide Narrow Wide Wide Narrow Wide Wide Narrow Wide Wide Narrow Wide

15.7–17.7 13.8–15.7 Fluid 1 15.1–17.0 13.2–15.1 Fluid 1

11.8–13.8 9.8–11.8 Fluid 3 11.3–13.2 9.4–11.3 Fluid 3
7.9–9.8 5.9–7.9 Fluid 2 7.5–9.4 5.7–7.5 Fluid 2
3.9–5.9 2.0–3.9 3.8–5.7 1.9–3.8
0.0–2.0 0.0–1.9

Fig. 5-31. Typical velocity maps and corresponding displacement patterns for three fluids, calculated by
a 2D simulator (from Ladva et al., 2001). Upper left: Stable displacement of Fluid 1 (mud) by Fluid 2 (pre-
flush), in turn displaced by Fluid 3 (cement slurry). Upper right: Viscous channel; the mud in the channel
is flowing. Lower left: Static channel; the mud in the channel is truly static. Lower right: Unstable dis-
placement caused by large unfavorable buoyancy forces. Reprinted with permission of SPE.

the fluid properties are varied and gravitational forces favorable density ratios improve fluid displacement in
exceed viscous forces by increasing buoyancy and keep- both the radial and azimuthal directions, provided the
ing the Reynolds number low. Eventually, the interface buoyancy number is large enough and the Reynolds
becomes steady. Consequently, buoyancy forces signifi- number remains low.
cantly improve the displacement efficiency by draining CFD tools are also used for the related problem of
more of the displaced fluid film at the wall. They also hole cleaning (King et al., 2000); their use will continue
observed that secondary flows appear on both sides of to grow.
the interface. The displaced fluid from the wall region is Understanding the effects of combinations of density
moved toward the center and, behind the interface, dis- and viscosity differences remains a subject of academic
placing fluid from the center is moved towards the wall. research (Lajeunesse et al., 1999), as is investigation of
For eccentric annuli, Szabo and Hassager observed cases in which the fluids are non-Newtonian and have
that a sufficiently great density difference promotes the same density (Lindner et al., 2000). Thus, the indus-
azimuthal flow, thereby increasing the relative velocity try does not have a good understanding of field condi-
in the narrow side. This effect seemed to make the dis- tions that combine non-Newtonian rheology, density dif-
placement efficiency independent of eccentricity. From ferences, and nonuniform gap width (for eccentric
these results, Szabo and Hassager (1997) concluded that annuli); however, recent work of Bittleston et al. (2002)

Chapter 5 Mud Removal 173

and Pelipenko and Frigaard (2004) provide directly which they are not. For upward mud flow, Eqs. 5-13 and
useful results on the stability of mud channels in such 5-14 show the necessary conditions.
conditions.
If ρmud < ρcem ,
2
5-3.5 Laminar mud displacement criteria ⎡⎛ d p ⎞ ⎤ 2 ⎛ dp⎞
Besides these computational models, Lockyear et al. ⎢⎜⎝ ⎟⎠ ⎥ + D + 2 ⎜⎝ ⎟⎠ Dcos α
⎣ d z f⎦ dz f
(1989), Couturier et al. (1990), and Ryan et al. (1992) then: > 1.
proposed various rules, based on the analysis of a 1D Y
(5-13)
model, that can be routinely used. In short, these rules
are based on an analysis of forces far from the displace- ⎛ dp⎞
ment front. These authors’ key objective was to combine ⎜⎝ ⎟⎠ + D cos α
dz f
density and viscosity effects and define design criteria If ρ mud > ρ cem , then: > 1. (5-14)
that would prevent or minimize mud channels, as well as Y
ensure that potentially immobile mud on the narrow Eq. 5-15 will lead to downward mud flow.
side is mobilized.
In a mud-channel situation, the fluids are flowing at ⎛ dp⎞
different speeds. Couturier et al. (1990) defined a rule to ⎜⎝ ⎟⎠ + D cosα
dz f
prevent mud-channel growth. The velocity of the mud on <1 (5-15)
the narrow side must be greater than or equal to the Y
cement velocity on the wide side. In these expressions, (dp/dz)f is the pressure gradi-
ent, D = (ρcem – ρmud)g, with Lmin being the minimum
⎛ dp⎞ ⎛ dp⎞
⎜⎝ d z ⎟⎠ (
+ ρ mud g cos θ < ⎜ ⎟
⎝ dz ⎠
) (
+ ρ cem gcos θ ) annular width. Y is the yield force of the displaced fluid
[2(τy)mud/Lmin] and α the deviation angle. The ratio
f - mud f - cem
(P + D cos α)/ Y is called the displacement number.
(5-10) Kroken et al. (1995) made an additional suggestion
for horizontal sections. Under such conditions, the
For the mud, the pressure gradient (dp/dz)f-cem casing frequently lies along the bottom of the borehole
assumes a slot width equal to the gap size on the narrow wall. They recommended allowing the dense cement
side of the annulus. For the cement, the slot width is slurry (which usually channels on the top side of
equal to the gap size on the wide side. the annulus) to fall by gravity onto the narrow side, dis-
According to Lockyear et al. (1989) and Couturier et placing the mud. For this situation to happen, the fluid
al. (1990), static mud on the narrow side will be forced velocities must be low enough for inertial forces to be
to move if the pressure exerted by the displacing fluid less than buoyancy forces. This leads to the following
flowing on the large side creates a wall shear stress on relationship.
the narrow side that exceeds the mud yield stress. This
reasoning leads to the following expressions for the min- (
gdo ρ cem − ρ mud ) > 1. (5-16)
imum pressure gradient—(dp/dz)f-min . 2
ρ mud v
⎛ dp⎞
=
( ) mud
2 τy
(5-11) In this equation, do is the hole diameter and v– is the
⎜⎝ d z ⎟⎠ ⎛ L ⎞ fluid velocity. It is important to note that Kroken et al.
f - min
Lmin ⎜ 1 − min ⎟ (1995) assumed that the mud was not gelled. In this
⎝ 4 ro ⎠
approach, fluid yield stresses are ignored. Therefore,
for Lockyear et al. (1989). this calculation cannot be used if the mud is gelled on
the narrow side. This is confirmed by 2D numerical sim-
⎛ dp⎞
=
2 τy( ) mud
(5-12)
ulations using the Bittleston et al. (2002) model.
Figure 5-32 shows that lowering the displacement rate
⎜⎝ d z ⎟⎠ Lmin enhances the displacement of the spacer (which has a
f - min
low yield stress). As a result, a channel of static mud is
for Couturier et al. (1990). left on the narrow side.
Ryan et al. (1992) proposed another set of expres- For horizontal wells, Jennings (1995) proposed using
sions that includes buoyancy effects, distinguishing the a pair of spacers and a pair of cement slurries. Each fluid
cases in which gravity forces are favorable from those in pair consists of a high-density fluid followed by a low-

174 Well Cementing

 1,400 the 1960s, with the introduction of cement dispersants
that allowed turbulence at achievable pump rates
1,500 (Chapter 3). In 1964, Brice and Holmes reiterated the
need for turbulent flow and suggested that the annular
1,600
space should be in contact with the turbulent displacing
Depth fluid for a sufficient time. However, it is difficult to define
1,700
(m) an optimum contact time from their data (Fig. 5-19).
1,800 Since then, some engineers have claimed that a 4-min
contact time is sufficient, while for others a 10-min
1,900 interval is necessary. From the Brice and Holmes data
2,000
(1964), 8 min appears to be a reasonable guideline; how-
ever, such an assumption would ignore other important
Wide Narrow Wide Wide Narrow Wide factors such as casing movement and centralization. The
Mud Viscous spacer turbulent-flow displacement technique has since gained
Tail slurry Chemical wash wide acceptance, because it has greatly increased the
Lead slurry success rate of cement jobs in many areas. For instance,
Raymondeau et al. (1991) reported excellent results
Fig. 5-32. Simulation of laminar displacement in a horizontal section,
showing the effect of displacement rate. Left: 5 bbl/min [800 L/min]; when the preflushes and the cement slurry were
right: 0.3 bbl/min [50 L/min]. pumped at rates that ensured turbulent flow all around
the eccentered annulus. For these jobs, the minimum
contact time was set at 5 min.
density fluid. The high-density fluid displaces the mud Today, there is a clear agreement that low-viscosity,
on the low side, while the lighter fluid fills channels low-density fluids such as water, diesel oil, or a chemical
remaining on the high side of the horizontal section. wash effectively remove mud, provided certain condi-
To summarize this review of models and criteria, one tions are met. Terms such as erosion, drag, mixing, or
can classify these models according to their complexity. dilution are used to explain the underlying mechanisms.
Actually, the fundamentals behind this practice are
■ One-dimensional models provide semiquantitative
not well understood. Clearly, turbulent flow alone is not
rules or criteria. They are limited to describing simple sufficient. In pilot-scale experiments, replacing a viscous
steady-state situations but have been successfully fluid flowing in laminar flow by a thin fluid pumped at
used for routine operations. the same rate usually leads to a sharp pressure-gradient
■ Two-dimensional models provide a reasonable analy- decrease. However, displacing a laminar fluid by a tur-
sis of either axial-radial flow (e.g., predicting the size bulent fluid at an equal friction-pressure drop will lead
of the mud layer along the walls) or axial-azimuthal to better cleaning (Clark and Carter, 1973). Note that
flow, i.e., mud channels. Early models suffered from shear stresses generated by unweighted fluids even at
various limiting assumptions that restricted their use very high Reynolds numbers are lower than those of vis-
to wells with limited deviation and relatively narrow cous spacers pumped at the same conditions. Table 5-6
annuli. In addition, the erosion of static layers was shows, for different annular sizes and rates, the
not modeled. The latest models do not suffer from Reynolds numbers, friction pressures, and wall shear
these limitations and are the ideal tools to perform stresses for water, oil, chemical washes, and weighted
parametric studies combining all the relevant forces spacers.
acting together. Hence, the conditions for efficient turbulent flow are
■ Three-dimensional models allow a good understand- more difficult to understand than those for laminar flow.
ing of the various phenomena but, owing to their com- Mechanisms other than the simple hydrodynamic forces
plexity, they remain research tools today. As available considered above are at play. Some of them are detailed
computational power continues to increase, such in Section 5-2.4.6. In this section, three types of static
tools will eventually enter the mainstream. materials are distinguished: gelled mud, mud filtercake,
and solids layers. Their properties differ widely, and they
can be found in different parts of the annulus. Thus, dif-
5-3.6 Turbulent flow and erosion mechanisms ferent mechanisms may be required to remove them.
As discussed earlier, Howard and Clark (1948) achieved
good mud-displacement efficiencies when using cement
slurries in the upper-laminar and turbulent-flow regimes.
Turbulent-flow displacement became commonplace in

Chapter 5 Mud Removal 175

Table 5-6. Reynolds Numbers, Friction-Pressure Drops, and Wall Shear Stresses Calculated for Three Displacement Fluids and Three Different Centered Annuli,
in Both Système International (SI) and U.S. Units.†
Newtonian Wash—Density 1,000 kg/m3, Viscosity 1 cP
ID OD Flow Rate = 2 bbl/min [318 L/min] Flow Rate = 5 bbl/min [795 L/min] Flow Rate = 10 bbl/min [1,590 L/min] Flow Rate = 15 bbl/min [2,385 L/min]
(in.) (in.)
NRe (dp/dz)f τw NRe (dp/dz)f τw NRe (dp/dz)f τw NRe (dp/dz)f τw
– (psi/1,000 ft) (lbf/100 ft2) – (psi/1,000 ft) (lbf/100 ft2) – (psi/1,000 ft) (lbf/100 ft2) – (psi/1,000 ft) (lbf/100 ft2)
4 51⁄2 27,964 8.8 4.0 69,909 44.2 19.9 139,818 151.3 68.1 209,727 312.5 140.6

7 8.5 17,139 3.8 1.7 42,848 18.6 8.4 85,695 63.4 28.5 128,543 130.3 58.6

9.625 12.25 12,144 0.4 0.3 30,361 1.9 1.5 60,721 6.43 5.1 91,082 13.2 10.4
(m) (m) NRe kPa/m Pa NRe kPa/m Pa NRe kPa/m Pa NRe kPa/m Pa
0.102 0.140 27,964 199.7 1.9 69,909 999.8 9.5 139,818 3,422.5 32.6 209,727 7,068.9 67.3

0.178 0.216 17,139 85.3 0.8 42,848 420.7 4.0 85,695 1,434.1 13.7 128,543 2,947.5 28.1

0.244 0.311 12,144 8.8 0.1 30,361 43.0 0.7 60,721 145.5 2.4 91,082 298.6 5.0
Turbulent Newtonian Spacer—Density 1,319 kg/m3, Viscosity 18 cP
ID OD Flow Rate = 2 bbl/min [318 L/min] Flow Rate = 5 bbl/min [795 L/min] Flow Rate = 10 bbl/min [1,590 L/min] Flow Rate = 15 bbl/min [2,385 L/min]
(in.) (in.)
NRe (dp/dz)f τw NRe (dp/dz)f τw NRe (dp/dz)f τw NRe (dp/dz)f τw
– (psi/1,000 ft) (lbf/100 ft2) – (psi/1,000 ft) (lbf/100 ft2) – (psi/1,000 ft) (lbf/100 ft2) – (psi/1,000 ft) (lbf/100 ft2)
4 51⁄2 2,049 19.5 8.8 5,123 118.2 53.2 10,246 384.1 172.8 15,368 771.4 347.1

7 8.5 1,256 12.0 5.4 3,140 50.7 22.8 6,280 166.8 75.1 9,419 332.7 149.7

9.625 12.25 890 1.6 1.2 2,225 4.0 3.1 4,450 17.4 13.7 6,674 34.5 27.2

(m) (m) NRe kPa/m Pa NRe kPa/m Pa NRe kPa/m Pa NRe kPa/m Pa
0.102 0.140 2,049 441.3 4.2 5,123 2,763.8 25.5 10,246 8,688.6 82.8 15,368 17,449.5 166.2

0.178 0.216 1,256 270.5 2.6 3,140 1,146.9 10.9 6,280 3,773.1 35.9 9,419 7,525.9 71.7

0.244 0.311 890 35.7 0.6 2,225 89.6 1.5 4,450 393.6 6.6 6,674 780.4 13.0
Power Law Laminar Spacer—Density 1,557 kg/m3, n = 0.49, k = 0.0071 lbf-s/ft2 [0.34 Pa-sn]
ID OD Flow Rate = 2 bbl/min [318 L/min] Flow Rate = 5 bbl/min [795 L/min] Flow Rate = 10 bbl/min [1,590 L/min] Flow Rate = 15 bbl/min [2,385 L/min]
(in.) (in.)
NRe (dp/dz)f τw NRe (dp/dz)f τw NRe (dp/dz)f τw NRe (dp/dz)f τw
– (psi/1,000 ft) (lbf/100 ft2) – (psi/1,000 ft) (lbf/100 ft2) – (psi/1,000 ft) (lbf/100 ft2) – (psi/1,000 ft) (lbf/100 ft2)
4 51⁄2 1,777 26.3 11.8 7,088 82.7 37.2 20,187 233.1 104.9 37,236 435.8 196.1

7 8.5 848 20.7 9.3 3,384 33.0 14.8 9,639 111.6 50.2 17,780 205.1 92.3

9.625 12.25 285 5.8 4.5 1,137 9.0 7.1 3,238 12.74 10.0 5,972 27.9 22.0
(m) (m) NRe kPa/m Pa NRe kPa/m Pa NRe kPa/m Pa NRe kPa/m Pa
0.102 0.140 1,777 594.0 5.7 7,088 1,870.7 17.8 20,187 5,272.9 50.2 37,236 9,858.1 93.9

0.178 0.216 848 467.6 4.5 3,384 746.5 7.1 9,639 2,524.5 24.0 17,780 4,639.5 44.2

0.244 0.311 285 130.3 2.2 1,137 204.3 3.4 3,238 288.2 4.8 5,972 631.1 10.5
† In almost all situations, the shear stress for a turbulent wash is less than that for a viscous laminar spacer.

Several experimental observations have been ference on the removal efficiency in both turbulent
described that help explain the basic cleaning mecha- and laminar flow, but they did not provide details
nisms. about the basic mechanisms. However, gravity insta-
■ Gravity instabilities bilities were most likely present. Gravity instabilities
are also observed in numerical simulations (Fig. 5-31,
Lockyear et al. (1989) observed the behavior of a
lower right).
heavy fluid on the narrow side of the annulus being
■ Shear instabilities
bypassed by a lighter fluid. The heavy fluid slumped
down and moved toward the wide side of the annulus. An interface between two Newtonian fluids flowing at
It was eventually entrained and absorbed into the different velocities is intrinsically unstable. This is the
light fluid (Fig. 5-33). Ryan et al. (1992) proposed well-known Kelvin-Helmholtz instability in which
Eq. 5-15 to quantify the threshold conditions allowing waves form (Fernando, 1991). Gabard (2000)
this phenomenon. Such instabilities lead to observed a similar interface instability between a
azimuthal flow. Similar slumping can eventually take static non-Newtonian fluid and a flowing Newtonian
place when the density ratio is reversed. The slump- fluid (Fig. 5-34). At present, there is no clear under-
ing occurs in the direction of flow. standing of these instabilities for non-Newtonian
Tehrani et al. (1993) observed similar gravity- fluids. Note that these instabilities are thought to orig-
driven instability in laminar flow. Jakobsen et al. inate from the drag stress at the interface of the two
(1991) observed marked effects of fluid-density dif- fluids. This stress is not included in most 2D models

176 Well Cementing

 Fluid A can cause detachment. This mechanism is similar to
backpulse cleaning used in other industries.
Shear flow over a bed of particles creates drag and
lift forces on individual particles. This can lead to par-
ticle fluidization and entrainment by the flow. This is
the basic mechanism used for modeling cuttings-bed
erosion and transport (Doron and Barnea, 1993).
Abrasion of external mud filtercake in turbulent
flow is unlikely under cementing conditions. The
presence of large particles is required, a condition
found in gravel-packing treatments (Johnson et al.,
1992; Becker and Gardiner, 2000).
Fluid B
■ Chemical and physico-chemical effects
(i) (ii) (iii)
Chemical and physico-chemical mechanisms include
Fig. 5-33. Illustration of slumping caused by density differences the diffusion of dispersants into the static layer (lead-
during displacement. Fluid A is denser than Fluid B (from ing to the weakening of the structure), a decrease of
Lockyear et al., 1989). Reprinted with permission of SPE. interfacial tension by surfactants, and a change of
wettability of solid particles by the action of deter-
gents. Degradation of polymers in the static material
by oxidizers or enzymes is also possible; however, this
method is not presently used in cementing opera-
tions. It has become a standard practice in comple-
tion operations (Ali et al., 2001).
In summary, various turbulent cleaning mechanisms
may destabilize a channel of gelled mud or a solids layer.
Once the interface is disturbed, it becomes easier for the
particles to detach and be carried by the flow. Turbulent
flow ensures that the particles are carried all around the
annulus. Chemicals present in the displacing fluids dif-
fuse into the mud and accelerate its effective dispersion.
The variety of possible mechanisms explains why
using a single parameter such as shear stress or
Fig. 5-34. Example of viscous finger formed when a low-viscosity
Reynolds number is too simplistic. However, a few empir-
fluid—colored in red—pushes a viscous transparent fluid in a ical expressions have been proposed to describe erosion.
tube (upward displacement). In this example, the fluids have the ■ Martin et al. (1978) defined an erosion index that is
same density. Although initially the finger is smooth and stable, it equal to the local shear stress in laminar flow and
becomes unstable after some time, allowing the layer of thick fluid
to be displaced eventually (from Gabard, 2000). proportional to the square of the shear stress in tur-
bulent flow. They said that for gel breakdown, the ero-
sion index should be at least 25% higher than the mud
gel strength, but no experimental data were pre-
(see Section 5-3.4). These models are thus intrinsi- sented to support this definition.
cally unable to reproduce this class of instabilities.
■ Haut and Crook (1981) defined a mud mobility
■ Turbulent eddies and local pressure bursts factor and found an experimental correlation of this
In turbulent flow, eddies are carried with the flow, factor with the flow velocity and the displacement
and they create pressure fluctuations. The pressure efficiency. Later, Smith and Ravi (1991) extended
varies at any given time and in any given place. The the range of operational parameters by comparing
pressure variations are transmitted to the static layer various drilling fluids. Their general conclusion was
as fluctuating stresses that may be sufficient to flu- that “[muds with] similar mud mobility factors do
idize gelled layers and solids beds (Plett, 1985). not necessarily exhibit the same filtercake-erosion
These pressure fluctuations may eventually be characteristics.”
large enough to create an instantaneous negative dif-
ferential pressure across the mud filtercake, which

Chapter 5 Mud Removal 177

■ Ravi et al. (1992) used a calculation similar to that of [0.16 to 0.36 g/cm3] are sufficient to create azimuthal
Lockyear et al. (1989) to deduce the shear strength of flow through a slumping mechanism. When considering
a mud filtercake (Eq. 5-11). This value was in the field constraints, the best turbulent cleaning fluids are
same order of magnitude as that calculated from the low-viscosity preflushes such as base oil or water
theoretical equation, but no direct measurement of pumped at the highest possible velocities. If possible,
the filtercake shear strength was performed. Ravi et dispersants should be added to thin the mud, and deter-
al. defined an erodability index proportional to the gents should be added to leave solid surfaces water-wet.
reciprocal of the measured yield strength. Only large-scale phenomena can be modeled by CFD
These approaches highlight the problems encoun- codes. These codes require careful modeling of the
tered when describing complex erosion phenomena azimuthal-pressure field. The evaluation of the other
using simple parameters. Nevertheless, present qualita- potential effects described above is restricted to experi-
tive understanding allows identifying the relevant mech- mental testing and empirical quantification. It is essen-
anisms and drawing a phenomenological picture of tur- tial to make sure the test methods mimic or reproduce
bulent displacement. From here, relevant parameters downhole phenomena. Various methods have been pro-
can be identified despite the absence of rigorous quanti- posed that are described in Appendix B. One of the
tative criteria. approaches is to study each of the physico-chemical
effects separately, using a specific apparatus. Various
■ At large length scales, thin washes usually channel
cleaning solutions can thus be compared.
through the mud because of their small friction-pres- None of the mechanisms presented above concerning
sure gradient. However, the large density difference the efficiency of turbulent-flow-displacement occur
between the narrow and wide sides of the annulus instantaneously. Therefore, a minimum contact time is
allows gravity slumping of heavy mud, widening the required to achieve complete mud removal. In the
channel of wash. absence of a complete understanding of the effect, and
■ At intermediate-length scales, the turbulent eddies the difficulty of deriving it from laboratory experiments
act by diluting the mud when mixed with the wash. (the length scale being too short), field experience
The eddies also carry detached mud particles into the should dictate a reasonable value when available.
main flow. At the wash/static-mud interface, the fluc- For the turbulent-flow displacement technique to be
tuating hydrodynamic forces destabilize the inter- successful, several criteria must be met.
faces and fluidize solids beds. This is an erosion ■ The displacing fluid must be sufficiently thin for the
mechanism. critical pumping rate to be achievable with field
■ At small scales, chemicals weaken the mud structure equipment. This implies that the viscosity of the dis-
(dispersing effect) and decrease interfacial forces. placing fluid should be much lower than that of the
This enables easier mixing and reduces the viscosity mud, at least under the specific flow conditions. Even
of the wash/mud mixture. Surfactants adsorbed on if turbulent flow exists all around the annulus, the
surfaces are replaced by detergents, providing water- displacing fluid may channel through the mud
wet surfaces. through a viscous fingering phenomenon. In this situ-
To summarize the previous discussion and experi- ation, various erosion mechanisms may help break
mental findings in terms of physics, the principal forces the mud channel.
at play in turbulent displacement are inertial and buoy- ■ The chemical and physical properties of the displac-

ant forces. Increasing inertia increases pressure and ing fluid must be carefully designed (Section 5-4). It
velocity fluctuations. Increasing buoyancy allows large- is of utmost importance for the displacing fluid to be
scale slumping. Viscous forces dampen both mecha- fully compatible with the mud. In addition, a
nisms. The intensity of turbulence is traditionally quan- weighted displacing fluid must be able to suspend the
tified by the parameter solids required to achieve the designed density on the
surface and under downhole conditions during place-
( v′ )2 , ment.
■ A single fluid usually cannot possess all the optimal
v physico-chemical requirements. A sequence of fluids
is therefore required.
the ratio of the square root of the mean square velocity
fluctuation, v′, to mean velocity, –v . In established turbu- Even if the above conditions can be fulfilled, there are
lent flow, this quantity is a constant. Therefore, the certain well conditions that can make this technique
higher the velocity, the higher the velocity fluctuations impractical or impossible.
will be. Empirically, density differences of 2 to 4 lbm/gal

178 Well Cementing

■ For an unweighted displacing fluid, the volume nec- These results show that the friction pressure, hence
essary to achieve a given contact time may be such the shear stress, does not fully characterize the cleaning
that pore pressure cannot be controlled. properties of the displacing fluid. In addition to this
■ For weighted displacing fluids, the critical pumping parameter, turbulent flow has specific features that lam-
rate for turbulent flow may exceed the capabilities of inar flow does not display (Section 5-3.6).
the available equipment, a reduction in flow rate may Lockyear and Hibbert (1988) and Lockyear et al.
occur when the displacing fluid rounds the shoe (1989) published some interesting experimental results
because of U-tubing, or the required volume of fluid that are supported by theoretical arguments. For effi-
may be cost-prohibitive. cient mud displacement in an eccentric annulus, the
■
friction pressure during the displacement should meet
Weak formations with low fracture gradients may not
the condition given by Eq. 5-11. This equation, which
be able to withstand the pressures associated with
assumes that the mud is flowing on the narrow side of
high displacement rates.
the annulus, was verified by their experimental results
■ Unstable formations may not be able to withstand the (Table 5-7 and Fig. 5-36).
high velocities without being eroded. Lockyear and Hibbert also claimed that the velocity
of the displacing fluid should be greater than zero on the
narrow side of the annulus far away from the interface,
5-3.7 Experimental investigations of immobile a condition McLean et al. (1967) found unnecessary. As
mud removal discussed in Section 5-2, satisfying such conditions does
Using a simulated borehole, Clark and Carter (1973) not guarantee optimal circulation efficiency, because
performed an interesting experimental study on the both fluids may flow completely around the annulus but
effect of high eccentricities on gelled mud removal by with a large interfacial-velocity difference between the
cement slurries. They observed poor displacement effi- narrow and the wide side. Lockyear and Hibbert (1988)
ciency when the cement slurry was pumped in laminar gave only a partial answer to this problem. For standoff
flow. Much better results were obtained when the fluids values of 50% in the absence of density differences, they
were pumped under partially turbulent flow conditions observed a sharp transition between severe and minimal
(Fig. 5-35) for the same calculated frictional pressure. channeling for an average Reynolds number of 1,500 for
They also observed that, for a given pressure drop, the the displacing fluid. This confirms that good displace-
results improved as the viscosity of the displacing fluid ment may be obtained, even in eccentric annuli, when
decreased. McLean et al. (1967) showed that turbulent the Reynolds number of the displacing fluid is in the
flow in itself does not reduce channeling. In some cases, upper-laminar or turbulent-flow range.
high shear stresses from laminar flow lead to reduced Tehrani et al. (1992) performed a series of carefully
mud channeling. controlled laminar flow experiments that demonstrated
the importance of density contrast and rheological hier-
archy. An interesting result was the observation of
Flow Pressure Drop No Pipe azimuthal instabilities that improve displacement.
Regime Calculation Method Movement These instabilities were further described by Tehrani et
Laminar Shear stress – Cement and mud al. (1993). Large density differences promote instabili-
shear rate density equal
Turbulent Reed slide rule (unless otherwise ties, while high flow rates have the opposite effect.
indicated) However, these azimuthal instabilities in laminar flow
100 ρcem – ρmud = 3 lbm/gal leave a fine strip of mud on the narrow side.
80 ρcem – ρmud = 3 lbm/gal Ryan et al. (1995) investigated mud cleanup in hori-
zontal wells. They concluded that increased flow rate
Displace- 60
ment had the clearest positive effect; however, little informa-
efficiency 40 tion was given about the fluids used. Van Vliet et al.
(%) (1995) studied filtercake removal in a slim hole with a
20
poor standoff. The best conditions leading to maximum
0 mud removal corresponded to a mud with low gel
0 20 40 60 80 100 120 140 160 180 200 strength and a thin filtercake. All of the displacing fluids
Friction pressure gradient for cement (psi/1,000 ft) were in laminar flow, and increasing the contact time of
the spacer had no effect.
Fig. 5-35. Effect of cement/mud density difference and cement rhe- Gabard (2000) studied how the layer of displaced
ology on displacement efficiency (from Clark and Carter, 1973).
Reprinted with permission of SPE. fluid left on the wall of a pipe can be displaced under
laminar flow. The condition that the shear stress of the

Chapter 5 Mud Removal 179

Table 5-7. Experimental conditions for various tests shown in Fig. 5-36†
Experiment B C D E G
Mud type KCI/P KCI/P KCI/P KCI/P KCI/P
μp /τy (cp/ lbf⁄100 ft2) 37/10 44/12 39/13 41/13 38/8
Density (sg) 1.62 1.62 1.62 1.63 1.69
10-s/10-min gel‡ 11/12 11/12 13 13 5
Gel during experiment 12 12 12 14 9

Spacer type Type A Type A Type A Type A Type B

μp /τy (cp/ lbf⁄100 ft2) 33/9 39/5 43/11 56/18 28/2
Density (sg) 1.62 1.62 1.62 1.63 1.69
10-s gel 6 7 9 16 1
Volume pumped (bbl) 6.0 5.0 5.3 6.4 6.0

Cement type Neat G Neat G Neat G Neat G Neat G

μp /τy (cp/ lbf⁄100 ft2) 21/42 33/40 21/20 33/37 47/38
Density (sg) 1.9 1.9 1.9 1.80 1.80
10-s gel 21 22 17 28 23
Volume pumped (bbl) 7.0 5.6 2.8 3.2 7.0

Deviation (º) 0 0 0 0 0
Mean standoff (%) 40 40 60 60 50
Narrow side gap at shoe (mm) 8 8 17 17 14
Casing size (in.) 7 7 7 7 7
Hole size (in.) 8.625 8.625 8.625 8.625 8.625

Displacement rate (bbl/min) 8.1 2.0 5.2 2.1 8.2

Maximum annular pressure 26.0 4.4 10.5 16.3 59
drop during displacement
(psi/100 ft)

Calculated minimum gap 1.9 12 5 4 1

to achieve full mud
displacement (mm)
Fluid in narrow side at shoe Cement Mud Cement Cement Cement
† From Lockyear and Hibbert, 1988. Reprinted with permission of SPE.
‡ Gel strength is given in lbf/100 ft2.

180 Well Cementing

 10

6
Distance
from shoe 5
(m)
4

0
–180 0 +180 –180 0 +180 –180 0 +180 –180 0 +180 –180 0 +180
Test C (°) Test B (°) Test E (°) Test D (°) Test G (°)

Fig. 5-36. Distribution of cement, spacer, and mud in the annulus for a number of tests (0° represents
the narrow side of the annulus)(from Lockyear and Hibbert, 1988). Reprinted with permission of SPE.

displacing fluid has to equal the yield stress of the static

fluid was found to be conservative. Careful investiga-
tions of the flow at the displacement front allowed
Allouche et al. (2000) to propose less-severe criteria,
which were verified by experiments (Fig. 5-30). Inter-
face instabilities were also observed that improved
cleaning in laminar flow conditions.
There are various ways to evaluate mud displacement
in the laboratory, such as visualization using model
fluids, monitoring the filtercake thickness (Ravi and
Beirute, 1994), and monitoring the shear bond between
the set cement and the casing (Sweatman et al., 1995).
Another approach is to allow the cement to set, cut the
annulus into slices, and measure the percentage of
annulus filled with cement (Fig. 5-37)(Kimura et al.,
1999).
This short overview of experimental studies shows
that immobile mud displacement in an eccentric annu-
lus most often begins by channeling of the displacing
fluid on the wide side. In laminar flow, provided the
fluids have a suitable density and viscosity ratio, the
interface can reach a steady state. As a result, the chan-
nel does not grow in length with time. If the channel
length grows, increasing the density ratio, the viscosity
ratio, or the flow rate tends to reduce the mud channel
width under laminar flow conditions. A turbulent dis-
placing fluid starts by channeling on the wide side.
Eventually, the mud channel may be destabilized and Fig. 5-37. Example of mud-displacement experiment in an eccentric
cleaned, provided contact time is long enough. If this annulus (Kimura et al., 1999). Reprinted with permission of SPE.
destabilization is not achieved, the mud channel is often
larger than in laminar flow.

Chapter 5 Mud Removal 181

5-3.8 Field studies 5-3.9 Effect of casing movement and
Deriving mud-removal information from field studies is casing hardware
difficult, because the available measurements are lim- Pipe movement during cement placement helps remove
ited. At present, the most direct measurement is ultra- the mud that would otherwise be trapped on the narrow
sonic imaging or sonic logging (Chapter 15). Under side of an eccentric annulus. The basic principle is the
appropriate conditions, mud channels and mud layers same as that during mud circulation; however, the
along the casing wall can be detected. physics involved is more complicated, and published
Brady et al. (1992) compared acoustic logs of wells models including the effect of casing movement are cur-
before and after applying the job-design criteria rently limited to the circulation process only.
described above (Eqs. 5-10 and 5-12) to validate these On the experimental side, McLean et al. (1967)
criteria. Terry (1993), Waters and Wray (1995), and reported a few conclusions concerning the effect of
Kelessidis et al. (1995) further validated these criteria, casing movement on mud displacement between imper-
using cement bond logs. meable walls. They observed that casing rotation is a
Jakobsen et al. (1991) presented three case studies more effective mud-removal technique than casing reci-
that highlight the benefit of turbulent flow when a low- procation. However, as mentioned earlier, they empha-
density wash can be used. Alternatively, if no wash is sized that lateral motion of the casing was not allowed in
used, a positive density difference and laminar flow can their experiments, even though it is likely to happen in
displace the mud on the narrow side in inclined sections. the field. When the pipe rotates at high speed, pure rota-
Kroken et al. (1995) presented other case studies show- tion becomes unstable, resulting in lateral motion.
ing the beneficial effect of a positive density difference Conditions leading to unstable movement are not fully
in highly deviated wells, combined with very low flow understood, but this movement clearly improves mud
rates. displacement (Sanchez et al., 1999). Lateral motion also
Ryan et al. (1992) presented other case studies occurs during pipe reciprocation in inclined wells (Fig.
involving the application of the rules given in Eqs. 5-13 5-12).
through 5-15 that lead to better cement placement. Mechanical devices such as scratchers, scrapers, and
McPherson (2000) described the efficiency of turbulent cable wipers (Chapter 11) also improve the efficiency of
flow together with pipe rotation to clean long horizontal the displacement process when used in combination
sections, provided large volumes of wash were used. with casing movement (Jones and Berdine, 1940; Teplitz
In 1993, Hoskins et al. presented a new approach to and Hassebroek, 1946). New hardware continues to be
predict mud removal efficiency based on historical data. developed with specific objectives such as inducing local
They analyzed field data from 15 wells. In eight wells, turbulent flow for better cleaning of horizontal wells
the mud was displaced in turbulent flow. Laminar flow (Kinzel and Martens, 1998) or wiping the casing surface
displacement was performed in the remaining wells. during reciprocation (Dillenbeck and Simpson, 1999).
Hoskins et al. built a neural-network model to predict These devices are attached to the pipe, and they con-
the cement-bond index as good, average, or bad, based tribute to the erosion of gelled or dehydrated mud that
on the design data. An important conclusion of this work would otherwise remain static in the annulus.
was that the variables are interdependent. However, the With the improvement of necessary equipment such
variables the researchers chose were not explicitly as rotating liner hangers, casing movement is now a
described in the paper. common practice that has been statistically demon-
When detailed databases of cementing jobs are avail- strated to be beneficial (Harris et al., 2001). Casing rec-
able, statistical analysis has been useful to pinpoint the iprocation has been successfully used in many critical
most important parameters leading to successful opera- operations (Kolthoff and Scales, 1982; Holhjem et al.,
tions. Using this method, Silva et al. (1996) determined 1982; Ghosh et al., 1999). Typical amplitudes for casing
that, for highly deviated wells, the most significant vari- reciprocation are on the order of 20 to 40 ft [6 to 12 m],
ables are the annular velocity, the Reynolds number, the a full cycle being completed every 1 to 5 min. These
mud yield strength, and the contact time. methods have an effect similar to knocking the casing
Harris et al. (2001) presented the results of a statisti- or downhole vibrators (Rakhimkulov and Strugovets,
cal analysis of a large database of shoe tests and high- 1990) to prevent gas migration problems (Chapter 9).
lighted a strong correlation of successful shoe tests with There are three main drawbacks associated with casing
casing reciprocation during cement placement. The dis- reciprocation.
placement rate does not appear to be a discriminating ■ Pipe reciprocation induces pressure surges and swab-
factor between success and failure. bing that may adversely affect well control, especially
when the annular clearance is small.

182 Well Cementing

■ There is a risk of the casing becoming stuck. ■ Mud-completion fluid displacement
■ The movement amplitude is reduced downhole The main objective of mud-completion fluid displace-
because of pipe stretching or buckling. Excessive ment is to fill the hole with a noncontaminated fluid
casing pull may be required, especially in highly devi- without destroying the competent filtercake lying
ated wells. against the formation wall.
Pipe rotation improves the quality of primary cement In these two applications, the key phenomena are the
jobs, specifically liner jobs (Landrum et al., 1985; hydrodynamic and chemical effects. In hole cleaning,
Buchan and Little, 1986), without presenting the above large particles are flowing and the active phenomena are
drawbacks. Rotary power tongs or power swivels are mainly based on mechanical and hydrodynamic effects,
used, and the rotation rate usually varies between 10 and such as pipe rotation and abrasion.
40 rpm. The key to the success of this technique is good The preceding discussion demonstrates that a large
torque control. For this reason, power swivels are pre- volume of work has been performed to better under-
ferred. stand mud removal from various vantage points.
Although not a common practice, it is worthwhile to Nevertheless, there is still no clear agreement about the
mention that some operators use both movements (rota- best cleaning methods to apply. Practices and recom-
tion and reciprocation) simultaneously with excellent mendations are still given after reviewing the fluids and
results. However, one must remember that casing move- their properties and requirements.
ment is not the panacea for all mud-displacement prob-
lems. Because the effects of casing movement have been
characterized only qualitatively, other methods for 5-4 Drilling fluids, spacers, and washes
improving primary cement jobs should not be omitted.
Because cement slurries are usually incompatible with
New types of external casing hardware, including per-
most drilling fluids, intermediate fluids called pre-
manent sensors such as fiber optics or perforation guns
flushes are pumped as buffers to prevent contact
(Dillenbeck and Cooper, 2000) create additional mud-
between them. Preflushes can be chemical washes that
removal challenges because good hydraulic isolation is
contain no solids or spacer fluids that contain solids and
required around these protrusions.
are mixed at various densities.
When an incompatible pair of fluids mix, a highly
5-3.10 Mud displacement in other oilfield viscous mass forms that can cause several problems.
applications—Summary and conclusions Cement can channel through the viscous mass.
The problem of mud displacement in primary cementing Unacceptably high friction pressures can develop during
is closely related to mud-mud displacement and mud- the cement job. Plugging of the annulus can result in job
completion fluid displacement. The correlation to hole failure. In all of these situations, zonal isolation is com-
cleaning and cuttings transport is weaker. Proper mud promised, and expensive remedial cementing may be
displacement requires displacing gelled mud, mud fil- required.
tercake, and solids beds. Each has a different focus— The use of preflushes would be unnecessary if the
gelled mud for primary cementing and mud-mud dis- mud were compatible with the cement slurry. This was
placement, mud filtercake for completion operations, one of the objectives of preparing drilling fluids with
and solids layers for hole cleaning. blast-furnace slag (Wilson et al., 1990; Nahm et al., 1995;
Daulton et al., 1995) (Chapter 7). Mud filtercakes con-
■ Mud-mud displacement
taining blast-furnace slags are cementitous and can
Severe compatibility problems may exist (e.g., improve bonding with the formation (Tare, 1997). How-
between an OBM containing a high concentration of ever, field use of blast-furnace slag muds is relatively
CaCl2 and a bentonitic WBM) (Courtney et al., 1999). rare, and well-designed preflushes remain the most com-
The mud can be displaced in successive steps as the monly used solution to prevent contact between the
drillstring is run in. Also, displacement can be per- cement slurry and the mud.
formed either down the annulus (reverse circulation) Compatibility is strongly dependent upon mud chem-
or down the pipe and up the annulus. The goal is to istry; therefore, this section begins with a brief discus-
recover the maximum amount of expensive OBM and sion of common drilling fluids.
to clean all surfaces.

Chapter 5 Mud Removal 183

5-4.1 Drilling fluids Fluid compatibility also depends on the cement-
The drilling fluid is intended to allow the operator to slurry composition. For example, Sweatman et al. (1995)
drill the well as safely and efficiently as possible. These reported that salt-saturated cement slurries are less
objectives involve controlling many mud parameters, interactive with OBMs than those made with fresh water.
including the rheology, density, lubricity, and inhibition The actual mud in a well is often significantly differ-
properties. This is accomplished by properly adjusting ent from one prepared in the laboratory, owing to aging,
the mud composition. There are many types of drilling cuttings contamination, and various treatments made
fluids, categorized as WBMs or invert-emulsion muds. while drilling. WBMs may contain a significant amount
Invert-emulsion muds are water-in-oil emulsions. The of oily components from various parts of the mud circuit,
external phase is an oil, such as diesel, mineral oil, syn- such as formation hydrocarbon, diesel fuel, grease, and
thetic aliphatic oil, or a synthetic ester. The internal pipe dope. Therefore, mud-cement slurry compatibility
phase is an aqueous fluid containing salts to control for- tests should be performed with representative samples
mation integrity. The emulsion stability is maintained by from the field. The compatibility-test procedures are
selecting proper surfactants. A detailed discussion of the described in Appendix B.
various mud types and chemistries is beyond the scope
of this chapter. For more information, the reader is 5-4.2 Preflushes
referred to specialized textbooks such as Darley and
Although the primary purpose of preflushes is to prevent
Gray, 1988.
commingling of the drilling fluid and cement slurry, they
It is worth noting that most muds are not designed for
may also aid mud displacement. As discussed previously,
easy displacement (Harder et al., 1992; Patel, 1998).
mud displacement occurs at various levels. In general,
Some muds are especially difficult to displace because of
fluid rheology and density determine the bulk displace-
their chemistry.
ment, while, on a smaller scale, the preflush chemistry
■ Polymer muds are more difficult to displace than ben- ensures cleaning of the annular walls. Preflushes should
tonite muds, especially when they contain compo- have three principal attributes. They must
nents designed to strongly adsorb onto mineral sur- ■ be compatible with both the mud and the cement slurry
faces.
■ present optimal rheology and density under downhole
■ Invert emulsions contain emulsifiers and oil-wetting
conditions to ensure good bulk mud displacement
surfactants that cover all polar surfaces.
■ have appropriate chemistry to clean solid surfaces
Before displacement, muds are treated or condi- and leave them water-wet.
tioned to aid displacement. The primary method is to
separate the cuttings from the mud by installing screens In addition, preflushes should not cause formation
on the shale shaker with finer mesh size and centrifug- damage by excessive filtration and changes of rock wet-
ing the fluid. When no such equipment is available, an tability (Gambino et al., 2001) (Chapter 6). Often, a
alternate method is to dilute and chemically treat the single fluid cannot meet all of these requirements. In
mud. The appropriate treatment depends on the mud such cases, multiple preflushes can be pumped in
type. Bentonite muds can be thinned by adding disper- sequence (Boyington et al., 1989).
sants, and polymer muds are diluted with water. Invert
emulsions can be diluted with the base oil and are usu- 5-4.2.1 Washes
ally loaded with emulsifiers to maintain or improve their Washes are preflushes with a density and a viscosity very
stability. close to that of water or oil. Consequently, they can
OBMs are often less compatible with cement slurries easily be pumped in turbulent flow. They act according
than WBMs. to the mechanisms described in Section 5-3.6: mud dis-
■ The aqueous phase usually contains high concentra- persion, tangential erosion of mud layers, and leaving
tions of calcium chloride and calcium hydroxide. the casing and formation surfaces water-wet for optimal
Such compounds have a strong gelling or accelerating cement bonding.
effect on cement slurries (Chapter 3). When WBMs are used, the simplest wash is fresh
■ Emulsifiers in the mud may adsorb on the cement water (Warembourg et al., 1980; Haut and Crook, 1981;
grains, inhibiting hydration (Harder et al., 1992, Smith and Crook, 1982; Sauer, 1987). However, for more
1993). efficient mud thinning and dispersion, chemical washes
that contain dispersants and surfactants are more com-
monly used (Evanoff and Cook, 1988). The dispersants

184 Well Cementing

are often similar to those used in cement slurries—poly- efficient surfactants, such as ethoxylated nonylphenols
naphthalene sulfonates (Wieland and Woods, 1975), lig- (Weigand and Totten, 1986), are no longer allowed off-
nosulfonates, tannates, or more environmentally shore. Therefore, biodegradable and less-toxic surfac-
friendly compounds like polycarboxylic acid derivatives. tants have been introduced, including alkyl glucosides
The surfactants help clean oily compounds adsorbed (Nilsson, 1996; Hill et al., 1997) and other sugar-base
onto solid surfaces. They are selected according to the surfactants. They are less efficient than the traditional
application. surfactants but provide an adequate balance between
When an invert emulsion is used as drilling fluid, the cleaning performance and toxicity.
chemical wash is either a mixture of water, mutual sol- Selecting the type and concentration of surfactant
vents, and surfactants or an oil wash followed by a water- and solvent is made through laboratory testing, using
base chemical wash (Motley et al., 1974; Bannister, various complementary methods adapted to the wash
1987b). objectives (mud thinning, tangential erosion, water-wet-
■ The oil is either the same as that in the drilling fluid ting, and compatibility). The test procedures are
or is a specific compound such as a terpene or ter- described in Appendix B.
pene derivative (Ray and Hines, 1999).
■ Mutual solvents are compounds that will incorporate
5-4.2.2 Spacers
water-base and oil-base compounds into a single
phase. They are normally added at a concentration Spacers are preflushes with carefully designed densities
between 1 and 10 vol%. A typical mutual solvent is and rheological properties (Warembourg et al., 1980).
ethylene glycol monobutyl ether. One of the simplest examples of a spacer is the scav-
enger slurry (Brice and Holmes, 1964), a low-density
■ The surfactants are mixtures of various compounds,
cement slurry with a low fluid-loss rate that can easily be
because a single product cannot achieve the various pumped in turbulent flow. Scavenger slurries have two
requirements defined above. In addition, these prod- principal drawbacks. They are frequently incompatible
ucts must be adapted to the base oil and the emulsi- with drilling fluids and are prone to severe sedimenta-
fier system of the invert emulsion. Anionic and non- tion. Spacers are more complicated chemically than
ionic surfactants are used to water-wet the casing and washes. Below is a description of the most common
formation, while oil-wetting surfactants such as qua- ingredients.
ternary fatty ammonium salts are used to clean oil- Viscosifiers are necessary to suspend the weighting
base drilling fluid from the walls (Motley et al., 1974). agent(s) and control the rheological properties. They
Nonionic surfactants are commonly classified accord- can be subdivided into two classes, water soluble poly-
ing to their hydrophilic-lipophilic balance (HLB). mers and clays.
Surfactants with low HLB values are more compatible ■ Polyacrylamides (Belousov et al., 1987)
with hydrophobic compounds and are used with low ■ Guar and guar derivatives (Weigand and Totten, 1986;
polarity base oils such as linear olefins. Surfactants with Wieland and Woods, 1975)
higher HLB values are used with polar base oils such as
■ Cellulose derivatives (carboxymethylcellulose,
esters. In Table 5-8, HLB values and their related and
typical applications are listed. hydroxyethylcellulose, hydroxymethylcellulose, and
Surfactants are frequently toxic to marine species; hydroxypropylcellulose) (Thomas, 1981; Weigand and
consequently, very stringent regulations for surfactants Totten, 1986; Bannister, 1987b).
exist in offshore areas. Many of the most common and ■ Various biopolymers, including

– xanthan gum. Since its introduction in 1964, xan-

than gum has been used extensively in the oil
Table 5-8. How the HLB Value of Surfactants Relates
to the Application
industry as a viscosifier and particular in drilling
fluids (Weigand and Totten, 1986; Sehault and
HLB Value Application Grebe, 1987).
3 to 6 Water-in-oil emulsion stabilizer – scleroglucan. This biopolymer has good particle
7 to 9 Wetting agent suspension properties and thus it has been advo-
cated for use in spacers (Parcevaux and Jennings,
8 to 18 Oil-in-water emulsion stabilizer 1992; Donche and Isambourg, 1994)
13 to 15 Detergent – welan gum. This polymer has become the most
15 to 18 Solubilizer commonly used for spacers because of enhanced
static and dynamic suspension properties (Powell

Chapter 5 Mud Removal 185

 and Seheult, 1991). Welan gum is a biofermented In addition to water-base spacers, other types of spac-
polymer (Whistler and BeMiller, 1993) that is very ers have been described.
compatible with both drilling fluids and cement ■ An original idea for the removal of OBMs has been
slurries. described by Oliver and Singer (1986)—a water-free
– diutan gum. This recently introduced water-soluble mixture of surfactants and an alcohol. Excellent com-
gum, useful in spacers (Navarrete et al., 2000), is a patibility with the mud and the cement slurry has
biofermented polymer obtained from a naturally been obtained.
occurring bacterial strain of the Sphingomonas ■ Hydroxypropylcellulose polymers may be used to pro-
genus. It is a more efficient viscosifier than welan vide both low viscosity to allow turbulent flow and
gum or xanthan gum and is stable to higher tem- particle transport requirements. Such materials impart
peratures. sufficient viscosity to suspend the weighting agent(s)
during mixing and pumping on the surface; however,
– nonpyruvylated xanthan gum. This recently intro-
when a critical temperature is reached during pump-
duced water-soluble gum, useful in spacers
ing downhole, the polymer is no longer soluble and
(Navarrete et al., 2000), is produced from
the viscosifying effect is lost, permitting turbulent-
Xanthomonas campestris. This polymer is soluble
flow displacement in the annulus (Bannister, 1981).
in high-density CaCl2 brines.
The turbulent action keeps the weighting material
■ Clays, such as bentonite, attapulgite, kaolinite, and suspended.
sepiolite (Beirute, 1981; Thomas, 1981; Weigand and ■ Carney (1974) described an emulsion spacer used to
Totten, 1986; Evanoff and Cook, 1988). Bentonite and
displace invert emulsion muds, together with an
sepiolite are often used in combination with biopoly-
extensive list of goals for the spacer including com-
mers (Carpenter et al., 1998)
patibility with muds and cements, no effect on
Dispersants enhance the compatibility of the spacer cement properties, tolerance to contamination, good
with WBMs and cement slurries and disperse the weight- fluid-loss properties, low viscosity and easy mixing.
ing agent in the spacer. The most common dispersant is ■ Sweatman et al. (1995) described settable spacers to
polynaphthalene sulfonate (Weigand and Totten, 1986; remove OBMs. The compositions contain blast-fur-
Guillot et al., 1986). nace slag (Chapter 7).
Fluid-loss control agents are usually water-soluble ■ Degni and Marcinkevicius (1999) presented field
polymers—guar gum (Wieland and Woods, 1975),
cases in which foamed spacers were used in depleted
poly(ethyleneimine) (Wieland and Woods, 1975), cellu-
areas. The spacers decreased the bottomhole pres-
lose derivatives (Weigand and Totten, 1986; Guillot et al.,
sure. The addition of nitrogen also raised the yield
1986), and polystyrene sulfonate (Guillot et al., 1986).
stress of the spacer and provided better mud dis-
Sometimes the same polymer functions as both a vis-
placement.
cosifier and fluid-loss control agent (Wieland and Woods,
■ Spacer sequences were developed for difficult-to-dis-
1975). The inorganic clays discussed above also have a
beneficial influence on fluid-loss control. place invert emulsion muds (Ali et al., 1998; Ray and
Weighting agent(s) are used to obtain the desired Hines, 1999; Gilmour et al., 2003), combining water-
spacer density—silica flour, fly ash, calcium carbonate, base suspensions and a nonaqueous terpene-base sol-
barite, hematite, ilmenite, and manganese tetraoxide vent.
(Thomas, 1981). Because of their relatively high viscosity, spacers are
Surfactants increase the compatibility of spacers most often flowing in the laminar flow regime during
with OBMs and leave the casing water-wet (Sauer, 1987). mud displacement. However, their composition can be
The same nonionic or anionic surfactants described optimized to decrease their viscosity without compro-
above for washes are usually appropriate. mising stability, allowing turbulent-flow placement.
Optionally, NaCl or KCl may be used to protect or pre- Designing a spacer fluid with optimal rheological prop-
vent the dissolution of massive salt formations or fresh- erties for a specific job is a tedious task. The rheology
water-sensitive shales (Wieland and Woods, 1975; Smith depends on many factors, including the concentration of
and Crook, 1982). Maroy (2002) proposed using quater- viscosifying polymer and weighting agent, temperature,
nary ammonium salts to prevent the swelling of native and the base-fluid composition (fresh water, seawater, or
clays. concentrated salt solution). To address this problem,
Théron et al. (2002) built a neural network model from
a database of fluid properties. One describes the neces-
sary rheological behavior, and the model provides a the-
oretical fluid composition that will meet the goal.

186 Well Cementing

5-5 Cleaning-fluid sequence—current The volume of fluid in Step 3 is adjusted according to
practices the effectiveness of Step 2. The minimum contact time
may vary from a few minutes to 10 min if channeling of
Krause (1986) clearly distinguished the three key objec- the spacer through the mud is likely. One would think
tives of mud displacement in the context of completion that chemical washes would not channel through the vis-
fluids. These objectives also apply to cementing. cous spacer because of their low density and turbulent-
1. Remove the mud from the annulus. flow regime; however, some job failures can be clearly
2. Keep incompatible fluids apart, or at least minimize attributed to chemical-wash channeling. Advanced dis-
their mixing. placement simulators such as that of Bittleston et al.
3. Remove all solids from the walls. (2002) can help design jobs to prevent such channeling
situations (Fig. 5-31). The empirical method to prevent
To achieve these goals, different pills are used, each gravity-driven instabilities is to pump low-density fluids
designed to perform a specific function. In short, viscous down the annulus and to pump heavy viscous fluids down
spacers separate incompatible fluids, and chemical the pipe and up the annulus. Unlike in completion oper-
washes clean the pipe walls. From there, various combi- ations, this is very rarely possible in primary cementing.
nations are possible, depending on the well require-
ments. Once the mud has been conditioned, the follow-
ing fluid sequences are recommended. 5-6 Other cement-placement problems
■ For a WBM: viscous spacer–seawater (or fresh
This chapter has shown that much effort is expended to
water)–chemical wash–viscous spacer optimize the behavior of fluids in the annulus. Ironically,
■ For an OBM: oil–viscous spacer–seawater (or fresh fluids that are designed for proper upward displacement
water)–chemical wash–viscous spacer in the annulus can behave improperly as they travel
Parlar et al. (2002) further justified the cleaning-fluid down the casing.
sequences and recommended four general steps. When no mechanical plugs are used to separate fluids
as they travel down the inside of the casing, commin-
1. The mud is conditioned (i.e., cleaned and thinned). If gling of the fluids naturally occurs at the interfaces. As a
it cannot be sufficiently thinned before displacement, result, slugs of mixed fluids eventually arrive at the shoe
the first fluid of the sequence acts as a diluent. For an as the sequence of fluids is pumped. The mixed fluids
OBM, it will be the base oil. Water is often sufficient may have rheological properties that could adversely
for WBMs. affect mud displacement in the annulus. In extreme
2. The bulk of the mud is displaced with a spacer that is cases, complete fluid swapping may occur inside the
more viscous than the mud, the push-pill. The dis- casing because of density differences. This effect
placement should be piston-like. becomes more important for large-diameter tubulars, for
3. The walls of the annulus (casing and formation) are which buoyancy forces predominate over viscosity
then cleaned in turbulent flow with a thin fluid con- forces. Two basic situations are distinguished below:
taining adequate surfactants. A key design require- purely laminar flow and turbulent flow.
ment for this step is the contact time. This fluid must For laminar flow in a vertical cylinder, the nonuni-
leave both the steel and formation walls water-wet. form velocity profile across a pipe causes the displacing
Turbulent flow may also contribute to complete fluid to finger at the middle. Griffin and Valko (1997)
removal of the “fluffy” part of the filtercake (Mathis et adapted the approach of Beirute and Flumerfelt (1977)
al., 2000). to this geometry for Newtonian laminar flow. Despite
4. Eventually, this dirty fluid (a mixture of various sur- these highly idealized conditions and the limitations of
factants and mud) is displaced by a second viscous the numerical method, their study clearly showed that,
spacer that prevents cement-slurry contamination. regardless of the relative fluid properties and flow con-
ditions, the displacement is not piston-like. If the casing
This basic sequence is often adjusted according to the is not vertical, slumping will occur, leading to an
compatibility of consecutive fluids. Depending on well increased length of intermixed fluids.
conditions and operators, priority may be given to viscos- Crawshaw and Frigaard (1999) solved the slumping
ity ratio or density ratio for spacers. For large vertical problem for situations in which the average of the flow
holes, the focus is on density hierarchy. For highly devi- rates across the pipe section (the mean flow) is zero.
ated holes, the viscosity ratio becomes the major require- This condition occurs during cement-plug placement
ment. A spacer combining high density and viscosity, (Chapter 14).
together with pipe rotation, remains the ultimate solu-
tion to clean and transport solids (Power et al., 2000).

Chapter 5 Mud Removal 187

 As soon as intermixing between two successive fluids shoe. As a result, contaminated cement may remain in
takes place, the interface is much less defined and may this region, and the set cement will be weaker than
become more stable. This phenomenon of interface sta- desired. This is commonly called “green cement.”
bilization by gradually varying the fluid properties was Hibbeler et al. (2000) proposed a solution to the situa-
demonstrated by Dumore (1964) for flow through porous tion when gelled mud remains underneath the float
media. These results have not been extended to flow collar: a fluid-diversion plate is inserted underneath the
through pipes, but Wilson (1991) and Chan (2001) used float-shoe valve, creating local turbulent flow that is suf-
the same principle to minimize fluid intermixing. ficient to clean the gelled mud.
Conversely, when the fluid viscosity reaches a maximum Brady et al. (1992) and Terry (1993) drew attention to
value at some mixing ratio—i.e., the fluids are incom- situations when the casing shoe is far above the bottom
patible—the general consensus is that the instability of the hole. After cement placement, the unfavorable
will be increased, as Dumore showed for other condi- density difference between the tail slurry and the mud
tions. left in the shoe track may lead to fluid swapping, hence
Intermixing can be minimized if one of the fluids is in poor cementing of the casing shoe (bad cement log and
turbulent flow. In turbulent flow, the volume of inter- shoe test). This phenomenon is detailed in Chapter 14.
mixed fluids is much smaller. Debacq et al. (2001 and Practical solutions include placement of a viscous pill or
2003) showed that, for limiting conditions of zero mean cement plug below the casing shoe or running the casing
flow and large enough turbulence, the volume of mixed as close to the bottom of the hole as possible.
fluid obeys the Taylor turbulent dispersion law (Taylor,
1954). The Taylor dispersion law states that the concen-
tration in the mixing front follows an error function. The
width of this front, Δw, increases with the square root of
5-7 Qualitative recommendations
time or volume pumped, Δw = Dt, with D being the dis- Mud removal is a complex problem. The principal causes
persion coefficient. Unpublished experimental results of poor mud displacement during primary cementations
(J.P. Crawshaw and N. Quisel, 2000) show that this law were identified more than 40 years ago. Since then,
provides a reasonable estimate for the mixing length in extensive research has been performed, and this work
a pipe flow situation. Thus, Taylor turbulent dispersion has led to the development of cementing practices that
law provides a lower bound for the volume of mixed are effective in the majority of situations.
fluids. ■ Mud gel strength, yield point, and plastic viscosity
The volume of mixed fluid can also be minimized by should be reduced to minimum values before remov-
using an inner string or stab-in assembly. These configu- ing the drillpipe. However, one must be careful not to
rations achieve two goals: minimizing the pipe internal impair the mud’s ability to suspend the weighting
volume and maximizing the fluid velocity so that turbu- agent.
lent flow occurs (Al-Buraik et al., 1998). Nonetheless, ■ The best possible centralization should be obtained.
the best method to minimize the mixing is to use at least Software is available to determine the optimal design.
two mechanical plugs. Boyington et al. (1989) described ■ In cases in which mud removal difficulties are
production casing cement jobs in which four wiper plugs
expected (e.g., hole irregularities, high gel strength,
were used. A limitation to the number of wiper plugs is
poor fluid-loss control, and poor centralization), the
the need to either interrupt the cement job to reload the
pipe should be equipped with scratchers, scrapers, or
cement head or use special cement heads that can con-
cable wipers. Pipe movement should also be planned.
tain three plugs (Fraser et al., 1996; Chapter 11). At pre-
■ Before the preflushes are pumped, sufficient time
sent, the use of three-plug heads is limited to special
situations such as deepwater wells. should be allowed to circulate at least two annular
As fluids travel down the inside of the casing, a thin volumes of mud at the highest rate possible without
layer of the displaced fluid may remain on the steel sur- losing returns. A better procedure involves using trac-
face. This is particularly common with OBMs. Such ers to monitor the volume of circulatable mud and cir-
layers are not likely to be removed by hydrodynamics culating until this volume represents at least 85% of
alone. As the top wiper plug travels down the casing, it the hole volume.
removes the mud layer and a slug of mud gradually accu- ■ If the mud viscosity cannot be decreased sufficiently,
mulates. When the top wiper plug lands, the mud fills the mud and spacer should be separated by a mutu-
the space between the float collar and the casing shoe. ally compatible preflush. For OBM, the wash must
When the cement slurry breaks through the wiper contain a sufficient amount of surfactants to ensure a
plug, it is unlikely to remove all of the mud around the smooth transition to the water-base spacer.

188 Well Cementing

■ If possible, a chemical wash should be used. The 5-8 Acronym list
volume of wash should be sufficient to ensure a con- 1D One-dimensional
tact time of at least 8 min across the zone of interest.
■
2D Two-dimensional
The density and rheological properties of the spacer
should lie between those of the mud and lead slurry. 3D Three-dimensional
The spacer volume should correspond to at least 500 API American Petroleum Institute
ft of annular length. When these conditions are diffi- CFD Computational fluid dynamics
cult to achieve, more-detailed designs with a numeri- ECD Equivalent circulating density
cal simulator can more precisely determine the prop-
erties that provide the best displacement. HLB Hydrophilic-lipophilic balance
■ Mixing of fluids in the casing should be prevented by ID Inside diameter
the use of mechanical plugs. ISO International Organization for Standardization
For complex situations, a better understanding and OBM Oil-base mud
quantitative recommendations are required. No simple OD Outside diameter
rule can be defined that covers all processes. Two- SI Système International
dimensional numerical simulators can simulate laminar
STO Standoff
flow displacements with reasonable accuracy; however,
modeling of turbulent flow remains a challenge. WBM Water-base mud

Chapter 5 Mud Removal 189

6-1 Introduction
Cement-Formation Interactions
Gérard Daccord, Bernadette Craster, Hemant Ladva, and
Timothy G. J. Jones—Schlumberger
Gabriel Manescu—M-I SWACO

Ideally, the interactions between a cement slurry and

the formation it contacts are minimal. The following
■
6
When the formation contains certain clay minerals,
cement slurry-to-clay interactions may compromise
hole stability unless the chemical compositions of the
points describe how most engineers would prefer the drilling fluid, chemical wash, spacer, and cement
cement slurry and formation to behave. slurry are compatible. The formation mineralogy may
also affect the development of a competent bond with
■ The cement slurry is placed in the casing-formation
the set cement.
annulus and remains there. During placement, all of ■
the circulatable mud is removed and is replaced by When the formation contains soluble evaporite min-
uncontaminated cement. A negligible amount of fil- erals, it may dissolve unless the wellbore fluids are
trate enters the formation; no formation fluid enters modified to prevent or minimize dissolution.
■ Formations that contain gas hydrates can pose seri-
the annulus during slurry placement or the setting
period. ous challenges during well construction (Collett et
■ During placement, the cement slurry flows tangen-
al., 2000). Gas hydrates are solids composed of water
tially to the formation face and does not erode or ice and hydrocarbon gas. They are stable only under
destabilize the wellbore. The formation is sufficiently conditions of high pressure and low temperature. If
strong to resist the pressures exerted during cement- heated, gas hydrates will melt and liberate hydrocar-
slurry placement, and the slurry remains in the annu- bon gas. This large-volume expansion and release of a
lus during the setting period. flammable material can lead to hazards. Therefore,
the wellbore-fluid properties must be adjusted to pre-
■ The interactions between the cement and the forma-
vent destabilizing gas hydrates.
tion are limited to establishing a strong bond at the
interface to ensure perfect zonal isolation. This chapter discusses all of these interactions in
detail. The mechanisms of fluid loss and lost circulation
Unfortunately, in the real world, undesirable interac- are presented together with methods to control them.
tions between the cement and the formation can arise, The effect of formation mineralogy and mud filtercakes
leading to various problems. on bonding is also presented.
■ Owing to the pressure differential between the annu-
lus and the porous formation, fluid exchanges occur.
Filtration of the aqueous phase of the cement slurry 6-2 Fluid loss
into the formation can lead to partial or total dehy-
dration of the slurry and to formation damage as the When cement slurries are pumped against the forma-
fluid interacts with the rock. Deposition of a filter- tion, the pressure differential between the slurry and
cake on the formation face is generally beneficial in the formation leads to filtration. The aqueous phase of
limiting the volume of cement-slurry filtrate entering the cement slurry escapes into the formation, leaving
the formation; however, the filtercake may restrict the solids behind. Depending on the relative importance
the annular space, leading to increased friction pres- of erosional forces during fluid flow and sticking forces
sure and impaired pressure transmission. Influxes of caused by filtration, the solids can form an external fil-
formation fluid into the annulus can lead to other tercake along the formation wall or remain suspended in
undesirable phenomena such as gas migration. the cement slurry. A small amount of solids may also
enter the larger pores in the formation, creating an
■ When the formation is naturally fractured or the
internal filtercake.
annular pressure is high enough to create new frac- During primary cementing, the cement slurry flows
tures, the cement slurry may escape into the forma- along the formation wall, and a dynamic tangential fil-
tion, leading to lost circulation. tration process takes place. In most cases, drilling mud,

Well Cementing ■ Chapter 6 Cement-Formation Interactions 191

chemical washes, and spacers have encountered the for- tercake is incompressible (i.e., constant permeability),
mation before the cement slurry; thus, some filtration the filtrate volume is also a function of the square root of
into the formation has already occurred. Later, when the differential pressure across the filtercake (Outmans,
pumping ceases, a static filtration period takes place. 1963).
During remedial cementing, the filtration is largely
static.
Insufficient fluid-loss control may be responsible for V=A
( )
2 k fc Δp t
, (6-2)
primary cementing failures owing to excessive increases μ filt × K dep
in slurry viscosity during placement, annular bridging, or
accelerated pressure declines during the waiting-on- where
cement (WOC) period. In addition, invasion of cement A = filtration surface area
filtrate into the formation can cause damage and reduce kfc = filtercake permeability
production (Bannister and Lawson, 1985; Economides Δ p = differential pressure
and Nolte, 2000; Yang and Sharma, 1991; Hill et al., t = time
1997). On the other hand, fluid loss can have some posi- V = cumulative filtrate volume
tive effects, such as improving cement-to-formation μ filt = filtrate viscosity.
bonding and increasing the formation fracturing pres-
Cement filtercakes are incompressible when
sure. However, these usually do not outweigh the draw-
deposited under a sufficiently high differential pressure
backs.
(Hook and Ernst, 1969), which varies from formation to
Fluid loss-control agents have been added to well
formation. If the cement filtercake is compressible, k
cement slurries for decades, and the industry has long
may vary with time. It is also important to note that the
recognized that they can significantly improve the qual-
filtercake may consist of more than one layer. There may
ity of both primary and remedial cement jobs. Various
be previously deposited filtercakes from the drilling fluid
simple fluid-loss criteria have long been used to justify
and spacer (Sherwood, 1993).
the level of fluid-loss control required to achieve good
Static filtration is relatively easy to simulate in the
cementing results (Christian et al., 1976). However, field
laboratory. The American Petroleum Institute (API) and
validation of these criteria has been difficult. More
the International Organization for Standardization
recently, studies have evaluated how much fluid-loss
(ISO) define standard testing procedures and equipment
control is required for a given situation.
(Appendix B). In a pressurized test cell, the aqueous
In this section, basic information concerning static
phase of the cement slurry passes through a metal
and dynamic cement-slurry filtration is presented. Then,
screen. A filtercake of cement solids forms on the screen.
fluid loss in the context of primary and remedial cement-
This test allows one to measure various parameters.
ing is discussed. Finally, field measurements of fluid loss
■ Volume of filtrate versus time
are covered.
■ Chemical composition of the filtrate fluid (e.g., ionic
and polymeric contents)
6-2.1 Static filtration
■ Properties of the filtercake (e.g., thickness, porosity,
If the filtercake is assumed to be incompressible, permeability, and shear strength)
the fluid-loss volume will be proportional to the filter-
■ Effects of filtration pressure on the above properties
cake volume. Therefore, the fluid-loss volume per unit
area, V⁄A, is proportional to the filtercake thickness, hfc ■ Deposition constant, Kdep

(Eq. 6-1). All of these properties can also be correlated with the
K dep × V cement-slurry composition, especially the type and con-
h fc = (6-1) centration of fluid-loss additive (Chapter 3).
A Several studies have described the effect of the slurry
composition on filtercake permeability, filtrate viscosity,
Kdep is an experimentally determined proportionality and Kdep.
constant, also called the “deposition constant” by
■ Hook and Ernst (1969) found that k decreases as the
Binkley et al. (1958). Kdep values have been measured by
various researchers and found to vary between 1.0 and concentration of fluid-loss additive increases.
2.5 (Christian et al., 1976; Cook and Cunningham, 1977; ■ Desbrières (1988) determined that, in the presence of

Desbrières, 1988). polymeric fluid-loss additives, the filtercake perme-

During static filtration, the filtrate volume varies with ability easily decreases by a factor of 100, while the fil-
the square root of time (Eq. 6-2). If one assumes the fil- trate viscosity increases by a factor of no more than 5.

192 Well Cementing

■ The particle-size distribution of the cement-slurry observed no significant cement-slurry filtration-rate
solids has a profound effect on the filtercake perme- increase when the temperature was increased
ability. Latex-modified cements form thin filtercakes (Fig. 6-1).
with very low permeability. Engineered particle-size
The erosion of a dehydrated mud filtercake by lami-
(EPS) cement slurries have naturally low fluid-loss
nar hydrodynamic forces is clearly impossible because of
rates owing to the low water content and optimized
the high yield stress of the mudcake. Depending on the
packing of the filtercake (Boisnault et al., 1999;
filtration pressure, Cerasi et al. (2001) measured yield-
Chapter 7).
stress values of 0.014 psi [100 Pa] for OBM filtercakes
■ Baret and Boussouira (1989) described emulsion and between 0.14 to 14.5 psi [103 and 105 Pa] for water-
cement slurries with reduced fluid loss. They attrib- base-mud (WBM) filtercakes. The yield stresses of
uted this effect to capillary forces. However, the cement filtercakes are similar to those of WBM filter-
reduced fluid loss may also be caused by the deforma- cakes (unpublished data, Daccord, 2002). Dehydrated
bility of oil droplets, as observed with oil-base muds muds have yield stresses too high to be removed by the
(OBM) (Aston et al., 2002). drag force of a fluid flowing past them (Ravi et al., 1992).
■ Cement filtercakes have much lower water content However, the mudcake can be partially removed with
than the original cement slurry. Consequently, once casing movement, or by using devices such as scratchers
the filtercake hardens, its strength is higher than that or jets (Chapter 11) attached to the casing string. Highly
of the cement slurry. turbulent flow can substantially erode mud filtercakes
(Beirute et al., 1991), but the underlying mechanisms
remain unknown.
6-2.2 Dynamic filtration In the special case of air-drilled holes, no mud filter-
Dynamic filtration is much more complex than static fil- cake exists when the cement slurry contacts the forma-
tration and has been studied extensively in the context tion; consequently, filtration is entirely controlled by the
of hydraulic fracturing fluids (Clark and Barkat, 1990; cement slurry (Brown and Ferg, 2003). Spacer fluids
Constien et al., 2000) and drilling fluids (Fordham et al., (described below and in Chapter 5) are often pumped
1988; Fordham and Ladva, 1992; Osisanya and Griffith, ahead of the cement slurry to establish fluid-loss control.
1997), but research concerning the dynamic fluid-loss
behavior of cement slurries is sparse.
Initially, dynamic filtration is similar to static filtra-
tion. Eventually, however, erosion limits the filtercake 0.0125
growth (Hook and Ernst, 1969). In addition, the mud fil-
tercake limits the cement fluid loss. The most dominant 0.0100
features of dynamic filtration are summarized below. Fluid-loss
■ Filtercake buildup is not fully reversible. If a filter- rate 0.0075
cake is deposited while pumping at a low flow rate, (cm3/min/cm2)
increasing the flow rate will not remove it entirely 0.0050
(Bannister, 1978). This has also been observed with
drilling fluids; it is known as “irreversible particle 0.0025
Mud–water–cement
attachment” (Fordham et al., 1988). Mud–spacer–cement
■ The filtercake is mechanically eroded. Nevertheless, 20 40 60 80
owing to limited particle penetration into the forma- Temperature (ºC)
tion, sufficient material remains to reduce the fluid-
loss rate of a subsequent fluid. Fig. 6-1. Effect of temperature and fluid sequence upon dynamic
fluid-loss rates. Reprinted with permission of SPE.
■ Once the dynamic filtration regime is reached, the fil-
tration rate is constant. Bannister (1978) measured
cement-slurry filtration rates through a mud filter-
cake of the order of 10–6 m/s (filtration rate per
unit area—0.006 cm3/min/cm2). Similar filtration-
rate reductions have been verified by Haberman et al.
(1992). After pumping a laminar flow spacer contain-
ing a cellulose-base fluid-loss additive between
the mud and the cement slurry, Bannister (1978)

Chapter 6 Cement-Formation Interactions 193

6-2.3 Fluid loss in primary cementing Cook and Cunningham, 1977). In an extreme situa-
Cement slurries are often preceded by chemical washes tion, the filtercake may bridge and plug the annulus,
and spacers, which have two principal objectives. resulting in job failure.
■ Finally, the filtrate that enters the formation inter-
1. Prevent direct contact of the cement slurry with the
mud acts with the formation minerals and modifies the
rock properties and the saturation profile. The fil-
2. Help remove the maximum amount of circulatable trate may also interact with the oil. All of these
mud (Chapter 5) effects are commonly known as formation damage
The second objective is achieved through both hydro- (King, 1966; Bannister and Lawson, 1985; Jones et al.,
dynamic forces and chemical interactions. The chemical 1991).
washes often contain mud thinners or clay stabilizers Once the cement slurry is placed and pumping has
(Maroy, 2002; Rueda et al., 2003). Spacers have a more stopped, subsequent static filtration is dependent upon
complex role. They are designed to remove the maxi- the way pressure is maintained in the annulus. At this
mum amount of mud from the formation face and also to point, further fluid loss affects the ability of the cement
create a filtercake. The filtercake minimizes spacer fil- sheath to provide zonal isolation (Christian et al., 1976;
tration and eventually helps control cement-slurry fluid Cook and Cunningham, 1977; Sabins et al., 1984a and
loss (Crinklemeyer et al., 1976). Jones et al. (1991) con- 1984b; Bannister and Lawson, 1985; Chenevert and Jin,
sidered spacer fluid loss to be of secondary importance, 1989; Daccord et al., 1992; Prohaska et al., 1994).
because the contact time with the formation is short. As Along with the negative consequences discussed
chemical washes and spacers are pumped past the for- above, fluid loss has a few positive effects.
mation, one can expect the mud-filtercake properties
■ The increased solids content of the slurry (i.e., its par-
and the filtration rate to change.
Filtrate invasion can be reduced by chemical reac- tial dehydration) leads to higher compressive
tions that deposit an impermeable skin on the formation strength, shorter transition time, lower permeability,
wall. For example, Whitfill and Whitebay (1990) pro- and less hydration shrinkage.
posed using a mixture of partially hydrolyzed polyvinyl ■ The presence of a hard cement filtercake on the for-

acetate and potassium silicate. An impermeable precip- mation face can improve the bond between the
itate forms when this mixture encounters divalent ions. cement sheath and the formation.
Moving the casing or liner during mud conditioning ■ The increased cement-slurry viscosity can aid dis-
and cement placement is recommended to help displace placement of the drilling mud when viscous forces are
the mud. When the casing is reciprocated, the swabbing greater than turbulent ones.
pressure may lift large parts of the external mudcake,
leaving the rock face bare; however, the internal filter- In addition to the above technical considerations, the
cake remains in place. Ladva et al. (2001) established cost of fluid-loss additives can be sufficiently high to play
that a very small negative differential pressure is suffi- a large role in deciding the level of fluid-loss control.
cient to break or lift a mudcake. Fluid-loss additives commonly double the cement-slurry
From an operational point of view, the cement slurry cost.
must be placed at the designed depth and at the
designed flow rate without abnormal pumping pressures 6-2.3.1 Fluid-loss criteria
and with minimal fluid exchange with the formation.
Common fluid loss-control guidelines, known as fluid-
Fluid loss may compromise these objectives in various
loss criteria, have largely been developed from field
ways.
experience, not theoretical models. Fluid-loss rates are
■ Fluid loss increases the proportion of solids in the
usually measured using the API static test (Appendix B),
flowing slurry. The increased solids content modifies even though many of the design criteria for fluid-loss
the slurry properties. Slurry density (Moran, 1986), control are linked to dynamic filtration rather than static
slurry viscosity, and friction pressure increase (King, filtration. For example, Hartog et al. (1983) defined a
1966; Christian et al., 1976; Beirute, 1988). Turbulent general maximum fluid-loss rate of 200 mL/30 min for oil
intensity decreases, shortens the thickening time wells and 50 mL/30 min for gas wells.
(Baret, 1988), and decreases the total volume (Moran, The 50-mL/30 min limit was also adopted by Christian
1986). et al. (1976), based on field results. On the other hand,
■ If cement solids are deposited and form a filtercake Dillenbeck and Smith (1997) showed that, for a specific
that does not erode, the annular clearance is reduced, gas field, no fluid-loss control was necessary to obtain
leading to increased friction pressure (Beirute, 1988; good cement jobs.

194 Well Cementing

 In subsequent sections, each of the major effects dhole
cited above are detailed, and quantitative criteria are dcsg
attached to them, following the approach of Daccord and
Baret (1994).

q Mud
6-2.3.2 Effects of fluid loss on
cement-slurry properties Dtoc
In this section, a hypothetical primary-cementing situa- Dpl
tion is considered in which a cement slurry flows along a
permeable formation. As shown in Fig. 6-2, the hypo-
thetical well is vertical, and there are no preflushes Cement slurry
between the mud and the cement slurry. Only water can
enter the permeable formation, and all of the cement
particles remain in the slurry and do not form a dynamic hpl vfilt Permeable
filtercake. The fluids are assumed to be incompressible. formation
This simplified problem allows one to derive quanti-
tative fluid-loss criteria and rank the relative impor- v0
tance of each, although the direct use of these criteria
Weak layer
is limited by our knowledge of dynamic filtration. In
Table 6-1, numerical values are given for each phenome-
non described, based on specific well data.
Fig. 6-2. Schematic of the well geometry.

Table 6-1. Numerical Values and Predicted Effects of Various Phenomena for a Specific Simple Well Condition
Situation Result
Mud filtercake thickness ~1 mm, permeability ~1 μD Equation 6-7
Cement filtercake thickness ~1 mm, permeability ~100 μD Filtration velocity ~1 × 10–6 m/s
Differential pressure = 140 psi [10 bar]
Filtrate viscosity = 1 cp

Hole size = 10 in., casing size = 7 in. Equation 6-8

Length of permeable formation = 300 ft [91 m] Relative volume of slurry lost = 0.38%
Length of cement above the permeable zone = 1,000 ft [305 m] Variation of top of cement depth = 3.8 ft [1.2 m]
Pump rate = 8 bbl/min [1,270 L/min]

Slurry density = 15.8 lbm/gal [1,900 kg/m3] Equation 6-9

Slurry porosity φcs = 50% Relative slurry-density increase = 0.18%
Cake porosity, φfc = 30% Equation 6-10
Maximum admissible filtration velocity (vfilt)max = 7.5 × 10–5 m/s

Mud density = 12 lbm/gal [1,440 kg/m3] Equation 6-11

Hydrostatic pressure increase = 0.8 psi [5.6 kPa]

The solids content increases by a relative amount of 0.9%, Equation 6-12

which leads to a relative slurry viscosity increase of 4.5%. For a maximum admissable viscosity increase of 30%,
the maximum admissible filtration velocity is 6 × 10–6 m/s.

Formation porosity accessible to filtrate, φ = 10% Equation 6-16

Filtrate penetration depth, L = 4.3 in. [11 cm]

Chapter 6 Cement-Formation Interactions 195

 At any depth z, the mass conservation equations for
the two phases, water and cement (volume fractions fwV (
v = v0 1 – vdf )
and fcV, respectively), are written as follows.
( )
fwV h pl =
fwV 0 – vdf h pl( )
π dhole × v filt +
π⎡
(
d )2 − ( dcsg ) ⎤⎥⎦ ( wVdz ) = 0
2 d f ×v 1 – vdf (h ) pl
4 ⎢⎣ hole
( ) (⎡ )
4 d × v filt × h pl
2 ⎤ d ( fcV × v)
hole
with vdf h pl =
dhole ) − ( dcsg ) ⎥
π⎡ (6-6)
⎢
4⎣
( 2

⎦ dz
= 0,
(
v0 ⎢ dhole ) − ( dcsg ) ⎥⎦
2 2⎤

⎣
fwV + fcV = 1
Practical use of these equations requires knowledge
(6-3) of the filtration velocity, vfilt. Three situations are possi-
ble:
where
■ competent mud filtercake is present, upon which a
dhole = hole diameter
cement filtercake is deposited
dcsg = casing diameter
■ only cement filtercake is present
v = slurry velocity in the annulus
v0 = slurry velocity below the permeable formation ■ there is no filtercake.
vfilt = filtration velocity into the formation
No filtercake can be formed when the formation per-
z = vertical coordinate.
meability is so low that the particles cannot stick to the
For a constant filtration velocity, the conservation formation surface (Fordham et al., 1988). In this case,
equations can be written in a simplified form. the filtration rate is entirely determined by the forma-
tion permeability, according to Darcy’s law. When filter-
v0 +
(
d fwV × v ) = 0, cakes are present, the filtration velocity obeys the fol-
dvdf lowing equation.

(
d ⎡⎣ 1 – fwV v ⎤⎦ )
=0
v filt =
1
×
Δp
(6-7)
(6-4) ( h fc ) mud ( h fc )cem μ filt
dvdf +
( k fc ) mud ( k fc )cem
with
The term hfc ⁄ kfc, the ratio of cake thickness to its per-

)
vdf ( z =
( ) .
4 dhole × v filt × z
(6-5)
meability, is the cake resistance to flow. In Eq. 6-7, both
a mudcake and a cement cake are considered with
v ⎢( d ) − ( d ) ⎥
⎡ ⎤ 2 2
respective resistances equal to (hfc)mud ⁄ (kfc)mud and
0 hole csg
⎣ ⎦ (hfc)cem ⁄ (kfc)cem.
vdf is a dimensionless coordinate but can also be consid- Slurry volume
ered as a dimensionless filtration rate. Effectively, for z
Based on the above analysis, the volume of slurry in the
equal to the permeable-layer thickness (hpl), vdf is equal
annulus is reduced by an amount equal to the total fil-
to the ratio of the total filtration rate to the slurry pump
trate volume.
rate. With the slurry velocity and water-volume fraction
(slurry porosity) given by v0 and fwV0, respectively, the v filt × h pl
slurry velocity and porosity at the top of the permeable
layer are given by the integration of the above system of
(
ΔV = πdhole Dpl − Dtoc ) v0
. (6-8)

equations.
Slurry density
At the end of cement placement, the density of the slurry
above the permeable zone is increased to the following
value:

⎣ (
ρ = ⎡ρ0 − vdf × ρ w ⎤ ×
1
⎦ 1 – vdf
. ) (6-9)

196 Well Cementing

where 160
ρ = density of slurry above permeable zone
ρ0 = initial slurry density 140
ρw = density of water.
The upper slurry-density limit corresponds to total 120
slurry dehydration. In this limiting situation, the solid- Thickening
time 100
volume fraction reaches a maximum value, equal to the (min)
solid-volume fraction of the cement cake, ffc, in a first
approximation. Equation 6-6 allows one to quantify this 80
limit and derive a maximum admissible filtration veloc-
ity, an important fluid-loss guideline. 60
Density (lbm/gal)
16.4 16.3 16.2 16.1 16.0 15.9 15.8 15.7 15.8

(v )
fcV – vdf fcV – f fc 40
f fc = or filt max = (6-10) 38 39 40 41 42 43 44 45 46 47
1 – vdf 1 – f fc
Water concentration (% BWOC)

Hydrostatic pressure increase Fig. 6-3. Variation of slurry thickening time with water/cement ratio.
Owing to both the decrease of slurry volume (Eq. 6-8)
that is replaced by an equal volume of mud and its
increase in density (Eq. 6-9), the hydrostatic pressure is 120
Experimental data
increased by an amount equal to 100 Equation 6-12

( )( )
80
Δ p = ρ mud − ρ w Dpl − Dtoc gvdf , (6-11) Viscosity 60
(cp)
40
where
20
ρmud = mud density
ρw = water density 0
g = acceleration of gravity. 31 36 41 46
Water/cement ratio (% BWOC)
When the formation has a low fracturing pressure and
additional hydrostatic pressure creates a risk of induc- Influence of Water/Cement Ratio on Rheology
ing losses into the formation, this equation can be used 100 Test run at 80°F with neat cement slurries
as another fluid-loss guideline. 90
80
Slurry thickening time 70
Decreasing the water/cement ratio accelerates cement 60
thickening (Fig. 6-3). The amount by which the thicken- Yield Class G
value 50
ing time varies is strongly dependent on slurry density (Pa) 40
and composition. In critical situations, laboratory tests
30
should be performed to determine the sensitivity of the
20
thickening time to variations in slurry density.
10 Class H
Slurry viscosification 0
36 38 40 42 44 46 48 50 52 54 56 58 60
As the water/cement ratio decreases, both the viscosity
and the yield stress of a slurry increase (Fig. 6-4). This Water concentration (% BWOC)
is a general law for suspensions, and many theoretical Fig. 6-4. Slurry rheology versus water/cement ratio.
or empirical models describe this dependence.
Equation 6-12 is the simplest model that describes the
variation of the high-shear-rate slurry viscosity, μslurry, −2
with the effective solid volume fraction, fsV (Quemada, ⎡ fsV ⎤
μ slurry = μ s 0 ⎢1 – ⎥ , (6-12)
1998). ( fsV)max is the maximum solid volume fraction in
⎣ ( fsV )max ⎦

where μs0 = viscosity of the suspending liquid.

Chapter 6 Cement-Formation Interactions 197

 This law shows that the slurry becomes a solid—the 6-2.3.3 Annular restriction and bridging
viscosity is infinite—when the effective solid-volume In the presence of a mudcake, filtration is limited by the
fraction reaches a maximum value. It is therefore equiv- mudcake; therefore, a significant cement filtercake is
alent to Eq. 6-10, although the values for the maximum unlikely to form under dynamic conditions. The filtration
solid-volume fractions are slightly different. Figure 6-4 velocity is too small to allow cement particles to stick to
shows experimental data from which an empirical the wall without being immediately eroded by the flow.
( fsV)max value can be derived. In turn, using Eq. 6-6, a However, if the mud filtercake is damaged or the flow is
maximum admissible filtration velocity is defined. stopped, a cement filtercake will form. The fluid-loss
guideline that is derived from this mechanism corre-
Dynamic placement pressure
sponds to a cake thickness that is equal to the annular
An increase in slurry viscosity leads to increased friction gap. Using Darcy’s law, the cement-filtercake thickness
pressure in the part of the wellbore above the permeable can be calculated (Baret, 1988). To a first approxima-
zone. Therefore, the placement pressure also increases. tion, one can assume that the mudcake is incompress-
Another quantitative fluid-loss guideline may be ible, whereupon the thickness of the cement cake is
extracted that requires more detailed calculations. An given by
example is shown in Fig. 6-5.

Concluding remarks ( h fc )cem = K dep

V (t ) = ( k fc )cem × ( h fc ) mud
A ( k fc ) mud
It is clear that, as long as a competent mudcake governs
the filtration velocity and that the well conditions ⎛ 2 ⎞
approximate those used above, none of the above conse- ⎜ 2Δp × K dep × ⎡⎣( k fc ) mud ⎤⎦ ⎟
quences is likely to perturb cement placement. The vari- ⎜ 1+ 2
× t – 1⎟ .
ations given in Table 6-1 are very small. ⎜ μ filt × ( k fc )cem × ⎡⎣( h fc ) mud ⎤⎦ ⎟
⎝ ⎠
However, if the mudcake is damaged and the cement
cake controls filtration, the filtration velocity will
(6-13)
increase by one or two orders of magnitude. In that case,
these consequences are likely to affect cement place- In the absence of a mud filtercake, the cement-filter-
ment. The increase of slurry viscosity and corresponding cake thickness is deduced from Eq. 6-2:
dynamic pressure are often the most critical effects.
These effects become especially important for narrow
annuli (Waters and Wray, 1995). ( h fc )cem = K dep
V (t )= 2( k fc )cem × K dep × Δ p
× t.
A μ filt

5,400 (6-14)
This filtercake thickness is therefore made equal to
5,300
the annular gap, (hfc) cem = (dhole – dcsg) ⁄ 2, to obtain
(kfc)cem Kdep ⁄μ filt. These properties are related to the
5,200 API/ISO static filtration test (Eq. 6-2), and yield the
Fracturing pressure
Bottomhole maximum filtration pressure and time that will lead to
pressure 5,100 annular bridging. Equation 6-15 gives the guideline
(psi) when considering a cement cake alone, in terms of max-
5,000 imum API/ISO filtrate volume (Appendix B).

( )
2
4,900 ⎡ d –d ⎤
A
× t API ⎢ ⎥
hole csg
(vdf)max Δ pt = Δ pAPI × API
⎢ 2 K dep V API ⎥
4,800 ⎣ ⎦
( )
0 0.05 0.1 0.15 0.2 0.25 0.3
Dimensionless filtration velocity ⎛V ⎞ dhole – dcsg Δ pAPI × t API
or ⎜ API ⎟ =
Fig. 6-5. Calculated bottomhole pressure versus filtration rate. The ⎝ AAPI ⎠ max 2 K dep Δ pt
intercept of the curve with the fracturing pressure gives the maxi-
mum filtration rate that can be tolerated. (6-15)

198 Well Cementing

 A similar relationship can be derived for the case cement slurries are used to achieve turbulent flow, the
when a mud filtercake is present. Numerical results are viscosity increase lowers the turbulent energy of the
shown in Fig. 6-6. Case 1 corresponds to a long filtration fluid and its mud-removal capability. Whether this
time under a high differential pressure. The cement fil- decrease in turbulence can be offset by an increase in
tercake is controlling the behavior; therefore, the API wall shear stress should be evaluated on a case-by-case
fluid-loss rate must be less than 50 mL/30 min to prevent basis.
annular bridging. Case 2 is for a smaller differential
pressure. As soon as the mudcake resistance is higher
than 0.5 mm/μD, the mud filtercake controls the behav- 6-2.3.5 Formation damage and
ior, and no fluid-loss control is required for the cement formation invasion depth
slurry. These typical mudcake resistance-values corre- A simple formation-damage and fluid-loss guideline is
spond to competent mudcakes. based on the maximum admissible penetration depth of
cement filtrate, although a distinction must be made
between filtrate-invasion depth and actual damage
6-2.3.4 Mud removal (Vidick and Reid, 1997). For example, assuming a piston-
The complexity of mud-removal mechanisms prevents like displacement into the rock, the penetration depth,
deriving simple fluid-loss guidelines. Parametric studies L, of the filtrate for a cement cake filling the entire
require using numerical simulators such as those annular gap is given by
described in Chapter 5. From a qualitative standpoint,
partial dehydration of the cement slurry will increase ⎡ 2 ⎤
the friction pressure and wall shear stress, improving ⎢ ⎛ dcsg ⎞ ⎥
d 1− ⎜ ⎟
the mud-removal capability. Figure 6-7 shows an exam- L = hole ⎢ d ⎥ (6-16)
ple of an eccentered casing in a deviated well. The image 2 ⎢ 1 + ⎝ hole ⎠ − 1 ⎥
⎢ φ × K dep ⎥
on the left side shows the presence of a mud channel on ⎣ ⎦
the narrow side of the eccentered annulus. When the vis-
cosity of the cement slurry increases (right side of where φ = the formation porosity. After perforating, if
figure), the mud channel is significantly reduced. the perforation channel extends beyond this depth, then
This beneficial effect only appears when mud removal no major effect of cement filtrate on production is
is governed by viscous forces (i.e., relatively narrow expected.
annuli). For larger annuli and smaller deviation, gravity
forces usually predominate, and no significant effect of
the slurry-viscosity increase is expected. Also, when thin

10,000 10,000
API fluid loss, case 1
API fluid loss, case 2
Permeability, case 1
1,000 Permeability, case 2 1,000

Maximum
API fluid-loss Cement
volume for 100 100 filtercake
cement permeability
slurry (µD)
(mL)
10 10

1 1
0 0.5 0.1 1.5 2
Mud filtercake resistance (mm/µD)

Fig. 6-6. Predicted maximum cement fluid-loss volume and cake permeability versus mudcake resistance for
two static cake buildup conditions. Case 1: filtration time = 6 h, pressure = 1,015 psi [70 bar]. Case 2: filtration
time = 3 h, pressure = 145 psi [10 bar] (figure adapted from Baret, 1988). Reprinted with permission of SPE.

Chapter 6 Cement-Formation Interactions 199

 4,600 4,600

4,800 4,800

5,000 5,000

5,200 5,200

5,400 5,400
Depth
(ft) 5,600 5,600

5,800 5,800

6,000 6,000

6,200 6,200

6,400 6,400

Wide Narrow Wide Wide Narrow Wide

Fig. 6-7. Effect of increasing cement-slurry viscosity on mud removal. The example is 14-lbm/gal mud in an
inclined eccentered annulus (shown in brown), plastic viscosity (PV) = 40 cp, yield point (YP) = 20 lbf/100 ft2
[9.6 Pa]) displaced by a cement slurry of identical density. Left: Cement PV = 60 cp, YP = 10 lbf/100 ft2 [4.8 Pa].
Right: Cement PV = 100 cp, YP = 15 lbf/100 ft2 [7.2 Pa].

Several possible formation-damage mechanisms have 6-2.3.6 Pressure transmission during WOC
been described (Cunningham and Smith, 1968; Yang and Preventing gas migration requires an integrated
Sharma, 1991; Hill et al., 1997), including clay swelling, approach, and cement fluid loss is only one of the rele-
particles plugging pores in the formation, precipitation vant parameters. A thorough presentation of the subject
of calcium silicates from the cement filtrate, dissolution- is available in Chapter 9. A number of related factors
precipitation of minerals from the formation, and decon- control the fluid-loss properties of the cement slurry
solidation of the formation. Jones et al. (1991) con- (Christian et al., 1976; Sabins et al., 1984a and 1984b;
ducted studies with Berea sandstone cores to correlate Rae et al., 1989). For example, pressure reduction
the type of formation damage with the composition of during the WOC period is directly linked to volume
the wellbore fluid. Severe permeability reduction was reduction and fluid loss. Sabins et al. (1984a and 1984b)
observed with freshwater spacers and latex-modified considered the annulus to be a closed system, and
cement slurries. Potassium chloride (KCl) was generally allowed the fluid compressibility, c, to compensate for
beneficial in that it reduced clay swelling, although the the lost volume. The corresponding pressure decrease
filtrate volume was greater. However, Dillenbeck et al. was calculated by the following equation.
(1991) noticed that too much KCl in cement slurries
caused gelation and increased fluid-loss rates. Gambino V filt
et al. (2001) studied formation damage by OBM filtrate Δp = , (6-17)
and cement filtrate. The OBM filtrate induced severe cVann
fines migration and wettability alteration. The cement
where
filtrate induced little formation damage.
Vann = the annular volume.
Equating this pressure reduction to the initial overpres-
sure gives a maximum allowable fluid-loss volume for the
cement slurry, V, which can be related to the API/ISO
fluid-loss properties (Eq. 6-2).

200 Well Cementing

 A more detailed description of pressure reduction Binkley et al. (1958) presented a detailed model that
must include the additional effects of cement yield reasonably represents perforation plugging. Also, labora-
strength and shrinkage, as well as wellbore geometry. tory testing can be designed to more closely resemble
Chenevert and Jin (1989) and Daccord et al. (1992) pre- downhole conditions (Hook and Ernst, 1969). These
sented simulation results that clearly showed the effect studies have resulted in widely-accepted fluid-loss guide-
of cement fluid loss. These sophisticated models show lines for perforation plugging.
that, even with excellent fluid-loss control, cement fil- For other applications, quantitative criteria can be
tration may remain a critical parameter if the slurry defined in a manner similar to primary cementing. For
properties are not optimized (e.g., if the slurry has a long example, Eq. 6-18 (derived from Eq. 6-15) gives an esti-
transition time). mate of the API/ISO fluid-loss rate required for a cement
The fluid-loss properties of the cement slurry may cake to bridge a mud channel of thickness h.
have an indirect effect on gas migration prevention. For
example, Christian et al. (1976) reported that, if fluid- V API h Δ pAPI × t API
loss control is achieved by adding a polymer that viscosi- = × . (6-18)
AAPI K dep Δ pt
fies the interstitial water, then gas percolation through
the cement matrix is likely to be hindered.
If the downhole filtration conditions are similar to
those of the high-pressure API test, then VAPI ~ AAPI h⁄
6-2.3.7 Comparison between various Kdep. For a 1-mm thick channel and Kdep = 2, VAPI = 1 mL.
phenomena and conclusions Clearly, in this case, extremely low API fluid-loss rates
The quantitative comparison between the various crite- are required for the slurry to successfully fill the channel
ria defined above clearly shows that, when a risk of gas without bridging it. Practically, the best way to achieve
migration exists, cement-slurry fluid loss is a critical such low fluid-loss rates without increasing the slurry
parameter. The use of quantitative criteria is of limited viscosity has been through the use of EPS slurries
interest when considering cement placement because (Farkas et al., 1999; Chapter 7).
the downhole conditions are usually poorly understood,
especially the filtration properties of the mudcake. 6-2.5 Field measurements
Several field problems have been interpreted in terms of
slurry dehydration or annular bridging. This uncertainty Direct measurement of the effects of fluid loss during
justifies the use of fluid-loss additives to limit potential primary cementing jobs is difficult. Variations in slurry
risks. properties, surface pressure, and flow rate are small.
The situation is slightly better for remedial cementing
operations, in which surface-pressure measurements
6-2.4 Fluid loss during remedial can be interpreted in a useful way. These two situations
cementing operations are considered below.
Fluid loss has long been recognized as a critical para- During primary cement jobs, the simultaneous occur-
meter in squeeze cementing (King, 1966). However, this rence of many phenomena prevents extracting the con-
is only one of several important parameters, and suc- tribution of cement-slurry fluid loss in an unambiguous
cessful remedial cementing requires one to consider way (Evanoff and Cook, 1988). Despite this problem,
many other factors (Chapter 14). In this chapter, the dis- Beirute (1988) proposed a method to diagnose six poten-
cussion is limited to fluid loss. For perforation plugging, tial cement job problems by analyzing surface parame-
fluid-loss control is required to achieve proper filtercake ters. Two of these problems are of relevance: flow restric-
buildup in perforation channels. However, cement-slurry tions and slurry dehydration. In both situations, the
fluid loss affects other phenomena. free-fall periods would be shorter than normal. For a
■ The filtrate that enters the formation may lead to for-
downhole restriction, the slope of the pressure curve
mation damage and may modify the formation stress should be normal, while it should be steeper for slurry
conditions. Consequently, the fracturing pressure is dehydration.
also affected (Marangos, 1989). The situation is more favorable when the pressure
decline can be monitored after pumping has stopped.
■ When the objective of the squeeze operation is to fill
Two primary factors contribute to short-term pressure
voids or mud channels behind the casing, excessive variations, namely compressibility and fluid loss, as
cement-slurry filtration will lead to rapid dehydration quantified by Eq. 6-17. This analysis is commonly per-
and premature pressure buildup. In these situations, formed during fracturing treatments to determine the
a low to extremely low fluid-loss rate is required.

Chapter 6 Cement-Formation Interactions 201

fluid-loss coefficient of the fluid (Nolte, 1983; Gulrajani operations (Goins, 1952). Lost circulation must not be
and Nolte, 2000) as well as for a pressure integrity test confused with fluid loss, which has been previously
(Postler, 1997). described. Figure 6-8 depicts how the fluid-loss process
Haberman et al. (1992) used a similar technique to is more related to primary porosity, whereas lost circula-
specifically determine fluid loss from drilling fluids and tion can occur in formations with both primary and sec-
cement slurries. Two methods provided very consistent ondary porosities. Lost circulation is a problem that is
results: interpretation of the pressure decline and mea- best attacked before performing the cement job.
surement of the pump rate required to maintain a con- Therefore, the treatment of lost circulation during
stant surface pressure. Measurements were performed drilling is included in the following discussion.
before and after cementing. As a general conclusion, it
was found that the fluid-loss rate of the drilling fluid was
lower than that measured in the laboratory and that the 6-3.1 Consequences of lost circulation
cement-slurry fluid loss was mainly controlled by the Lost circulation can be an expensive and time-consum-
drilling-fluid filtercake. ing problem. During drilling, this loss may vary from a
gradual lowering of the mud level in the pits to a com-
plete loss of returns. The major consequences of lost cir-
6-3 Lost circulation culation are shown in Table 6-2.
To effectively solve lost circulation with the correct
Lost circulation (or lost returns) is defined as the total technique, it is necessary to know the severity of the
or partial loss of drilling fluids or cement slurries into losses, the type of lost circulation zone, and the drilling
highly permeable zones, cavernous formations, and frac- history of the well just before the losses occurred.
tures or fractures induced during drilling or cementing

Diameter of solids in the drilling mud > 3 × pore diameter = cake building and fluid loss
Primary
porosity
Pore diameter > 3 × diameter of solids = seeping/mud invasion
Porosity

Secondary Void diameter > 3 × diameter of solids = lost circulation

porosity

Fig. 6-8. Fluid loss versus lost circulation.

Table 6-2. Major Consequences of Lost Circulation

Drilling Cementing Completion and Workover
Loss of mud Reduced annular coverage Loss of completion fluid

Lost time Casing corrosion Lost time

Poor cement job Poor zonal isolation Formation damage

Reduced safety Reduced safety Reduced safety

Stuck in hole Lost reserves

Wasted casing string Loss of well

Failure to reach target depth

Blowout and kill operations

Downhole blowouts

Environmental incident

202 Well Cementing

6-3.2 Classification of the severity of losses If the well becomes stable, one then calculates the
A standard severity classification for lost circulation is hydrostatic pressure required to fill the well. If losses
shown in Table 6-3. persist, one begins spotting conventional lost circulation
material (LCM) pills and progresses to plugs if conven-
tional treatments are unsuccessful. Owing to the reduc-
tion of hydrostatic head, the well must be monitored
Table 6-3. Severity Classification for Lost Circulation
closely at all times for influx of fluids. It may be possible
Type of Losses Severity in some areas to continue drilling if the fluid cost is low
Seepage (minor) < 10 bbl [1.5 m3]/hr and pressures are manageable.

Partial (medium) 10 to 100 bbl [1.5 m3 to 15 m3]/hr

6-3.2.4 Total losses
Severe (massive) 100 to 500 bbl [15 m3 to 75 m3]/hr
In case of total losses, the return flow stops immediately.
Total (complete) Unable to keep the hole full Treatment begins by spotting conventional LCM pills
and progresses to plugs if conventional treatments are
unsuccessful.
An underground blowout occurs when a fluid or gas
6-3.2.1 Seepage losses (1–10 bbl/hr flows from one zone into an upper weaker or hydrauli-
[1.5 m3 to 15 m3/hr]) cally fractured zone. This is usually indicated by unsta-
ble surface-pressure readings. This phenomenon is prob-
Seepage takes the form of very slow losses. Sometimes ably the most serious lost circulation condition. Not only
the losses can be in the form of filtration to a highly per- is there a loss of fluid, but a dangerous well-control situ-
meable formation, an extreme case of fluid loss. Seepage ation also exists. Casing placement is critical to avoid
losses can also be confused with the volume of cuttings total losses.
removed at the surface. One must not confuse these two
completely different occurrences. If seepage losses are
suspected, the bit must be pulled off bottom and the 6-3.3 Classification of lost circulation zones
mud volumes checked with and without circulation. All Lost circulation occurs by one of two mechanisms.
mixing equipment and nonessential solids removal
■ Natural losses. Whole fluid or cement is lost to for-
equipment should be turned off and baseline values
recorded. mations that are highly permeable, unconsolidated,
Once it is established that fluid is being lost, one must fractured, cavernous, or vugular.
decide whether to tolerate or cure this situation. ■ Induced losses. The mud or fluid is lost because

Depending on the cost of the drilling fluid, rig time, or excessive induced pressure hydraulically fractures
both, one might decide to continue drilling with seepage the formation.
losses. If formation damage or stuck pipe are potential It is common to classify lost circulation zones into five
problems, one should attempt to cure the losses before categories.
proceeding with drilling.
■ Unconsolidated formations

■ Highly permeable/low-pressure formations (depleted

6-3.2.2 Partial losses (10–100 bbl/hr) zones)
Partial losses are more voluminous than seepage losses; ■ Natural fractures or fissures
therefore, the fluid cost becomes more crucial in the
■ Induced vertical or horizontal fractures
decision to drill ahead or combat the losses. Drilling
with partial losses can be considered if the fluid is not ■ Cavernous and vugular formations

expensive and the pressures are within operating limits. Seeping losses can occur with any type of lost circu-
lation zone when the mud solids are not sufficiently fine
6-3.2.3 Severe losses (100–500 bbl/hr) to seal the formation face. Partial losses frequently
occur in highly permeable gravels, small natural frac-
When severe losses are encountered, regaining full cir-
tures, or as a result of fracture initiation. Complete
culation is mandatory. This can be accomplished by
losses are usually confined to long gravel sections, large
pumping a lower-density fluid down the annulus
natural fractures, wide induced fractures, or cavernous
(drilling mud, water, or other lightweight fluid) and
formations.
monitoring the volumes required to fill the well.

Chapter 6 Cement-Formation Interactions 203

6-3.3.1 Unconsolidated formations 6-3.3.4 Induced fractures
Coarse, unconsolidated formations can have a suffi- If the borehole pressure exceeds the formation parting
ciently high permeability for whole mud or cement pressure, open fractures will be created, permitting the
slurry to invade the formation matrix (10–100 D). For loss of mud or cement. There are three typical circum-
whole mud or cement slurry to be lost, the average par- stances when this can occur.
ticle size of solids in the mud or cement slurry must be ■ An immovable mud ring may develop in the annulus
less than or equal to one-third the size of the formation owing to fluid loss. The resulting circulating-pressure
opening (Barkman and Davidson, 1972). These losses increase may initiate a hydraulic fracture.
are normally confined to shallow wells or surface holes. ■ When one is drilling through an undercompacted for-
Rates of loss can vary from seepage to total losses. In the
mation, typically found offshore, fractures can occur.
event that losses are total, it is sometimes common prac-
■ When one is drilling from a mountaintop, the over-
tice to continue drilling if a sufficient supply of water is
available and there are no environmental or well-control burden pressure is low, and fracturing occurs easily.
concerns. Well irregularities, high mud weight, and rough han-
One reason for preventing shallow mud losses is that dling of drilling tools may also help induce fractures.
unconsolidated formations may wash out, forming a Simpson et al. (1988) suggested that lost circulation
large, unstable cavity that could collapse from the over- caused by fracture initiation is more common when
burden and rig weight. In some areas, it may be more using oil-base instead of water-base mud. They believed
common to drill with air, mist, and foamed or aerated this arises from the failure to consider the compressibil-
muds to prevent losses. Unconsolidated formations are ity of the oil under downhole conditions. They also
typically found at shallow levels and normally consist of observed that induced fractures do not “heal” readily
sands or gravel; however, they can also occur in shell when OBM is present.
beds or reef deposits. When partial losses occur with WBM, an accepted
practice is to let the hole “soak.” The mud is allowed to
6-3.3.2 Highly permeable or depleted formations rest in the hole. The resulting mud filtration allows the
fractures to be filled with mud solids and often permits
To permit the penetration of whole mud or cement full circulation to be restored with no mud-weight reduc-
slurry, the formation permeability must be greater than tion. However, filtration from OBM is often too slow to be
10 D; however, significant seepage losses can occur in helpful. Once fractures are initiated with an OBM, frac-
lower-permeability consolidated sandstones. Producing ture extension can be expected until the borehole pres-
formations in the same field or general vicinity may sures are reduced or the fracture openings can be
cause subnormal (depleted) formation pressure because sealed.
of the extraction of formation fluids. Loss of mud to
these formations requires that the passages be suffi-
ciently large and connected to allow entry of whole mud 6-3.3.5 Cavernous and vugular formations
and that the mud pressure must exceed the formation Large voids or caverns are sometimes encountered when
pressure. Seepage losses to severe losses can often lead drilling through certain limestone and dolomite forma-
to differentially stuck drillpipe. Such depleted reservoirs tions as well as the soluble caprock of salt domes.
are found at any depth. Sudden and complete losses are typical of this type of zone.

6-3.3.3 Natural fractures or fissures 6-3.4 Lost circulation while drilling

Hard, consolidated formations may contain natural frac- According to Messenger (1981), it is possible to classify
tures that take mud when penetrated by the drillstring. the available lost circulation solutions into three main
The overburden must be self-supporting for a horizontal categories:
fracture to exist, but this is not the case for a vertical
■ bridging agents in the drilling fluid
natural fracture. To widen a horizontal fracture, the
overburden must be lifted. Vertical fractures will propa- ■ surface-mixed systems

gate when the fracture pressure is exceeded. A sudden ■ downhole-mixed systems.

loss of returns in hard, consolidated formations often
indicates the presence of natural fractures. There is an optimal technique for solving each partic-
ular type and severity of a lost circulation problem.
Figure 6-9 shows a lost circulation decision tree to
address lost circulation problems.

204 Well Cementing

 Lost circulation

Stop drilling and

observe levels.

Yes Are there losses

No
on the surface?

Locate and stop losses. Yes Is the well flowing?

No
Record shut-in
pressure. Kill the well.
Measure rate
of loss.

Seepage losses (<10 bbl/hr) Partial losses (10–100 bbl/hr) Total loss of returns

Oil- or synthetic- Water-base Oil- or synthetic- Water-base Oil- or synthetic- Water-base

base mud mud base mud mud base mud mud

Spot pill: Spot pill: Spot pill: Spot pill:

Graphite Fine mica Medium fiber Medium mica
Medium and coarse Fine fiber Graphite Medium fiber
calcium carbonate Fine granular LCM Coarse calcium carbonate Coarse calcium carbonate
Medium calcium carbonate Medium granular LCM

No success No success

No success Spot pill: Spot pill:

Medium or coarse fiber Medium mica
Graphite Medium and coarse fiber No success
Coarse calcium carbonate Coarse calcium carbonate
Spot plug:
Medium granular LCM Medium granular LCM
Polymer crosslinked pills Spot plug:
Reverse gunk or cement Polymer crosslinked pills
Gunk or cement

Fig. 6-9. Lost circulation treatment decision tree.

Chapter 6 Cement-Formation Interactions 205

6-3.4.1 Bridging agents in the drilling fluid Table 6-4. Optimizing the Bridging Material According to the Formation
Permeability
When the loss of mud is first detected, immediate con- Sand Grain Size Diameter of Pore Approximate Bridging
sideration should be given to the possibility of reducing (in. [μm]) Throat Opening Permeability Particle Size
and maintaining the mud weight at the minimum neces- (μm) (mD) (μm)

sary to control the formation pore pressure. Reduced 0.00025 [6.46] 1.0 1 0.33
mud pressure will help combat losses no matter what 0.00125 [31.67] 4.9 24 1.63
types of formations are exposed. A continuing partial 0.0015 [38.10] 5.90 35 1.97
loss of returns is indicative of seepage and can usually be
0.0017 [43.18] 6.68 45 2.23
solved by decreasing the equivalent mud circulating
0.0021 [53.34] 8.26 68 2.75
density or by adding LCMs to the drilling mud. The
equivalent mud circulating density can be reduced by 0.0024 [60.96] 9.44 89 3.15
decreasing the mud weight and/or adjusting its down- 0.0029 [73.66] 11.40 130 3.80
hole rheological properties. According to their physical 0.0035 [88.90] 13.76 189 4.59
nature and their mechanism of action, LCMs can be clas-
0.0041 [104.14] 16.12 260 5.37
sified into five different groups:
0.0049 [124.46] 19.27 370 6.42
■ granular
0.0058 [147.32] 22.80 520 7.60
■ lamellar (or flake-like)
0.0069 [175.26] 27.13 740 9.04
■ fibrous
0.0082 [208.28] 32.24 1,040 10.8
■ mixed
0.0097 [246.38] 38.14 1,460 12.7
■ encapsulated fluid-absorbing particles.
0.0116 [294.64] 45.61 2,080 15.2

0.0138 [350.52] 54.26 2,940 18.1

Granular LCMs
0.015 [381.00] 58.97 3,480 19.7
Granular LCMs are rigid materials that are spherical in
0.0164 [416.56] 64.48 4,160 21.5
nature. They form two types of bridges—one at the for-
mation face and one within the formation matrix. The 0.0195 [495.30] 76.67 5,880 25.6

latter type of sealing is preferred because the particles 0.0232 [589.28] 91.22 8,320 30.4
are not easily dislodged by pipe movement in the well- 0.0276 [701.04] 108.5 11,800 36.2
bore. The effectiveness of granular LCMs depends pri- 0.0328 [833.12] 128.9 16,600 43.0
marily on selecting the proper particle-size distribution,
0.0390 [990.60] 153.3 23,500 51.1
with larger particles first forming a bridge across or
within the void and the smaller particles bridging the 0.046 [1,168.00] 181.0 32,800 60.3

openings between the larger particles (Gatlin and 0.055 [1,396.00] 216.0 46,700 72.0
Nemir, 1961). This process continues until the void 0.065 [1,650.00] 255.0 65,000 85.0
spaces become smaller than the drilling-mud solids. The 0.078 [1,980.00] 307.0 94,200 102
problem finally becomes one of filtration. A blend con-
0.093 [2,361.00] 365.0 133,000 122
taining a range of particle sizes is most commonly used.
Such systems are usually more successful in high-solids- 0.110 [2,793.00] 432.0 187,000 144
ratio systems such as cement slurries. 0.131 [3,326.00] 515.0 266,000 172
In 1976, Abrams showed that the median particle size 0.156 [3,960.00] 613.0 376,000 204
of the bridging additive should be slightly greater than or
0.185 [4,697.00] 727.0 529,000 242
equal to one-third the median pore size of the void.
0.221 [5,610.00] 868.0 753,000 289
Assuming a relatively homogeneous sandstone forma-
tion in which the sand grains are of similar size, it is pos- 0.263 [6,677.00] 1,034.0 1,070,000 345
sible to predict the required bridging-agent particle size 0.312 [7,921.00] 1,226.0 1,500,000 409
to bridge the pore throats of the formation matrix. These
values are given in Table 6-4. In addition, Abrams deter-
mined that the minimum concentration of bridging
solids is 5% by volume of solids in the final mud mix. More recently, Dick et al. (2000) wrote a software
However, Abrams’ rule only indicates the amount of application that determines the optimal blend of bridg-
material necessary to commence plugging. It does not ing agents according to the maximum pore size and the
address an ideal packing sequence for optimal sealing. formation permeability. Instead of using Abrams’ rule,

206 Well Cementing

the software is based on the ideal packing theory (IPT) Mixed LCMs
(Kaeuffer, 1973). The work of Dick et al. saw the first Blends of granular, flake, and fibrous LCMs are also
practical application of IPT to oilfield use. effective. These LCMs also provide a gradation of parti-
The IPT takes a graphical approach to determine the cle sizes as well as a variation of material types for seal-
optimal concentration of bridging material for given for- ing different classes of lost circulation zones.
mation characteristics. This theory can be applied to Howard and Scott (1951) performed a series of exper-
both lost circulation pills and water- or oil-base drill-in iments comparing the fracture sealing capacity of vari-
fluids and is particularly applicable to systems used to ous types of LCMs with their concentration in drilling
drill or seal permeable sand, limestone, or fractured mud (Fig. 6-11). They found that granular LCMs were
shale. This is a field-proven theory that has been applied more effective than the laminar or fibrous materials for
successfully in all of the above applications. For exam- sealing larger fractures. Table 6-5 is a list of typical com-
ple, the ideal particle-size distribution based on a maxi- mercial materials, their particle-size distributions, and
mum pore-throat size of 133 μm is shown in Fig. 6-10. the normal concentrations used.
The figure also shows the particle-size distributions for Nayberg and Petty (1986) performed a laboratory
various commercial bridging materials. study comparing the effectiveness of fibers, flakes, gran-
ules, and thermoset rubber in controlling mud losses to
Fibrous LCMs simulated medium-size (0.13 in. [3.3 mm]) fractured for-
Fibrous materials are best for controlling losses to mations. They claimed that a blend of medium- and fine-
porous and highly permeable formations because they grained (10- to 200-mesh) particles of thermoset rubber
are able to form a mat-like bridge over the pore openings. performed better than the conventional LCMs. An inter-
The mat reduces the size of the openings to the forma- esting observation was that granular LCMs sometimes
tion, permitting the colloidal particles in the mud to exhibited a “channeling” phenomenon. When a high
rapidly deposit a filtercake. pressure differential and an insufficient mud-solids con-
centration existed, a bridge at the formation face, or
Lamellar LCMs within the formation matrix, could not develop.
The conventional LCMs tend to be supplied in three
Lamellar LCMs are flat, layer-like materials with limited grades—fine, medium, and coarse.
or no rigidity that are also designed to bridge and form a
■ Under most circumstances, fine materials will pass
mat on the formation face. These LCMs also provide the
best results when treating losses to permeable and through the shaker screens and stay in the mud.
porous formations. ■ Medium materials tend to be screened out of the mud
but most likely will not plug jets or other downhole
tools.

100.00
90.00
80.00

Cumulative 70.00
percent 60.00
below
particle 50.00
size 40.00 Ideal PSD utilizing
indicated largest pore size
on x-axis 30.00 CaCO3 (VF)
20.00 CaCO3 (F)
CaCO3 (M)
10.00 CaCO3 (C)
Closest mix with CaCO3
0.00
0 5 10 15 20 25
Square root of particle size (μm)

Fig. 6-10. Ideal particle-size distribution of bridging agent for a formation with a 133-μm pore size.
Particle-size distributions (PSDs) of various commercial bridging agents are also shown (from Dick, 2000).

Chapter 6 Cement-Formation Interactions 207

 20
A typical
16 lamellar
material

12
Material
concentration A typical A typical
(lbm/bbl) 8 fibrous granular
material material

0
0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.18 0.2
Largest fracture sealed (in.)

Fig. 6-11. Effect of concentration of LCMs when sealing fractures (after Howard and Scott, 1951).
Reproduced courtesy of the American Petroleum Institute.

Table 6-5. Typical LCMs†

Material Type Description Concentration Largest Fracture Sealed (in.)
(lbm/bbl)
0 0.04 0.08 0.12 0.16 0.20
Nutshell Granular 50%– 3⁄16 + 10 mesh 20
50%–10 + 100 mesh

Plastic Granular 50%–10 + 100 mesh 20

Limestone Granular 50%–10 + 100 mesh 40

Sulfur Granular 50%–10 + 100 mesh 120

Nutshell Granular 50%–10 + 16 mesh 20

50%–30 + 100 mesh

Expanded perlite Granular 50%–3/16 + 10 mesh 60

50%–10 + 100 mesh

Cellophane Lamellar 3
⁄4-in. flakes 8

Sawdust Fibrous 1
⁄4-in. particles 10

Prairie hay Fibrous 1

⁄2-in. fibers 10

Bark Fibrous 3
⁄8-in. fibers 10

Cottonseed hulls Granular Fine 10

Prairie hay Fibrous 3

⁄8-in. particles 12

Cellophane Lamellar 1
⁄2-in. flakes 8

Shredded wood Fibrous 1

⁄4-in. fibers 8

Sawdust Fibrous 1⁄16-in. particles 20

† After Howard and Scott (1951). Reproduced courtesy of the American Petroleum Institute.

208 Well Cementing

■ Coarse materials can plug off any downhole assem- is stopped (Chapter 4). In practical terms, thixotropic
blies except open ended drillpipe. systems are fluid during mixing and displacement but
rapidly form a rigid, self-supporting gel structure when
Other specialty materials are graded according to parti-
pumping ceases. When a thixotropic slurry enters a lost
cle size or the size of the opening through which they
circulation zone, the velocity of the leading edge
will pass.
decreases and a gel structure starts to form (Chapter 7).
The first patent concerning the use of encapsulated
As the gel strength develops, resistance to flow increases
particles to control lost circulation was that of
until the entire zone is plugged (Childs et al., 1985).
Armentrout (1958). The technique consists of encapsu-
Such systems are very effective for solving severe lost cir-
lating bentonitic particles within a water-insoluble poly-
culation in naturally fractured formations.
meric coating, through which a small hole is drilled.
When the encapsulated bentonite is pumped down the
6-3.4.2.2 High-fluid-loss squeezes
wellbore, water from the mud seeps through the hole.
The bentonite swells and ultimately ruptures the coat- When a high-fluid-loss slurry is squeezed into the loss
ing. The swollen bentonite then seals the voids in the zone, it readily dehydrates so that solids pack the frac-
lost circulation zone. Walker (1987) followed this by tures and form a seal. A typical high-fluid-loss slurry con-
describing a technique in which the lost circulation tains a mixture of diatomaceous earth, bridging agents,
additive is a highly water-absorbent polymer encapsu- and barite suspended in either water or oil.
lated by a protective coating. The particle size of the These slurries are effective for sealing induced frac-
encapsulated particles varied from 0.1 μm to 5 mm. The tures in which external bridging is not paramount and
casing can be a material that slowly dissolves when in when it is important to achieve a high pressure drop in
contact with the wellbore fluid or a waxy substance that the fracture. The mud solids should provide the neces-
melts at a temperature between the bottomhole static sary fines for bridging. In low-porosity, fractured forma-
and circulating temperatures. The polymer then absorbs tions, 10 lbm/bbl [0.03 kg/L] of fine fibrous LCM is usu-
water, forming a semisolid, nonflowing mass that seals ally added. Coarse or granular LCMs should not be added
the zone. The water-absorbent polymers include alkali because they may prevent ingress of the slurry into the
metal polyacrylates or saponified copolymers of a vinyl fracture. Coarse or granular LCMs may also act as a
ester, which have the capacity to absorb more than 100 proppant if they invade the fracture.
times their weight of water. Another patent by Hydrostatic pressure is often sufficient to seal the
Delhommer and Walker (1987) described a very similar loss zone. A light squeeze pressure (100–300 psid
technique for oil-absorbing polymers, enabling the use of [4.79–14.4 kPa]) may be applied to open up and then
such systems in OBM. seal fractures. High-fluid-loss slurries can be pumped
through the drillbit jets.
6-3.4.2 Surface-mixed systems 6-3.4.2.3 Polymer-base crosslinked systems
When adding LCMs to drilling fluids does not control lost Crosslinking is the linking of two independent polymer
circulation, other types of fluids are often prepared at chains by a crosslinking agent that spans or links two
the wellsite and pumped downhole to seal the formation. chains. Polymeric plugs (or “pills”) are a blend of poly-
mers, crosslinking agents, and fibrous LCMs designed to
6-3.4.2.1 Cement plugs plug naturally fractured or vugular formations. They can
Neat cement slurries are effective for solving seeping or be activated by a crosslinking agent, time, temperature,
minor losses and offer the advantage of providing high or shearing at the bit (Quinn et al., 1999; Caughron et al.,
final compressive strengths. Slurries with a limited 2002). After setting, the pills can be rubbery, ductile, and
degree of fluid-loss control can be used to solve seeping, spongy soft. Crosslinked pills are solid but tend to have a
partial, or total losses; they may contain a mixture of rubbery consistency with little compressive strength.
clays, diatomaceous earth, and LCMs. One selects the The shear strength is high enough to support a fluid
LCM particle size according to the severity of the losses. column but low enough to allow removal by washing or
Low-density cement systems can be used for any type of jetting. The setting time is fully controllable by using
lost circulation problem. They have the added advantage either a retarder or accelerator that is based upon the
of reducing the hydrostatic pressure. temperature of the thief zone or bottom hole. Most plugs
Thixotropic cements are also used as cement plugs. can be weighted up to 16.0 lbm/gal and are thermally
Thixotropy is a term used to describe the property exhib- stable up to 300°F [149°C]. Care should be exercised in
ited by a system that is fluid under shear (i.e., pumping or near the producing interval, because the plugs cannot
or agitation) but develops a gel structure when the shear be degraded and produced back. Different types of poly-

Chapter 6 Cement-Formation Interactions 209

mer-based crosslinked pills (PCP) are engineered for
specific thief zones and lost circulation mechanisms
(Table 6-6).
The blend consists of medium- to high-molecular-
weight crosslinking polymers and sized fibrous materials.
When added with a biopolymer and activated with a com-
bination of crosslinking agent, time, and temperature, it
produces a rubbery, ductile, spongy, set gel (Fig. 6-12).

6-3.4.2.4 Other surface-mixed systems

Systems that do not contain Portland cement usually
involve a gelling agent with an activator. After a given
period of time, or because of a temperature increase, the
components react to form a nonflowing mass. The advan-
tage of such systems is the ability to predict when the
mixture will change from a liquid to a solid. In general,
they are most applicable to partial lost circulation prob-
lems in high-permeability sandstones or for sealing
microfissures.
Sharp (1966) first described the use of an aqueous
solution of sodium silicate and urea that, at tempera-
tures above 145°F [63°C], reacts to form a hydrosol of
silicic acid. With time, the hydrosol converts to a silica
gel within the formation, providing a firm structure that
is essentially impervious to fluid. The gel time may be Fig. 6-12. Polymer-base crosslinked pill (PCP) (from Ivan et al., 2002).
preselected by varying the relative concentrations of the
reactants; however, the downhole temperature must
exceed 145°F [63°C] for the reaction to proceed at a Smith (1986) used reducing sugars, such as lactose
useful rate. This feature permits the preblending of the and fructose, as thermally responsive activators for sili-
mixture several hours before the operation commences. cate solutions. For applications in which the well tem-
Elphingstone et al. (1981) described the use of halo- perature is below 120°F [49°C], the addition of small
genated hydrocarbons, more specifically sodium amounts of a reactive salt (such as calcium chloride)
trichloroacetate, as an activator for aqueous silicate was suggested to provide short gelling times without
solutions. The addition of silica flour (325-mesh) was having to increase the concentration of the reducing
recommended to increase the viscosity of the final gel. sugar.

Table 6-6. Description of Polymer-Base Crosslinked Pills†

Product Name Description
PCP for large natural fractures and PCP for large natural fractures or vugular zones is a one-sack blend of polymers, crosslinking
vugular lost circulation zones agents, and medium-to-coarse fibrous LCMs. When activated with time and temperature, PCP
produces a soft-to-medium strength, rubbery, ductile, spongy, set gel that effectively prevents
loss of fluid to the formation. The LCM in PCP packages is a fibrous cellulose type containing
a wide variation of particle sizes.

PCP for induced fractures and porous For induced fractures or matrix losses (i.e., microfractured network or permeable formations)
and/or permeable formations the PCP has been reengineered for maximum penetration and higher strength. In this case,
the blend consists of medium-molecular-weight crosslinking polymers and fine-sized fibrous
materials. When added with biopolymer and activated with a combination of crosslinker, time,
and temperature, PCP produces a medium-to-hard strength, rubbery, ductile plug. The LCM
in this variety of PCP comprises specially sized and concentrated fibrous cellulose containing
a mixture of fine particle sizes to plug deep fractures, faults, and matrices. The crosslinking
agent is packaged separately. Thus, the plug can be mixed ahead of time and will not cross-
link. The crosslinker is added just before pumping.
† from Ivan et al. (2002).

210 Well Cementing

 For OBM, additives such as organophilic clays, soaps, As the temperature rises, swelling starts to take place
asphaltic materials, and mineral fibers have been used and eventually a stable system forms when equilibrium
for many years to impart gel strength. These materials is achieved. In the presence of shear and temperature,
have various shortcomings, most notably limited thermal this process takes place much faster (Stage 3). When the
stability. Patel and Salandanan (1985) developed an system is totally activated, it remains stable even if the
improved system that comprises a polymeric oil gellant temperature drops (Stage 4).
that imparts thixotropic properties to oil-base drilling Yearwood et al. (1988) and Vidick et al. (1988)
fluids. The gelling agent is thermally activated, forming described the use of an internally activated low-viscosity
a gel structure as soon as the temperature exceeds silicate solution that, depending on the fluid design and
140°F [60°C]. The structure melts at 250°F [121°C]. the temperature, gels rapidly. The final gel is strong and
Figure 6-13 depicts the activation of the gelling agent by permanent, with very little free water at temperatures
a combination of temperature and shear. The gelling up to 355°F [180°C]. To demonstrate the sealing capac-
mechanism of the material involves a swelling of the ini- ity of this system, a series of laboratory experiments was
tial agglomerates and a gradual release of the individual performed using core plugs with different permeabili-
oligomer chains (Stage 2). The oligomers associate with ties. The results demonstrated that, once the gel has
other particulate materials in the system to form a formed in the formation matrix, the system is able to
thixotropic fluid. withstand differential pressures greater than 1,500 psi/ft
[0.24 Pa/m] (Table 6-7).

Stage 1 Stage 2 Stage 3 Stage 4

Agglomerated Swollen and Activated Activated
deagglomerated
Time, shear, Time,
Shear
warming shear

Solvent Heat Cool

swelling down
a b c d
Fig. 6-13. Activation of OBM gelling agent.

Table 6-7. Performance of Internally Activated Silicate System in a Core-Flow Test†

Core Nature Average Saturating Test Extrusion Pressure
Permeability Fluid Temperature Resistance for 1 ft
to Water (D) (°F [°C]) of Plugged Core (psi)
20/40 fracturing sand 6 Fresh water 105 [40.5] 1,200

Porous sandstone 2 Fresh water 105 [40.5] >1,500

Porous sandstone 2 Diesel oil 105 [40.5] >1,500

Porous sandstone 2 Brine 105 [40.5] >1,500

Fissured limestone 1 Fresh water 140 [60.0] >1,500

Fissured limestone 1 Brine 140 [60.0] >1,500

20/40 fracturing sand 6 Fresh water 175 [79.4] >1,500

20/40 fracturing sand 6 Brine 175 [79.4] >1,500

† Yearwood et al. (1988). Reprinted with permission from the Petroleum Society of CIM.

Chapter 6 Cement-Formation Interactions 211

 Vidick et al. (1988) presented an equation to relate A borate crosslinked, xanthan-base polymer pill
the gelling time of the silicate systems to the active (Powell et al., 1991) also can act as a bridging material.
matter content and the temperature. Equation 6-19 Modified starch polymers are also added as fluid-loss
helps not only to predict the gelling time at a particular additives. Further, sized borates are used for additional
temperature for a given active matter content but also to bridging to reduce losses while the plug sets. The
calculate the gel time variations resulting from slight crosslinking time can be controlled by using accelera-
bottomhole temperature fluctuations. tors or retarders. The plug is acid soluble, more than 90%
being dissolved on contact with a 15% HCl solution.
⎛ – Eact ⎞ An acid-soluble cementitious product is Sorel
⎜ RT ⎟
Kx ⎝ ⎠
tgel = K T × x ×e (6-19) cement, a mixture of magnesium oxide, magnesium chlo-
ride, and water (Chapter 7). Alsdorf and Dittmar (1987)
where pointed out that this type of cement is not applicable at
elevated temperatures, because control of the setting
Eact = Activation energy (kJ/mol)
time is difficult. They recommended the use of a mater-
KT = A constant (function of the temperature)
ial containing about 60% by weight of ground milk-of-
Kx = Coefficient related to the active matter content
lime grit, plus calcium or magnesium chloride and water.
R = Gas constant (8.3143 J/mol K)
By varying the percentage of the calcium or magnesium
tgel = Gelling time (min)
chloride, it is possible to vary the thickening time from 1
T = Temperature (K)
to 4 hr at temperatures up to 195°F [90°C]. The maxi-
x = Active matter content (% by volume).
mum compressive strength is obtained after 5 to 24 hr.
The values for Kx and KT at four temperatures are The final product is completely soluble in 5% hydrochlo-
given in Table 6-8. ric acid.

Table 6-8. Values for Kx and KT at Different Temperatures†

6-3.4.3 Downhole-mixed systems
Downhole-mixed systems consist of two or more fluids
Temperature Kx KT that, upon making contact in the wellbore or the lost cir-
(°F [K]) (min)
culation zone, form a viscous plug or a precipitate that
104 [313] –16.94 e62 seals the zone. It is common practice to prevent the
140 [333] –17.14 e61 mixing of the fluids until they are in front of the lost cir-
culation zone, either by pumping a spacer or by pumping
176 [353] – 3.84 e18 one fluid down the drillstring while the other fluid is
266 [403] –11.34 e51
simultaneously pumped down the annulus. These sys-
tems are not suitable for total lost circulation situations,
† Vidick et al. (1988). in which the actual displacement rates are not known,
because it is very difficult to control the mixing of the
fluids.
In cases in which the lost circulation zone is also a For partial losses, Iljas (1983) found better success by
zone of interest, either for production or injection pur- using mud-diesel-oil-bentonite (M-DOB) plugs instead of
poses, it may be necessary to design the plugging mater- LCMs. M-DOB plugs are a combination of diesel oil and
ial for eventual removal during the completion of the bentonite and are sometimes called “gunk plugs.” When
well. Such removable systems are generally acid soluble, this mixture contacts water or WBM, it forms a mass
consisting of bridging agents slurried in a viscous fluid, with high gel strength. Soft, medium, and hard plugs may
crosslinked polymer plugs, or cementitious materials. be formed by controlling the proportions of the ingredi-
Typical bridging materials include ground calcium car- ents. The M-DOB slurry is pumped down the drillpipe,
bonate particles with diameters ranging from 0.0003 in. and the mud down the annulus.
[8 μm] to 0.01 in. [250 μm]. They are used at concentra- M-DOB plugs suffer from several drawbacks.
tions up to 10 lbm [4.5 kg]/bbl of carrying fluid. ■ They break down with time.
Assuming a relatively homogeneous sandstone formation ■ They are difficult to apply in long openhole intervals.
in which the sand grains are of similar size, it is possible
■ When losses are severe, it is impossible to achieve a
to predict the required particle size of calcium carbonate
to form a bridge in the pore throats of the formation reliable pumping rate down the annulus; therefore,
matrix, thereby reducing the loss of fluid. These values the degree of mixing cannot be controlled.
are given in Table 6-6. Alternatively, one can use the IPT ■ They develop no compressive strength.

software described earlier in this chapter.

212 Well Cementing

 Gaddis (1975) increased the gel strength of the M- tions, with small freshwater spacers as separating fluids.
DOB plug by blending a water-soluble polymer with the The entire sequence is then followed by a Portland
bentonite in diesel oil. On contact with water, the poly- cement slurry.
mer hydrates, and the clay flocculates to form a stiff Quinn et al. (1999) presented a novel shear-sensitive
cement-like plug. For severe losses, Messenger (1981) plugging fluid that can be pumped through the drill bit,
and Iljas (1983) suggested a better version of the curing losses instantly. The fluid is an invert emulsion.
M-DOB plug—the mud-diesel-oil-bentonite-cement The continuous phase is a suspension of crosslinker par-
plug (M-DOB2C). The advantage of this system is the ticles in oil, and the internal phase contains a high con-
development of compressive strength. The ratios of mud centration of polymer. As the emulsion passes through
and DOB2C required to produce mixtures of various the bit nozzles, the pressure drop causes the emulsion
hardnesses are shown in Table 6-9. Many downhole- to “flip” and exposes the polymer to the crosslinker
mixed systems use a combination of two or more sur- (Fig. 6-14). The result is a rigid gel that can be designed
face-mixed systems to provide an effective plugging to generate a solid structure from 1 to 10 min after pass-
material. For example, an M-DOB plug can be followed ing though the bottomhole assembly.
by a cement plug, thereby improving its strength and
durability.

Table 6-9. Hardness of Different Combinations of Mud with

DOB2C†
Mud DOB2C Hardness
1 1 Soft 500 psi
[35 bar]
1 1.5 Medium Emulsion Set gel

1 2 Medium hard Water phase containing

high concentration
1 2.5 Hard of polymer
Oil phase
1 3 Very hard
Crosslinker particle
† Iljas (1983).

Fig. 6-14. Mechanism of invert emulsion system for curing lost

circulation problems (from Quinn et al., 1999).
In 1972, Biles described a technique for sealing highly
permeable channels. A sodium silicate solution is
allowed to mix with a solution containing divalent This gelling mechanism ensures accurate placement
cations, forming a precipitate. This technique success- at the loss zones and is more reliable than pills activated
fully sealed permeable formations, but the precipitate by downhole temperature or downhole fluid interaction.
was not sufficiently strong to seal naturally fractured for- After placement, the set gel is stable for several weeks
mations. Russell (1983) refined this technique by under downhole conditions, allowing sufficient time to
employing a preflush of extended Portland cement drill and complete the section.
slurry, followed by a sodium silicate solution and then a Cement can also be incorporated into the emulsion
neat or thixotropic cement slurry. Both laboratory and system. Such systems initially form a high-viscosity gel.
field results showed a dramatic strength improvement The cement sets later, providing long-term lost circula-
when the sodium silicate and cement slurry intermixed. tion control.
This was apparently caused by the high availability of
calcium ions from the cement and the instantaneous
dehydration of the cement slurry owing to the reaction.
Murphey (1983) proposed the use of potassium sili-
cate instead of sodium silicate because the latter may
tend to gel prematurely when mixed with brine. He
described a common practice for solving total lost circu-
lation in fractured and cavernous formations—pumping
alternating batches of silicate and divalent cation solu-

Chapter 6 Cement-Formation Interactions 213

6-3.5 Methods for controlling massive lost 6-3.5.3.2 Mixed metal hydroxide and mixed
circulation metal oxide systems
Mixed metal hydroxides (MMH) are synthetic inorganic
6-3.5.1 “Floating mudcap” method compounds designed to be electron deficient with a high
The “floating mudcap” method can be used to complete aspect ratio (i.e., narrow edge and high surface area on
a section in which massive losses are occurring. This the face) (Burba et al., 1988). Furthermore, the MMH
method involves keeping the annulus full of weighted crystals are manufactured to be smaller than a typical
mud to control formation pressures while drilling ahead bentonite platelet. When the MMH and bentonite are
with water, which is continually lost to the formation. mixed in an aqueous medium, a strong association
The density of the fluid in the annulus must be high occurs. This results in a mud with exceptional suspen-
enough to maintain well control, but not so high that the sion abilities (Fig. 6-15). Such muds can also prevent
fluid induces further losses. If possible, the casing and cure circulating losses. Good results have been
should be set just above the loss zone, because this achieved in areas where other methods to control lost
makes control of the mudcap considerably easier. circulation have failed (Fraser and Aragão, 2001).
Because of the safety implications and potential loss of
well control, it is necessary to use well-trained and expe-
rienced rig crews for this relatively risky method.

6-3.5.2 Drilling “blind”

When caverns are too large to realistically fill with mud
or seal off, an established practice is to continue drilling
until competent beds are encountered and then to set
the casing. This procedure is commonly known as
“drilling blind.”

6-3.5.3 Using specialized drilling fluids

6-3.5.3.1 Foamed and aerated mud systems
Foamed muds (Kuru et al., 1999) and foamed cement
can be applied successfully in massive loss situations.
However, foams should only be used where formation
pressures are low and the formation is competent,
because there is little fluid pressure to support the well-
bore wall.
Another technique uses “microbubbles” (Growcock et
Fig. 6-15. Suspension performance of drilling fluid containing mixed
al., 2004) that are noncoalescing and can be recircu- metal hydroxide.
lated. These unusual bubbles, called “aphrons,” are
made to occur within the mud system without injecting
external air or gas. Aphrons exist as independent
spheres with a gas or air core encapsulated by a multi- Mixed metal oxide (MMO) systems are the successor
ple-layer film. This film is the key to maintaining the to mixed metal hydroxides (Santarelli et al., 1992) and
bubble strength that allows aphrons to function as a mixed metal oxide and hydroxide (MMOOH) systems.
bridging agent. A surfactant is used to form the aphrons, Electrostatic interactions between bentonite platelets
build the multilayer bubble walls, and create interfacial with negative faces and MMO crystals with positive
tension to bind the aphrons into a network capable of charges lead to a network of bentonite/MMO complexes
creating downhole bridging. that imparts a thixotropic nature to the fluid (Fig. 6-16).
An MMO crystal is 10 times smaller than a bentonite
platelet, and has 6 to 7 times its charge density. MMO
systems operate under the same principles as their pre-
decessors but are more cost-effective than either MMH
or MMOOH systems.

214 Well Cementing

 Bentonite MMO Bentonite/MMO complex

and gives

Fig. 6-16. Schematic diagram of MMO/bentonite association.

6-3.6 Lost circulation during cementing The downhole pressures exerted on lost circulation
Before initiating a conventional primary cementing zones can also be decreased by using mechanical
operation, any lost circulation problem should be elimi- devices such as stage collars or external casing packers
nated or significantly reduced by the techniques (ECPs). Stage collars permit the casing string to be
described above. If this is not possible, or if losses are cemented in two or three stages, lowering the dynamic
anticipated during the primary cementing job, there are and hydrostatic pressures (Chapters 11 and 12).
two options for remediation, as described by Nayberg To reduce the risk of cement fallback if losses do
and Linafelter (1984). The first is to maintain the down- occur, a special stage collar with a packoff adaptation
hole pressure during the job below the maximum equiv- can be used. When expanded, this stage collar provides
alent mud circulating density by reducing the density of a seal between the casing and the formation to prevent
the cement slurry, minimizing the height of the cement downward fluid movement. Cement baskets can also be
column, or limiting the casing and annular friction pres- placed just below the stage collar to provide the same
sures during the placement of the cement slurry. The effect. Turki and Mackay (1983) described the place-
second option is to pump a plugging material as a spacer ment of ECPs immediately above the lost circulation
in front of the cement slurry, add LCMs to the cement zone to reduce the hydrostatic pressure. A typical appli-
slurry itself, or use special additives that impart cation would be a two-stage job with an ECP just above
thixotropic properties to the cement slurry. The third the lost circulation zone and a stage collar just above the
option is to use a combination of techniques, which is ECP. After the first stage is performed, the ECP is
often necessary when trying to prevent cement losses to expanded to seal the annulus, preventing the transmis-
highly fractured or vugular formations. sion of hydrostatic pressure to lower zones (Fig. 6-17).
However, if the size of the hole is larger than anticipated,
the ECP may fail to provide a perfect seal because of
6-3.6.1 Downhole pressure reduction insufficient lateral expansion.
Computer simulators can calculate the estimated down- Turki and Mackay (1983) also mentioned the “hydro-
hole pressures at any particular depth in the well and at static cementing technique” for attempting to obtain
any time during the cementing operation (Chapter 12). zonal isolation across cavernous lost circulation zones. A
This enables the operator to know (for a particular well conventional first-stage job is performed, followed by
completion) exactly which cement-slurry parameters pumping a predetermined quantity of cement slurry
and job procedures are required to prevent lost circula- down the annulus. Most of the slurry is lost to the cav-
tion and maintain adequate hydrostatic pressure in front ernous formation. However, after the hydrostatic pres-
of permeable zones. The most relevant parameter is the sure of the cement slurry equilibrates with the formation
cement-slurry density, which may be reduced by adding pressure of the lost circulation zone, a portion will
one or more cement extenders. Chapter 3 provides a remain in the annulus. When the cement sets, the
detailed discussion of extenders and the optimal slurry- cavern is bridged, and cement exists at some height
density range for each. above the cavern. The application of this technique was
The rheological properties of cement slurries may recommended only when lost circulation cannot be sig-
also be adjusted to reduce friction pressures during nificantly reduced by conventional means or when open
placement. This is especially critical in narrow annuli in holes are excessively washed out.
which viscous slurries can cause very high friction pres-
sures. Another technique mentioned by Nayberg and
Linafelter (1984) is to lighten the hydrostatic column
above the top of the cement by injecting nitrogen into
the mud.

Chapter 6 Cement-Formation Interactions 215

 cium ions are present in the formation, a second pre-
flush of a calcium chloride solution can be pumped. The
133⁄8-in. casing shoe high concentration of calcium ions in the cement slurry
at 2,500 ft ensures immediate sealing of the formation.

6-3.6.3 LCMs for cement slurries

Nayberg and Petty (1986) and Turki and Mackay (1983)
Second-stage cement agreed that the effectiveness of LCMs in cement slurries
to surface is limited to minor or partial losses in highly permeable
formations, and that they are not for solving total lost
circulation in naturally fractured or cavernous forma-
Nonpacker tions. When LCMs are used in the cement slurry, care
multistage tool
must be taken to ensure that the materials are inert to
the cement composition. Also, the size and concentra-
ECP assembly tion of the materials should be selected to avoid bridging
or plugging of the downhole equipment. The morpholo-
Loss zone gies of the materials are the same as those used in
First stage cement
drilling fluids. The authors are not aware of any reported
use of encapsulated additives in cement slurries for solv-
First stage shutoff baffle
ing lost circulation problems. Table 6-10 is a list of LCMs
95⁄8-in casing shoe for cement slurries, their properties, and typical effec-
at 4,500 ft tive concentrations. These LCMs are also described in
Chapter 3.
Fig. 6-17. Cementation using an ECP. The most common LCMs for cement slurries are of
the granular type, designed to bridge at the formation
face or within the matrix. Gilsonite, a naturally occur-
ring asphaltine hydrocarbon with a particle size in the
6-3.6.2 Preflushes range 8–60 mesh, is widely used. Gilsonite is not suitable
Murphey (1983) described the use of a potassium sili- for high-temperature applications because of its low
cate solution as a preflush to enable the formation to melting point (220°F [104°C]). Crushed coal, with a
support a greater than normal hydrostatic pressure. The standard mesh size in the range 14–200 and a melting
preflush penetrates the highly permeable formations, point of approximately 1,000°F [538°C], is applied in the
permitting contact with calcium ions in the formation same manner as gilsonite, and can be used in high-tem-
and resulting in the formation of a gel. If insufficient cal- perature wells. Shells from walnuts, pecans, and other

Table 6-10. Materials Commonly Added to Cement Slurries to Control Lost Circulation
Type Material Nature of Particles Amount Used Water Required
Granular Gilsonite Graded 5 to 50 lbm/sk 2 gal/50 lbm

Perlite Expanded ⁄2 to 1 ft3/sk

1
4 gal/ft3

Walnut shells Graded 1 to 5 lbm/sk 0.85 gal/50 lbm

Coal Graded 1 to 10 lbm/sk 2 gal/50 lbm

Lamellar Cellophane flake Flaked ⁄8 to 2 lbm/sk

1
none

Fibrous Nylon Short fibers 1 ⁄8 to ⁄4 lbm/sk

1 1
none

Glass Long fibers 2 to 3 lbm/sk none

216 Well Cementing

nuts are also available in fine, medium, and coarse 6-3.6.4 Thixotropic cement systems
grades; however, care should be exercised at concentra- The self-supporting property of thixotropic cements is
tions above 4 lbm/sk to avoid the plugging of downhole useful across formations with low fracture gradients.
equipment. When ordinary slurries pass over a weak zone, the
Cellophane flakes with diameters of 3⁄8 to 3⁄4 in. [9.5 to increase in hydrostatic pressure can cause formation
19 mm] are the most common flake material. At con- breakdown. As a result, the top of the cement falls to a
centrations above 2 lbm/sk, bulk loading and mixing of point below the desired level of fill-up. Thixotropic slur-
the cement slurry becomes extremely difficult. ries do not fall back because some of the hydrostatic
Nylon and polypropylene fibers have been used as pressure is transmitted to the formation face and casing
LCMs in cement slurries. As the cement slurry enters walls. Several thixotropic cement compositions exist,
the lost circulation zone, the fibers associate to form a and their chemistries are described in Chapter 7.
mat that promotes cement filtercake development.
However, their use has been limited by problems of
mixing them into cement slurries and their tendency to 6-3.6.5 Foamed cement systems
plug pump plungers and float equipment. Very often, the Owing to their low density, foamed cement systems are
fibers are seen floating on the slurry surface and forming often used as an alternative to conventional cement sys-
“fur balls.” tems to solve lost circulation problems (Nayberg and
More recently, silica-base fibers have been developed Petty, 1986). The design and application of foamed
as LCMs for cement slurries (Messier et al., 2002; Low et cement is discussed thoroughly in Chapter 7.
al., 2003; El-Hassan et al., 2003). They are very flexible
and readily mix into cement slurries. The fibers are sup-
plied in small bundles called “tows.” When dispersed in 6-3.7 Lost circulation design software
an aqueous medium, the tows dissociate and readily An innovative software program (Bruton et al., 2004) has
form a fibrous network (Fig. 6-18). been developed to assess operator and well-specific lost
The silica fibers have been used successfully in both circulation problems and link them to available solu-
primary and remedial cementing applications. No diffi- tions. The software tool focuses on existing resources,
culties with field equipment have been encountered, such as lost circulation products, procedures, geome-
even in coiled tubing applications. chanical models, logs, and mineralogy/geology analyses
to create project-specific plans and solutions.
The software contains four key modules designed to
(a) consider well-specific data, (b) determine the most
likely thief zones, (c) locate the position of potential
thief zones relative to key depths of interest, such as the
casing/liner shoe, and (d) identify the most appropriate
preventative and treatment measures (Fig. 6-19). The
goal is to minimize and prevent lost circulation problems
rather than develop an after-the-fact cure. The integra-
tive preplanning process analyzes offset well histories
and formation data to identify risk zones and also gath-
Fig. 6-18. Photographs of silica-fiber “tows” (left) and a dispersed- ers information on the exact fracture and pore size as
fiber network (right). well as fracture density. Afterward, detailed interval-
specific solutions are developed for stopping losses
encountered while drilling or tripping in.

Chapter 6 Cement-Formation Interactions 217

 a. Gathering and processing all specific well data b. Identifying the thief zone(s)

c. Developing lost-circulation prevention measures d. Identifying the best lost circulation solutions

Fig. 6-19. The four key modules of the lost-circulation software.

6-3.8 Lost circulation—conclusions the wrong zone. Many loss zones thought to be at the
Lost circulation problems, whether during drilling or bit are actually farther up the hole at the first point of
cementing, can be solved if the correct technique is loss.
applied for each individual case. Choosing the correct ■ LCMs and techniques must be systematically
solution from the wide variety of available remedies matched to the type and severity of the loss zone. For
described above can be a difficult task; however, certain example, using LCMs in the drilling mud to stop total
general guidelines can be followed. Messenger (1981) losses to a vugular limestone will almost never work.
summarized the most important factors to consider. One has a much better chance for success with a com-
■ The location of the loss zone must be determined bination of surface- and downhole-mixed systems
accurately; otherwise, the remedy will be placed in with low densities, thixotropic behavior, and good
strength development.

218 Well Cementing

■ Consulting records of prior experience with lost cir- Force required
culation in a particular field often points the way to to initiate movement
an effective solution.
■ Above all, careful prejob planning can prevent the
occurrence of lost circulation. It is important to
obtain, if possible, sufficient well information to per-
form a computer simulation of the cement job Core
(Chapter 12).

Cement
6-4 Cement-formation bonding
The ability of a cement sheath to provide zonal isolation
is directly related to the following properties: Contact area
■ the cement/casing interface
■ the bulk cement
■ the cement/formation interface.
Cementing recommendations are frequently based on
the compressive or tensile strength of set cement, and
many governmental regulatory bodies impose minimum
strength requirements for well cements. The assumption Force
Shear bond strength =
is that a material satisfying certain strength require- Contact area
ments will provide an adequate bond to the casing and
Fig. 6-20. Shear bonding.
formation. However, field and laboratory experience has
shown that this assumption is not always valid.
In a wellbore, shear bond and hydraulic bond are two
criteria often considered for effective zonal isolation and filtercakes on the bond strength. A detailed discus-
along the cement/casing and cement/formation inter- sion of their work is beyond the scope of this chapter;
faces. Shear bonding mechanically supports the pipe in however, the major findings are summarized in the fol-
the hole and is determined by measuring the force lowing discussion.
required to initiate pipe movement in a cement sheath More recently, work has been performed by Ladva et
(Fig. 6-20). This force, divided by the cement/casing al. (2004) to better characterize the cement-formation
contact surface area, yields the shear-bond strength. bond. Instead of working with sandstone and limestone,
Hydraulic bonding blocks the migration of fluids in a Ladva et al. concentrated on formations that contain
cemented annulus. It is usually measured by applying unstable clay minerals. Such formations have long been
pressure at the pipe/cement or pipe/formation interface known to be the most problematic during well construc-
until leakage occurs (Fig. 6-21). For zonal isolation, tion, particularly drilling.
hydraulic bonding is more important than shear bonding. This section first presents a background discussion of
In 1961, Evans and Carter published the results of an the cement/casing bond, then moves on to the subject at
extensive study that investigated shear and hydraulic hand: the cement-to-formation bond. The Evans and
bonding of cement systems to casings and formations. A Carter work with sandstones and limestones is discussed
variety of Portland cement systems was tested that con- first, followed by the Ladva et al. study with clay-bearing
tained various additives such as pozzolanic extenders, formations. The section concludes with a discussion of
bentonite, retarders, and fluid-loss additives. Latex-mod- work performed in the civil-engineering industry to
ified cements and resin cements were also tested. The determine how Portland cement slurries interact and
formation materials were sandstone and limestone. bond with formations.
Evans and Carter also evaluated the effects of mud films

Chapter 6 Cement-Formation Interactions 219

 Pressure

Hydraulic pressure
to initiate leakage
Resin

Cement slurry
Core

Cement Mudcake

Mudcake
Formation core

Core
Pressure

(a) (b)

Fig. 6-21. Hydraulic bonding to pipe (a) and formation (b).

6-4.1 Cement-to-casing bonding strength variations were observed when the test condi-
The shear-bond tests performed by Evans and Carter tions deviated from this ideal.
investigated the effects of pressure, pipe finishes (e.g., The effects of pipe-surface finish variations, as well as
mill varnishes), uncoated pipes (wire-brushed, rusty and pipe wetting with WBM and OBM, are shown in Table 6-11.
sand blasted), and WBM and OBM. A loss of shear-bond strength was readily noticeable
Evans and Carter found a correlation between when a thin sheath of mud was present at the pipe-
cement compressive strength and the shear-bond cement interface. OBM was more detrimental to the
strength. The data, shown in Fig. 6-22, were generated cement bond than the WBM. Lower bond strengths were
after the cement slurry had been set against clean, dry measured with new pipe coated with mill varnish. Higher
pipe with no internal pressure applied inside the pipe bond strengths were observed when the pipe surface was
during the setting process. Significant shear-bond- roughened by rust, wire brushing, or sandblasting.

300

200
Bonding strength
(psi)

100

0
0 500 1,000 1,500 2,000 2,500 3,000 3,500 4,000
Compressive strength (psi)

Fig. 6-22. Shear-bond strength versus cement compressive strength (from Evans and Carter, 1961).
Reproduced courtesy of the American Petroleum Institute.

220 Well Cementing

Table 6-11a. Bonding Properties of New and Used Pipe†, ‡ Table 6-11b. Effect of Mud Type on Bonding Properties
of New and Used Pipe†, ‡
Casing Type Time Hydraulic Shear
(days) Bond (psi) Bond (psi) Mud Type Time Hydraulic Shear
(days) Bond (psi) Bond (psi)
New 8 hr – 10
New Pipe
Used (rusted) 8 hr – 53
Water-base mud 2 175–225 46
New 1 300 79
Dry 2 375–425 284
New (sandblasted) 1 500 123
Used (Slightly Rusty) Pipe
Used (slightly rusty) 2 500–700 182
Oil-base mud 2 – 75
Used (wire brushed) 2 500–700 335
Water-base mud 2 – 174
New (sandblasted) 2 500–700 395
Dry 2 – 182
Used (rusted) 2 500–700 422 † From Evans and Carter (1961). Reproduced courtesy of the American Petroleum Institute.
‡ API Class A cement cured at 80°F; casing size of 2 in. inside 4 in.; cement sheath of 0.812 in.
† From Evans and Carter (1961). Reproduced courtesy of the American Petroleum Institute.
‡ API Class A cement cured at 80°F; casing size of 2 in. inside 4 in.; cement sheath of 0.812 in.

Table 6-12. Effect of Curing Temperature and Pressure on

Pressurizing the casing during the setting process is Hydraulic-Bond Strength to Pipe†, ‡
detrimental to the shear bond. The loss of bond strength
Curing Conditions Failure
is a function of the pressure and the pressurization time Pressure
(Fig. 6-23). Time Temperature Pressure (psi)
The effect of drilling mud on hydraulic-bond strength (°F) (psi)
to pipe was investigated. As shown in Table 6-11, the 24 80 0 500
hydraulic bond was reduced by the mud coating. Unlike
the shear-bond strength, the hydraulic bond strength 48 80 0 500
improved when the curing was performed under pres- 24 120 0 500
sure (Table 6-12).
24 140 2,000 600

24 120 3,000 800

24 80 0 600
† From Evans and Carter (1961). Reproduced courtesy of the American Petroleum Institute.
‡ The cement is API Class A; the pipe was 2-in. diameter inside 4-in. diameter and 10-in. long.

3,000
7 in.–9 3⁄4 in. (2 hr)
2,500
51⁄ 2 in.–85⁄8 in. (2 hr) 7 in.–9 3⁄4 in. (1 hr) 51⁄ 2 in.–85⁄8 in. (2 hr)
2,000
Closed-in
pressure 1,500
(psi) 5 1⁄ 2 in.–8 5⁄8 in. (24 hr)
1,000
7 in.–93⁄4 in. (24 hr)
500

0
0 25 50 75 100 125 150 175 200
Shear bond (psi)

Fig. 6-23. Shear bond versus casing pressurization (from Evans and Carter, 1961). Cement was cured for
24 hr at 150ºF—51⁄2 in. cemented inside 85⁄8 and 7 in. cemented inside 93⁄4. Closed-in pressure was held for
1, 2, and 24 hr. Reproduced courtesy of the American Petroleum Institute.

Chapter 6 Cement-Formation Interactions 221

6-4.2 Cement-to-formation bonding (sandstone reveals that squeezing generally has a beneficial effect
and limestone) on the bond strengths. The notable exception is the
Evans and Carter used Indiana limestone and Berea resin cement. Because the resin cement had no solids to
sandstone to evaluate the shear and hydraulic pressure deposit as a filtercake, it is logical that squeezing would
necessary to break the bond between the cement/forma- have little effect.
tion interface. The permeability of the Indiana lime- Figures 6-27 and 6-28 show the same comparison as
stone was 1 mD, and that of the sandstone was 100 mD. Figs. 6-25 and 6-26, with the exception that a thin layer
Figure 6-24 shows the shear-bond strength, compres- of mud was left on the walls of the test chamber. A reduc-
sive strength, and hydraulic-bond strength measured for tion in bond strength was noted when mud was left on
six cement systems placed against dry limestone cores. the walls, even though the compressive strength of the
The sandstone results were not shown because failure cement was the same.
occurred within the core rather than at the bonded Figures 6-25 through 6-28 also show that higher
interface. hydraulic-bond strengths were obtained against the less-
The next four plots contain data collected after the dense sandstone formation. The reason for this is that
cores had first been exposed to a water-base drilling cement slurry can be dehydrated more efficiently
mud. The cement slurry was either squeezed against the against a more permeable formation, resulting in a
formation core at a pressure of 100 psi [4.8 kPa] or higher-strength cement.
allowed to set unpressurized. The mud filtercake was As discussed earlier, clay-bearing formations are
either removed from the formation face or allowed to more problematic during the well construction process.
rest undisturbed. The work of Evans and Carter did not address such for-
Figure 6-25 shows the results when the formation mations. The remainder of this section is devoted to
walls were cleaned and the cement was squeezed. work with clay-bearing formations, notably shales. The
Despite the cleaning of the formation walls, the shear discussion begins with a short review of clays, followed
and hydraulic bond strengths are lower than those by information about how various drilling fluids and
shown in Fig. 6-24. cement slurries interact with shales. Finally, related
Figure 6-26 shows the results when the formation research performed in the civil engineering industry is
walls were cleaned but the cement was squeezed. presented.
Comparison of these results with those in Fig. 6-25

API Class A cement

Pozzolan X cement
2% bentonite

Pozzolan Y cement
2% bentonite
API Class A cement
12% bentonite
0.5% lignin retarder
API Class A cement
1% low fluid loss

Latex cement

Shear bond
Resin cement Compressive strength
Hydraulic bond
10 100 1,000 10,000 100,000
Shear bond, compressive, and hydraulic bond strength (psi)

Fig. 6-24. Bonding properties of cement to dry limestone (from Evans and Carter, 1961).
Reproduced courtesy of the American Petroleum Institute.

222 Well Cementing

 API Class A cement

Pozzolan X cement
2% bentonite

Pozzolan Y cement
2% bentonite
API Class A cement
12% bentonite
0.5% lignin retarder
API Class A cement
1% low fluid loss

Latex cement Shear bond

Compressive strength
Hydraulic bond
Resin cement Sandstone
Limestone
10 100 1,000 10,000 100,000
Shear bond, compressive, and hydraulic bond strength (psi)

Fig. 6-25. Bonding properties of cement to formation with cement squeezed and walls cleaned
(from Evans and Carter, 1961). Reproduced courtesy of the American Petroleum Institute.

API Class A cement

Pozzolan X cement
2% bentonite

Pozzolan Y cement
2% bentonite
API Class A cement
12% bentonite
0.5% lignin retarder
API Class A cement
1% low fluid loss

Latex cement Shear bond

Compressive strength
Hydraulic bond
Resin cement Sandstone
Limestone
10 100 1,000 10,000 100,000
Shear bond, compressive, and hydraulic bond strength (psi)

Fig. 6-26. Bonding properties of cement to formation with cement not squeezed and walls cleaned
(from Evans and Carter, 1961). Reproduced courtesy of the American Petroleum Institute.

Chapter 6 Cement-Formation Interactions 223

 API Class A cement

Pozzolan X cement
2% bentonite

Pozzolan Y cement
2% bentonite
API Class A cement
12% bentonite
0.5% lignin retarder
API Class A cement
1% low fluid loss

Latex cement Shear bond

Compressive strength
Hydraulic bond
Resin cement Sandstone
Limestone
10 100 1,000 10,000 100,000
Shear bond, compressive, and hydraulic bond strength (psi)

Fig. 6-27. Bonding properties of cement to formation with cement squeezed and walls not cleaned
(from Evans and Carter, 1961). Reproduced courtesy of the American Petroleum Institute.

API Class A cement

Pozzolan X cement
2% bentonite

Pozzolan Y cement
2% bentonite
API Class A cement
12% bentonite
0.5% lignin retarder
API Class A cement
1% low fluid loss

Latex cement Shear bond

Compressive strength
Hydraulic bond
Resin cement Sandstone
Limestone
10 100 1,000 10,000 100,000
Shear bond, compressive, and hydraulic bond strength (psi)

Fig. 6-28. Bonding properties of cement to formation with cement not squeezed and walls not cleaned
(from Evans and Carter, 1961). Reproduced courtesy of the American Petroleum Institute.

6-4.3 Properties of clays ing proportions. Clays are all essentially hydrous magne-
Seventy percent of footage drilled while constructing sium or aluminum silicates. Although a clay may consist
a wellbore is through shale, a type of rock that contains of a single clay mineral, there are usually several mixed
clay minerals. Clay is a rock term and, like most rocks, with other minerals such as feldspar, quartz, carbonates,
it is composed of a number of different minerals in vary- and micas.

224 Well Cementing

 Clay minerals belong to the group of silicate minerals
known as phyllosilicates. In phyllosilicates, the SiO4
tetrahedra are arranged in sheets such that the Si/O
ratio is 2:5. This structural unit is sometimes called the
“siloxane sheet” (Figure 6-29).

H20 H20

= Oxygen
= Hydroxyl
= Silicon
= Aluminum
Fig. 6-30. The structure of montmorillonite.

ally large cell dimensions; consequently, the minerals of

this group show unrivalled capacity for swelling when
Fig. 6-29. Si2O5 sheet structure. Large spheres represent oxygen, wetted. One of the most common smectite minerals is
and small spheres represent silicon.
montmorillonite (bentonite contains mainly montmoril-
lonite) (Chapter 3), used as an extender and/or thick-
ener in drilling fluids and cement slurries.
The siloxane sheet is not electrically neutral and The negatively charged clay surfaces do not have an
cannot form a stable structure alone. To balance the equal affinity for all cations but exhibit selectivity depend-
charges, the siloxane sheets combine with sheets of ing on such factors as ionic charge, ionic size, and state of
hydroxyl ions (OH–) coordinated by magnesium ions hydration. The order of increased affinity for the alkali
(forming brucite) or aluminum ions (forming gibbsite). metal ions is Li+ < Na+ < K+ < Rb+ < Cs+. For the alka-
The resulting structures are very open, allowing other line earth ions, the order is Mg2+ < Ca2+ < Sr2+ < Ba2+. In
species (e.g., iron, potassium, and sodium) to substitute general, cations that are least preferred by the surface
for magnesium or aluminum. The properties of clays (such as Na) are associated with limited particle aggre-
depend on how the various sheets are arranged and the gation and extensive swelling, while strongly bound ions
nature of the ionic substitutions. (such as K+) promote extensive aggregation and limited
Owing to their open structure, clays are sensitive to swelling. Organic anions and cations can also influence
contact with foreign fluids containing electrolytes. To stability.
maximize stability from both a structural and electrical The adsorption of water by clays can occur either by
point of view, a clay will exchange anions and cations hydration of the crystal surfaces or hydration of the
with the electrolyte (Swartzen-Allen and Matijevic, exchange cations. Some values for the cations found in
1974). The clay is then said to have a certain cation natural montmorillonites are given in Table 6-13.
exchange capacity, usually quoted as the number of Sedimentary rocks such as shales are formed when
chemical equivalents per unit weight of dry clay. ocean floor sediments are buried and subjected to over-
Among the most sensitive clays are those of the smec- burden pressures and moderately high temperatures
tite group. A structural diagram is shown in Fig. 6-30. (Pettijohn, 1975). The composition of shales is varied,
Molecular water containing free cations is sandwiched because these deepwater sediments are formed by abra-
between two Si2O5–Al2(OH)6–SiO5 layers. The essen- sion processes (mainly silt), weathering (residual clays),
tially uncharged layers may expand in this way to unusu-

Chapter 6 Cement-Formation Interactions 225

Table 6-13. Characteristic Sizes of Cations Commonly separate the drilling fluid and the cement slurry and to
Found on Montmorillonite Surfaces† make the formation surface receptive to bonding with
Ion Dehydrated Charge Hydrated cement. Nevertheless, it would be prudent to assume
Ion Density Ion that the shale in the near-wellbore region is more per-
Diameter (charge/Å2 ) Diameter meable than the far-field shale.
(Å) (Å) Shale has the capacity to limit the movement of
Sodium (Na+) 1.90 0.088 11.2 charged species through its pores but not stop their
movement completely (Schlemmer et al., 2003). Silicate
Potassium (K+) 2.66 0.045 7.6 muds and OBMs can enhance the ability of shale to
Magnesium (Mg2+) 1.30 0.376 21.6 exclude ions, forming osmotic membranes on the shale
surface or close to the wellbore wall (Mody et al. 2002;
Calcium (Ca2+) 1.90 0.176 19.0 Bol et al., 1994). By setting the water activity in the
† Mitchell (1993). Reprinted with permission of John Wiley & Sons, Inc. drilling fluid lower than that of the shale, silicate or oil-
based drilling fluids can dewater shale. Consequently,
the near-wellbore region is less porous and permeable
and chemical and biochemical additions. Shales can than the far-field shale. For silicates, this effect is
contain as much as 40 wt% of clays and are sometimes enhanced by the deposition of silicate precipitates. A sil-
classified as shales or mudstones depending on the ori- icate-base fluid reduces the permeability of the shale to
entation of the clay minerals. Shales are highly oriented, ions and water (Bailey et al., 1998; Van Oort et al., 1996).
whereas mudrocks are randomly oriented. As burial The silicate fluid works by converting sodium silicate to
depth increases, the abundance of less-reactive clays calcium silicate, or by precipitation of polymeric sili-
(illite and kaolinite) increases and that of the montmo- cates as the pH drops below 10 at the shale surface.
rillonite clays such as smectite decreases. A typical shale In the wellbore, shales are often in contact with
does not exist, although shales are sometimes classified drilling fluids for several days before cementing. Ladva
from soft to brittle as the smectite content decreases et al. (2004) pretreated Oxford shale cores with various
(Weaver, 1989). types of drilling fluids. It can be seen in Fig. 6-31 that the
shale cores swell in the partially hydrolyzed polyacry-
lamide (PHPA) and a low inhibitive glycol fluid (poly-
6-4.4 Interactions between drilling ethylene glycol or KCl) and shrink in the silicate-based
fluids and shales and OBM fluids. This behavior is connected to the water
During the well-construction process, the rock forma- activities of the muds, which are greater than 0.95 for
tions are exposed to drilling fluids. The drilling fluids the PHPA and glycol mud but 0.83 for the silicate and
can interact with the surfaces of clay-bearing formations OBM systems.
and alter their water content and permeability. The authors found no detailed studies in the litera-
Erosion of the wellbore wall during the drilling ture that discuss the quality of primary cementing jobs
process is prevented by careful drilling fluid selection. after using specific drilling fluids. However, some papers
Drilling fluids fall into two main categories—water-base do allude to increased success of primary cementing jobs
and oil-base. The water-base systems can contain poly- after using silicate drilling fluids (Cardno et al., 1997).
mers such as acrylamides and glycols or silicates. In The advantage of silicate mud over polymer mud for pri-
some cases the degree of clay swelling can be reduced by mary cementing was described in a study of the Lennox
the addition of salts. These water-base fluids work by field, Liverpool Bay. The silicate system appeared to
adsorbing organic compounds onto the clay surfaces or improve hole gauge sufficiently to allow good primary
by the formation of osmotic membranes. cementing on the 95⁄8-in. section without the problem of
Sherwood and Bailey (1994) investigated the depth of gas migration. However, no detailed scientific study was
invasion of water from a viscosified KCl solution performed to confirm the incorporation of silicate from
(≈0.5 M) into predrained Oxford shale cores. The ingress the drilling fluid into the cement matrix. A patent by
was of the order of 10 mm after a 48-hr circulation Mondshine (1977) proposes pretreating the wellbore
period, resulting in a layer of swollen shale on the sur- with a silicate fluid. However, there is no conclusive evi-
face of the laboratory-scale wellbore. At present, it is not dence of improved bonding. There is also some evidence
known if such a layer would remain on the wellbore wall provided by Dillenbeck and Nelson (1992) that cement
during the cementing stage. During a primary cementing slurries containing potassium salts or amines are useful
operation, a spacer or wash fluid is frequently used to in front of shale sections.

226 Well Cementing

 0.03

0.02
OBM
Polyethylene glycol/KCl
Axial PHPA
swelling 0.01 Silicate
strain,
dimensionless

0.01
0 20 40 60 80
Time (hr)

Figure 6-31. The free swelling of Oxford shale cores in various drilling fluids.

6-4.5 Interactions between cement 2

slurries and shales 1.9
1.8 Valve open
Filtration of cement filtrate into shale sections can lead
1.7
to further disturbance of the shale. The extent of well-
bore-wall damage is largely controlled by the following 1.6
Pore Valve closed
factors: pressure 1.5
(MPa) 1.4
■ initial hydrostatic pressure relative to the shale pore
1.3
pressure
1.2
■ initial water activity in the cement slurry relative to
1.1
that in the shale pore. 1
Once the cement has set and the internal pore pres- 0 0.2 0.4 0.6 0.8 1 1.2 1.4
sure has dropped, a positive suction pressure is possible, Normalized time†
drawing water from the swollen shale back into the †t
max = 7.85 hr
cement.
As discussed in Chapter 2, Portland cement hydration Fig. 6-32. Pore pressure profile for a cement slurry with (valve
is an exothermic process. When the temperature-profile open) and without (valve closed) access to an external fluid reser-
voir (Appleby and Wilson, 1996). Reprinted with permission from
maxima are reached and the cement slurry begins to set, Elsevier.
the internal pore pressure begins to drop. Along with the
pore-pressure reduction, internal shrinkage occurs.
Such shrinkage can cause problems such as wellbore
Cement hydration is a dissolution-precipitation
microannuli and annular fluid migration (Chapter 9). It
process. Ions are available in solution before they are
is important to note at this stage that the extent of the
incorporated into the hydration products. Table 6-14
pore-pressure drop depends on the external supply of
shows the solution concentration of certain ionic species
water to the cement slurry. Appleby and Wilson (1996)
after 20 min of hydration (Michaux et al., 1989a and
examined the pore-pressure decrease with and without
1989b).
access to excess fluid, as shown in Fig. 6-32, in which
valve open and valve closed represent the presence and
absence of excess water. This ability of cement to dewa-
ter adjacent surfaces without altering its external
dimensions could lead to debonding at a shale or filter-
cake surface.

Chapter 6 Cement-Formation Interactions 227

Table 6-14. The Ionic Strength in Solution After 20 min 0.03
Catoosa
of Cement Hydration at 77°F [25°C] Oxford
Species Concentration (mmol/L)
Ca2+ 21.1 0.02

Na+ 117 Axial

swelling
K+ 422 strain
0.01
SO42– 201

OH– 168

Cl– 6 0
0 2 4 6 8 10
pH 13.10 Time (hr)

Fig. 6-33. Swelling of shale cores in synthetic cement filtrate.

A simple approximation, such as that shown in

Eq. 6-20, can be used to estimate the water activity, aw, Acrylic plastic spacer
in a cement slurry containing single (u) and complex PTFE cylinder Cement Acrylic plastic lid
(m) species.

ln aw = −0.018 K os (∑ j
1
u+ ∑ 1 m
j
) (6-20)
44.0 mm

Assuming an osmotic coefficient (Kos) of 1, the calcu-

lated solution water activity is 0.98 after 20 min of hydra-
tion and 0.96 after 3 hr of hydration at 25°C. These Shale 25.4 mm
approximate calculations agree with those carried out by
48.0 mm
van Breugel (1992). In fact, van Breugel suggests that the
water activity in the interfacial transition zone (100 μm) Fig. 6-34. A schematic drawing of the shear-bond (push-out) testing
between an aggregate and cement remains high for at cell.
least 10 hr after mixing. This high water activity, coupled
with hydrostatic pressure, will increase the ingress of
fluid into shale sections during well construction The acrylic plastic lid and base were sealed to the
(Heathman et al., 2003). polytetrafluoroethylene (PTFE) cell walls using silicone
The potential for shales to swell when exposed to a rubber on the outer surface.
cement slurry was studied by Ladva et al. (2004). After the curing period, the base, lid, and acrylic plas-
Unconfined linear swelling tests were performed with tic spacer were removed. The cell walls and cement ring
two shales: Oxford shale (17 wt% smectite) and Catoosa were supported while a brass rod (0.8 in. [20 mm] in
shale (6 wt% smectite). As discussed earlier, smectite diameter) was used to load the shale core. The maxi-
readily swells in the presence of water. After exposure to mum force required to push out the shale at a rate of
a synthetic cement filtrate with the composition shown 0.5 mm/min was measured. The shear-bond strength was
in Table 6-14, the Oxford shale swelled more than the calculated by dividing the maximum force by the area of
Catoosa shale (Fig. 6-33). the shale/cement interface.
Ladva et al. (2004) also performed shear-bond or push- A typical shear test profile, presented in Fig. 6-35,
out tests between a neat 15.8 lbm/gal Class G cement and shows the maximum load that the interface could
the Oxford and Catoosa shales. A cell was designed to withstand before rupturing. After 8 days curing at
measure the shear-bond strength between the shale core 185°F [85°C], the shear-bond strength was 79.8 lbf/in2
or filtercake and set cement (Fig. 6-34). A freshly cored [0.55 MPa]. During the shear-bond test, the untreated
shale plug with a diameter of 1 in. [25.4 mm] and a nom- Oxford shale core was pushed out of the cement ring, but
inal length of 0.8 in. [20 mm] was placed centrally on the a film of shale remained on the inside of the cement.
acrylic plastic base. An acrylic plastic spacer was then This suggests that the shear-bond strength of the
placed on top of the shale core, and the cement slurry shale/cement interface was greater than the shear
was poured around the shale to the top of the spacer. strength of the shale itself.

228 Well Cementing

 800
poverburden Cement pressure
p
Bond strength
Gas outlet
600

Interfacial
deformation Rubber
Force 400
(N) Rock
pcon Cement

200 Bottom

0
0 5 10 15 20
Position (mm) Gas inlet Gas inlet

Figure 6-35. The push-out profile of Oxford shale from cement

(cured at 85°C for 8 days). Fig. 6-36. Schematic diagram of the small-wellbore simulator.

Table 6-15. Shear-bond Strength Values† the drilling-fluid exposure, cement can be set under an
Cure Time (days) Pressure (MPa) applied pressure that is independent of the overburden
pressure on the rock.
1 0.48
The experimental sequence consisted of the following
4 0.98‡ steps.
■ An Oxford shale core is drained at 1,450 psi [10 MPa]
5 0.79‡
for 120 hr and moved to the wellbore simulator.
† Catoosa shale cured for different times at 20°C.
‡ Failure of cement ring ■ An homogeneous stress of 10 MPa is applied with mud
flow at a temperature of 68°F [20°C] for 48 hr.
■ The sample is depressurized, the mud is removed, a

The adhesion of Catoosa shale or Oxford shale to cement slurry is added, and the core is repressurized
cement cured for 24 hr at 20°C was not measured because at 10 MPa and 167°F [75°C] for 120 hr.
the sample fell apart at the interface on handling. ■ At the end of this period, the core is depressurized
over a period of 1 hr.
6-4.6 Bonding between cements and shales ■ The core is removed from the cell, allowing one to

treated with various drilling fluids observe the cement/shale interface and perform mea-
surements.
In another set of experiments, Ladva et al. (2004) used a
small wellbore simulator (SWBS) that allows rocks to be In the SWBS, Oxford clay was treated with various
exposed to flowing drilling fluid under pressure. The sim- drilling fluids for 48 hr. The drilling fluids were typical
ulator is depicted in Fig. 6-36. The rock sample has a silicate muds and low inhibitive glycol muds. After set-
diameter of 6 in. [150 mm] and a length of 8 in. [200 mm]. ting cement against the pretreated shales, postmortem
Along the axis of the rock is a predrilled central wellbore analyses were performed. The analyses showed that the
that is 1 in. [25.4 mm] in diameter. The rock sample is water content of the shale at the cement interface was
contained within a steel pressure vessel that operates at greater when the sample was pretreated with low inhib-
a pressure of 4,570 psi [31.5 MPa]. The outer cylindrical itive glycol mud than with a silicate mud. Thus, the
surface of the rock is surrounded by a rubber sleeve effect of drilling fluids on the water content of the near-
through which a radial confining stress may be applied. wellbore region persists after cementing. Also, when the
The top platen supplies the overburden pressure. shale was pretreated with silicate mud, the failure
Temperatures up to 176°F [80°C] can be maintained occurred at the cement-shale interface. Failure
while operating under pressure. Drilling fluid is circu- occurred in the shale itself when the glycol mud was
lated down the central wellbore under pressure. After used and when there was no pretreatment.

Chapter 6 Cement-Formation Interactions 229

 In small-scale push-out tests, the position of the fail- between 149°F and 185°F [65°C and 85°C]. The samples
ure plane also varied with the drilling fluid pretreatment were cored out of the Boom clay after test periods of 12
(72 hr). However, pretreatment with drilling fluid did or 18 months. Sections of the interface were analyzed by
not substantially increase the shear-bond strength. electron microprobe and X-ray diffraction. Observations
Results from such tests are presented in Fig. 6-37. from the HADES study using ordinary Portland cement
(OPC) plugs are summarized below (Read et al., 2001).
In the case of OPC samples exposed for only 12
months, different zones were apparent with mineralogi-
Glycol cal and textural changes reflecting transport of calcium,
magnesium, aluminum, iron, silicon, and sulfur. These
PHPA changes occurred across a region extending from 100 to
250 μm either side of the interface.
Silicate The clay was depleted in aluminum, silicon, and mag-
nesium but enriched in calcium relative to the
OBM
untreated clay. The outward diffusion of calcium was
evident (~200 μm) from the precipitation of calcite
No mud
in the clay. Also, the mineral hydrotalcite,
0 0.1 0.2 0.3 0.4 Mg6Al2(CO3)(OH)16 • 4H2O, formed in the initial 10- to
Pressure (MPa) 20-μm zone. In samples stored at 185°F [85°C], a frac-
ture was observed at the interface of the altered and
Fig. 6-37. Shear-bond measurements of interface between cement unaltered clay zones.
and Oxford shale treated with various drilling fluids. The cement layer was leached in appearance and had
reduced calcium but increased silicon, sulfur, aluminum,
and magnesium concentrations relative to the unaltered
6-4.7 Cement-to-formation bonding studies cement. After 1 year, increased porosity observed at the
performed outside of the petroleum industry interface was attributed to portlandite dissolution.
Ideally, samples of set cement and the adjacent forma- Unfortunately, no details of the mechanical strength
tion should be recovered from oil wells. Recovery would of the interface were given. However, the existence of an
give us a better understanding of the chemical nature of interfacial transition zone in construction concrete is
the interface under downhole conditions during the life- well documented.
time of the well. In view of the lack of such samples, it is Concrete can contain OPC and granular materials
interesting to examine studies from the nuclear and con- such as calcined shale, limestone, or quartz (Mehta and
struction industries. Monteiro, 1986; 1987). In one study, Portland cement
There are other situations in which cement or alka- was set against quartz. After 28 days (at an unknown
line fluids are in contact with clays, namely, in construc- temperature), a continuous thin film of calcium hydrox-
tion cements, nuclear waste sites, and a few naturally ide formed at the surface, and the cement was very
occurring repositories. At present, two lines of research porous. Below this film, other hydration products such
exist. In one, preset plugs of cement are inserted into as calcium silicate hydrate and ettringite appeared.
boreholes drilled in clay formations. In the other, clay Compared to the bulk concrete, the large crystal size in
crystals are treated with alkaline fluids at elevated tem- the transition zone (30 μm) was caused by the high
perature. These studies are an important source of infor- water/cement ratio, which also accounts for the high
mation when trying to predict and modify the condition porosity. Another source of weakness in this particular
of the cement-rock interface in a wellbore over time. sample was the presence of Hadley grains (Barnes,
For disposal, low-level radioactive waste is sometimes 1975), which are hollow cement particles from which
encased in cement blocks and lowered into clay beds. the solids have been dissolved. These form because of
There is a concern that over time alkaline fluids near the high local water-to-cement ratios.
clay surface will alter its effectiveness as a barrier to the The cement/limestone interface was found to be the
encased nuclear waste. Studies by the nuclear industry same as that of the quartz experiment after a 28-day
to increase their understanding of the cement-to-forma- curing period (Mehta and Monteiro, 1988). At later stages
tion interface are ongoing. At the High Activity Disposal (56 days or more), evidence of strong chemical bonding
Experimental Site (HADES), an underground laboratory between the aggregate and the bulk cement paste was
at Mol, Belgium (Glasser, 2001; Read et al., 2001), cement inferred from the fact that some of the cement paste
plugs were precured and then placed in cored-out holes adhered to the limestone surface. Microscopic evidence
in Boom clay. Samples were maintained at temperatures suggested the carbonate surface is etched by the cement

230 Well Cementing

and extra precipitated products are present in addition to bility and plasticity of such zones increase the difficulty
the calcium hydroxide crystals on the cement side. of obtaining a successful primary cementation. The
In the construction industry, concrete can contain cement slurry can dissolve large quantities of formation
expanded shale (Yun and Qingli, 1992; Schneider and material, resulting in a modification of performance
Chen, 1992). This shale bears little relation to the (Ludwig, 1951). Plastic salt zones can encroach upon
hydrated shale found in wellbores. Expanded shale is the casing before the cement sets. Nonuniform forma-
prepared by grinding clay, mixing it with additives and tion movement exerts point loading on the casing string,
heat-treating it at 2,190°F [1,200°C]. The resulting prod- sometimes resulting in casing failure and collapse
uct contains elliptical particles that adsorb water and (Cheatham and McEver, 1964). Salt cements are used
can form chemical bonds with the hydrated lime in the routinely to reduce the severity of these problems; how-
concrete. The pore throats in the expanded shale ever, some controversy exists regarding their efficacy
become filled with hydrated cement, leading to (Chapter 7).
increased strength. The first recorded use of salt in well cements was
One approach to increasing the interfacial bond during the 1940s, when wells were completed across salt
strength is to reduce the porosity of the interfacial tran- domes in the U.S. Gulf Coast. Later, this became stan-
sition zone. Alteration of the transition zone by adding dard practice in the Williston basin (North Dakota and
silica or organics to the cement has been investigated Montana in the United States) and certain areas in the
(Caliskan, 2003; Kim and Robertson, 1998). Caliskan North Sea. The concentration of NaCl usually varied
studied silica addition and reported a correlation from 18% to 37% by weight of water (BWOW). While such
between the reduction in the thickness of the interfacial practices prevented the dissolution of the formation, the
transition zone and an increase in SiO2 content and high salt concentrations impeded the performance of
bond strength. Kim et al. (2000) added polyvinyl alcohol other cement additives, especially dispersants and fluid-
to the cement formulation placed at the limestone inter- loss additives, which were originally developed for fresh-
face. They attributed the bond-strength increase to the water systems. In addition, the high salt concentrations
porosity reduction in the interfacial transition zone and tended to overretard the cement system; thus, formation
the inhibition of calcium hydroxide production. encroachment and casing damage could occur before
Reactions between minerals and alkaline fluids will the cement set. Two approaches have been followed to
affect the barrier properties (including permeability) solve these difficulties: eliminating salt from the cement
and the chemical properties (e.g., sorption potential system and developing additives that are compatible
and cation exchange capacity) of the host rock with salt cements.
(Hodgkinson and Hughes, 1999). These rocks are likely Salt-free cement (Goodwin and Phipps, 1982), or
to be altered by the dissolution and/or precipitation cements containing very low salt concentrations (3%
reactions. Ion exchange may occur in clays and may play BWOW) have been successfully applied in the Williston
a significant part in the alteration of smectite materials basin (Bryant, 1984) and the Middle East (Ismail and
(Jefferies et al., 1988; Johnston and Miller, 1984; Khalaf, 1993). No casing collapse was reported with such
Koomareni and White, 1981). systems, compared to a 20% failure rate with salt-satu-
rated cements. To prevent excessive dissolution of the
formation, the cement-slurry displacement rates were
6-4.8 Cement-to-formation bonding conclusion low.
This section has summarized what is known in the petro- Ford et al. (1982) proposed an intermediate approach.
leum and civil-engineering industries about the Cement systems containing 18% NaCl BWOW, in combi-
cement/formation interface. The quality of the bond at nation with holding the casing in tension during the
this interface is affected by many variables, including cement job, improved the success rate of primary cement
the formation composition, drilling-fluid composition, jobs in the Williston basin.
cement-slurry composition, temperature, and pressure. The above approaches may improve initial results;
Clearly, more research will be required to achieve a sat- however, considering the previously discussed long-term
isfactory understanding that will lead to more reliable effects of ionic disequilibrium, cement failure may ulti-
zonal isolation. mately occur. The rate of ionic diffusion is determined by
the difference in salt concentration between the cement
and formation as well as the permeability of the cement.
6-5 Cementing across evaporite zones In addition, Rae and Brown (1988) revealed that con-
The presence of salt domes and massive evaporite tamination of a freshwater cement system by as little as
sequences has long been problematic for drilling, com- 10% salt could reduce the thickening time by 30%,
pletion, and long-term production. The high water solu- increase the slurry viscosity by 100%, and increase the

Chapter 6 Cement-Formation Interactions 231

fluid-loss rate by nearly 500%. Yearwood et al. (1988) supporting the well. Conditions vary by location, but typ-
confirmed these findings. ically water depths greater than 1,500 ft are conducive to
Today, it is possible to design cement systems with hydrate formation, and the hydrate stability zone can
appropriate performance throughout the salt-concentra- extend as deep as 2,000 ft below the mudline. However,
tion range. Cement systems containing from 5 to 37% in most deepwater areas, including the Gulf of Mexico,
NaCl (BWOW), with excellent placement characteris- most of the sediments encountered contain less than
tics, appropriate thickening times, and acceptable com- 10% of pore volume hydrates and historically do not rep-
pressive-strength development are now commonplace. resent a serious drilling hazard. Before drilling, a geo-
Such systems are described in Chapter 7. hazard survey is normally performed to locate hydrate-
Selecting the appropriate salt concentration in a rich areas, such as seeps and large active faults. This
cement slurry is largely a function of the local formation allows the operator to avoid hazardous areas (personal
characteristics, operational limitations, and success communication, J.B. Bloys, 2004).
ratios. There is no industry consensus regarding slurry Very little has been published about the interactions
design for salt-zone cementing. At present, anecdotal between cement systems and gas hydrates. Special
evidence shows that the majority of salt zones are cement systems with low heats of hydration, described
cemented with systems containing between 8 to 18% in Chapter 7, have been recommended to help prevent
NaCl (BWOW). In addition, scrupulous attention is given hydrate dissociation in arctic areas where hydrate con-
to proper hole conditioning and good cementing prac- centrations near the base of permafrost can be as high
tices. as 80% (Collett et al., 2000).
Blast-furnace slag, combined with Portland cement
or drilling fluids, has also been used recently to cement
salt zones. This application is also discussed in 6-7 Acronym list
Chapter 7.
API American Petroleum Institute
BWOW By weight of water
ECP External casing packer
6-6 Gas hydrates
EPS Engineered particle-size
Gas hydrates are a special combination of two common
substances, water and natural gas. If these meet under HADES High Activity Disposal Experimental Site
conditions in which the pressure is high and the tem- IPT Ideal packing theory
perature low, they join to form a solid, ice-like substance. ISO International Organization for
Vast volumes of shallow sediments below the deep ocean Standardization
mud line and in arctic regions are conducive to hydrate LCM Lost circulation material
formation. The basic hydrate unit is a hollow crystal of
water molecules with a single gas molecule trapped M-DOB Mud-diesel-oil-bentonite
inside. The crystals fit together in a tight latticework. At M-DOB2C Mud-diesel-oil-bentonite-cement
typical deepwater seabed temperatures around 39°F MMH Mixed metal hydroxides
[3.8°C] methane hydrates can form at any pressure over MMO Mixed metal oxide
about 550 psi [3.8 MPa].
The petroleum industry began to take an interest in MMOOH Mixed metal oxide and hydroxide
gas hydrates in the 1930s, when gas-hydrate formation OBM Oil-base mud
was found to be a cause of pipeline blockages. Since OPC Ordinary Portland cement
then, most industry efforts related to gas hydrates have PCP polymer-based crosslinked pills
been directed toward preventing them or hindering
their accumulation into blocking masses. However, in PHPA Partially hydrolyzed polyacrylamide
recent years, interest in exploiting gas hydrates for their PSD Particle-size distribution
hydrocarbons is becoming more intense. PTFE Polytetrafluoroethylene
For operators drilling in deep water, encountering PV Plastic viscosity
naturally occurring gas hydrates during drilling can pose
a well-control problem if large amounts enter the bore- SWBS Small wellbore simulator
hole and dissociate. Furthermore, circulation of warm WBM Water-base mud
fluid within the wellbore can increase the temperature WOC Waiting on cement
in the surrounding hydrate-rich sediments, leading to YP Yield point
hydrate melting and destabilization of the sediments

232 Well Cementing

As well cementing technology has advanced, special
cement systems have been developed to address specific
Special Cement Systems
Erik B. Nelson, Bruno Drochon, and Michel Michaux—Schlumberger,

7-1 Introduction
Tom J. Griffin—Griffin Cement Consulting LLC
7
during the static period, and the pressure required to
move the fluid does not change with time. In the case of
a thixotropic fluid, the yield point is exhibited only upon
problems. This chapter presents cement technologies the withdrawal of shear stress. If there is a lapse of time
specific to problems such as slurry fallback, lost circu- after the withdrawal of shear, a force greater than
lation, microannuli, cementing across salt formations, the yield point will be required to move the fluid, as
and corrosive well environments. Some of these solu- indicated in Figs. 7-1, 7-2, and 7-3. The difference
tions were introduced more than 50 years ago, yet between the “gel strength” and the yield point gives
they are still commonly used today. a measure of the “degree of thixotropy” of the fluid.
Other systems have been developed with special per-
formance characteristics such as flexible cements and
drilling fluids that convert to a hardened cementitious
product. This chapter also contains a thorough discussion
of foamed cements and cement systems with multimodal
particle-size distributions. Increasing shear rate
Special technologies also exist for problems such as Gel strength
annular fluid migration and high well temperatures. Shear stress
These are presented in Chapters 9 and 10.
Decreasing shear rate

7-2 Thixotropic cements

A thixotropic system is fluid under shear, but develops Yield point
a gel structure and becomes self-supporting when at
Shear rate
rest (Shaw, 1970). In practical terms, thixotropic
cement slurries are thin and fluid during mixing and Fig. 7-1. Generalized rheological behavior of thixotropic fluids.
displacement but rapidly form a rigid, self-supporting
gel structure when pumping ceases. Upon reagitation, the
gel structure breaks and the slurry regains fluidity. Then,
Thixotropic cement systems have several important
upon cessation of shear, the gel structure reappears and
applications. They can be used in wells in which exces-
the slurry returns to a self-supporting state. This type of
sive fallback of the cement column is a common occur-
rheological behavior is continuously reversible with truly
rence (Wieland et al., 1969). Such wells have weak zones
thixotropic cements.
that fracture under low hydrostatic pressure, allowing
As a rule, thixotropic slurries behave as Bingham
the cement slurry to invade the formation. Self-support-
plastic fluids under stress (Chapter 4); consequently,
ing cements reduce the hydrostatic pressure on the
their behavior is defined by a yield value (τy) and a
formation as gel strength increases and prevent fallback.
plastic viscosity (μp) (Clement, 1979). τy is a theoretical
Another important application is the prevention of
value concerning the behavior of a fluid under condi-
lost circulation during placement (Chapter 6). When a
tions of shear. With thixotropic slurries, τy would be
thixotropic slurry enters the thief zone, the velocity of
the shear stress necessary to initiate movement, i.e.,
the leading edge decreases and a gel structure begins to
measured at zero shear rate.
develop. Eventually, the zone becomes plugged because
For a nonthixotropic fluid, the yield value remains the
of the increased flow resistance. Once the cement sets,
same whether the shear rate is increasing or decreasing.
the zone is effectively consolidated.
There is no change in the physical structure of the fluid

Well Cementing ■ Chapter 7 Special Cement Systems 233

 because, after repeated stops, excessive pump pressure
may be required to resume movement. To minimize the
risk of job failure, most operators prefer to avoid pump
shutdowns when placing thixotropic cement slurries.
The use of thixotropic cements has decreased in recent
years owing to the development of low-density cement
systems that provide excellent zonal isolation (Sections
7-9 and 7-10).
Several thixotropic cement systems currently exist.
Thin when mixed Rigid when pumping stops The chemistry and special operational considerations of
each are described below.

7-2.1 Clay-base systems

Portland cement systems containing water-swellable
clays such as bentonite develop high gel strength
and exhibit some degree of thixotropic behavior
(Messenger, 1980). Such systems have also been shown
to control gas migration in certain circumstances
(Chapter 9). The bentonite concentration and the slurry
Fluid again when force Thin when pumping is
is applied resumed density can be varied from 0.05% to 2.0% by weight of
cement (BWOC) and 11.5 and 21.0 lbm/gal [1,381 to
Fig. 7-2. Thixotropic behavior. 2,521 kg/m3], respectively.

7-2.2 Calcium sulfate-based systems

Pressure required The most widely used material to prepare thixotropic
to break circulation cement slurries is calcium sulfate hemihydrate
_
Flow rate (CaSO4 • 1⁄2H2O or, in cement notation, CSH1⁄ 2) (also
Pump called
_ plaster of paris). When added to Portland cement,
pressure CS_H1⁄ 2 first hydrates to form gypsum (CaSO4 • 2H2O or
CSH2 ) and then reacts with tricalcium aluminate (C3 A)
to form the calcium sulfoaluminate hydrate mineral
No Pump pressure ettringite (Chapter 2). The chemical equation for the
flow
reaction is shown below (Kalousek, 1973).
Time
HO
Fig. 7-3. Pump pressure and flow rate for a thixotropic fluid. 3CaSO 4 i 2H2 O + 3CaO i Al 2 O3 ⎯⎯
2
→
3CaO i Al 2 O3 i 3CaSO 4 i 32H2 O
(ettringite) (7-1)
Other uses for thixotropic cement systems include
the following: Ettringite occurs as needle-shaped crystals deposited
■ to repair split or corroded casing on the cement-grain surfaces. The ettringite crystals
■ as lead slurries for remedial cementing when it is promote greater physical association between the
difficult to obtain a squeeze pressure (Spangle and cement particles, resulting in the formation of a loose
Calvert, 1972; Welch et al., 1990) (Chapter 14) network or gel. Upon agitation, the network is easily dis-
■ as a grout, to quickly immobilize the slurry rupted, and the slurry returns to a fluid state.
Calcium sulfate hemihydrate can be used with most
■ to prevent gas migration in certain situations
Portland cements to prepare thixotropic _ cements.
(Chapter 9). Depending upon the cement, the optimal CSH1⁄2 concen-
Thixotropic cement slurries have another notable tration varies between 8% and 12% BWOC. Cements
characteristic. After each static-dynamic cycle, the with a C3 A content of less than 5% should be used with
gel strength and yield point tend to increase. During caution, because insufficient ettringite may crystal-
cementing operations this could pose a problem lize to impart thixotropy. The water requirement for

234 Well Cementing

slurries containing calcium sulfate hemihydrate is with Portland cements that contain less than 5% C3 A. It
higher than that for conventional systems; consequently, can be supplied in liquid form, which is convenient for
the slurry densities are lower. Representative data for offshore operations.
such systems are presented in Table 7-1. These soluble sulfates react with calcium hydroxide
in the cement slurry to form gypsum in situ, and ettrin-
gite formation then proceeds. The global reaction is
Table 7-1. Slurry Properties and Performance of shown below.
Thixotropic Slurries Containing Calcium Sulfate HO
Hemihydrate Al 2 (SO 4 )3 + 6Ca(OH)2 ⎯⎯
2
→
System 1 2 3 4 5 3CaO i Al 2 O3 i 3CaSO 4 i 32H2 O (7-2)
Density (lbm/gal) 15.6 14.9 14.6 14.6 14.2
This reaction proceeds much faster than that involving
Water (%BWOC) 46 60 64 64 70 calcium sulfate hemihydrate. Aluminum sulfate is a pow-
erful cement accelerator, and a strong, irreversible gel
Water (gal/sk) 5.20 6.78 7.20 7.20 7.90
structure would develop if it were added alone. Iron (II)
Calcium sulfate 0 12 10 12 10 sulfate, a weak cement retarder, is included in the system
hemihydrate (%BWOC) to inhibit the aluminum sulfate and preserve thixotropy
throughout the pumping time. Because of the fast kinet-
CaCl2 (%BWOC) 0 3 2 3 0
ics of this system, very little ettringite is formed after the
Yield (ft3/sk) 1.18 1.48 1.50 1.54 1.60 cement sets. Thus, significant cement expansion does
not occur at curing temperatures above 100°F [38°C].
Well conditions (°F)
BHCT 113 60 80 80 125
BHST 170 70 95 95 200 7-2.4 Crosslinked cellulose polymer systems
Thickening time at BHCT 4:00+ 3:10 2:08 1:50 3:15
Thixotropic cements can prepared by the addition of
(hr:min) water-soluble crosslinkable polymers and a crosslinking
agent (Childs et al., 1985). Hydroxyethylcellulose,
Compressive strength (psi) at BHST carboxymethylhydroxyethyl cellulose, polyvinyl alcohol,
7 hr 350 100 485 500 950 and various sulfonate polymers can be crosslinked with
18 hr 1,750 –† 1,165 1,250 1,750 certain titanium or zirconium chelates. The optimal
24 hr 2,900 500 1,295 1,350 2,200 polymer-crosslinker combination, and the relative con-
96 hr – 1,150 – – – centrations of each, vary with the well temperature.
† not available

7-3 Expansive cement systems

Thixotropic cements containing calcium sulfate Good bonding between the set cement and casing and
hemihydrate are incompatible with many fluid-loss between the set cement and the formation is essential for
additives. Jones and Carpenter (1991) overcame this effective zonal isolation. Poor bonding limits production
problem by adding 1.5% BWOC polyvinyl alcohol to the and reduces the effectiveness of stimulation treatments
cement slurry. (Chapter 1). Communication between zones can be caused
Calcium sulfate hemihydrate systems have additional by inadequate mud removal, poor cement/formation
attributes besides thixotropy. Such systems are highly bonding, expansion and contraction of the casing resulting
sulfate resistant, because the C3 A is effectively neutralized from internal pressure or thermal stress, and cement
(Chapter 2). Also, after setting, ettringite continues contamination by drilling or formation fluids. Under such
to form; as a result, a significant amount of bulk expan- circumstances, a small gap or “microannulus” is frequently
sion occurs within the cement matrix. This phenome- present at the cement/casing or the cement/formation
non, and the benefits derived from it, are addressed in interface (Moran et al., 1991; Baumgarte et al., 1999).
detail later in this chapter. Cement systems that expand slightly after setting are
a proven means of sealing microannuli and improving
primary cementing results. The improved bonding is
7-2.3 Aluminum sulfate/iron (II) sulfate system the result of mechanical resistance or tightening of
A blend of Al2 (SO4)3 and FeSO4 also relies upon the the cement against the pipe and formation. However,
formation of ettringite to impart thixotropy to cement expansive cement systems are not a substitute for the
slurries (Nelson, 1983). This blend was developed for use good cementing practices that lead to maximum mud

Chapter 7 Special Cement Systems 235

removal (Chapter 5). Expansive cements cannot cure 0.25
the problems associated with mud that is left on the
0.20
casing or formation surfaces.
From Chapter 2, the reader may recall that Portland 0.15
Linear
cement manufacturers limit the amount of certain expansion
Ettringite-based expansive
cement system (14.8 lbm/gal)
alkaline impurities to avoid expansion of the set cement, (%) 0.10
a condition called “unsoundness.” In an unrestrained
0.05
environment such as a road or building, expansion of Neat Portland
the set cement can result in cracking and failure. In a 0 cement (15.8 lbm/gal)
wellbore environment, however, the cement is restrained 0 5 10 15 20 25 30
by the casing and, when competent, the formation; Time (days)
consequently, once the cement has expanded to elim-
inate void spaces, further expansion reduces internal Fig. 7-4. Comparison of expansion between neat Portland cement
cement porosity. It is important to note that expanding and an ettringite-based expansive cement system.
cements must be more flexible than the formation; other-
wise, the cement will not expand toward the casing, and
a microannulus may form. Ettringite is not stable at higher temperatures and
converts to a more dense calcium sulfoaluminate
hydrate and gypsum according to the following chemical
7-3.1 Ettringite systems equation (Bensted, 2001).
Most expansive well cement systems rely upon the for-
mation of ettringite, discussed in the preceding section, 2Al(OH)4– + 3SO2– 2+ –
4 + 6Ca + 12OH ⎯→
⎯
after the cement has set. Ettringite crystals have a
3CaO i Al 2 O3 i CaSO 4 i 12H2 O +
greater bulk volume than the components from which
they form; consequently, expansion occurs because of 2CaSO 4 i 2H2 O + 16H2 O (7-3)
the internal pressure exerted upon crystallization.
Currently, there are four commercial expanding cement
systems that rely on ettringite formation. 7-3.2 Salt cements
Type K cement is a blend of Portland cement, calcium Cement slurries containing high concentrations of NaCl,
sulfate, lime, and anhydrous calcium sulfoaluminate Na2SO4, or both were among the earliest expansive well
(Klein and Troxell, 1958). This cement is composed cements (Carter et al., 1965). After setting, cement
of two separately burned clinkers that are ground expansion occurs because of internal pressure exerted
together. Type K cement systems typically expand by by the crystallization of the salts within pores and
0.05% to 0.20%. by chlorosilicate and chlorosulfoaluminate reactions
Type M cement is either a blend of Portland cement, (Smith, 1976). Typical expansion performance of such
refractory calcium aluminate cement (Chapter 10), and systems at ambient conditions is shown in Fig. 7-5. These
calcium sulfate, or an interground product made with systems are equally effective at temperatures up to
Portland cement clinker, calcium aluminate cement 400°F [204°C].
clinker, and calcium sulfate (Root and Calvert, 1971). Other applications of salt cements are presented
Type S cement is a commercially prepared blend of later in this chapter (Section 7-5).
high-C3 A Portland cement with 10.5% to 15% gypsum.
The fourth and most common method of preparing
an ettringite-based expansive cement is to add calcium 0.4
sulfate hemihydrate to a Portland cement containing 10% sodium sulfate + saturated salt water
at least 5% C3 A. This formulation is similar to Type S; 0.3
however, because the blend is prepared as needed
Linear Saturated salt water
before a cement job, shelf life is not a concern. As dis- expansion 0.2 18% salt water
cussed in the previous section, such systems are also (%)
thixotropic. If this is not desired, the thixotropy can be 0.1
defeated by adding a cement dispersant. The expan- 10% salt water Fresh water
sion performance of Portland cement–calcium sulfate 0
hemihydrate systems is illustrated in Fig. 7-4. 0 60 120 180 240
A major limitation of ettringite-based systems is their Time (days)
inability to provide significant expansion at curing tem-
peratures above about 170°F [76°C] (Bour et al., 1988). Fig. 7-5. Expansion of salt cement systems.

236 Well Cementing

7-3.3 Aluminum powder to be of practical benefit. The MgO concentration required
Zinc, magnesium, iron, and aluminum powders can be for adequate expansion varies between 0.25% and 1.00%
used to prepare expansive cements (Carter et al., 1965). BWOC, depending on temperature. Figure 7-6 shows the
Finely powdered aluminum reacts with the alkalis in the expansion performance of a Class G cement system con-
cement slurry to produce tiny bubbles of hydrogen gas. taining 1.0% MgO BWOC and illustrates that the amount
This system is effective in shallow well applications, of expansion increases with increasing temperature.
because the expansive pressure of the bubbles is greater
than the formation pressure. The performance of such
systems is illustrated in Table 7-2. 2

500°F [260°C]

Table 7-2. Effect of Pressure on Expansion of Cement

Systems Containing Powdered Aluminum†, ‡ Linear 300°F [149°C]
expansion 1
Aluminum (%) Volume Expansion (%) (at 80°F) (%)
Curing Pressure 200°F [93°C]
0 psi 3,000 psi
175°F [79°C]
0.05 11.84 0.712

0.10 17.90 0.917 7 14 21 28 2 3 4

(days) (months)
0.25 24.00 1.64
Curing time
0.50 56.51 2.64
Fig. 7-6. Expansion of cement containing 1% BWOC calcined MgO.
1.00 57.19 5.17
† After Carter et al. (1965). Reprinted with permission of SPE.
‡ At 80° F.

7-4 Freeze-protected cements

Permafrost zones in Alaska, northern Canada, and Siberia
The rate of hydrogen liberation is strongly affected by
present unique cementing difficulties. Permafrost is
the fineness and concentration of aluminum, tempera-
defined as any permanently frozen subsurface forma-
ture, and pressure. Thus, careful slurry design is necessary
tion. The depths of such formations vary from a few
to obtain optimal results. More recently, the pressurization
meters to 2,000 ft [600 m]. Below the permafrost, the
effect of aluminum powder systems has been applied to
geothermal gradients are normal. Permafrost sections
prevent gas migration (Chapter 9).
vary from unconsolidated sands and gravels containing
ice lenses to ice-free, consolidated rock.
7-3.4 Calcined magnesium oxide During drilling and completion, a permafrost forma-
The hydration of magnesium oxide to magnesium hydrox- tion must not be allowed to thaw. Melting can cause the
ide provides an expansive force within the cement thawed earth to subside, particularly in the upper 200 ft
matrix. The hydrated material occupies more space than [60 m] of the well (Thorvaldson, 1962). The cement
that of the original ingredients. system should have a low heat of hydration and be able
to develop sufficient compressive strength (without
MgO (periclase) + H2 O ⎯→
⎯ Mg(OH)2 (brucite) (7-4) freezing) at temperatures as low as 20°F [–3°C]. Casing
strings must be cemented to surface, or a nonfreezing
The MgO must be calcined at very high temperatures, fluid placed in the annulus, to prevent casing damage
between 2,012° and 2,372°F [1,100° and 1,300°C]; oth- owing to the expansion of water upon freezing.
erwise, the hydration occurs before the cement sets, and Conventional Portland cement systems are not satis-
no significant cement expansion is observed (Spangle, factory in permafrost conditions, because they freeze
1988; Ghofrani and Plack, 1993; Rubiandini, 2000). before developing sufficient compressive strength. It is
Cement systems containing MgO provide excellent possible to add salt, alcohol, or other freeze-depressing
expansive performance at curing temperatures as high materials to the mix water; however, these additives
as 550°F [288°C]. However, at temperatures below about have adverse effects upon the quality of the set cement
140°F [60°C], the hydration reaction proceeds too slowly (Morris, 1970). Several types of cement systems perform

Chapter 7 Special Cement Systems 237

successfully in this severe environment: (1) calcium Table 7-4. Typical Compressive-Strength Data for a 50:50
aluminate cement, (2) gypsum-Portland cement blends Gypsum-Portland Cement Blend
(Benge et al., 1982), and (3) ultrafine Portland cement Temperature Compressive Strength (psi)
(Harris, 1991). (°F)
As described in Chapter 10, calcium aluminate At 16 hr At 24 hr At 3 days At 7 days
cement is a special-use material of limited production. It 20 300 350 1,100 2,000
is used to cement in situ combustion thermal wells. Such
40 450 450 1,350 1,950
cements also set and gain strength rapidly at low and
near-freezing temperatures (Maier et al. 1971). Fly ash 60 700 800 1,400 2,450
is often added as a diluent to reduce the cement’s heat
of hydration and for economy. The typical performance 80 750 1,000 1,850 2,650
of 50:50 fly ash-calcium aluminate cement systems is
shown in Table 7-3.
Table 7-5. Compressive Strength of a 50:50 Gypsum-
Portland Cement Blend After Freeze-Thaw Cycling
Table 7-3. Performance of 50:50 Calcium Aluminate-Fly Ash
Cement Systems† Day °F psi Day °F psi

Sodium Slurry Slurry Curing Curing Time (hr) 1 40 860 14 100 2,750
Chloride‡ Weight Volume Temperature
8 16 24 2 40 970 15 50 3,100
(%) (lbm/gal) (ft 3) (°F)
0 14.8 0.95 20 –§ – 355 3 30 1,250 16 15 3,480

5 14.9 0.97 20 – – 310 4 15 1,450 17 50 2,850

10 15.0 0.96 20 – – 145 5 50 1,790 18 100 2,820

0 14.8 0.95 25 20 45 90 6 100 1,990 19 160 2,740

5 14.9 0.97 25 NS†† 210 495 7 160 2,100 20 160 2,680

10 15.0 0.96 25 NS NS NS 8 160 2,270 21 160 2,690

0 14.8 0.95 40 50 658 1,560 9 160 2,360 22 160 2,670

0 14.8 0.95 50 115 2,250 2,475 10 160 1,980 23 160 3,380

0 14.8 0.95 60 145 1,800 2,900 11 160 2,520 24 160 2,750

† 0.53 ft3 (3.96 gal) water per 74 lbm of blend 12 160 2,420 25 160 2,710
‡ Based on weight of mixing water
§ Not measured
13 160 2,460 26 100 3,000
†† Not set

Gypsum-Portland cement blends, with sodium chloride Ultrafine Portland cements generally have Blaine
as a mix-water freezing depressant, are used extensively finenesses greater than 10,000 cm2/g, about three times
for permafrost cementing. The gypsum sets and gains greater than conventional Portland cement (Section
strength rapidly at freezing temperatures and prevents 7-12). It is well known that early compressive strength is
the slower-setting Portland cement from freezing. Such directly related to the surface area of the cement grains.
cement systems also have a lower heat of hydration than Greater surface area causes an accelerated reaction rate
that of calcium aluminate cement; therefore, they are because of the increased availability of reaction sites
particularly applicable to unconsolidated permafrost (Frigione and Marra, 1976). Figure 7-7 shows the com-
formations. The typical performance of a 50:50 blend of pressive strength development of an ultrafine cement
gypsum and Portland cement, with 12% NaCl by weight of system mixed with fresh water and seawater, compared
water (BWOW), is shown in Table 7-4. The effect of with a gypsum-Portland cement blend at 40°F.
freeze/thaw cycling upon compressive strength is illus-
trated in Table 7-5. No degradation of strength is observed.

238 Well Cementing

 1,000 ■ During cementing across massive salt formations or
water-sensitive zones, cement slurries containing
Ultrafine cement large amounts of salt prevent dissolution of the salt
in seawater
formation and prevent clay swelling.
Compressive
strength 500 Ultrafine
(psi) cement
7-5.1 Salty water as mixing fluid
Gypsum-Portland
cement blend In the absence of fresh water, brackish water or seawa-
ter is frequently used for mixing cement slurries. Such
0
0 1 3 7
waters are advantageous because of their availability and
economy.
Time (days)
Brackish waters from ponds and other surface water
Fig. 7-7. Compressive strength development of ultrafine Portland bodies vary significantly and should be thoroughly
cement system compared to a gypsum-Portland cement blend tested in the laboratory before use on location. The most
(from Harris, 1991). Slurry density is 12.2 lbm/gal and temperature is important ions to monitor are Cl–, SO42–, Ca2+, Mg2+, and
40°F [4°C]. Reprinted with permission of SPE. various organic compounds resulting from the decompo-
sition of plant material. Such impurities have significant
effects upon the performance of Portland cement
systems, including gelation, overretardation, or both
7-5 Salt cement systems (Kieffer and Rae, 1987). All laboratory cement-slurry
design experiments should be performed with a sample
Cement systems that contain significant quantities of of the location water.
NaCl or KCl are commonly called “salt cements.” Salt Seawater is the basic mixing fluid for offshore
has been used extensively in well cements for three cementing operations. Lyman and Fleming (1940) and
principal reasons. McIlhenny and Zeitoun (1969) characterized seawaters
■ In certain areas (e.g., offshore), salt is present in the from various locations and, as shown in Table 7-6, found
available mix water. them to be reasonably uniform. Smith and Calvert
■ Salt is a common and inexpensive chemical that, (1974) confirmed seawater to be suitable for preparing
when used as an additive, can modify the behavior of well cements and stated that the performance is “pre-
the cement system. dictable to a safe degree.”

Table 7-6. Seawater Analyses†

Components Gulf of Cook Trinidad Kharg Gulf of Sable Sea-Rite North Standard
(mg/L) Mexico Inlet, Island Suez Island Lake Sea Seawater
Alaska Production ASTM D-1141
Chloride 19,000 16,600 19,900 23,000 22,300 18,900 19,952 17,970 19,359

Sulfate 2,500 2,000 2,580 3,100 3,100 2,260 2,738 2,810 2,702

Bicarbonate 127 140 78 171 134 140 144 181 142

Carbonate 12 0 27 24 11 –‡ – – –

Sodium and 10,654 9,319 11,170 13,044 12,499 10,690 11,276 10,270 11,155
potassium

Magnesium 1,300 1,080 1,300 1,500 1,570 1,199 1,326 1,270 1,297

Calcium 400 360 408 520 464 370 419 390 408

Total dissolved 33,993 29,499 35,283 41,359 40,078 33,559 35,600 32,890 35,169
solids

pH 8.2 8.0 8.3 8.2 8.2 7.3 – – 8.2

Specific gravity 1.026 1.023 1.027 1.031 1.030 1.022 – – 1.025

† Reprinted with permission of SPE.
‡ Not measured

Chapter 7 Special Cement Systems 239

 Comparative laboratory testing has identified the (Slagle and Smith, 1963). In addition, laboratory studies
following effects of seawater upon the performance of have shown significant formation permeability reductions
Portland cement systems: as a result of exposure to low-salinity fluids (Hewitt,
■ reduced thickening time (Table 7-7) 1963; Jones, 1964; Mungan, 1965).
■ higher fluid-loss rate
Slagle and Smith (1963) tested the visual integrity of
clay formations after immersion into cement slurries
■ higher early compressive strength at low tempera-
of varying salinity. Salt-saturated cements were most
tures (Table 7-7) compatible with formations containing montmorillonite,
■ slight dispersing effect illite, and chlorite. However, NaCl concentrations as
■ higher shear-bond strength low as 10% BWOW were often sufficient to prevent
■ increased tendency for slurry foaming during mixing.
significant damage. Cunningham and Smith (1967)
showed that saline cement filtrate restricts the swelling
As discussed in Chapter 3, the presence of salt and migration of water-sensitive clays. Lewis et al.
depresses the ability of bentonite to extend a cement (1987) demonstrated improved bonding between salt
slurry. Thus, either prehydration of the bentonite or the cements and sensitive formations.
use of attapulgite is necessary (Smith and Calvert, 1974). Beach (1982) demonstrated that the cement slurry
salinity must be chosen with care. Beach observed that
significant long-term deterioration occurred when the
Table 7-7. Thickening-Time and Compressive-Strength ionic concentration of the cement was not comparable
Behavior of Cement Mixed with Fresh Water and Seawater to that of the formation. Disequilibrium causes ionic
Thickening Time Compressive Strength
diffusion, which disrupts the Portland cement binder.
(hr:min) (psi) In the same vein, Economides and Nolte (2000) recom-
at 6,000 ft at 100°F after 24 hr mended that cement slurries for sensitive formations
should contain a minimum of salt (in equilibrium with
Class A mixed 2:32 1,780
with fresh water the formation salinity), exhibit sufficient fluid-loss
control to minimize cement filtrate invasion, and not be
Class A mixed 2:05 2,150 overdispersed (to minimize invasion by a large amount
with seawater of free water).

7-5.4 Cementing across massive salt formations

7-5.2 Salt as a cement additive The presence of salt domes and massive evaporite
Salt is an extremely versatile cement additive that has sequences has long been problematic for drilling, com-
been used in well cements for more than 50 years. pletion, and long-term production. The high water solu-
Depending upon its concentration in the slurry, salt can bility and plasticity of such zones increase the difficulty
behave as an accelerator or a retarder (Chapter 3). Salt of obtaining a successful primary cementation. The
is also used to disperse cement slurries (Chapter 3), cement slurry can dissolve large quantities of formation
induce cement expansion (Section 7-3.2), and prepare material, resulting in a modification of performance
freeze-protected cements (Section 7-4). (Ludwig, 1951). Plastic salt zones can also encroach
upon the casing before the cement sets. Nonuniform
formation movement exerts point loading on the casing
7-5.3 Cementing across shale and bentonitic string, sometimes resulting in casing failure and collapse
clay formations (Cheatham and McEver, 1964). Salt cements are used
Approximately 87% of petroleum reservoirs contain clay routinely to reduce the severity of these problems; how-
minerals and silica fines (Hill, 1982). Drilling through and ever, some controversy exists regarding their efficacy.
cementing across such formations changes the chemical The first recorded use of salt in well cements was
environment surrounding the clays and may induce during the 1940s, when wells were completed across salt
destabilization, clay swelling, or fines migration, resulting domes in the U.S. Gulf Coast. Later, this became standard
in formation damage. For this reason, freshwater cement practice in the Williston basin (North Dakota and
slurries are not appropriate for primary cementing Montana, USA) and certain areas in the North Sea. The
across certain shale or bentonitic clay formations. This concentration of NaCl usually varied from 18% to 37%
problem was first identified when remedial cementing BWOW. While such practices prevented the dissolution
across such formations was found to be more successful of the formation, the high salt concentrations impeded
if saline formation waters were used to mix the slurry the performance of other cement additives, especially

240 Well Cementing

dispersants and fluid-loss additives that were originally In addition, Rae and Brown (1988) revealed that con-
developed for freshwater systems. In addition, the high tamination of a freshwater cement system by as little as
salt concentrations tended to overretard the cement 10% salt could reduce the thickening time by 30%,
system; thus, formation encroachment and casing damage increase the slurry viscosity by 100%, and increase the
could occur before the cement set. Two approaches have fluid-loss rate by nearly 500%. Yearwood et al. (1988)
been followed to solve these difficulties: eliminating salt confirmed these findings.
from the cement system and developing additives that Research has been performed to develop salt-saturated
are compatible with salt cements. cement systems (37.2% NaCl BWOW) without the disad-
Salt-free cement (Goodwin and Phipps, 1982), or vantages discussed earlier. In 1978, Messenger patented
cements containing very low salt concentrations (3% the use of certain hydroxycarboxylic acids as dispersants
BWOW), have been successfully applied in the Williston for salt cement slurries. Fluid-loss additives for salt
basin (Bryant, 1984) and the Middle East (Ismail and cement systems have been developed by Chatterji and
Khalaf, 1993). No casing collapse was reported with Brake (1982), Nelson (1986), Fry et al. (1987), and Bair
such systems, compared to a 20% failure rate with salt- et al. (2002).
saturated cements. To prevent excessive dissolution of the As a result, it is possible to design cement systems
formation, the cement-slurry displacement rates were low. with appropriate performance throughout the salt-
Ford et al. (1982) proposed an intermediate approach. concentration range. Cement systems containing from
Semisaturated cement systems (18% NaCl BWOW), in 5% to 37% NaCl (BWOW), with excellent placement char-
combination with holding the casing in tension, acteristics, appropriate thickening times, and acceptable
improved the success rate of primary cement jobs in the compressive-strength development were reported by Rae
Williston basin. and Brown (1988), Yearwood et al. (1988), Whisonant,
The above approaches may improve initial results; et al. (1988), and Brothers and deBlanc (1989). Typical
however, considering the previously discussed long-term performance data are presented in Tables 7-8 and 7-9.
effects of ionic disequilibrium, cement failure may Successful field results using proprietary cement system
ultimately occur. The rate of ionic diffusion is deter- compositions have been reported in various locations
mined by the difference in salt concentration between around the world.
the cement and formation and the permeability of the
cement (Kumar et al., 1987).
Experiments performed by P. Drecq (unpublished Table 7-8. Typical Compressive-Strength Performance
data, 1987) and van Kleef (1989) illustrated that low of Proprietary Salt Cement Systems
displacement rates would not necessarily prevent signif-
NaCl Cellulose/ BHST Density Compressive
icant formation dissolution, because salt zones are not (%BWOW) Organic (°F) (lbm/gal) Strength
eroded by cement slurries during placement. Three salt Acid at 3,000 psi
blocks of equal dimensions were submerged for 60 min (%BWOC)
8 hr 24 hr
in cement slurries with various salt concentrations.
The slurry temperature was 140°F [60°C], and slight 5 –† 200 15.8 1,850 2,600
agitation was provided. As shown in Fig. 7-8, significant 15 – 200 15.8 1,700 2,300
salt erosion was observed, except when the cement was
salt-saturated. 30 – 200 16.6 2,020 2,650

30 0.8/0.1 200 16.6 0 2,500

30 – 200 16.6 1,350 2,100

30 0.8/0.1 200 16.2 0 2,040

30 – 230 16.2 1,450 2,150

† not applicable

Fig. 7-8. Salt block appearance after 60 min of exposure to cements

with various salinities (1: saturated; 2: 18% BWOW; 3: 7.5% BWOW)
(after unpublished data by P. Drecq, 1987).

Chapter 7 Special Cement Systems 241

Table 7-9. Rheology and Fluid-Loss Performance of Proprietary Salt Cement Systems
NaCl Cellulose/Organic Acid BHCT Fluid Loss Rheological Properties at BCHT
(%BWOW) (%BWOC) (°F) (mL/30 min)
Yield Value Plastic Viscosity
(lbf/100 ft2) (cp)
30 0.8/0.1 120 178 22 244

30 –† 120 64 0.2 33.1

30 0.8/0.1 160 285 31.3 105

30 – 160 52 0.1 17.5

† Not applicable

Today, selecting the appropriate salt concentration in a As discussed in Chapter 2, absolute-volume shrinkage
cement slurry is largely a function of the local formation and internal volume reduction are a result of Portland
characteristics, operational limitations, and success cement hydration. Upon cement setting, stresses build
ratios. There is no official industry consensus regarding within the matrix, resulting in the formation of microc-
slurry design for salt-zone cementing. At present, anec- racks (Fig. 7-9). The propagation of the cracks lowers the
dotal evidence shows that most salt zones are cemented tensile capacity of the set cement and increases its per-
with systems containing between 8 to 18% NaCl (BWOW). meability. In latex-modified systems (Fig. 7-10), the latex
In addition, scrupulous attention is given to proper hole particles coalesce to form a plastic film that surrounds
conditioning and good cementing practices. and coats the C-S-H phase. Because of its elasticity and
Blast-furnace slag (BFS) combined with Portland high bonding strength, the latex bridges the microcracks
cement or drilling fluids has also been used recently and limits their propagation; as a result, the tensile
to cement salt zones. This application is discussed in strength of the set cement increases and the permeability
Section 7-8.2.2. decreases.

7-6 Latex-modified cement systems

Latex is a general term describing an emulsion polymer.
The material is usually supplied as a milky suspension
of very small spherical polymer particles (200 to 500 nm
in diameter), often stabilized by surfactants to improve
freeze/thaw resistance and prevent coagulation when
added to Portland cement. Most latex dispersions contain
about 50% solids. A wide variety of monomers, including
vinyl acetate, vinyl chloride, acrylics, acrylonitrile, ethyl-
ene, styrene, and butadiene, is emulsion-polymerized to
prepare commercial latexes.
The first use of latexes in Portland cements occurred
in the 1920s, when natural rubber latex was added to
mortars and concretes. Since then, latex-modified con-
cretes have become commonplace because of the follow-
ing improvements in performance (Ohama, 1987):
■ improved pumpability

■ decreased permeability Fig. 7-9. Photograph of microcracks in set Portland cement (from
Kuhlmann, 1985). Reprinted with permission from Elsevier.
■ increased tensile strength

■ reduced shrinkage

■ increased elasticity

■ improved bonding between cement/steel and

cement/formation interfaces.

242 Well Cementing

 7-6.2 Early latex-modified well cement systems
In 1958, Eberhard and Park patented the use of vinyli-
dene chloride latex in well cements. Improved perfor-
mance was observed for systems containing up to 35%
latex solids BWOC. Later, polyvinyl acetate latex was
identified as a suitable material (Woodard and Merkle,
1964). The preferred concentration of latex solids varied
from 2.5% to 25% BWOC. The polyvinyl acetate system
has been used successfully for many years; however, it is
limited to applications at temperatures less than 122°F
[50°C] bottomhole static temperature (BHST).

7-6.3 Styrene butadiene latex systems

An improvement in latex cement technology occurred
when Parcevaux et al. (1985) identified styrene butadiene
latex as an effective additive for the prevention of annular
Fig. 7-10. Photograph of latex-modified Portland cement, magnified
gas migration (Chapter 9). Additional refinements were
1,200 times (from Kuhlmann, 1985). Reprinted with permission from made by Sault et al. (1986).
Elsevier. Styrene butadiene latexes impart the same beneficial
effects described above; however, they are effective at
temperatures as high as 350°F [176°C]. Fig. 7-11 is a plot
7-6.1 Behavior of latexes in well cement slurries of API/ISO fluid-loss rate versus latex concentration for
The use of latex in well cements occurred much later. various well cement slurries. The results illustrate that
In 1957, Rollins and Davidson reported improved per- normal-density neat slurries require less latex to
formance when latex was added to the mix water. In achieve a given fluid-loss rate. More latex is required for
addition to the attributes mentioned above, adding slurries containing extenders or weighting agents, espe-
latex provided the following additional benefits: cially those with a lower solids content (extended with
■ better bonding to oil-wet and water-wet surfaces sodium silicate). Figure 7-12 illustrates the decreased
■ less shattering when perforated
volumetric shrinkage observed with a latex-modified
Portland cement system cured at 158°F [70°C].
■ increased resistance to contamination by well fluids

■ lowered fluid-loss rate

■ improved durability. 3.0

Neat: 15.8 lbm/gal
When latex is added as part of the liquid phase of a Barite: 18 lbm/gal
Bentonite: 13.3 lbm/gal
Portland cement system, the resulting slurry has a normal 2.5 Fly ash: 13.6 lbm/gal
color and consistency; however, because of the solids con- Sodium silicate: 13 lbm/gal
tent of the latex, such slurries contain 20% to 35% less water.
After curing, the set product consists of hydrated cement 2.0
particles connected by a film of latex particles (Kuhlmann, Styrene
butadiene
1985). The film of latex particles imparts the physical latex
and chemical properties described above (Parcevaux (gal/sk) 1.5
and Sault, 1984; Drecq and Parcevaux, 1988). While the
slurry is still liquid, the latex particles impart excellent
rheological properties because of a lubricating action. In 1.0
addition, the latex particles provide excellent fluid-loss
control by physically plugging small pores in the cement
filtercake (Drecq and Parcevaux, 1988) (Chapter 3). 0.5
0 50 100 150 200 250 300
API/ISO fluid loss (mL/30 min)

Fig. 7-11. Fluid-loss rates of latex-modified cement slurries (185°F

[85°C]).

Chapter 7 Special Cement Systems 243

 7 Portland cement systems are not compatible with
Neat cement
6 strong inorganic acids such as sulfuric, hydrochloric,
and nitric acid. In such environments, organic polymer
5
Absolute
cements, usually epoxy-based, must be used to provide
4 sufficient chemical resistance (Cole, 1979). Such sys-
volume
shrinkage 3 tems are also known as “synthetic cements.”
(%) Latex-modified cement Epoxy cements are prepared by mixing an epoxy resin
2
such as bisphenol-A (Fig. 7-13) with a hardening agent.
1 Depending upon the desired end properties, the hard-
0 ening agent can be an anhydride, aliphatic amine, or
0 6 12 18 24 polyamide (Sherman et al., 1980). A solid filler such as
Time (hr) silica flour is often used to build density, reduce cost,
and act as a heat sink for the exothermic reaction that
Fig. 7-12. Absolute volume shrinkage of normal-density Portland occurs during the curing period. Depending upon the
cements (from Parcevaux and Sault, 1984). Reprinted with permis-
sion of SPE.
circulating and static well temperatures, various cata-
lysts and accelerators can also be added to control the
placement and setting times.

7-7 Cements for corrosive environments

Set Portland cement is a remarkably durable and forgiving CH3 OH CH3
O O
material, but it has its limits. In a wellbore environment, CH2 CHCH2 O C OCH2CHCH2 O C OCH2CH CH2
Portland cement is subject to chemical attack by certain n
formations, migrating fluids, and substances injected from CH3 CH3
the surface. As discussed in Chapter 10, saline geothermal
brines containing CO2 are particularly deleterious to the Fig. 7-13. Chemical structure of bisphenol-A.
integrity of the set cement. Cement durability is also a
key consideration in wells for chemical waste disposal
and for enhanced oil recovery by CO2 flooding. Epoxy resin cement systems are characterized by
their corrosion resistance and high compressive and
shear-bond strength. They are compatible with strong
7-7.1 Cements for chemical waste disposal wells acids and bases (up to 37% HCl, 60% H2SO4, and 50%
Zonal isolation is of paramount importance in chemical NaOH) at temperatures up to 200°F [93°C] during
waste disposal wells. If not properly confined, injected extended exposure periods. Epoxies are also resistant
waste fluids could contaminate freshwater aquifers and to hydrocarbons and alcohols, but not to chlorinated
corrode the exterior of the casing. To ensure zonal isola- organics or acetone. Typically, the compressive strengths
tion throughout the life of such wells, the cement and range between 8,000 and 10,000 psi [56 to 70 MPa], and
the tubular hardware in the well must be chemically shear-bond strengths can be as much as 9 times higher
resistant to the waste fluids (Runyan, 1974). than those of Portland cement (R.A. Bruckdorfer,
The chemically resistant casings used in waste disposal unpublished data, 1985).
wells include modified polyester and epoxy fibercast, Until recently, to prevent premature setting and to
or metal alloys such as Carpenter 20,† Incoly 825,‡ and enhance cement bonding, nonaqueous spacers such as
Hastalloy G.§ The cement systems are chosen for compat- gelled oil, diesel, or alcohol were required on all epoxy
ibility with the injected waste material. cement jobs. The use of such preflushes is costly and
Modified Portland cements are generally appropriate time consuming. In 2001, Eoff et al. introduced a water-
for disposal wells involving weak organic acids, sewage compatible epoxy cement system. Instead of bisphenol-A,
waters, or solutions having a pH of 6 or above. The dura- the epoxy sealant is the diglycidyl ether of cyclohexane
bility of the set cement is improved by adding pozzolans, dimethanol. The hardening agent is a mixture of diethyl-
increasing the density by addition of dispersants, toluenediamine and tris(dimethylaminomethylphenol).
employing controlled particle-size technology, or adding The performance of this system is unaffected by water
liquid latexes to the slurry. These methods substantially contamination, up to 25% by volume.
reduce the permeability of the set cement.

† Trademark of CRS Holdings, Inc.

‡ Trademark of International Nickel Co.
§ Trademark of Haynes International, Inc.

244 Well Cementing

7-7.2 Cements for enhanced oil recovery by CO2 Instead, the most practical approach is to lower the
flooding degradation rate of Portland cement systems.
Corrosion owing to CO2 in production operations is well The cement matrix permeability can be reduced by
documented (Newton and Hausler, 1984), and studies of lowering the water-to-cement ratio, employing controlled
Portland cement-based well cement corrosion by CO2 particle-size cement technology, adding latex, and adding
have been conducted by Onan (1984) and Bruckdorfer pozzolanic materials. As discussed in Chapter 3, pumpable
(1986). It is well known that carbon dioxide–laden Portland cement slurries with densities up to 18.0 lbm/gal
waters can destroy the structural integrity of set [2,161 kg/m3] can be prepared with the addition of a dis-
Portland cements (Biczok, 1967). The basic chemistry persant. After setting, the water permeability of such sys-
describing this process is as follows. tems is usually less than 0.001 mD; consequently, invasion
of carbon dioxide–laden water is inhibited, and the rate

(1) CO2 + H2 O  
 H2 CO3  + –
 H + HCO3 of corrosion is slowed. The addition of pozzolans (such as
fly ashes) and latex, as well as controlling the particle-
(7-5) size distribution, also reduces set-cement permeability
(Chapter 3) and effectively eliminates Reaction 2 above.
(2) Ca(OH)2 + H+ + HCO3– ⎯→
⎯ CaCO3 + 2H2 O When such measures are taken, the corrosion rate can
be reduced by as much as 50%.
(7-6) BFS-Portland cement blends reportedly show
improved resistance to CO2 (Cowan et al., 1994a). This
(3) C-S-H phase + H+ + HCO3– ⎯→
⎯ application is discussed further in Section 7-8.2.3.
CaCO3 + amorphous silica gel (7-7) Elastomeric composites based on styrene-butadiene
rubber have also been introduced as cements for CO2-
In Step 1, approximately 1% of the dissolved carbon injection wells (Onan et al., 1993; Section 7-11).
dioxide reacts with water to form carbonic acid. As The long-term efficacy of the modified Portland
the carbon dioxide–laden water diffuses into the cement systems in CO2-flood wells remains to be seen.
cement matrix, the dissociated acid is free to react with Such systems may only postpone the inevitable dete-
the free calcium hydroxide (Step 2) and the C-S-H phase rioration. More research is needed to develop truly
(Step 3). As carbon dioxide–laden water continues to stable, yet economically viable, cements for this diffi-
invade the matrix, other equilibria are established. cult environment.

(4) CO2 + H2 O + CaCO3 ⎯→

⎯ Ca(HCO3 )2
(7-8) 7-8 BFS systems
The use of BFS systems in well cementing is not a new

(5) Ca(HCO3 )2 + Ca(OH)2 
 2CaCO3 + H2 O concept; indeed, patents covering slag-cement composi-
(7-9) tions for wellbores appeared as early as 1958 (Harmeen
and Stuve, 1958). BFSs can be used alone as a cementi-
In the presence of excess carbon dioxide (Step 4), tious material or blended with Portland cements (called
calcium carbonate is converted to water-soluble calcium Portland slag cements or slag cements). Slags are also
bicarbonate, which can migrate out of the cement matrix. used to convert drilling fluids into a cementitious material.
In Step 5, the dissolved calcium bicarbonate can react In this case, chemical activators are required to speed
with calcium hydroxide, forming calcium carbonate and up slag hydration kinetics.
water. The liberated water can then dissolve more calcium The use of BFS for well cementing has increased sig-
bicarbonate. The net result is a leaching of cementitious nificantly since the early 1990s. This section discusses
material from the cement matrix, an increase of porosity the production and chemistry of BFS, as well as the use
and permeability, and a decrease of compressive strength. of BFS in various well-cementing applications.
Downhole, this translates to a loss of casing protection
and zonal isolation. 7-8.1 Production, composition, and hydration of BFS
Carbon dioxide corrosion of Portland cements is ther- BFS is a byproduct of steel manufacture. In the opera-
modynamically favored and cannot be prevented. An easy tion of a blast furnace, the iron oxide ore is reduced
solution to this problem would be a synthetic cement; by means of coke to metallic iron, while the silica and
unfortunately, such systems are not economically feasi- alumina constituents combine with the lime and mag-
ble for most CO2-flooding or CO2-sequestration projects. nesia to form a molten slag that collects on top of the

Chapter 7 Special Cement Systems 245

molten iron in the furnace. The slag issues from the 7-8.2 BFS-Portland cement blends
blast furnace as a molten stream at a temperature of During the past decade, the use of BFS for well cementing
2,462°–2,822°F [1,350°–1,550°C]. If allowed to cool has increased significantly, notably for squeeze applica-
slowly, the slag crystallizes into a material having tions (Chapter 14). When BFS is used in combination
virtually no cementitous properties. If rapidly cooled to with Portland cement, special chemical activators are
below 1,472°F [800°C], it forms a glass that is a latent not generally needed, because two slag-hydration activa-
hydraulic cement. Cooling is usually effected by spraying tors are already present: gypsum (sulfate activation) and
droplets of the molten slag with high-pressure water jets. portlandite, Ca(OH)2.
This produces a wet, sandy material that, when dried and The granulated BFS is blended or interground with
finely ground, is called ground granulated blast-furnace Portland cement or a clinker-gypsum mixture. The relative
slag and often contains more than 95% glass. The slag con- proportions of Portland cement and BFS vary widely;
sists of a silicate network containing calcium, magnesium, the BFS content can exceed 80%, but less than 50% is
and aluminum cations. The crystalline components, more common. These cements are used extensively in
which occur as inclusions in the glass, are generally civil engineering and are readily available worldwide.
composed of merwinite (C3MS2) and melilite (a solid The principal hydration products of BFS-Portland
solution of gelhenite, C2 AS, and akermanite, C2 MS2). cement blends are similar to those of pure Portland
The slag composition depends on the original iron ore cements, but the quantities of portlandite liberated
and can vary widely. However, the slag composition pro- upon hydration are lower. When the BFS concentration
duced from a particular plant is generally consistent in the blend exceeds 60%, the portlandite content peaks
unless the ore source changes. Table 7-10 shows the and then decreases with hydration time. The mean bulk
compositions of slags produced in various countries. lime-to-silica (C/S) ratio of the calcium silicate hydrate
(C-S-H) phase is lower than that found in pure Portland
cement slurries. For a blend with 50% BFS, C/S values in
Table 7-10. Oxide Composition (Percentage) of BFS† the region of 1.5 are obtained, compared to 1.8 when
Source CaO SiO2 Al2O3 MgO Fe2O3 MnO S BFS is not present.
United Kingdom 40 35 16 6 0.8 0.6 1.7
BFS is a pozzolanic material (Chapter 3) that can
react with portlandite to form additional C-S-H at tem-
Canada 40 37 8 10 1.2 0.7 2.0 peratures below 230°F [110°C]. At higher temperatures,
France 43 35 12 8 2.0 0.5 0.9
the C-S-H phase converts into other crystalline calcium
silicate hydrate phases (Chapter 10).
Germany 42 35 12 7 0.3 0.8 1.6 Slag cements are generally more resistant to chemi-
cal attack than pure Portland cements. The principal
Japan 43 34 16 5 0.5 0.6 0.9
benefits are
Russia 39 34 14 9 1.3 1.1 1.1 ■ better sulfate resistance

South Africa 34 33 16 14 1.7 0.5 1.0 ■ slower diffusion of chloride and alkali ions through
the cement matrix
United States 41 34 10 11 0.8 0.5 1.3
■ lower set-cement permeability.
† From Moranville-Regourd (2001). Reprinted with permission from Elsevier.
These performance improvements are the result of
changes in the cement-matrix microstructure. When BFS
To ensure complete hydration, the Blaine fineness of is present, the pore sizes in the set-cement matrix are
ground slag is generally around 4,000–5,500 cm2/g. The smaller, and the capillary porosity (pore radii above 300 Å)
densities typically vary between 2.88 and 2.96 g/cm3, is lower. The total capillary porosity can be reduced by 25
depending on their chemical composition. to 30% when BFS is 75% of the Portland cement mixture.
Unlike Portland cement, which requires only the The pore-size reduction can be attributed to the
addition of water to set, BFS is a latent hydraulic material absence of large portlandite crystals. In addition, when
that requires chemical activation, especially at low the proportion of slag in the blend is increased, the
temperatures. Slag activators include caustic soda C-S-H–phase microstructure becomes progressively less
(sodium hydroxide, NaOH), soda ash (sodium carbonate, fibrillar and more foil-like. The foils encapsulate
Na2CO3), sodium silicate, sodium sulfate, calcium sulfate, some of the pores, making them inaccessible to fluids
lime, or mixtures thereof. Portland cement is also an and reducing the set-cement permeability. The lower
effective activator (Section 7-8.2). In all cases, the C-S-H amount of portlandite, which is more soluble in water
phase (Chapter 2) forms, with aluminum, magnesium, than the C-S-H phase, improves the resistance of the set
iron, and sulfate ions incorporated in its structure. cement to sulfate and chloride attack.

246 Well Cementing

 In addition to mud-to-cement technology, slags and Portland cement slurry can negatively affect rheology,
slag cements are used in the following well cementing thickening time, fluid-loss control, and compressive
applications: strength. However, in systems containing BFS, sodium
■ squeeze cementing using microfine cements chloride can considerably increase the compressive
■ salt-saturated cement slurries
strength of the resultant set cement (Table 7-11).
■ CO2-resistant cements

■ foamed cement Table 7-11. Effect of Salt-Saturated Mix Water on

■ improvement of construction grade cements in well Compressive Strengths of BFS Cement Systems
cementing. Slag Water Compressive Strength (psi [MPa])
Each application is briefly described below. NEWCEM† Fresh 566 [3.9]

NEWCEM Salt 977 [6.7]

7-8.2.1 Squeeze cementing
KOCH‡ Fresh 965 [6.7]
When a casing leak occurs or a poor bond exists between
the cement and the casing or formation, a competent KOCH Salt 1,453 [10]
repair is necessary to restore wellbore integrity † NEWCEM from Blue Circle Cement Company; Blaine specific surface area: 5,500 cm 2/g
(Chapter 14). Squeeze cementing is a common method. ‡ KOCH from Koch Minerals; Blaine specific surface area: 10,040 cm2/g

Conventional cement systems are sometimes ineffective,

because the fissure or channel to be squeezed is too
narrow for the cement grains to penetrate. Cowan et al. also recommended drilling the wellbore
Small-particle-size cement systems have been devel- with salt-saturated drilling fluids containing BFS and
oped for squeeze cementing. These so-called microfine converting the drilling fluid into a settable cement
cements include pure Portland cements, BFS, and very slurry by adding activators, additional slag, or both. In
often BFS-Portland cement blends (Clarke, 1988; Meek this case, slag-based cements bond to the casing partic-
and Harris, 1991; Ewert et al., 1991). The Blaine fineness ularly well owing to the absence of shrinkage upon
of microfine cements varies between 6,000 cm2/g and curing as well as compatibility with drilling fluid
16,000 cm2/g, compared to about 3,200 cm2/g for Class G remaining on the casing surface (Section 7-8.3). The
cements. The median diameter of cement particles is strength and ductility of slag cements is also improved
2–3 μm, compared to 15–18 μm for Class G cements. by inclusion of crosslinkable polymers such as polyacry-
The diameter of the largest cement particles must be lamide in the cement slurry.
less than 30 μm and preferably below 10–15 μm. For
high-temperature wells, in which the BHST is above
230°F [110°C], silica fume or very fine silica is added 7-8.2.3 CO2 resistant cements
instead of silica flour to prevent strength retrogression, Slag cements are more resistant to carbon dioxide and car-
because the particle size of silica flour is too large. bonic acid than Portland cements (Cowan et al., 1994a).
Effective retarders have been developed to retard Portland cements gradually degrade while exposed to
microcements at high temperatures. Brothers (1994) carbon dioxide because of the transformation of calcium
described a methylene phosphonate derivative that pro- silicate hydrates into calcite and silica gel (Section 7-7). In
vides predictable retardation up to about 245°F [118°C]. addition, carbonic acid can interfere with the setting of
Rodrigues and Lindsey (1995) described another phos- Portland cement slurries. The improved resistance of slag
phonate compound that performs reliably at up to 400°F cements is most likely caused by the lower matrix perme-
[204°C]. ability.

7-8.2.2 Salt-zone cementing 7-8.2.4 High-pressure, high-temperature gas wells

The use of BFS for well cementing across salt formations BFS cements have been used successfully to cement
was patented by Cowan et al. (1994b). As discussed in high-pressure, high-temperature gas wells in South
Section 7-5, large quantities of sodium chloride are often Texas, USA (Sweatman et al., 1995). They are especially
added to the mix water (up to saturation, 37% BWOW) suitable for obtaining shorter waiting-on-cement times
when Portland cement slurries are used. This practice at the top of cement, where temperatures are lower than
prevents dissolution of the salt formation that could BHCTs. Slag systems have excellent shear-bond
result in poor bonding between the cement and for- strengths under simulated downhole conditions, and
mation. Adding such high concentrations of salt to a previously broken shear bonds seem to regenerate or

Chapter 7 Special Cement Systems 247

heal with time. The shear bond provided by the slag compressible and have excellent fluid-loss control, short
system improves with time at elevated temperatures and transition times, and good thermal insulation properties.
has much better bond regeneration properties than A thorough discussion of foamed cements is presented
Class H cement (Table 7-12). The superior rebonding in Section 7-10.
or healing performance of slag may be linked to the
development of internal stresses caused by a densified
microstructure and other mechanisms that are not yet 7-8.2.6 Steam injection
well understood. Slag systems have also maintained Blanco et al. (1999) performed a comparative study of
zonal isolation after exposure to high pressures during BFS cements and Portland cements for steam injection
hydraulic fracture treatments. wells. After enduring thermal shocks to simulate the
steam-injection process, the BFS systems exhibited
better thermal stability and compressive strength, and
lower water permeability, than the corresponding
7-8.2.5 Foamed cements Portland cement systems. The BFS cements expanded
Chatterji et al. (1998) showed that slags could be used in after thermal shock. As a result, the cement bond logs
foam cement. The slag has a Blaine fineness of 5,900 indicated superior zonal isolation compared to those
cm2/g, and is activated with sodium carbonate. The completed with Portland cement. The authors con-
slurry density varies from 6 to 16 lbm/gal [720 to 1,921 cluded that BFS systems are more suitable for steam
kg/m3], depending on foam quality. The cement com- injection wells. More information about cementing
positions are particularly suitable for primary steam-injection wells is presented in Chapter 10.
cementing operations because they are lightweight and

Table 7-12. Ambient and High-Temperature Shear Bonds†

Slurry Composition Salt-Saturated Class H Freshwater Slag Cement
lbm/bbl lbm lbm/bbl lbm
Class H 325 94 0 0

Slag 0 0 344 87

Hematite 38 11 120 30.5

Silica flour 113 32 69 17.4

Salt 49 18.7 0 0

Organic additives 9.6 2.8 6.9 1.7

Surfactant 0 0 7.0 1.8

Soda ash 0 0 2.1 0.5

Fresh water 173 50.2 168 168

Water requirement, gal/sk 6.0 6.0 5.1 5.1

First Test Second Test Average First Test Second Test Average
Shear bond, psi (280°F/1,000 psi)
Initial, 8-day 152 158 155 122 158 140
Retest, 14-day 109 90 100 191 164 178
Change (%) –28 –43 –36 57 4 28
Regenerated‡ 36 44 40 117 118 118

Lower section shear bond, psi

Ambient conditions 442 460 451 560 588 574
† For 17-lbm/gal [2,141 kg/m3] salt-saturated Class H and freshwater slag-cement slurries measured on high-pressure,
high-temperature shear-bond tester (from Sweatman, 1995). Reprinted with permission of SPE.
‡ Regenerated shear bond = 14-day shear bond + 8-day skin

248 Well Cementing

7-8.2.7 Treatment of construction cements ical properties of construction-grade cement such
Mueller et al. (1995) showed that blending BFS with that the compressive strengths and thickening times
construction-grade Portland cements can produce a meet API and International Organization for
high-quality well cement. The blends have fluid-loss, Standardization (ISO) standards. Construction-grade
free-water, and compressive-strength properties cements could not otherwise be used for well cementing
comparable to, and sometimes better than, Portland because of their unpredictability from brand to brand
cements (Table 7-13). With construction-grade cements, and their failure to meet the API/ISO standards. The lower
a slag/Portland cement blend produces a more linear cost of construction-grade cement decreases the cost
response to lignosulfonate retarders (Di Lullo Arias, of well cementing operations, especially in remote
1996). The BFS concentration can vary from 10% to areas to which API/ISO oilwell cements must be
200% BWOC. The BFS alters the chemical and phys- imported.

Table 7-13. Performance of Some Construction Cements Modified with BFS†

Slurry Designs Composition
1 2 3 4 5 6 7 8
Gas Slurry Lead Tail Lead Lead Lead Lead Tail
BFS:PC ratio‡ 0.6:1 0.6:1 1:1 1:1 0.4:1 0.4:1 0.8:1 0.4:1§

PVAP†† (%BWOW) 15 5 5 –‡‡ 18 10 – 3

LMS§§ (gal/sk) 2.5 – – – – 2.5 – –

Dispersant (gal/sk) 0.05 – 0.12 – 0.1 0.25 – 0.15

Liquid Na Silicate (gal/sk) – – – – 0.5 0.35 0.6 –

BASS††† (gal/sk) 0.5 0.5 – – – – – 0.3

CaCl2 (%BWOC) 2.5 0.5 – 3.5 4 1.5 2.5 –

Density (lbm/gal) 11.7 11.7 15.8 13 13.4 13.4 11.5 15.6

BHST (°F) 145 136 220 120 88 106 192 500‡‡‡

BHCT (°F) 115 103 160 96 84 140 145 250

Test Results
Thickening time (min) 375 390 248 370 202 161 260 300

Fluid loss (mL/30 min) 25 186 94 na§§§ 32 39 na 252

Free fluid (%) Trace 0.5 Trace 0.5 Nil Nil 0.4 Trace

Compressive strength (psi/24 hr) 1,150 370 3,150 400 425 2,350 1,220 3,650

μp /τy 18/10 14/9 56/14 22/8 24/9 24/10 10/3 64/10

† From Mueller et al. (1995). Reprinted with permission of SPE.
‡ Density of BFS = 5 lbm/ft3
§ Portland cement = ASTM Type I

†† Polyvinyl acetate phthalate

‡‡ Not present

§§ Liquid microsilica

††† Blend of aluminum salts and surfactant

‡‡‡ Producing steam temperature

§§§ not applicable

Chapter 7 Special Cement Systems 249

7-8.3 Cementitious drilling fluids Table 7-14. Typical Performance of Mud-to-Cement
Since the early 1990s, there has been considerable inter- Systems†, ‡
est in converting drilling fluids into a cementitious Component or Property 560 m2/kg 1,100 m2/kg
material for well cementing. Such technology has several Fineness BFS Fineness BFS
potential advantages (Cowan et al., 1992). Initial density of mud (lbm/gal) 10.3 10.3
■ Better mud removal
Amount of slag added (lbm/bbl) 300 300
■ Less effect of mud contamination on the performance
of the well cement, owing to greater compatibility Amount of sodium hydroxide
added (lbm/bbl) 10 6
between the two fluids
■ Potential solidification of both the mud filtercake and Amount of sodium carbonate
undisplaced mud added (lbm/bbl) 10 6
■ Reduction of mud disposal volume (Nahm et al., Final density of mixture (lbm/gal) 13.3 13.3
1993)
Thickening time at 90°F 5:00+ 5:00+
Two principal systems have been developed. The first
Free water at 80°F (mL) 1.5 1.5
involves treating the mud with special copolymer disper-
sants and accelerators and then adding cement (Wilson et Compressive strength, psi, at 80°F after:
al., 1989; 1990; Bloys et al., 1991). The second consists of 8 hr 185 Not set
mixing BFS and a water-base mud that has been diluted,
12 hr 450 100
conditioned, and pretreated with a chemical activator for
the BFS (Hale and Cowan, 1991; Cowan and Hale, 1994; 24 hr 945 1,695
Hale, 1996; Cowan et al., 1996; Hale and Cowan, 1997). 48 hr 1,370 2,690
Other patented systems include drilling fluids comprising † From Cowan et al. (1992).
condensed silica fume activated by alkalis (Terry et al., ‡ For solidification of salt-saturated/starch mud
1994; Onan et al., 1994), oil-base mud conversion using
BFS (Nahm and Wyant, 1993), and drilling fluids contain-
ing monomers that are polymerized by chemical activation
or irradiation (Novak, 1988). In all of these cases, elaborate 7-8.3.1 Mud-to-cement process
and time-consuming pretreatment of the mud is required The mud-to-cement process involves several steps. First,
before adding cementitious material. the drilling fluid to be converted is isolated. Second, the
During the last decade, BFS-drilling fluid systems drilling fluid is diluted with water to obtain a specific
have been used in cementing operations, including final slurry density or for rheological control. In most
primary cementing jobs, temporary abandonment plugs, cases, the mud must be diluted by 60% or more to accom-
and sidetracking plugs. Up to 500 lbm [227 kg] of BFS modate slag into the system (Benge and Webster, 1994).
can be added per barrel of pretreated drilling fluid. This This minimizes the volume of mud that can be converted.
system has a wide application range. Slurry densities Third, other additives such as retarders, dispersants, and
can vary from 10 to 20 lbm/gal [1,200 to 2,400 kg/m3], slag activators are added. Finally, the treated drilling
and the application temperature can vary from ambient fluid is transferred to the cementing unit, slag is added,
to about 392°F [200°C]. Hale and Cowan (1991) found and the mixture is then pumped down the well.
that, unlike Portland cement, BFS does not significantly
affect mud properties. BFS can also be added during
drilling to ensure that the mud filtercake contains a 7-8.3.2 Mud-to-cement performance
cementitious material and will harden once chemically Song et al. (2000), Leimkuhler et al. (1994), Nahm et al.
activated (Table 7-14). (1995), and Dahua et al. (1998) reported that adding
BFS to mud can not only provide good-quality cement
jobs in normal wells, but also is very useful for solving
gas-migration and lost-circulation problems. Using
cement bond logs as the principal criterion, these authors
reported that the success rate of wells completed with
mud-to-cement systems was frequently better than that
of conventional cement systems.

250 Well Cementing

 Several limitations surrounding this technology have 7-9 Engineered particle-size distribution
been reported (Saasen et al., 1994, Benge and Webster, cements
1994; Mueller and Dickerson, 1994; Sabins et al., 1996).
In laboratory testing, microfractures have been observed 7-9.1 Introduction
in the set cement. The cause of the microfractures is As discussed in Chapter 3, cement-slurry properties
unclear. It is also unknown whether the microfractures depend on many variables: the amount and properties
occur in the well in which the cement is confined. Benge of solids (including cement), the additives, the amount
and Webster (1994) observed that, unlike set Portland and type of mix water, temperature, and pressure.
cement, set drilling fluids containing slag have no fibers Achieving optimal cement performance can be difficult
or crystals connecting the grains. They concluded that the at the extremes of the slurry-density range. At slurry
set material is less deformable and has a greater tendency densities above 17.5 lbm/gal [2,101 kg/m3], designing a
to crack. However, Silva et al. (1997) did not observe stable and pumpable slurry can be difficult. In the low-
microcracks after testing at high temperature and pres- density range [less than 14.0 lbm/gal (1,680 kg/m3)]
sure. They concluded that BFS–drilling fluid systems achieving rapid compressive strength development and
can be used successfully in high-temperature conditions. high final compressive strength can be challenging. In
Mueller and Dickerson (1994) showed that overacti- conventional well cement systems, the slurry density is
vation of slag-based systems produces several detrimental traditionally adjusted by varying the water-to-solids ratio
effects, including abnormal thickening-time response, or by adding weighting agents or low-density extenders.
excessive heat buildup, high mixing viscosities, and This section presents a more sophisticated design
diminished compressive strength owing to stress cracking. method that considers the particle-size distribution of
They suggested that, when the activator system contains the solids in the cement slurry.
sodium hydroxide, stress-cracking and brittleness result. The civil engineering industry uses the concept of
Another important issue is the design complexity particle packing to prepare high-performance concrete
of BFS-based mud-to-cement systems. Unlike a conven- (de Larrard, 1989; de Larrard and Sedran, 1994). By
tional cement system, the composition of a drilling fluid increasing the packing density of solids in the dry blend,
can vary widely as drilling progresses. The composition of the amount of water necessary to prepare a pumpable
the cuttings changes as the drill bit penetrates different slurry decreases (Mooney, 1951). As a result, the set
formations, each with different chemical environments, cement is stronger and less permeable. This technology has
and chemicals are continually added by mud engineers recently been adapted for well cementing applications.
to maintain the desired properties. Such chemicals can
strongly affect thickening time, rheology, and com-
pressive strength. Therefore, to properly design a 7-9.2 Engineered particle-size concept
BFS–drilling fluid cement system, the testing labora- In conventional cements, the slurry and set-cement
tory must receive a mud sample that is representative properties depend on the slurry density (water-to-solids
of the material that will be available for conversion. For ratio) and the additives. For engineered particle size
this reason, long before the cementing treatment, a (EPS) cement systems, the slurry and set-cement prop-
large volume of mud is usually isolated from the active erties depend on the following:
system, allowing the laboratory sufficient time to design ■ properties of the solids
an appropriate slurry.
■ packing volume fraction (PVF or fpV) of the solids
The addition of BFS to drilling muds can substantially
increase fluid viscosity and result in severe flocculation. ■ solid volume fraction (SVF or fsV) of the resulting

Conventional drilling-fluid dispersants do not necessarily slurry.

disperse cement slurries. However, several suitable dis- In a powder, the PVF is defined as the volume of space
persants have been patented (Bloys et al., 1994; occupied by the solid particles (the absolute volume)
Carpenter and Johnson, 1999). A styrene sulfonic acid- divided by the total volume of solid particles plus the
maleic anhydride copolymer, recently developed by void space between them (the bulk volume).
Hou and Liu (2000), is an effective dispersant for mud-
to-cement technology. (Va ) solids
f pV = (7-10)
Vbulk

Chapter 7 Special Cement Systems 251

 The packing of a powder is a purely geometrical Vsolids
phenomenon. Therefore, the PVF depends only on the fsV = (7-12)
Vslurry
size and the shape of the particles. A perfect arrange-
ment of spheres with the same size (compact hexagonal
packing) has a PVF of 0.74 (Cumberland and Crawford, The volume of mix water divided by the total slurry
1987). A random packing of the same spheres has a volume is the slurry porosity (φ).
PVF of 0.64. In other words, the solids occupy 64% of Vwater
the bulk volume, and the void space occupies 36% of the φ= (7-13)
bulk volume. Vslurry
Higher PVFs can be achieved by preparing powders
with solids that have more than one particle size (i.e., a As shown in Eq. 7-14, the slurry density (ρslurry) is
polydisperse mixture). The smaller particles fit in the adjusted by varying the porosity for a given blend of
void spaces between the larger ones. From geometry, if solids with density ρb (Maroy and Baret, 1994; Baret et
the smaller particles are more than 10 times smaller, al., 1996; Noik et al., 1998).
and they are all the same size (i.e., monodisperse), they
can fill 64% of the void. Hence, the PVF would be )
ρ slurry = ρ b (1 – φ + φρmw (7-14)

f pV =
0.64 + ( 0.36 × 0.64 ) = 0.87. (7-11)
Or, expressed in terms of SVF, pmw = mix water
1.0 density.

The PVF can be maximized by using coarse, medium, ( )

ρ slurry = ρ b × fsV + 1 – fsV ρ mw . (7-15)
and fine particles in specific volumetric ratios. As shown
in Fig. 7-14, the fine particles fit in the void space In addition to affecting slurry density, the SVF influ-
between the medium-size particles, and the medium- ences the cement-slurry and set-cement properties in
size particles fit in the void space between the coarse the following ways.
particles. For two consecutive granulometric classes, the
■ Thickening time (the lower the SVF, the longer the
order of magnitude between the mean particle diameter
(d50) of each class should ideally be between 7 and 10. thickening time)
In such cases, the PVF increases to 0.95. The PVF cal- ■ Compressive strength (the higher the SVF, the higher

culation should also take the particle shapes into the compressive strength)
account, but this parameter is not considered in most ■ Stability (the higher the SVF, the better the slurry
mathematical models. stability)
The solid volume fraction (SVF) is the ratio between ■ Fluid loss (the higher the SVF, the lower the fluid-loss
the volume of solids in a slurry and the total slurry rate)
volume (solids + mix water).
■ Rheology (the higher the SVF, the higher the slurry
viscosity)

7-9.3 EPS slurry design

The SVF of a neat, 15.8-lbm/gal [1,897 kg/m3] Class G
slurry is 0.41 (or, φ = 0.59). Maximizing the PVF of the
solids reduces the amount of water required to prepare
a pumpable slurry. In most cases the optimal SVF in
EPS slurries is 0.55 to 0.60 (or, φ from 0.40 to 0.45).
Within such a narrow SVF range, EPS slurry design relies
on only two parameters: the particle-size distribution
and the specific gravities of the blend components. As
a result, EPS cement systems have several advantages
over conventional cement systems.
■ The slurry rheology is significantly less dependent on
the slurry density.
Fig. 7-14. Illustration of PVF concept. ■ The compressive strength and permeability of the set
cement are significantly less dependent on the slurry
density.

252 Well Cementing

■ The slurry is more stable. Table 7-16. Slurry Properties of EPS Cement Systems†, ‡
■ Better fluid-loss control is achieved. Density (lbm/gal) 20 8.7

Porosity (%) 42 40
7-9.3.1 Slurry density Yield point (lbf/100 ft2)
To achieve a given slurry density, engineers choose a blend
After mixing 15 9
of solids with a given specific gravity (Eq. 7-16).
At 185°F 9 14

ρb =
⎡ ρ
⎣ ( ) ⎤
slurry – φρ mw ⎦
. (7-16)
Plastic viscosity (cp)

1– φ After mixing 255 213

where At 185°F 181 163

ρb = blend density. 10-min gel strength (lbf/100 ft2) 19 32

Table 7-15 presents some common cement-slurry 1-min gel strength (lbf/100 ft2) 11 20
solids with various specific gravities, organized according
to their position in the trimodal-particle-size scheme. API/ISO free fluid (mL/250 mL) 0 0
Assuming that the particles in the blend are spheri- Fluid-loss rate (mL/30 min at BHCT †) 28 12
cal, it is possible to achieve PVFs as high as 0.88. The † At 8.7 and 20 lbm/gal [1,044 to 2,400 kg/m3] slurry densities
available selection of particles allows one to prepare ‡ Bottomhole circulating temperature
stable EPS slurries at densities from 8.0 to 23.0 lbm/gal
[960 to 2,761 kg/m3].

7-9.3.3 Slurry stability: free water and segregation

Table 7-15. Cementing Materials with Various Particle Free water and segregation arise from particle interaction
Sizes and Specific Gravities in the cement slurry (Chapter 3; Appendix B). Often, to
Fine Medium Coarse achieve a certain slurry density, coarse particles with a
low specific gravity (e.g., ceramic microspheres) or a very
Light (< 2 sg) – Glass Cenospheres
bubbles Organic polymer high specific gravity (e.g., hematite) must be used. Such
particles particles can be very difficult to stabilize in conventional
cement slurries. As shown by Stokes law (Eq. 7-17), the
Medium Microsilica Cement CaCO3 settling velocity of a particle is more dependent on its
(> 2 sg < 3 sg) Silica flour Silica sand
size than its specific gravity.
Heavy (> 3 sg) Manganese Hematite Ilmenite
tetraoxide
(
⎡ g × ρ – ρ × d2 ⎤ )
v= ⎣ ⎦,
L
(7-17)
18μ L

7-9.3.2 Slurry rheology where

The rheology of a concentrated suspension depends on d = particle diameter
the packing behavior of the particles. At PVFs as high as g = acceleration of gravity
0.88, the resulting slurries have low plastic viscosities v = settling velocity
across the slurry density range (Table 7-16).
For a given SVF, τ y decreases as the concentration μL = viscosity of liquid medium
of dispersant increases. This is also true for conventional ρ = specific gravity of particle
slurries (Chapter 3). Using a dispersant is important ρL = specific gravity of liquid medium.
in EPS slurries to prevent agglomeration of the fine
particles. The τ y of most EPS slurries is usually less than For example, the specific gravities of hematite and
20 lbf/100 ft2. silica sand are 4.8 and 2.6. According to Stokes law, for a
given particle size, the hematite particle would settle
about twice as fast as the silica particle. However, for
a given particle density, if particle size is increased to
500 μm from 1 μm, the settling rate increases by a factor
of 250,000.

Chapter 7 Special Cement Systems 253

 Stokes law considers a single particle that can fall at very low slurry densities, EPS slurries develop more
unobstructed through the liquid medium. A cement slurry compressive strength than foamed cements. The lower
is a concentrated suspension that tends to behave as a permeability of EPS slurries, combined with the presence
porous solid that may shrink under the force of gravity. of inert fillers in the blend, also improves the chemical
Such behavior is also called hindered settling, and this resistance of the set cement. Figure 7-17 illustrates the
effect increases with the particle concentration or SVF. resistance of various cement systems to mud acid (a
As the SVF approaches 60% in EPS slurries, the hindered mixture of HCl and HF). The EPS blends are signifi-
settling effect is more pronounced compared to con- cantly more resistant, particularly when latex is present
ventional slurries; consequently, it is easier to prepare a in the formulation.
stable system.
In concentrated slurries of uniformly sized spheres,
the settling velocity is given by the following equation. 3,000
Bentonite extended

( ) n 2,500 Sodium silicate extended

vs = v 1 – fsV , (7-18) Cenosphere extended
2,000 EPS blend
Compressive
where strength 1,500
n = an experimentally determined value (usually 5) (psi) 1,000
(Richardson and Zaki, 1954) 500
v = single particle settling velocity
0
vs = settling velocity in concentrated slurry. Cement systems

Fig. 7-15. 24-hr compressive strengths of 12.0-lbm/gal [1,440 kg/m3]

7-9.3.4 Additives cement systems.
Although EPS cement systems have a high solids
content, most of the particles are chemically inert.
Compared to conventional cement systems, the cement 3,000
Foamed cement
concentration is very low—between 20 and 50% by 2,500 EPS blend
volume of blend. Consequently, the amounts of retarders
2,000
and dispersants in the cement slurry are also lower. Owing Compressive
to the lower amount of water in EPS systems and the strength 1,500
ability of solids to achieve maximum compactness, the (psi) 1,000
amount of fluid-loss-control agent required to achieve a
500
given fluid-loss rate is roughly half that of conventional
slurries (Baret et al., 1996; Villar et al., 2000). 0
8 9 10
Density (lbm/gal)
7-9.4 Performance of EPS cement systems
Fig. 7-16. 24-hr compressive strengths foamed cements and EPS
The slurry porosity reduction allowed by blend optimi- cements at various densities.
zation has a beneficial effect on the properties of the
set material, particularly at the upper and lower ranges
of the slurry-density scale (Moulin et al., 1997). 25
Conventional: 15.8 lbm/gal
20 Latex-modified: 15.8 lbm/gal
7-9.4.1 Low-density systems EPS blend: 12.1 lbm/gal
Latex-modified EPS blend: 11.9 lbm/gal
The slurry porosity of conventional low-density cement Acid 15
systems is typically high, and such slurries often contain solubility
extenders such as bentonite and fly ash (Chapter 3). The (wt%) 10
higher water-to-cement ratio can lead to slurry-stability
problems, and the compressive strength and permeability 5
are usually compromised.
0
The reduced slurry porosity of EPS slurries ensures
Cement systems
early compressive-strength development and a higher
ultimate compressive strength (Fig. 7-15) (Revil and Fig. 7-17. Solubility of various cement systems in mud acid. Weight
Jain, 1998; Rucker et al., 2001). Figure 7-16 shows that, loss (%) after 24 hr exposure at ambient temperature.

254 Well Cementing

 Some low-density EPS systems are tailored for low- 1,200
Sodium silicate extended:
temperature applications such as deepwater wells. The 12.0 lbm/gal
1,000
reactivity is enhanced by using microcement as fine Cenosphere + 2% CaCl2:
particles. The high reactivity of the microcement signif- 15.8 lbm/gal
800 EPS blend: 12.0 lbm/gal
icantly improves compressive-strength development 24-hr
Compressive 600
(Michaux et al., 1998; Piot et al., 2001) (Fig. 7-18). strength
Similar properties have been achieved using a high-alu- (psi) 400
mina cement. Such systems develop compressive
strengths above 3,000 psi [21 MPa] within 24 hr at 40°F 200
[4°C] (Villar et al., 2000). 0
Cement systems
7-9.4.2 High-density systems Fig. 7-18. 24-hr compressive strength of various cement systems at
Conventional heavyweight slurries are prepared by 40°F [4°C].
adding weighting agents, decreasing the amount of
cement and water, and adding a dispersant to maintain
pumpability. Because a minimum amount of cement is 500
required to achieve sufficient compressive strength, Conventional slurry: 19 lbm/gal
400 Optimized blend: 19 lbm/gal
there is a limitation on the upper slurry density. In addi-
tion, such slurries tend to be unstable. Because the EPS
Friction 300
design method allows the slurry viscosity to remain low
pressure
at high SVFs, more cement can be added to blend. As a (psi/1,000 ft) 200
result, high compressive strength can be maintained
(Jain et al., 2000; Pokhriyal et al., 2001). 100
The performance of high-density EPS cement sys-
0
tems is illustrated in Figs. 7-19 and 7-20. Figure 7-19 1 3 5
compares the friction pressure that is observed when
Pump rate (bbl/min)
19.0-lbm/gal [2,280 kg/m3] conventional and EPS
cement systems are pumped through pipe at various Fig. 7-19. Friction pressure versus pump rate for high-density con-
pump rates. Less friction pressure develops when EPS ventional and EPS cement systems.
systems are pumped.

400 5,000 11,000

10,500
360 4,500 10,000
9,500
320 4,000 9,000
8,500
8,000
280 3,500
Compressive strength (psi)

7,500
Curing temperature (°F)

Curing pressure (psi)

7,000
240 3,000 6,500
6,000
200 2,500 5,500
5,000
160 2,000 4,500
4,000
120 1,500 3,500
3,000
2,500
80 1,000 2,000
1,500
40 500 1,000
500
0 0 0
0:00 5:00 10:00 15:00 20:00
Time (hh:mm)

Fig. 7-20. Compressive strength development of 23-lbm/gal [2,761 kg/m3] EPS cement system.

Chapter 7 Special Cement Systems 255

 Use of heavy particles allows EPS systems to achieve (1982) reported useful properties at densities as low
slurry densities up to 23 lbm/gal [2,761 kg/m3]. The as 3.5 lbm/gal [0.420 kg/m3]. Slaton (1981) reported the
water-to-cement ratio is still low, ensuring a rapid devel- use of foamed cement at densities as low as 5.0 lbm/gal
opment of compressive strength. [600 kg/m3] in situations in which compressive strength
and permeability were not critical.
Foamed cement has several advantages in addition to
7-10 Ultralow-density cement systems its low density:
■ relatively high compressive strength developed in a
Formations that have a low fracturing gradient or are
highly permeable, vuggy, or cavernous pose difficult reasonable time
cementing situations. Such formations are often unable ■ less damaging to water-sensitive formations (Bozich
to support the annular hydrostatic pressure exerted by a et al., 1984; Bour and Vennes, 1989)
conventional cement slurry. Some formations will not ■ lower chance of annular gas flow (Tinsley et al., 1980;
even support a column of water. The density of con- Hartog et al. 1983)
ventional cements mixed with water always exceeds ■ ability to cement past zones experiencing total
8.33 lbm/gal [1,000 kg/m3]. In reality, the density of
losses.
conventional systems with acceptable properties usually
exceeds 11 lbm/gal [1,320 kg/m3]. Therefore, there are Also, because the gas has little effect on placement
situations in which it is impossible to perform a suc- properties such as thickening time, the system density can
cessful cement job with a conventional slurry. be adjusted during the cement job by simply changing
Ultralow-density cements provide a solution to such the gas concentration.
problems. In general, the ultralow-density category refers The low density of foamed cements reduces losses
to systems with densities less than about 10 lbm/gal to potential producing zones, and increased well pro-
[1,200 kg/m3]. ductivity may result (Colavecchio and Adamiak, 1987).
More recent applications of foamed cements include
controlling shallow flows below the mudline in deepwater
7-10.1 Microsphere and EPS systems wells, deterring compaction damage in soft formations,
Hollow glass microspheres and cenospheres are frequently and resisting damage from external stresses placed on
used as extenders to prepare ultralow-density cement the cement sheath.
systems. In addition, ultralow-density EPS systems that
incorporate microspheres and cenospheres are commonly
used. Such systems are described in detail elsewhere 7-10.2.1 Foam stability and structure
in this textbook (Chapters 3 and 10, and Section 7-9 in The stability of foamed cement is affected by the foam-
this chapter). ing agent, the quantity of gas, the chemical and physical
composition of the slurry, thermodynamic factors, and
the mixing method and conditions. Stable foams exhibit
7-10.2 Foamed cements spherical, discrete, disconnected pore structures with a
Foamed cements are coarse dispersions of a base cement clearly defined cement matrix. Unstable foams have
slurry, a gas (usually nitrogen), a foaming surfactant, nonspherical and interconnected pores, caused by the
and other materials to provide foam stability. The base rupture and coalescence of gas bubbles. Such unstable
cement slurry is usually a conventional 15–16 lbm/gal foams have a sponge-like structure and develop lower
[1,800–1,920 kg/m3] system. The density of nitrogen is, compressive strength, higher permeability, and inferior
for all practical purposes, 0 lbm/gal. Therefore, the bonding properties.
foam density is adjusted by varying the nitrogen concen- Foams are categorized by their quality (Qfoam), or
tration. Although foamed cement was first used by the the ratio of the volume occupied by the gas to the total
construction industry more than 60 years ago, its first volume of the foam (expressed as a percentage) (Eq.
application in well cementing occurred in 1979. Foamed 7-19).
cementing technology has been evolving ever since.
Foamed cements are generally less expensive than V gas
Q foam = × 100 (7-19)
systems containing glass microspheres or cenospheres V foam
(Edmondson and Benge, 1983); however, special equip-
ment is required at the wellsite to inject nitrogen or air As the foam quality varies, two structural situations
into the base slurry. Foamed cement can be mixed at occur. Concentrated foams are mostly gas phase and
lower densities than conventional microsphere systems consist of polyhedral gas cells separated by thin liquid
and yet maintain acceptable properties. Montman et al. films. Dilute foams consist of nearly spherical bubbles

256 Well Cementing

separated by thick liquid films. Foamed cement belongs The most common method to prepare foamed cement
to the second category, with a quality not exceeding 80% at the wellsite is to mix a base cement slurry with all the
and usually less than 50%. additives except the surfactants and then inject the sur-
Foamed cement is a compressible fluid; consequently, factants and the gas as the slurry is being pumped down-
owing to hydrostatic-pressure variations, the foam quality, hole (Chapter 13). In addition to the pressure at which
and therefore the density, changes as the foam circulates foam is generated, the foam-generating equipment can
in the well. Qfoam decreases and density increases as influence the stability and bubble-size distribution of
the foam moves from the surface to the bottom of the the foamed cement.
casing. As the foam moves back up the annulus, Qfoam The solids play a significant role in foam stabilization
increases and density decreases. The density can be (Davies et al., 1981). When a solid particle adheres to a
predicted as a first approximation by considering the bubble, it inhibits bubble coalescence and enhances
compressibility laws and the solubility of nitrogen in the foam stability (Sharma et al., 1982). The foam stability
base slurry. is related to the size of the particle and its wettability.
Foamed cement is a three-phase system (gas/liquid/ The manner by which solid particles are retained at a
solid), with many phenomena occurring at the inter- liquid/gas interface is analogous to the adsorption of
faces. This system is in constant evolution because of the solute molecules. In both cases, work is required to
reorganization of gas bubbles that may grow, shrink, or transfer the material off the surface and into the bulk
coalesce, and because of the chemical reactions that solution. This work of desorption is the phenomenon
occur in the base cement slurry. Foams are difficult to that confers thermodynamic stability to the foam. In the
characterize because they are shear history-depen- case of finely divided solids, Ross (1969) reported that
dent fluids, and their texture is strongly affected by the stability is improved by a reduction in the particle size.
mixing procedure. Foamed cements made under large- Also, particle size is significant because it affects the
scale field conditions, with high shear rates and high sedimentation rate.
pressure, have been found to be more stable than Thermodynamic properties also affect foam stability
foamed cements made under laboratory conditions (de Rozières and Ferrière, 1991). Some of the principal
(Davies et al., 1981). More recently, de Rozières and thermodynamic parameters are listed in Table 7-17. For
Ferrière (1991) showed that foamed cement generated more detailed discussions of foam thermodynamics, the
in the laboratory at 1,000 psi [7 MPa] is more stable and reader is referred to the following references: Davies
has a narrower bubble-size distribution than slurries and Rideal, 1963; Harris, 1985; Manev, 1974; Monsalve
generated at atmospheric pressure (Fig. 7-21). and Schechter, 1984; and Sanchez, 1987.

Atmospheric pressure 1,000 psi (7 MPa)

0.25 0.25

0.20 0.20

0.15 0.15
Incremental Incremental
volume volume
(mL/g) 0.10 (mL/g) 0.10

0.05 0.05

0.00 0.00
–3 –2 –1 0 1 2 3 –3 –2 –1 0 1 2 3
Log diameter (µm) Log diameter (µm)

Fig. 7-21. Structures and bubble-size distributions of foamed cements prepared at two pressures (from de Rozières and Ferrière, 1991). Foamed
cement prepared at 1,000 psi [7 MPa] is more stable than that prepared at atmospheric pressure. Reprinted with permission of SPE.

Chapter 7 Special Cement Systems 257

Table 7-17. Thermodynamic Parameters that Affect The duration of foam stability must be longer than
Foam Stability the setting time of the base slurry (Hengst and Tressler,
Stabilizing Phenomena Destabilizing Phenomena 1983). In addition to the base slurry, foaming agent, sta-
bilizers, and additives, one must use a gas that is inert
Laplace diffusion Gravity drainage with respect to the cement properties (e.g., nitrogen)
Plateau border suction Surfactant concentration gradient and a foam generator with sufficient energy and mixing
action (Chapter 13).
Marangoni effect Gibbs elasticity

Surface viscosity 7-10.2.2.2 Foamed cement properties

Laboratory testing of foamed cement under simulated
downhole conditions is difficult. Because of the pressure
and temperature dependence of the foam volume,
7-10.2.2 Foamed cement design curing a foamed cement at high pressure and tempera-
ture requires different equipment than that used for
Early in the development of foamed cementing, suitable conventional slurries (Appendix B). Work by de Rozières
base-slurry compositions to prepare foamed cements and Ferrière (1991) showed that varying the pressure
were limited. Today, with improved surfactants and and mixing conditions during foam preparation can
foam generating equipment, virtually any base slurry can strongly affect compressive strength and permeability.
be foamed. In addition, more sophisticated laboratory This is presumably caused by differences in the bubble-
testing equipment is available (Appendix B). size distribution of the resulting foam.
7-10.2.2.1 Cement, foaming agents, stabilizers, and additives
Stability
A stable base slurry is a prerequisite for a stable foamed The stability must be tested to ensure that the gas will
cement system. The same mechanisms that produce free not break out of the slurry. If the gas coalesces and the
fluid or solids segregation will contribute to the destabi- size of the bubbles increases, gas pockets form and rise in
lization of foamed cements. the cement column, resulting in uncemented sections or
The selection of the base slurry density depends on the channels in the well. A simple test to evaluate stability
required set-foamed-cement properties. Normal-density involves slicing a column of set foamed cement into
base slurries will lead to higher compressive strengths. wafers of equal size. The weight of each wafer should be
However, such slurries require larger volumes of gas to the same in a stable system.
achieve a given foam density; therefore, the resulting
permeabilities will be higher. Conversely, less dense base Compressive strength and permeability
slurries (prepared using lightweight filler materials)
require less gas and will produce foamed cements with When the density of conventional cement slurries is
lower permeabilities; however, the compressive strengths reduced by adding water or other extenders, the amount
will be lower. of cementitious material is diluted. Because of the large
To select suitable foamers and stabilizers for cement, density difference between water and gas, much less gas
one should consider the following criteria: is required to reduce the density by an equal amount.
Using gas results in less dilution and, therefore, less
■ safety and handling considerations
impact on cement properties. Consequently, the physical
■ compatibility properties of foamed cements are similar to those of
■ effect on the cement strength and permeability conventional lightweight cements that are 2 to 4 lbm/gal
■ stability [240 to 480 kg/m3] heavier (Tanner and Harms, 1983;
Smith et al., 1984).
■ efficiency
Compressive strength and permeability are frequently
■ cost. tested using slurries containing the same volume
The chemicals used to generate and stabilize the percent of gas that would exist in the slurry under down-
foamed cement must be effective at elevated tempera- hole conditions but cured at atmospheric pressure. De
tures and pressures and in the highly alkaline, calcium- Rozières and Ferrière (1991) showed that the results
containing aqueous phase of a cement slurry. Common can be misleading, and they developed equipment
foaming agents include ethoxylated alcohols and quater- to generate and cure foamed cement under higher pres-
nary ammonium salts of fatty acids. Common stabilizers sures (Figs. 7-22 and 7-23).
include polyglycol ethers and sulfate salts that impart To generate foamed cement, nitrogen is injected into
some thixotropy to the base slurry (Section 7-1). the lower chamber while the base slurry is pumped
through the mixing chamber and the test-cell loop.

258 Well Cementing

 Four-way valve Nitrogen is injected until the desired test pressure is
attained. Foam quality is determined by the ratio of
slurry to the total volume in the circulating loop. The test
Manometer cell can be used to cure specimens for testing compressive
Circuit of strength, permeability, or other properties. The foamed
pressurized cement slurry can also be injected into a fluid-loss cell.
air A
Starting with normal-density Class C and Class G
B cement base slurries, de Rozières and Ferriere (1991)
prepared foamed cement and determined the compres-
C sive strengths after curing them 72 hr at 80°F [27°C].
Piston Check valve
D From each base slurry, foams with broad and narrow
bubble-size distributions were generated. The results,
Test cell shown in Table 7-18, clearly show the benefits of having
a broad bubble-size distribution. Throughout the
Nitrogen Magnetic slurry-density range, a broad bubble-size distribution
drive leads to higher compressive strength. The effect of bubble-

Table 7-18. Performance of Foamed Cements Prepared

at Various Densities and Bubble-Size Distributions†
Class C Base Slurry
Curing time of 72 hr at 27°C
Fig. 7-22. Schematic diagram of foamed cement generator (from de
Rozières and Ferrière, 1991). Reprinted with permission of SPE. Broad BSD‡ Narrow BSD
Density Compressive Permeability Compressive Permeability
(g/cm3) Strength (mD) Strength (mD)
(MPa) (MPa)
1.77 20.8 0.050 – –

1.46 14.0 0.0076 8.2 1.25

Mold for 1.22 9.4 0.017 6.8 5.7

cylindrical
sample 1.08 7.4 0.054 5.7 25.7

0.90 4.7 89 3.2 1.99

0.61 1.5 58,890 1.6 6,800

Class G Base Slurry

Three molds Curing time of 72 hr at 27°C

for cubes Broad BSD Narrow BSD
Density Compressive Permeability Compressive Permeability
(g/cm3) Strength (mD) Strength (mD)
(MPa) (MPa)
1.80 33.9 0.025 – –

1.45 13.6 0.039 10.2 3

1.27 10.0 0.160 6.9 14.5

1.12 6.3 44 5.3 173

Fig. 7-23. Schematic diagram of test cell to cure foamed cements
(from de Rozières and Ferrière, 1991). Reprinted with permission of 0.86 3.2 12,022 2.8 676
SPE.
0.61 2.7 5,321 1.9 3,730
† From de Rozières and Ferrière, 1991. Reprinted with permission of SPE.
‡ Bubble-size distribution

Chapter 7 Special Cement Systems 259

size distribution is less clear. At densities above about bond to compressive strength is higher for foamed
1.0 g/cm3, foams with broad bubble-size distributions have cements and increases with nitrogen concentration.
lower permeabilities. At densities below about 1.0 g/cm3, These data are presented in Table 7-19.
foams with a narrow bubble-size distribution have
lower permeabilities. Thickening time
Chekiri (1978) reported that perforation of foamed Among the tests performed on foamed cement, thicken-
cements with qualities above 40% tends to cause exces- ing time is the most difficult to perform and the least
sive fracturing. As a rule, the permeability also increases conclusive. To be valid, this test should be performed
dramatically when the quality exceeds 40%; however, under simulated downhole conditions, and the foam
this depends on the additives, type of foamer, and curing should be mixed in a manner comparable to what occurs
conditions (Aldrich and Mitchell, 1975; Smith et al., on location. Thus, ideally, the slurry should be prepared
1984). in a pressurized mixer and transferred under pressure to
the pressurized consistometer. The thickening time test
Mechanical properties involves measuring the evolution of slurry viscosity
Foamed cement has a lower Young’s modulus than (Appendix B). Because of the particular rheological
conventional cements (Deeg et al., 1999). To achieve a behavior of foam, the shear field in the consistometer is
lower Young’s modulus with conventional cements, one not uniform. A large part of the foam remains static,
must add large amounts of water, resulting in lower com- while the small amount that is sheared is destabilized.
pressive strength. With foamed cement, the impact on Calorimetry experiments performed under static
compressive strength is lower. Cements with lower conditions at atmospheric pressure showed that the
Young’s moduli are less susceptible to failure when foam quality does not influence the hydration kinetics
exposed to the common mechanical stresses associated (de Rozières and Ferrière, 1991) (Fig. 7-24). A calori-
with well operations. A detailed discussion of the metric thermogram is not equivalent to the thickening
mechanical properties of well cements is presented in time. These results only demonstrate that the cement
Chapter 8. hydration process is not affected by the presence of gas
in the system.
Shear bond Instead of testing foamed systems, a common proce-
Davies et al. (1981) reported that foamed cement can dure is to measure the thickening time of the base slurry
undergo a bulk expansion before setting. In some situa- containing the additives, surfactants, and stabilizers.
tions, this can result in improved bonding (Slaton, This method gives a reasonable estimate of the working
1981). This hypothesis is supported by indirect evidence time for the foamed slurry (Davies et al., 1981; McElfresh
from the improved bond logs obtained from wells and Boncan, 1982).
cemented with foamed cement and can be explained as
an effect of pressure maintenance by the compressed gas Fluid loss
in the cement. As the cement loses hydrostatic pressure Introducing a gas to a liquid medium significantly
during gelation, the gas pressure maintains tight contact reduces the rate at which the liquid will flow through
between the cement and the casing or formation. porous media (Anderson, 1975). De Rozières and
Smith et al. (1984) reported that foamed cement at Ferrière (1991) evaluated foamed cements, with and
7.9 lbm/gal [948 kg/m3] develops higher shear-bond without fluid-loss additives, and found that the fluid-loss
strength than a 12% bentonite cement at 12.7 lbm/gal rates were lower when gas was present (Fig. 7-25).
[1,520 kg/m3]. They also found that the ratio of shear

Table 7-19. Compressive Strength and Shear-Bond Strength of Conventional and Foamed Cements†, ‡
Composition Density Compressive Strength Shear Bond Strength Ratio of Shear to
(lbm/gal [g/cm3]) (psi [MPa]) (psi [MPa]) Compressive Strength (%)
Class G 15.8 [1.90] 4,200 [29.0] 403 [2.8] 9.6

Class G + 12% bentonite 12.7 [1.53] 772 [5.0] 70 [0.5] 9.5

40% gas 9.5 [1.14] 873 [6.0] 118 [0.8] 13.5

50% gas 7.9 [0.95] 571 [3.9] 97 [0.7] 17.1

† From Smith et al. (1984). Reprinted with permission of World Oil.
‡ Cured 24 hr at 80°F (27°C), then 24 hr at 176°F (80°C)

260 Well Cementing

 56 200
Base slurry
52 Foam quality = 18%
Foam quality = 27% 160
48 Foam quality = 50%
Foam quality = 71%
44 120
Temperature Fluid-loss
(°C) 40 volume
(mL) 80
36
40 Base slurry
32 Foam 20%
Foam 37%
28 Foam 50%
0
0 10 20 30 40 50 60 70 80
0 5 10 15 20 25 30
Time (hr)
Time (min)
Fig. 7-24. Effect of foam quality on setting kinetics of foamed
Fig. 7-25. Effect of foam quality on fluid-loss behavior of foamed
cements (from de Rozières and Ferrière, 1991). Reprinted with per-
cements (from de Rozières and Ferrière, 1991). Reprinted with per-
mission of SPE.
mission of SPE.

Thermal and electrical conductivity Table 7-20. Resistivity of Foamed and Conventional
Short et al. (1961) reported that foams have lower Cements†
thermal conductivity, because of the presence of gas Cement type Density Specific Resistivity‡
voids and the lower amount of solids. Nelson (1986) (lbm/gal [g/cm3]) (ohm-cm)
reported that the thermal conductivity of cement Conventional 15.7 [1.88] 1.25 × 104
systems is roughly proportional to slurry density, regard-
less of whether the cement was foamed. These data are Foamed 9.3 [1.11] 1.4 × 10 4
presented in Fig. 7-26. † From Smith et al. (1984). Reprinted with permission from World Oil.
Studies of the resistivity of foamed cement indicate ‡ ASTM D-257

that the electrical conductivity is similar to that of

conventional cements (Smith et al., 1984) (Table 7-20).

Cement density (lbm/gal)

0.9 5.8 6.7 7.5 8.3 9.1 10.0 10.8 11.6 12.5 13.3 14.1 15.0 15.8

0.8

0.7

Thermal 0.6
conductivity Conventional
0.5 low-density
⎛ BTU ⎞
⎜⎝ hr × ft × °F ⎟⎠ systems
0.4
Foamed cements and
0.3 microsphere systems

0.2

0.1
0.6 0.7 0.8 0.9 1.0 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9 2.0
Cement density (g/cm3)

Fig. 7-26. Cement density/thermal conductivity relationship (from de Rozières and Ferrière, 1991). Reprinted with permission of SPE.

Chapter 7 Special Cement Systems 261

Rheology high fluid pressures and temperatures inside the casing
The rheological behavior of foams is unlike that of other during testing, perforating, hydraulic fracturing, or fluid
fluids (Chapter 4). The differences arise from many production (Bosma et al., 2000). Another stress condi-
factors. Foams are compressible fluids; they are hetero- tion results from exceedingly high pressures that occur
geneous and have variable properties under shear. Foams inside the cement sheath because of thermal expansion
are dynamically unstable, shear-history-dependent fluids of the interstitial fluid. A third condition involves
in which the bubble structure is continuously destroyed tectonic movement of the formation. When such stresses
and rebuilt. For these reasons, rotational viscometers are exerted on set cement in the wellbore, the set
with a fixed amount of sample are unsuitable (Heller and cement can fail in the form of radial circumferential
Kuntamukkula, 1987). The rotational shear disturbs the cracking of the cement matrix or by a breakdown of
bubble network, often resulting in foam collapse. the cement/casing or cement/formation bonds. Such
Continuous flow-tube viscometers are more suitable failures compromise zonal isolation and can lead to
for testing foams, despite the fact that foams are com- severe well problems. Thus, the well cementing industry
pressible, are non-Newtonian, and will never attain a has recognized the need for highly resilient and flexible
steady state (Reidenbach et al., 1986; Harris and well-cement compositions that can withstand the
Reidenbach, 1987; Mueller et al., 1990). Viscosity, den- stresses outlined above.
sity, and flow rate will vary continuously as the system As explained in Chapter 8, a detailed analysis of the
pressure changes along the tube. As a result, the mechanisms leading to failure of the cement sheath
equations to calculate shear stress and shear rate at showed that the risk of rupture is directly linked to the
the wall require corrections (Harris and Reidenbach, tensile strength of the set cement and is attenuated
1987). In addition, low-pressure viscosity measurements when the ratio of tensile strength to Young’s modulus is
of foamed cements may not be representative of field increased. Young’s modulus characterizes the flexibility
conditions. Such problems are minimized if tests are of a material. Thus, to increase the tensile strength to
performed at the pressures encountered in the well and Young’s modulus ratio, the set cement should have a low
at low differential pressure. Young’s modulus (Thiercelin et al., 1997).
While studying foamed fracturing fluids, Harris and
Reidenbach (1987) found that the bubbles tend to reach
a small, uniform size at high energy levels. At low energy
7-11.1 Slurry density reduction
levels, the bubbles coalesce and drain, forming large Conventional slurry density reduction, using extenders
nonuniform bubbles. Harris and Reidenbach also that accommodate additional water (e.g., sodium silicate
observed that the frictional pressure drop for foamed and bentonite), increases the flexibility of the set cement
fluids can increase twofold from the time a foam is (Table 7-21). Although the flexibility is increased, the
created until equilibrium is reached. effects on compressive strength and permeability are
To the best of the authors’ knowledge, no routine rhe- detrimental.
ological measurement is made on foamed cement slur-
ries. However, visual observations of foamed cements 7-11.2 Flexible particles
flowing in plastic tubes have shown that, at low flow
rates, the foams flow as a rigid plug, moving on a thin Another possibility to modify the elasticity of the set
film of water next to the wall (Princen, 1982). cement is to incorporate flexible particles in the slurry
design. In the construction industry, incorporating
ground particles from rubber tires in concrete improves
toughness, durability, and resilience (Eldin et al., 1993).
7-11 Flexible cements The rubber aggregates are used in highway construction
As discussed in Chapter 8, the mechanical properties for antishock properties, as a sound-absorbing material in
of well cements have become a topic of considerable antinoise walls, and in earthquake-resistant buildings.
interest. Today, in addition to the traditional unconfined- Adding ground rubber particles to well cements has
compressive-strength measurement, tensile strength, been practiced for more than 30 years as a method to
Young’s modulus, and Poisson’s ratio are frequently improve resistance to shocks during perforating (F.E.
considered during the cement-design process. The Hook, unpublished data, 1971). More recently, improve-
methods employed to measure these parameters are ments have been made with optimized rubber-particle-
discussed in detail in Appendix B. size distributions. Karimov et al. (1985) proposed using
During the life of a well, the set cement can fail particles in the 4/20 mesh range to improve impact
because of shear and compressional stresses. There are and bending strength. Brothers and de Blanc (1998) pro-
several stress conditions associated with cement-sheath posed adding ground-rubber-tire particles in the 10/20
failures. One such condition is the result of relatively mesh to 20/30 mesh size range. The set cements

262 Well Cementing

Table 7-21. Mechanical Properties of Conventionally Extended Set Cements†
Slurry Density Tensile Strength, TS Young’s Modulus, E TS/E Compressive Strength Water Permeability, k
(lbm/gal [kg/m3]) (psi [MPa]) (psi [MPa]) (× 1,000) (psi [MPa]) (mD)
15.8 [1,897] 605 [4.23] 944,000 [6,600] 0.64 51,230 [36.6] 0.001

14.0 [1,680] 479 [3.35] 538,000 [3,759] 0.90 3,270 [22.9] 0.008

12.0 [1,440] 86 [0.60] 72,100 [504] 1.19 450 [3.21] 0.138

† After Le Roy-Delage et al. (2000). Reprinted with permission of SPE.

exhibit improved elasticity, ductility, and expansion Table 7-22. Mechanical Properties of Set Cements
properties over a wide density range—14.3 to 19.0 lbm/ Containing Rubber Particles†
gal [1,720 to 2,280 kg/m3]. Le Roy-Delage et al. (2000) Slurry Tensile Young’s TS/E Compressive
described cement systems containing from 30% to 100% Density Strength, Modulus, E (× 1,000) Strength
BWOC rubber particles in the 40/60 mesh range. As (lbm/gal TS (psi [MPa]) (psi [MPa])
shown in Table 7-22, the resulting set cements are [kg/m3]) (psi [MPa])
more flexible. 12.0 [1,440] 93 [0.65] 73,900 [516.7] 1.26 412 [2.88]
EPS cement systems (Section 7-9) have been formu-
13.1 [1,570] 140 [0.99] 137,800 [963.6] 1.04 819 [5.73]
lated using flexible particles as one of the components
(Le Roy-Delage et al., 2000). The particles have the 13.6 [1,630] 207 [1.45] 189,000 [1,320] 1.11 1,130 [7.93]
following characteristics:
14.1 [1,690] 237 [1.66] 240,000 [1,678] 1.01 1,782 [12.46]
■ Particle size: ≤500 μm

■ Young’s modulus: <5,000 MPa, preferably <2,000 MPa 15.2 [1,820] 390 [2.73] 460,900 [3,223] 0.86 2,890 [20.21]
■ Poisson’s ratio: >0.3. 16.4 [1,970] 323 [2.26] 864,000 [6,042] 0.76 3,934 [27.51]
† After Le Roy-Delage et al. (2000). Reprinted with permission of SPE.
Thermoplastics, like polyamide, polypropylene, and
polyethylene, or polymers, like styrene divinylbenzene
or styrene butadiene, are compatible with these perfor-
mance criteria. Because the specific gravities of these
materials fall between 0.9 and 1.2, they can also reduce the
cement system density. Table 7-23 presents mechanical-
properties data from set cements containing various
amounts of flexible particles.

Table 7-23. Mechanical Properties of EPS Set Cements Containing Flexible Particles†
Flexible Particle Slurry Density Tensile Strength, Young’s Modulus, E TS/E Compressive Strength,
(volume %) (lbm/gal [kg/m3]) TS (psi [MPa]) (psi [MPa]) (× 1,000) CS (psi [MPa])
Styrene divinylbenzene (25) 14.0 [1,680] 365 [2.52] 521,400 [3,595] 0.72 4,860 [33.5]

Styrene divinylbenzene (30) 12.1 [1,450] 160 [1.10] 194,200 [1,339] 0.84 1,930 [13.3]

Polyamide (25) 14.0 [1,680] 406 [2.80] 374,200 [2,580] 1.09 4,050 [27.9]

Polypropylene (19) 14.0 [1,680] 329 [2.27] 426,600 [2,941] 0.78 3,130 [21.6]

Polypropylene (24) 13.7 [1,640] 381 [2.63] 438,000 [3,020] 0.88 3,810 [26.3]

Styrene butadiene (25) 14.2 [1,700] 299 [2.06] 302,400 [2,085] 1.00 2,100 [14.5]

Polyethylene (25) 13.6 [1,630] 306 [2.11] 299,800 [2,067] 1.04 3,320 [22.9]
† After Le Roy-Delage et al. (2000). Reprinted with permission of SPE.

Chapter 7 Special Cement Systems 263

7-11.3 Elastomeric composites 1,600
Neat slurry
Onan et al. (1993) described a composite material based 1,400 Microribbon slurry
on styrene-butadiene rubber latex. Various fillers, such
as carbon black, calcium carbonate, silica, and quick- 1,200
setting Portland cement can be added to adjust the
mechanical properties of the composite. This system has 1,000
been used to cement CO2-injection wells and multilat- Force 800
eral junctions (Xenaxis et al., 1997). Typical mechanical (N)
performance data are shown in Table 7-24. 600

400
7-11.4 Fibers
Adding fibers or ribbons to a cement matrix also 200
improves flexural strength. Nylon fibers have been used 0
for many years for this purpose (Chapter 3). More 0 0.2 0.4 0.6 0.8 1
recently, Le Roy-Delage et al. (2000) and Baret et al. Deflection (mm)
(2002) described the addition of metallic microribbons
to improve impact resistance, toughness, and tensile Fig. 7-27. Typical load-deflection curve comparing neat and micro-
strength. The principal applications of this system are ribbon cement systems.
kickoff plugs and multilateral junctions.
The performance of the metallic ribbon system is
illustrated in Fig. 7-27, which shows a load-deflection
curve recorded during a flexural test. The plot shows the
amount of force required to bend or deflect the set-
cement sample a given distance. The results show that
a neat slurry fails completely after being deflected
less than 0.1 mm. The microribbon slurry was able to
bear a load after a deflection of nearly 1 mm.

Table 7-24. Mechanical Properties of Elastomeric Composite Systems†

Elastomeric Composite Slurry Density Tensile Strength, Young’s Modulus, TS/E Compressive Strength, Air Permeability,
(%BWOC) (lbm/gal) TS (psi) E (psi) (× 1,000) CS (MPa) k (mD)
Neat 15.7 200 700 0.28 –‡ 0.383

67% thermatomic carbon 14.0 720 4,100 0.18 – 0.014

80% CaCO3 12.0 910 5,200 0.18 – 0.029

130% CaCO3 14.0 1,080 10,000 0.11 – 0.007

300% silica 12.1 110 5,300 0.02 105 0.012

150% silica flour, 14.0 210 8,200 0.03 245 0.001

25% quick-setting cement

67% quick-setting cement 14.6 180 9,100 0.02 505 –

77% quick-setting cement 14.0 470 23,000 0.02 985 –

113% quick-setting cement 14.0 215 200,000 <0.01 732 –

150% Class A cement 14.7 150 25,000 0.01 795 –

† After Onan et al. (1993). Reprinted with permission of SPE.
‡ Not measured

264 Well Cementing

7-12 Microfine cements 6
Microfine cements are composed of very small particles
(generally 4 to 15 μm). Therefore, the surface area is 5
very large (500 to greater than 1,000 m2/kg). The most 140°F
4 160°F
common microfine cements are very fine Portland
cements (Ewert et al., 1991; Bensted and Barnes, 2002); 116°F
however, slag cements may also be incorporated (Clarke Retarder 3
(% BWOC) 155°F
and McNally, 1993; Section 7-8). The advantage of such
cements is their improved ability to penetrate and flow 2
through tight spaces and porous media. The Portland 102°F
cement–based microfine cements generally behave sim- 1
ilarly to their conventional counterparts; however, owing
to their higher reactivity, additional gypsum or retarders 0
may be necessary to achieve predictable rheological per- 0 100 200 300 400
formance and setting. Thickening time (min)
The principal uses for microfine cements include the
following. Fig. 7-28. Thickening time of acid-soluble cement versus borate
retarder concentration. Reprinted with permission of SPE.
■ Squeeze cementing (Heathman and East, 1992)
(Chapter 14; Section 7-8.2.1)
■ Sealing casing leaks (Meek and Harris, 1991;
cements can be pumped similarly to conventional well
Lizak et al., 1992) (Chapter 14) cements. Once placed, the set cement can be readily
■ Cementing permafrost zones (Section 7-4) removed by acidizing with HCl.
■ Cementing shallow strings in deepwater wells In addition to solving lost circulation problems, acid-
soluble cements have been used as diverters to provide
More detailed discussions of the applications of micro- temporary zonal isolation and as kickoff plugs in weak
fine cements are found in the sections indicated above. formations (Chapter 14).

7-13 Acid-soluble cements 7-14 Chemically bonded phosphate

As discussed in Chapter 6, lost circulation during drilling ceramics
and cementing is a common problem. Most common
remedies, such as lost-circulation materials (e.g., flakes Chemically bonded phosphate ceramics (CBPCs) are
and fibers) and thixotropic cements, remain in the thief binders that fall between sintered ceramics (e.g., pot-
zone permanently. This is not a problem unless the thief tery, porcelain) and chemically bonded systems (e.g.,
zone is also the producing zone. Such situations include Portland cement). CBPCs are formed by acid-base reac-
gas-injection and gas-storage wells. Therefore, there is a tions between an acid phosphate (e.g., Mg, Ca, or Al) and
need for lost circulation materials that can easily be a metal oxide (e.g., MgO, CaO, or ZnO2) (Jeong and
removed after well construction is complete. Wagh, 2003).
Such a temporary system, based on magnesium oxy- Calcium aluminate–based CBPCs are being used
chloride (or Sorel) cement, was developed by Sweatman as cements for geothermal wells (Chapter 10).
and Scoggins (1990) and Vinson et al. (1992). Magnesium Magnesium potassium phosphates, originally devel-
oxychloride is made by mixing powdered magnesium oped to stabilize and encapsulate radioactive and haz-
oxide with a concentrated solution of magnesium chloride. ardous waste streams, have also found use as fast-setting
The principal binder phases are Mg3(OH)5Cl • 4H2O and well cements (Wagh and Brown, 1999) The cement
Mg2(OH)3Cl • 4H2O. The set-cement is very strong; how- is prepared by calcining MgO at 2,372°F [1,300°C].
ever, it readily dissolves when exposed to acid. The calcined MgO is then reacted with a KH2PO4 solu-
A modified Sorel cement for oilfield applications tion to form the binder.
consists of magnesium and calcium oxides, carbonates, MgO + KH2PO 4 + 5H2 O ⎯→
⎯ MgKPO 4 + 6H2 O
and sulfates. When added to seawater or a chloride brine,
magnesium oxychloride hydrates form. The thickening (7-20)
behavior can be modified by common retarders such as
borax (Fig. 7-28) and accelerators such as calcium chlo-
ride (Bensted and Barnes, 2002); therefore, such

Chapter 7 Special Cement Systems 265

 When the MgO is added to the KH2PO4 solution, the liquid cement premixes that can be stored in the liquid
resulting paste sets within 1 hr at ambient temperature. state almost indefinitely and activated at the wellsite
The system can be retarded by adding boric acid. The during the cement job. This technique can lead to
boric acid reacts with the acid phosphate and forms a tem- improvements in job efficiency and service quality in
porary coating of magnesium boron phosphate on the sur- both primary and remedial cementing operations (Rae
face of the MgO particles. Typical performance of a system and Johnston, 1996a and 1996b).
containing boric acid retarder is shown in Fig. 7-29. The storable slurries can be made from Portland
The performance is notable in that the ultimate com- cement, slag cement, or other blended cements, and
pressive strength is much higher than that of set can also contain other cement additives. Water is the
Portland cement. carrier fluid. A set retarder, such as a hydroxycarboxylic
acid, is necessary to prevent the setting of the slurry
during storage. To prevent settling of the cement solids
7-15 Storable cement slurries during storage, suspending agents such as polymers and
clays are added. At the wellsite, the storable slurry is
As described in Chapter 11, cement-slurry mixing in the
pumped to the cement mixer, where an activator is
well cementing industry essentially consists of combining
added to counteract the retarder and reinitiate cement
fine cement powder with mix water. The process
hydration. The preferred activator is sodium silicate.
involves cutting sacks of cement or transferring bulk
cement pneumatically to the mixing unit. Many parameters
must be controlled at the mixing unit, including slurry den-
sity, mixing energy, and mixing time. Density is arguably 7-16 Summary
the most difficult parameter to control. Key cement-slurry Table 7-25 summarizes the cement systems presented in
properties such as thickening time, rheology, free water, this chapter, indicating their principal uses, chemical
and compressive strength depend upon density. compositions, and mechanisms of action.
A cementing operation can be simplified by elimi-
nating the handling of dry powders. This concept was
realized by Rae and Johnston (1995), who introduced

160 13,000
140 12,000
120
11,000
100
Setting Compressive 10,000
time 80 strength
(min) (psi) 9,000
60
8,000
40
20 7,000

0 6,000
30 40 50 60 70 80 90 0 10 20 30 40 50
Temperature (°C) Time (hr)

Fig. 7-29. Setting characteristics of a magnesium potassium phosphate cement system (Wagh and Brown, 1999).

266 Well Cementing

Table 7-25. Cement System Summary
Cement System Principal Uses Chemical Compositions Mechanisms of Action
Thixotropic cements Lost circulation prevention Portland cements containing one
Slurry fallback prevention of the following additives
Gas migration prevention
Bentonite Increased gel strength

Calcium sulfate hemihydrate Formation of ettringite crystals

Aluminum sulfate/iron sulfate

Crosslinked cellulose polymer Increased gel strength

Expansive cements Improved cement/casing and Commercial expanding cements or Portland Formation of ettringite crystals
cement/formation bond cement containing calcium sulfate hemihydrate

Salt cements Internal pressure exerted by

crystallization of salt in pores

Portland cement containing aluminum powder Generation of hydrogen gas in situ

Portland cement containing Conversion of MgO (periclase) to

calcined magnesium oxide Mg(OH)2 (brucite)

Freeze-protected cements Cementing across permafrost zones Calcium aluminate cement Rapid strength development at
low temperatures

Gypsum-Portland cement blends Rapid strength development at low

temperatures; lower heat of hydration
than calcium aluminate

Ultrafine Portland cement High surface area increases hydration rate

Salt cement systems Cementing across salt zones Portland cements containing sodium Systems do not disturb salt-bearing
or sensitive formations chloride or potassium chloride at formations
concentrations up to saturation Systems do not disturb sensitive clays

Latex-modified cement Improved cement/casing and Portland cements containing one

systems cement/formation bond of the following latexes
Improved fluid-loss control
Gas migration prevention Polyvinylidene chloride Film formation on bonding surfaces

Polyvinyl acetate Latex particles plug small pores in

Styrene-butadiene copolymer cement matrix

Cements for corrosive Cementing chemical waste Epoxy-based cement systems Chemically inert to strong acids and
environments disposal wells Elastomeric composites bases

Cementing CO2 injection wells Pozzolanic cement systems Reduced cement-matrix permeability,
BFS systems improved chemical resistance

BFS systems Alternative to or supplement for BFS + activator (e.g., sodium hydroxide Formation of C-S-H phase, with extensive
Portland cement or Portland cement) incorporation of Al, Mg, Fe, and sulfate
Conversion of drilling fluid to cement BFS + activator mixed with drilling fluid in structure

EPS cement systems Systems with improved Cement blends with multimodal Particle-size distribution minimizes mix-water
placement and set-cement particle-size distributions concentration required to prepare pumpable
properties over a wide slurry- slurry; set cement is less permeable than
density range conventional systems

Ultralow-density cement Cementing across formations Cements containing glass or ceramic Low slurry density reduces hydrostatic
systems with low fracture gradients or microspheres pressure in wellbore and prevents
that are vuggy or cavernous Foamed cements formation breakdown

Flexible cement systems Improved resistance to stresses Cements containing flexible particles Flexible particles decrease Young’s modulus
induced by perforating, hydraulic and increase Poisson’s ratio
fracturing, and tectonic movement

Cements containing nylon or metallic Fibers act as reinforcement and improve

fibers flexural strength and toughness

Elastomeric composites Polymer matrix filler is flexible

Microfine cement systems Squeeze cementing, sealing casing leaks Portland or BFS cements with small particle Smaller particles more readily enter cracks
size (4 to 15 μm) and high surface area and formation pores
(500 to 1,000 m2/kg)

Acid-soluble cement Temporary solution for lost circulation Magnesium oxychloride (Sorel) cements Principal binder phase is readily removed
systems by contact with a strong acid (e.g., HCl)

Chemically bonded Fast-setting cements that develop high Magnesium potassium phosphate Acid-base reaction between an acid
phosphate ceramics compressive strength phosphate and a metal oxide

Storable cement slurries Eliminates handling of dry powders Portland, BFS or blended cements During cement job, concentrated slurry
during cementing operations slurried in aqueous solution containing is diluted with mix water containing
a strong cement retarder activator (e.g., sodium silicate)

Chapter 7 Special Cement Systems 267

7-17 Acronym list
API American Petroleum Institute
BFS Blast-furnace slag
BHCT Bottomhole circulating temperature
BHST Bottomhole static temperature
BWOW By weight of water
CBPC Chemically bonded phosphate ceramics
CS Compressive strength
EPS Engineered particle size
ISO International Organization for
Standardization
PVF Packing volume fraction
SG Specific gravity
SVF Solid volume fraction
TS Tensile strength

268 Well Cementing

8-1 Introduction
Until recently, the well cementing industry focused on
one principal mechanical property of set cement—
Mechanical Properties
of Well Cements
Marc Thiercelin—Schlumberger 8
field stress, but also from any change of downhole tem-
perature or pore pressure. Cement tensile strength,
elasticity, and ductility are mechanical parameters that
unconfined uniaxial compressive strength—to qualify a may be more important to long-term durability than an
cement design. The uniaxial compressive strength is arbitrary measure of cube compressive strength
determined by a cube-crushing test (Appendix B) and is (Thiercelin et al., 1997; Bosma et al., 1999; di Lullo and
used to estimate the ability of set cement to support Rae, 2000; Ravi et al., 2002).
casing and survive the perforation process. When com- Knowledge of the appropriate cement mechanical
bined with water- or air-permeability measurements properties allows one to characterize the cement defor-
(Appendix B), one can also estimate the cement’s abil- mation under applied downhole stresses and to predict
ity to provide zonal isolation and resist attack from for- whether the cement will be able to survive the stresses.
mation fluids. In this chapter, basic rock-mechanics concepts are pre-
Indirect ultrasonic techniques using tools like the sented in the context of well cements, followed by a dis-
ultrasonic cement analyzer (UCA) (Appendix B) allow cussion of how these concepts can be applied to design
one to monitor the evolution of strength with time. With cement systems that are appropriate for the anticipated
the UCA, one can determine the curing time required downhole environment.
for a cement system to attain a given compressive
strength.
In recent years, the well cementing industry has 8-2 Basic concepts
begun to concentrate on set cements’ ability to provide The basic concepts of stress and strain are presented in
zonal isolation throughout the lifetime of the well. This this section. These concepts have been presented previ-
was triggered by the observation that, even in situations ously in the context of rock mechanics analysis
in which the cement was properly placed and provided (Thiercelin and Roegiers, 2000). For further details, the
a good initial hydraulic seal, zonal isolation disappeared reader is referred to the classical works by Love (1927),
with time (Goodwin and Crook, 1992; Jackson and Timoshenko and Goodier (1970), and Muskhelishvili
Murphey, 1993). Zonal isolation loss has been attributed (1977).
to several causes. Stress and strain have distinct meanings to the engi-
■ Gas-migration problems that were not initially neer. For example, a normal stress, σ, can be viewed as
detected a normal force (to a surface) per unit area while linear
■ Loss of cement-bond log response with time
strain, ε, is the resulting deformation.
■ Fracturing the wrong zone during a stimulation
treatment 8-2.1 Stress
■ Extreme downhole temperature or pressure changes Consider a randomly oriented plane of area Δ A, cen-
■ Chemical attack tered on a Point P within a body. When a resultant force
Δ –F acts on the body (Fig. 8-1), the stress vector σ
– at
■ Pressure migration to shallower zones
that point is defined as
Laboratory and modeling studies have shown that the
principal cause of cement-sheath damage is the stresses ⎛ ΔF⎞
σ = lim ⎜ ⎟. (8-1)
induced by varying downhole conditions (Parcevaux and Δ A→0 ⎝ Δ A ⎠
Sault, 1984; Jutten et al., 1989; Goodwin and Crook,
1992; Jackson and Murphey, 1993; Bol et al., 1997; Therefore, this quantity is expressed as a force per
Thiercelin et al., 1997; Bosma et al., 1999). The stresses unit area. In classical solid mechanics, the tensile stress
arise from not only the wellbore pressure or the rock far- is taken to be positive by convention. However, in rock

Well Cementing ■ Chapter 8 Mechanical Properties of Well Cements 269

 ΔF
y
σ
n
σn n
τ
τxy

θ
σx

ΔA τxy x
σy
Fig. 8-1. Stress vector definition.

Fig. 8-2. State of stress in two dimensions.

mechanics, compressive stress is taken to be positive
because geologic forces are usually compressive in
nature. Because either convention can be used for the shear-stress components vanish. These are called
mechanical studies of well cements, it is important to the principal planes. The normal stresses associated
clearly state the convention being used. In this chapter, with these planes are called the principal stresses. In
tensile stresses are taken to be positive. two dimensions, expressions for the principal stresses
The resultant stress, σ, can be decomposed into a can be found by setting τ = 0 in Eq. 8-3 or, because they
normal component (σ n) and a shear component (τ). are the minimum and maximum values of the normal
The shear component tends to “shear” the material in stresses, by taking the derivative of Eq. 8-2 with respect
the plane Δ A. The normal component of stress acts to the angle θ and setting it equal to zero. In either case,
perpendicular to Δ A, and the shear component acts in one obtains the following expression for the value of θ at
the plane of Δ A. which the shear stress vanishes
An infinite number of planes can be drawn through a
given point, varying the values of σ n and τ. The stress 1 ⎛ 2τ xy ⎞
θ = arctan ⎜ ⎟. (8-4)
condition, therefore, depends on the inclination of the 2 ⎝ σy −σx ⎠
plane. Consequently, for a complete description of a
stress, one must not only specify its magnitude and The two principal stress components, σ1 and σ2, are
direction, but also the direction of the surface upon
1/ 2

( ) ( ) ( )
which it acts. 1 ⎡ 2 1 2⎤
In a two-dimensional situation, if σx, σy , and τxy are σ 1 = σ x + σ y + ⎢ τ xy + σ x − σ y ⎥ (8-5a)
known (Fig. 8-2), the stress state on any plane oriented 2 ⎣ 4 ⎦
at an angle θ from σx can be expressed as follows:
and
( 2
)( )(
σ n = σ x cos θ − 2τ xy sin θ + σ y sin θ 2
) (8-2) 1/ 2

( ⎡
) ( ) ( )
1 2 1 2⎤
σ 2 = σ x + σ y − ⎢ τ xy + σ x − σ y ⎥ (8-5b)
and
2 ⎣ 4 ⎦
where θ is given by Eq. 8-4.
⎡1
( )⎤
(
τ = ⎢ σ y − σ x sin 2θ ⎥ + τ xy cos 2θ .
⎣2 ⎦
) (8-3) If one generalizes this concept to three dimensions,
it can be shown that six independent components of the
stress (three normal and three shear components) are
These expressions are obtained by writing equilib- needed to define the stress unambiguously. The stress
rium equations of forces along the σ n and τ directions, vector for any direction of Δ A can generally be found
respectively (see Fig. 8-2). Note also that the moment by writing equilibrium-of-force equations in various
equilibrium implies that τxy is equal to τyx. There are directions. There are three principal planes for which
always two perpendicular orientations of ΔA for which the shear-stress components vanish and three principal

270 Well Cementing

stresses exist. Quantities that show this kind of behavior borhood of a point. For example, compatibility equations
are known as second-order tensors, and the state of ensure that the strained body remains continuous and
stress in a body is often referred to as the stress tensor. that no cracks or material overlaps will occur.

8-2.2 Strain
8-3. Cement behavior
When a body is subjected to a stress field, the relative
positions of points within it are altered. The body A cement sample, like any material, deforms when sub-
deforms. If the new positions of the points do not corre- jected to stress. Determining a relationship between stress
spond to a translation and/or a rotation (i.e., by rigid-body and strain is an important aspect of solid mechanics. This
motion), the body is strained. This strain along an arbi- relationship is called the constitutive equation of the
trary direction can be decomposed into two components: material under consideration, and various theories have
been developed to describe it in a simplified way. The sim-
■ an elongation, defined as
plest one is the theory of elasticity, which assumes a
unique relationship between stress and strain (and that
L* − L the behavior is reversible). This theory is usually sufficient
ε = lim (8-6)
L→0 L to analyze cement failure in tension or in compression at
ambient conditions. Other theories, such as the theory of
■ and a shear strain, defined as elastoplasticity, have been developed to take into account
nonreversible behaviors that are observed in materials
γ = tan Ψ ,( ) (8-7) before failure. Significant nonreversible behavior is
observed in cements subjected to confining pressure.
where Ψ is the change of angle between the two direc-
tions that, before straining, were perpendicular (Fig. 8-3).
Consequently, strain (being either a ratio of lengths 8-3.1 Stress-strain curve
or a change of angle) is dimensionless. If one assumes Fig. 8-4 presents a typical stress-strain relationship for
that the stresses are positive in traction, a positive lon- cement. The test is carried out under constant confining
gitudinal strain, ε, corresponds to an increase in pressure and constant axial strain rate (Appendix B).
length. Just as in the case of stresses, principal strains The sample is protected from the confining fluid by an
can be defined as longitudinal strain components acting impermeable flexible jacket. Measurements include the
on planes in which the shear strains have vanished. axial stress, the axial strain, and the radial strain. When
The analogy between stress and strain analyses is a confining pressure is applied to the sample, the origin
not completely valid; equilibrium equations must be of the stress-strain plots is usually translated to remove
satisfied by the stresses and compatibility equations by the influence of the hydrostatic loading on the stress
the strains. These relationships place some restrictions and strain (i.e., the axial stress is actually the differ-
on the local variation of stress and strain in the neigh- ential σa – pcon), where σa is axial stress and pcon is the
confining pressure.

O x
F

P A Strength
C
B
P′ Axial stress
pcon (compression)
A′ E

y A

Ψ Radial strain O Axial strain

B (expansion) (shortening)

Fig. 8-4. Typical stress strain-curve during compression. The axial

B′ stress is computed from the load. The symbol pcon denotes the confin-
ing pressure and is maintained constant during the test. Compressive
Fig. 8-3. Shear strain component. The points P, A, and B are adjacent stresses are taken as positive in this figure for practicality.
points. P’, A’, and B’ represent new positions after a displacement.

Chapter 8 Mechanical Properties of Well Cements 271

 During the initial stages of loading, from Points O when a compressive (negative) force, F, is applied to its
to A, the cement becomes stiffer. This nonlinear regime ends (Fig. 8-5). According to the definitions given in the
is probably caused by the closing of pores, although previous section, the axial stress applied to the sample is
experimental artifacts, such as the misalignment of the
sample surfaces with the loading platen of the machine, F 4F
σa = = , (8-8)
could also produce such an effect. A π d2
As the load increases, the stress/strain curve
becomes linear (from A to B). This is the portion of the where A is the surface area of the end of the sample. The
stress-strain curve in which the behavior of the cement axial strain is:
is nearly elastic. If unloading occurs in this region, the
strain returns nearly to zero, often along a different L* − L
εa = , (8-9)
path. When the strain follows different paths during L
loading and unloading, it is called hysteresis and indi-
cates that some energy is dissipated during this cycle. where tension is assumed to be positive. Linear elastic-
When the cement specimen is loaded beyond Point B, ity assumes a linear and unique relationship between
the yield point, the stress-strain behavior becomes stress and strain, in which strain returns to zero when
nonlinear and large deformations will eventually occur. the material is unloaded. In the case of a uniaxial test,
If the sample is unloaded in this region, permanent this means that:
strains at zero stress are observed. The Point C is the
σa = E × εa. (8-10)
maximum load that the cement can sustain under a
given confining pressure. Cement failure (i.e., when the The coefficient of proportionality, E, is the Young’s
sample loses its integrity) occurs around this point. modulus. The higher the value of the Young’s modulus,
After this, the cement might start losing its strength the less the sample will shorten for a given load. Some
because of the development of cracks. This is called the examples of Young’s moduli are given in Table 8-1.
postfailure region. Some cements, especially those with
high porosity and/or high elasticity, may not exhibit a
maximum peak stress at high confining pressure, but
A F
will continue to carry increasing stress (i.e., continue
to harden) because of compaction. Under such test
conditions, these cements do not fail.
Another interesting cement characteristic is the vol- dL
umetric strain, defined as the volume change divided by
the original specimen volume. For the triaxial test, the L
volumetric strain is εa + 2 εr , where εa is the axial strain L*
and εr the radial strain. Nonelastic volumetric strain can
show an increase of volume (called dilatancy when
cracks are created) or a decrease in volume because of
F
compaction.
The three important regions are therefore: Fig. 8-5. Sample deformation in the axial direction.
■ the elastic region, in which deformation is reversible

■ the plastic region, in which permanent deformation

occurs Table 8-1. Young’s Moduli of Various Materials
■ the postfailure region, in which the cement has lost Material Elastic Modulus (psi [MPa])
its integrity. Aluminum 10 × 10 6 [70,000]
An important stress value is the cement strength, Copper 16 × 10 6 [110,000]
which is the peak strength measured during the test.
These various regions and critical stresses are presented Steel 30 × 10 6 [210,000]
in more detail below. Window glass 10 × 10 6 [70,000]

Oilwell cement 0.14–1.4 × 10 6 [1,000–10,000]

8-3.2 Linear elasticity
Polyethylene 14–200 × 10 3 [100–1,400]
To introduce the theory of linear elasticity, consider a
cylindrical sample of initial length L and diameter d. Rubber 0.6–11 × 10 3 [4–80]
The sample will shorten along the loading direction

272 Well Cementing

 When the specimen is compressed in one direction, it These stress-strain relations can be generalized to
will not only shorten along the loading direction, but the full three-dimensional space by
also expand in the lateral directions (Fig. 8-6). This
σx ν
effect is quantified by the introduction of another con-
stant, the Poisson’s ratio, ν, defined as the ratio of the
εx = − σ +σz ,
E E y
( )
radial strain to the axial one: σy ν
ε
εy = − σ +σz ,
E E x
( )
ν=− r , (8-11)
εa and
σz ν
with
∗ ∗
εz = −
E E x
(
σ +σy . )
d −d r −r
εr = = , (8-12) 1 1 1
d r γ xy = τ xy ; γ yz = τ yz ; γ xz = τ xz , (8-13)
G G G
where r is the sample radius.
with the shear modulus, G, given by
E
G= . (8-14)
r
F
r* 2 (1 + ν )
In isotropic linear elasticity, only two elastic con-
stants are independent. For example, and as seen above,
the shear modulus G can be written as a function of E
and ν. Similarly, the bulk modulus, which is the coeffi-
L cient of proportionality between a pressure applied to
L* the sample and the volumetric strain, is
E
K= . (8-15)
F
3 (1 − 2ν )
Fig. 8-6. Sample deformation in the radial direction. Elasticity theory can be extended to nonlinear and
anisotropic materials. A nonlinear elastic material does
not have a linear relationship between stress and strain,
The negative sign is included to ensure that the but recovers all strain when unloaded. An anisotropic
Poisson’s ratio is a positive quantity in the most common material has properties that vary in different directions.
cases (i.e., when the axial strain and radial strain are of
different signs). A Poisson’s ratio of 0.5 means that the 8-3.3 Thermoelasticity
material is incompressible (rubber for example).
Conventional cements have a Poisson’s ratio of approxi- In a downhole environment, set cement will be subjected
mately 0.15. A foamed rubber or a sponge shows a very to temperature changes. For example, cement liberates
low value of Poisson’s ratio. heat during hydration, and the temperature will change as
In experiments to determine the Young’s modulus the set cement equilibrates to the surrounding formation
and Poisson’s ratio of cements, the sample must be sat- (i.e., bottomhole circulating temperature to bottomhole
urated with water, and the strain rate must be chosen static temperature).
appropriately. These precautions prevent the generation One of the most dramatic examples is the wellbore-
of capillary pressure and excessive pore pressures, temperature increase that occurs during steam injection
which would change the sample response. The measure- or geothermal well production, in which temperature
ment can be made at various pressures to determine the fluctuations can exceed 350°F [194°C]. If the wellbore
influence of confining pressure on the elastic response. temperature increases, heat will diffuse by conduction
through the casing then through the cement sheath and
the formation rock, and the temperature will progres-
sively increase in the wellbore region (Fig. 8-7).

Chapter 8 Mechanical Properties of Well Cements 273

 250 To calculate the temperature change with time, one
uses the conduction equation:
200

150 ∂2 T ∂2 T ∂2 T ρC ∂T
Temperature + + = × , (8-18)
(°C) ∂x 2
∂y2
∂z 2 K ∂t
100

50
where K is the thermal conductivity, ρ the density, and
C the specific heat. A temperature gradient will develop
0 and, as shown in Fig. 8-7, a heat front will move with
60 65 70 75 80 85 90 time.
Distance from well axis (mm)

Fig. 8-7. Temperature profile in the cement sheath at different 8-3.4 Poroelasticity
times, following a 200°C temperature increase in the wellbore: after The pore pressure can be expected to change in the
1 min. (blue), 10 min. (red), and 100 min (green).
cement sheath during the life of the well because, just
after the hydration and the resulting water consump-
tion, the cement pore pressure drops to a very low value
Large temperature changes can affect the mechanical (Chapter 9). Once the cement has set, fluid from the
properties of set cement through chemical transforma- formation will flow into the set cement to equilibrate
tions (see Chapter 10); however, the thermal dilation of the pressure, and the set-cement pore pressure will
the steel, cement, and rock can have a greater influence. increase. A rapid increase of wellbore pressure and
All of these materials expand as the temperature temperature will also lead to a pore-pressure increase.
increases and contract as the temperature decreases. Pore fluids in cement play an important role because
Damage may occur owing to nonuniform heating. A they support a portion of the total applied stress. Thus,
portion of the material that is being heated might be only the remainder of the total stress, the effective stress
prevented from expanding, while an unheated portion component, is carried by the cement matrix (Fig. 8-8).
might be subjected to forced expansion. This generates In 1923, Van Terzaghi first introduced the effective-
stresses in the set cement that may lead to failure or stress concept for one-dimensional consolidation of
debonding. To determine the thermoelastic behavior, soils, and proposed the following relationship:
such stresses must be added to those generated by
elasticity. Set cement, rocks, and tubulars will follow the σ ′ = σ + p pore , (8-19)
same behavior, although they will differ quantitatively.
The deformation caused by thermal effects in where σ is the total applied stress, σ′ is the effective
absence of stress is given by stress governing the consolidation of the material,
and ppore is the pore pressure (note that if compressive
ε = β T, (8-16) stresses are taken to be positive, the equation is
σ′ = σ – ppore). Biot (1941, 1956) later proposed a con-
where β is the coefficient of linear thermal expansion and sistent theory to account for the coupled diffusion and
Τ is the temperature. Therefore, for the stress-strain rela- deformation processes that are observed in elastic mate-
tionship: rials. For time-independent processes, this poroelastic
material behavior is similar to that of an elastic solid
σx ν
εx −βT = − σ +σz ,
E E y
( ) when the stresses in Eq. 8-14 are replaced by effective
stresses such that
σy ν
εy −βT = − σ +σz ,
E E x
( ) σ ′ = σ + αppore . (8-20)

σ This poroelastic constant, α, varies between 0 and 1,

ν
ε z −βT = z − σx + σy ,
E E
( ) and describes the “efficiency” of the fluid pressure
in counteracting the total applied stress. Its value
and depends on the pore geometry and the physical proper-
ties of the solid.
1 Another similarity is that fluids diffuse through set
γ rz = τ . (8-17) cement in a manner similar to temperature. Assuming
G rz

274 Well Cementing

 ■ An increase of pore pressure induces cement dilation.
■ A compression of the cement produces a pore pres-
sure increase if the fluid is prevented from escaping
the porous network.
ppore When the fluid is free to move, pore-pressure diffu-
ppore
sion introduces a time-dependent character to the
mechanical response of the set cement. The set cement
ppore will react differently according to the loading rate and
the ability of the cement matrix to accommodate the
resulting pore-pressure fluctuation.
ppore σ Hence, two limiting behaviors must be introduced:
ppore drained and undrained responses. One limiting case is
realized when a load is instantaneously applied to a
ppore porous cement. In this case the excess fluid pressure has
no time to diffuse; the medium will react as if it were
undrained. Undrained values of Young’s modulus and
ppore ppore Poisson’s ratio must be used in this situation. To determine
undrained values, one performs a compression test as
shown in Fig. 8-9a, in which the pore fluid is not allowed
to escape from the sample. On the other extreme, if the
pressurization rate is sufficiently slow and the diffusion
has ample time to drain excess pressure areas, the
Fig. 8-8. The cement matrix supports only a portion of the total cement will be less stiff, or more compliant. The stiffen-
stress, σ; the other portion is supported by the pore pressure
inside the cement.
ing effect during undrained deformation is greater if the
pore is filled with a relatively incompressible liquid,
rather than a relatively compressible gas.
To determine drained elastic values (which are the
that the set cement is incompressible and the pore fluid intrinsic elastic properties of the sample), one performs
is slightly compressible, the diffusion equation is a compression test as shown on Fig. 8-9b, in which the
pore fluid is allowed to escape from the sample and the
∂2 p pore ∂2 p pore ∂2 p pore φμ c ∂ p pore loading rate is low enough to prevent the generation of
+ + = × , (8-21)
∂x 2
∂y 2
∂z 2 k ∂t excess pore pressure in the sample.

with ppore the pore pressure, φ the porosity, μ the

viscosity of the fluid, k the permeability, and c the F F
fluid compressibility.
Permeability measurements use this diffusion
pgauge
equation (Appendix B). In this case one assumes that
the fluid is incompressible.

kA( Δ p ) pore pcon pcon

q= , (8-22)
μL

where q is the flow rate, A the sample cross section, and

L the sample length.
In reality, however, solid materials are compressible.
This compressibility, which changes the porosity of the a b
solid when stress and/or pore pressure is applied, means Fig. 8-9. In test a, the fluid cannot leave the sample and the pore
that any stress change induces a pore-pressure change. pressure is measured during the test. The test is undrained and the
One cannot dissociate the pore-pressure changes from deformations measure undrained elastic constants. In test b, the
the rock deformation. Two basic mechanisms highlight fluid can escape the sample and the pore pressure is maintained
this coupled behavior (Detournay and Cheng, 1993). constant (atmospheric in this case). The test is drained and the
deformations are used to measure drained elastic constants.

Chapter 8 Mechanical Properties of Well Cements 275

8-3.5 Plastic behavior is associated with a stress increment is the sum of an
Most set cements exhibit nonreversible deformations elastic component and a nonelastic component:
after unloading, or at least a nonunique relationship dε = dε el + dε pl , (8-23)
between stress and strain. This means that set cements
are not perfectly elastic materials, and a number of the- where dε is the total-strain increment, dε el is the
ories have been developed to model such behavior. They elastic-strain increment, and dεpl is the plastic-strain
include the theories of plasticity, damage mechanics, increment. Unlike the elastic-strain component, the
and time-dependent analysis (creep). As an example, plastic-strain component cannot be recovered during
the theory of elastoplasticity will be briefly described. unloading. To predict the plastic-strain increment one
Fig. 8-10 shows the stress-strain relationship of an needs a yield criterion that indicates whether plastic
ideal cylindrical elastoplastic sample. From O to A, deformation occurs, a flow rule that describes how the
the relation between stress and strain is linear, and the plastic strain develops, and a hardening law.
slope of the curve is the Young’s modulus, E. The The yield criterion is a relationship between the
stress-strain relationship does not change if the sample stresses that is used to define the conditions under
is unloaded in this region. This is the region in which which plastic deformation occurs. In three dimensions,
the theory of elasticity applies. Beyond Point A, the this is represented by a yield function that is a function
slope of the curve decreases. Moreover, if one unloads of the state of stress and a hardening parameter:
the sample in this region, say at Point B, the unloading
portion does not follow the same path as the loading
portion but is perfectly linear with a slope E. At zero
( )
f σ 1 , σ 2 , σ 3 , M = 0. (8-24)
stress, part of the deformation has not been recovered. The hardening parameter, M, determines the evo-
This unrecovered deformation represents the plastic- lution of the yield curve with the amount of plastic
strain component in the theory of elastoplasticity. The deformation of the material. For further details on
Point A is actually the initial yield stress of the cement. elastoplasticity, the reader is referred to Hill (1951) and
During reloading, the sample behaves as a perfectly Chen and Han (1988).
elastic solid up to the Point B, which is the new yield
stress. The increase of the yield stress with the increase
of plastic strain is called strain hardening; a decreasing 8-3.6 Creep of cement
yield stress with plastic strain is also possible, and is Creep is a general term that covers the time-dependent
called strain softening. A perfectly plastic material deformations of the set-cement matrix (not to be con-
is a material with no strain hardening or softening. As fused with the time-dependent deformation caused by
shown in this example, the yield stress is a function of pore fluid flow or temperature diffusion). All set
the loading history. cements present some creep behavior. Creep has been
In elastoplasticity, part of the strain is predicted thoroughly studied for concrete to prevent the failure
by the theory of elasticity. Any strain increment that of a concrete structure after a given amount of time,
but has been poorly studied in oilwell cements even at
ambient conditions.
As with plasticity, deformation caused by creep can
be added to the instantaneous elastic deformation
B ε = ε el + ε c ( t ), (8-25)
A
where εc(t) is the time-dependent deformation. This
deformation can be permanent (e.g., viscous defor-
Axial stress mation) or reversible (time-dependent elasticity).
Generally speaking, at low loads the deformation will be
time-dependent elastic, while at higher loads, viscous
deformation will dominate.
In general, the material creeps to reduce high shear
εpl εel
stresses. This process can be unstable, leading to mater-
O Axial strain ial failure. A typical example is the flow of salt zones. To
avoid high shear stresses at the wellbore wall, they tend
Fig. 8-10. Graphical representation of plastic deformation. to fill in and close the wellbore. Once the wellbore is
closed, the flow will stop.

276 Well Cementing

8-3.7 Set-cement strength and set-cement failure ing a limit beyond which instability or failure occurs. The
Set-cement failure occurs when cracks or discontinu- Terzaghi effective stress is used in the failure criterion
ities appear in the matrix. To achieve set-cement failure, σ ′ = σ + p pore . (8-26)
the cracks must be large enough to cause the separation
of fragments. The failure can be brittle or ductile Several criteria have been proposed in the literature
(Fig. 8-11). A brittle failure means that failure occurs in for various applications. The more popular criteria
the elastic section of the stress-strain deformation include the following.
curve. Cracks initiate and propagate very quickly. Brittle ■ The maximum tensile stress criterion states that
failure is often unstable because the elastic energy failure initiates as soon as the maximum effective
released by the cracking process is higher than the sur- principal stress reaches the tensile strength, Stens,
face energy consumed when cracks are created. In a of the material:
crushing test, this is characterized by shattering of the
set cement and fragments flying away. Set cements are σ 1′ = Stens . (8-27)
brittle in tension, under impact loading, and during a
compressive strength test at low confining pressure. ■ The Tresca criterion states that failure occurs when
Ductile failure means that failure occurs in the plastic the maximum shear stress, (σ1 – σ3)/2, reaches a
regime, when permanent deformation is created. Set characteristic cohesion value, Yco:
cements exhibit a ductile failure under confining pres-
sure. Ductile failure is more advantageous for cement- σ 1 − σ 3 = 2Yco . (8-28)
sheath integrity in a tectonic environment, because the
■ The Mohr-Coulomb criterion states that, for com-
set cement will deform significantly before being dam-
aged. In some cases the cement will be able to flow with- pressive failure, the shear stress tending to cause
out cracking. This is especially common with porous or failure is opposed by the cohesion of the material and
flexible cements at high confining pressure. by a factor analogous to the coefficient of friction
multiplied by the effective normal stress acting
across the failure plane.

)
τ = Yco − tan ( Φ σ n ′ , (8-29)

where Φ is the angle of internal friction and Yco is

the cohesion. The Mohr-Coulomb failure criterion
Ductile can be rewritten in terms of principal stresses to give
Axial stress σ3 at failure in terms of σ1 (recall that tensions are
Brittle positive, and under this convention, σ1 is the confin-
ing stress):

⎛ π Φ⎞
( )
σ 3 ′ = − σ c + tan 2 ⎜ + ⎟ σ 1′
⎝4 2⎠
(8-30)

where σc is the compressive strength.

Axial strain
This criterion shows an increase of compressive
Fig. 8-11. Brittle versus ductile failure. strength with an increase of confining pressure. As
shown above, the Tresca and Mohr-Coulomb criteria do
not include the influence of the intermediate stress, σ2.
To estimate the initiation of cement failure, one uses a Experimental evidence shows that, in many cases, it is a
failure criterion. A failure criterion is usually a relation- good approximation. However, other criteria include the
ship between the principal effective stresses, represent- effect of σ2.

Chapter 8 Mechanical Properties of Well Cements 277

8-3.8 Influence of confining pressure on set- 8-3.9 Postfailure behavior of set cement
cement behavior and failure There are cases when the set cement will be subjected
In a downhole situation, especially when the rock is to excessive loads and vibration (e.g., during window
creeping or when large tectonic stresses are present, drilling), leading to failure. In such cases, to ensure that
the cement will be subjected to confining pressure. the damaged set cement does not fall apart, it is impor-
Increased confining pressure leads to increased cement tant to control the postfailure behavior. As mentioned
strength, as predicted by the Mohr-Coulomb failure above, conventional set cements are brittle materials at
criterion and also by the transition from a brittle to low confining pressures. When cracks are initiated they
a ductile material. Fig. 8-12 shows an example of the can propagate very quickly. Unstable failure occurs
influence of the confining pressure on a cement sample. because the elastic energy released by the cracking
The tests were carried out as shown in Fig. 8-4, but each process is higher than the surface energy consumed by
test was performed at a different confining-pressure the creation of cracks. To control the failure, one must
value, pcon. The set-cement strength does indeed increase unload the cement sufficiently to release the elastic
with confining pressure, but in reality this increase energy. In a controlled experiment, the area below the
is not the linear one predicted by the Mohr-Coulomb stress-strain curve is exactly the energy per unit volume
criterion. An empirical law relating the cement strength consumed to propagate the crack. In a bending test
to the confining pressure might be preferred. (Fig. 8-13) or a direct-tension test, this energy per
One key feature of set cements under confining pres- unit volume is called toughness. Models have been
sure is that they quickly become fully plastic with strain developed to describe tensile-crack propagation based
hardening, and thus failure is no longer observed. on this type of relationship (Bosma et al., 1999).
Another characteristic is that, because cements are However, in difficult cases in which conventional
highly porous, they will compact quickly under confining set cements are so brittle that uncontrolled failure is
pressure (if the sample is drained). During compaction, inevitable, high-toughness cements are required. High-
the set-cement porosity decreases. This is the main toughness cements are designed so that the surface
cause of strain hardening, as the cement becomes area below the stress-strain curve is sufficiently great
stronger with decreased porosity (unless the porosity is to ensure that, when cracks propagate, they do so in
so high that the matrix collapses). a stable manner. Ideally the cracks do not extend or
coalesce; instead, small microcracks form.

0
Uniaxial
–10 17.24-MPa confinement

–30

Differential stress –40

(MPa)
–50

–60

–70
–50,000 –40,000 –30,000 –20,000 –10,000 0 10,000 20,000
Microstrain

Fig. 8-12. Radial and axial strain as a function of axial differential stress (axial stress – confining stress)
and confining pressure. Stress and strain values follow the convention that tensile stress is positive
(i.e., compression negative). Conventional neat system.

278 Well Cementing

 total chemical shrinkage is the volume reduction that
occurs as a consequence of cement hydration.
At 100% hydration, the total chemical shrinkage of
F Portland cement is about 6.25 cm3/100 g of cement
(Powers, 1958). The total chemical shrinkage can
be measured by placing cement paste in a container,
surrounding it with water, and monitoring the water
level versus time. This technique tends to underestimate
the total chemical shrinkage because it assumes that
the water has full access to the pores. This is not neces-
Flexural strength sarily the case, as the permeability of the matrix
decreases with time. Justnes et al. (1995) measured the
total chemical shrinkage of various Portland cement
Stress
compositions at ambient conditions (68°F [20°C] and
atmospheric pressure). After 48 hr the total chemical
shrinkage was found to be independent of the water-to-
cement ratio—about 2.17 cm3/100 g of cement for neat
E
Class G slurries (or 2.8 vol% for a water-to-cement ratio
Strain of 44 wt%). This difference between the results of
Fig. 8-13. Bending test or flexural test: The area below the stress-
Justnes and those of Powers indicates that at 48 hr the
strain curve measures the toughness of the sample. degree of hydration is only 2.17/6.25 = 0.35.
Once the cement begins to develop compressive
strength, the total chemical shrinkage can occur as
bulk shrinkage. Bulk shrinkage is the external volume
The most efficient way to promote high cement reduction that can occur during the hydration of Portland
toughness involves the addition of fibers or microribbons cement. It is a mechanical response to the stresses and
to the cement matrix (Chapter 7). Fibers resist breakage pore-pressure changes generated by chemical shrinkage.
when traversed by a microcrack and, up to a certain Bulk shrinkage can create a microannulus between
load, remain bonded to the cement matrix. This also the formation and the cement, preventing the cement
improves the tensile strength of the set cement. But the from fulfilling its role as a sealing material (Justnes
most dramatic effect is that the fibers hold the broken et al., 1994; Justnes et al., 1995). When the set cement
cement fragments together and prevent the generation shrinks, it moves toward the casing. Formation of a
of larger cracks (Xenakis et al., 1997). The toughness of microannulus can lead to gas leaks at the surface (even
set cements containing fibers can be significantly when production has ceased: Dusseault et al., 2000),
higher than that of conventional cement systems, even or incorrect leakoff-pressure estimation (Zhou and
at low fiber concentrations; however, its efficiency Wojtanowicz, 1999).
depends on the chemical nature of the reinforcement Bulk shrinkage can be measured by placing cement
(Park et al., 1999; Nataraja et al., 2000). High-toughness paste in a flexible membrane, sealing the membrane,
cements based on metallic microribbons have been and monitoring the volume variations of the membrane.
used as shock-resistant kickoff plugs (Babasheikh At the beginning of the experiment, the bulk shrinkage
et al.; 2003) (Chapter 14). curve matches that of the total chemical shrinkage.
Then the bulk-shrinkage curve flattens out as a rigid
structure starts to form, preventing the total collapse
8-3.10 Shrinkage and expansion of the material (Fig. 8-14). Total chemical shrinkage
Cement shrinkage and expansion are discussed in continues to occur, because cement hydration does not
Chapters 2 and 9; however, in the context of the present stop after the cement sets.
discussion, it is necessary to revisit the subject. A review Using similar techniques in which the cement paste
of the terminology surrounding this subject is useful to is isolated from the surrounding medium, several
prevent confusion. authors (Parcevaux and Sault, 1984; Chenevert and
The fundamental mechanism responsible for cement Shrestha, 1991; Bensted, 1991; Sabins and Sutton, 1991;
shrinkage is the total chemical shrinkage (following Justnes et al., 1994; De Rozières and Sabins, 1995;
the terminology of Justnes et al., 1994). As discussed Justnes et al., 1995; Justnes et al., 1996; Backe et al.,
in Chapter 2, the volume of the reaction products of 1997) have reported bulk shrinkage values of 0.5 to 5.0%
cement hydration is less than that of the reactants. The by volume. Although these results were obtained under

Chapter 8 Mechanical Properties of Well Cements 279

 2.5 enough to allow the external fluid to invade the pore
structure (Appleby and Wilson, 1996). Restated in solid
2.0 mechanics terms, the pressure and saturation of the
1.5
fluid in the cement pores must remain constant during
Shrinkage cement hydration to avoid bulk shrinkage.
(mL /100 g) 1.0 When cement has no free access to water (e.g.,
because of the presence of an impermeable membrane),
0.5 total chemical shrinkage leads to pore-pressure reduc-
tion and eventual pore collapse. Restated in solid
0
mechanics terms, the cement compacts if the external
0 10 20 30 40 50 60
total stress remains constant and the internal pore
Time (hr)
pressure decreases during cement hydration, leading
Fig. 8-14. Chemical shrinkage (green line) and bulk shrinkage (red to an increase of effective mean stress. Note that the
line) as a function of time for a cement with a water/cement ratio chemical shrinkage can be so high that pore saturation
by weight of 0.4 (after Justnes et al., 1995). Reprinted with permis- falls below 100%, and, because of capillary effects, a
sion from Thomas Telford Limited. negative pore pressure is observed.

a variety of temperature and pressure conditions, such a 8-4 Mechanical behavior of a cement
wide range of results is probably caused by the formation
of free water during the experiments (Justnes et al., cased wellbore
1996) and the failure to control the boundary conditions. 8-4.1 State of stress in the cement sheath
If bulk-volume variations are measured when the To determine whether a cement sheath will fail or debond
cement has access to additional water during the test in the annulus, one must calculate the state of stress.
(e.g., by measuring the dimensional variation of an The calculated stress is then entered into an expression
annular ring mold or cylindrical sleeve filled with the to determine whether failure is attained. To calculate
cement paste and placed in water), a bulk expansion as the state of stress, one must assume a deformation
high as 0.3% by volume is observed after the cement sets behavior (e.g., elasticity) and consider the various
(de Rozières and Sabins, 1995). Uncontrolled bulk applied loads at specific boundaries such as the
expansion can be as harmful as bulk shrinkage, because casing/cement and cement/formation interfaces. In some
it can disrupt the casing/cement interface (Beirute et cases, the influence of temperature and pore pressure
al., 1988; Baumgarte et al., 1999). To avoid this prob- must also be considered.
lem, one must ensure that the cement has a lower In recent years, various models have been devel-
Young’s modulus value than the surrounding rock oped to analyze the state of stress in the cement
(Baumgarte et al., 1999, Le Roy-Delage et al., 2000). (Thiercelin et al., 1997; Bosma et al., 1999; Gino di
In the context of the mechanical properties of well Lullo and Rae, 2000; Fleckenstein et al., 2000;
cements, the initial phase of shrinkage is not relevant Philippacopoulos and Berndt, 2002; Pattillo and
(Setter and Roy, 1978). During this phase, the cement is Christansen, 2002). They are based on analytical solu-
still a liquid slurry. During primary cementing, before tions, numerical solutions, or a combination of both.
the cement slurry begins to set, the top of the cement
column moves downward to compensate for the volume
reduction (Chenevert and Shrestha, 1991). The bulk 8-4.2 Modeling the cement sheath using
shrinkage that occurs after a rigid network of hydration thermoelasticity
products has formed and compressive strength begins to In this section the modeling of stresses in a cased
develop is relevant to the mechanical properties. wellbore containing a finite number of concentric
Previous studies of cement shrinkage have shown two casings is briefly described. A cross section of the well-
key behaviors that can be easily linked to porous elasto- bore is shown in Fig. 8-15. The stresses in the cement
plastic solid behavior controlled by effective stress are calculated assuming that casing, cement, and rock
(Thiercelin et al., 1998). When cement has free access to are thermoelastic materials. The casing/cement and
additional water, the external water flows into the cement/rock interfaces are also assumed to be either
cement pore space to compensate for the total chemical fully bonded or unbonded. Finally, it is assumed that the
shrinkage, and almost no bulk shrinkage is observed. In cement is under no internal “effective” stress after
some formulations, an expansion might even be observed. setting. This final assumption is obviously a strong
For this to occur, the cement permeability must be high

280 Well Cementing

 Depth of interest
Rock Cement
σθ
Section at depth of interest

Cement

σr r

Casing

Fig. 8-16. Radial stress and tangential stress.

Casing

σr ν
)
Fig. 8-15. A cross section of the well.
εr −βT = − σ +σz
E E θ
(
σ ν
simplification. It is known that, after placement, the
cement slurry unloads as its gel strength develops
εθ − β T = θ − σ r + σ z
E E
( )
σ ν
)
(Chapter 9). Field observations (Cooke et al., 1983;
Morgan, 1989) tend to confirm that the total stress in ε z − β T = z − σ r + σθ
E E
(
the cement drops to at least the hydrostatic pressure
1
given by the saturating formation fluid (mainly water), γ rz = τ rz , (8-31)
justifying the zero-effective stress assumption. However, G
in some cases the pore pressure can drop below the
hydrostatic pressure, especially in a casing-to-casing where E, ν, and G are respectively the Young’s modu-
configuration. One can imagine a variety of situations, lus, Poisson’s ratio, and shear modulus, and β is the
depending on the cement properties, cementing coefficient of linear thermal expansion.
procedure, formation permeability, and nature of The temperature distribution as a function of time is
the saturating fluid. Nevertheless, in the absence of obtained from the heat diffusion equation, which is
better information, this simplification is appropriate. expressed under the assumption that the initial and
Consequently, to study the cement behavior, only the boundary conditions do not depend on θ, as
variations of pressure, stress, or temperature that occur
∂2 T 1 ∂ T ∂2 T ρ C ∂ T
after the cement sets are considered. + + = , (8-32)
The geometry of the problem is axisymmetric, with ∂r 2 r ∂r ∂z 2 λ ∂t
the axis of symmetry being the wellbore axis, allowing
the use of cylindrical coordinates r, θ, and z. The sim- where λ is the thermal conductivity, ρ the density, and
plest situation is when the boundary and initial condi- C the specific heat.
tions (wellbore and far-field states of stress and temper- Plane strain is also assumed, meaning that there is no
ature) are independent of θ. The variables of interest are axial movement. This is usually a good assumption,
then the radial displacement; radial stress, σr; tangential although axial movement could develop when casing
stress, σθ; axial stress, σ z; the shear stress, τrz; and the sections are being heated and axial casing deformation
temperature, T (which in practice is the temperature dif- is not prevented at the surface.
ference from a reference state). The tangential stress is The stress model uses analytical solutions that have
a principal stress. The radial and tangential stresses are been presented in Thiercelin et al. (1997). The solution
shown in Fig. 8-16. The sign convention is that tensile is constructed with the conditions that the radial
stresses are positive. Thermoelasticity provides a linear displacements and radial stress are continuous across
. the interface between two materials and the radial
relationship between the strains ε r, ε θ, ε z, and γrz,
stresses, and temperature, T. stress is compressive. If the radial stress is tensile (or

Chapter 8 Mechanical Properties of Well Cements 281

above a small value that can be considered as cement 3.0
adhesion at the interface), debonding occurs, radial
displacement is discontinuous, and the radial stress is 2.5
set to zero. The inner surface of the inner casing has an
imposed radial stress condition, given by the variation of 2.0
wellbore pressure on its inner surface. Stress 1.5
Heat transport takes place by conduction, as described (MPa)
above. To calculate the evolution of temperature with 1.0
time, it is often more convenient to use a numerical
technique such as the finite-difference technique or the 0.5
finite-element technique. The influence of temperature
on the state of stress is a function of thermal expansion 0
of the various materials. Time is introduced by the heat- 60 65 70 75 80 85 90
diffusion process. Distance from well axis (mm)

8-4.3 Influence of wellbore pressure increase Fig. 8-18. Tangential stress in the cement sheath as a function of
distance from the wellbore.
The most damaging wellbore pressure increases often
occur during a pressure test of the casing. It is indeed
unfortunate that, by checking the casing integrity, one
can damage the cement sheath. An increase of mud respectively. The wellbore pressure increase is 2,900 psi
weight, a hydraulic fracturing treatment, or a perfora- [20 MPa]. The openhole diameter is 7 in. [178 mm], the
tion test can also generate large wellbore pressure casing outside diameter (OD) is 5 in. [127 mm], and the
increases. A wellbore pressure increase induces a casing weight is 23.20 lbm/ft [34.53 kg/m]. Because we
radial elastic expansion of the casing, which in turn are using linear elasticity with a single loading condi-
loads the cement. tion, the stress values are a linear function of wellbore
The effect of a pressure increase on the state of stress pressure. Thus, doubling the wellbore pressure results in
in the cement sheath is shown in Figs. 8-17 and 8-18, in doubling the value of the stresses in the cement.
which the radial and tangential stresses in the cement These figures show that, in this case, the radial stress
are shown as a function of the distance from the well- is compressive and the tangential stress is tensile. The
bore axis. In this case, the Young’s moduli of the steel, tangential stress is about half the absolute value of the
cement, and rock are 29.0 × 106 psi [200 GPa], 0.725 radial stress. Because cements are about 10 times
× 10 6 psi [5 GPa], and 1.45 × 10 6 psi [10 GPa], weaker in tension than in compression, the cement
respectively. The Poisson’s ratios are 0.27, 0.15, and 0.2, failure will occur in tension. Tensile failure appears
when the tensile stress is greater than or equal to the
tensile strength. The highest value of the tangential
1.0 stress is at the steel/cement interface; therefore, this is
where failure should first occur. In this case, cement
0.5 failure will correspond to the initiation and propagation
of tensile radial cracks, because tensile cracks propagate
0 perpendicular to the direction of the maximum tensile
Radial stress. If the wellbore pressure increases by 2,900 psi
stress –0.5 [20 MPa], the value of the tangential stress at the
(MPa) steel/cement interface can be used to calculate the
–1.0 tensile strength the cement must have to avoid failure.
The cement Young’s modulus has a strong influence
–1.5
on the cement sheath response. This is demonstrated in
–2.0
Fig. 8-19, which shows the tangential stress as a function
60 65 70 75 80 85 90 of distance from the wellbore axis. In this case the
Young’s modulus of the cement is 0.725 × 10 5 psi
Distance from well axis (mm)
[500 MPa]. This time the tangential stress in the cement
Fig. 8-17. Radial stress in the cement sheath as a function of is less tensile, even compressive, near the cement/rock
distance from the wellbore. interface. This is because of the mechanical support pro-

282 Well Cementing

 0.20 Table 8-2. Cement Failure as a Function of Cement
Properties During a Wellbore Pressure Increase
0.15
Formulation A (Conventional) B (Elastic)
0.10 Tensile strength (psi [MPa]) 566 [3.9] 305 [2.1]
Stress 0.05 Young’s modulus (psi [MPa]) 1,311,000 [9,041] 376,000 [2,594]
(MPa)
Poisson’s ratio 0.15 0.22
0
Tangential stress at 998 [6.88] 138 [0.95]
–0.05 wellbore (psi [MPa])

–0.10 Failure Yes No

60 65 70 75 80 85 90
Distance from well axis (mm)
Fig. 8-19. Tangential stress in the cement sheath as a function strength, and that cement failure in tension is expected.
of distance from the wellbore for a cement Young’s modulus of A less stiff cement containing flexible particles, which
500 MPa. gains support from the rock, is more appropriate
(Formulation B). In this case, the tangential stress does
not exceed the tensile strength of the cement, even
vided by the rock. In fact, the resistance of the cement though the actual tensile strength is less than that of the
sheath is increased, as this effect usually compensates conventional cement.
for the decrease of cement strength often associated
with a decrease of Young’s modulus. The mechanical 8-4.4 Influence of wellbore temperature increase
support is actually related to the ratio of the set
cement’s Young’s modulus to the rock’s Young’s modulus, The calculation of stresses on set cement caused by a
so a rock with a higher Young’s modulus will show similar temperature increase involves additional parameters
behavior in reducing the tensile stress in the cement. such as the density, specific heat, thermal conductivity,
This also happens when the set cement’s Poisson’s ratio and the coefficients of thermal expansion of the various
is higher. Obviously other issues, such as casing protec- materials. For simplification one can assume that the
tion and casing support, must be considered. coefficients of thermal expansion of the steel, cement,
These results demonstrate that, to determine and rock are the same: 1.3 × 10–5 K–1. The complete set
whether the cement will fail owing to a wellbore pres- of thermoelastic properties is given in Table 8-3.
sure increase, one must know the tensile strength of Other material parameters are the same as in the exam-
the cement and the elastic properties of the cement and ple in Section 8-4.3, except for the stiffness of the
the rock. The geometry of the cased wellbore is also cement, which, in this example, is higher than that of
an important parameter. For example, increasing the the rock.
casing thickness will decrease the tensile strength The radial and tangential stresses in the cement show
requirement. the same pattern as that during a wellbore pressure
A field example was described by Le Roy-Delage et al. increase, with the generation of high tensile tangential
(2000). A cemented cased section of a well, in which stress in the set cement. The stresses are first generated
the slurry was displaced with a low-density mud, was to by the rapid thermal dilation of the casing. Then, as heat
be subjected at a later stage to a pressure increase of diffuses through the set cement and eventually into
6,500 psi [44.8 MPa]. This pressure increase was the
consequence of a mud-density increase, required to drill
through a salt zone located at a deeper location. Table 8-3. Assumed Thermal Properties of Steel, Cement,
The openhole diameter was 12 in. with a 9 5⁄8-in., and Rock
53.5-lbm/ft casing. The Young’s modulus and Poisson’s
Steel Cement Rock
ratio of the rock were 3.6 × 106 psi [24,800 MPa] and
0.25, respectively. The mechanical properties of the Solid density (kg/m3) 8,000 1,900 2,100
cement formulations used for this application are shown Specific heat (J/kg-K) 500 2,100 1,900
in Table 8-2. The slurry-density requirement was 16
lbm/gal [1,920 kg/m3]. Table 8-2 shows that the tangen- Thermal expansion coefficient (K–1) 1.3 × 10–5 1.3 × 10–5 1.3 × 10–5
tial stress imposed on the conventional cement
Thermal conductivity (W/m-K) 15 1.0 1.0
(Formulation A) largely exceeds the cement’s tensile

Chapter 8 Mechanical Properties of Well Cements 283

the rock, stresses are generated by the thermal dilation cement density varied from 12 lbm/gal [1,440 kg/m3] to
of the cement. Fig. 8-20 shows the tangential stress 14 lbm/gal [1,680 kg/m3]. The operator required a low-
in the cement, 1 min after a 360°F [200°C] temperature permeability cement with a final compressive strength
increase in the well. The tangential stress is compressive of at least 1,500 psi [10 MPa]. As shown in Table 8-4,
near the steel/cement interface and tensile near the a conventional 14-lbm/gal [1,680 kg/m3] cement
cement/rock interface. This behavior is caused by the (Formulation C) would not support the stress generated
nonuniform thermal expansions of the materials. The by such a temperature increase. The thermal expansion
temperature in the region near the well is higher than of the casing imposes a tensile tangential stress well
that further away. Compressive tangential stress is gen- above the tensile strength of the cement sheath.
erated in this region because the thermal expansion is Moreover, because the well was shallow, the formation
confined by the cooler materials in the surrounding (sandstone) had a low Young’s modulus and would not
region. Similarly, tensile stress is generated in the sur- prevent large deformations of the cement sheath. The
rounding region that is being pressurized by the near- model shows when and where the tensile stress reaches
wellbore region. As temperature diffuses through the the most critical value in the cement sheath. In this case,
cement, the stress pattern will evolve. The cement the most critical location is 5.2 in. [132 mm] from the
achieves a more uniform temperature and is therefore wellbore axis. Various formulations were tested.
under less stress. Formulation D, containing flexible particles, was able to
provide very low stiffness while retaining sufficient com-
pressive strength and low permeability. Table 8-4 demon-
8 strates that this cement will not crack under the thermal
loading, because the tangential stress at the critical loca-
6 tion is always below the value of the cement’s tensile
strength.
4

Stress 2
(MPa) Table 8-4. Cement Failure as a Function of Cement
Properties During a Wellbore Temperature Increase
0
Formulation C D
–2 Tensile strength (psi [MPa]) 334 [2.3] 218 [1.5]

–4 Young’s modulus [psi [MPa]) 545,000 [3,759] 177,000 [1,221]

60 65 70 75 80 85 90
Poisson’s ratio 0.16 0.24
Distance from well axis (mm)
Maximum value of 799 [5.51] 149 [1.03]
Fig. 8-20. Tangential stress as a function of wellbore radius tangential stress (psi [MPa])
after 1 min.
Failure Yes No

This process is similar to a pressure increase, but is

more complex owing to the influence of the temperature 8-4.5 Influence of wellbore pressure and
diffusion in the set cement and in the rock and the temperature decrease
existence of a temperature gradient resulting in the non- A decrease of wellbore temperature and pressure
uniform thermal dilation of a given material. A detailed creates tensile stress at the cement/rock and
analysis as a function of time is therefore required. cement/casing interfaces, if the cement is assumed to
A second field case addresses one of the most severe be fully bonded to the rock and casing. The radial and
conditions a cement sheath would ever experience in a tangential stresses are then exactly the same as in
well: steam injection (Le Roy-Delage et al., 2000). For Fig. 8-17 and Fig. 8-18, but with a sign change (again a
this well, the temperature was increased to 400°F consequence of elastic behavior). Radial stress is now
[204°C] from 120°F [49°C]. The openhole diameter as tensile and, because the adhesion forces are often very
measured by the caliper was 13 in. [330 mm]. The casing small, debonding will occur. Whether debonding will
was 9 5⁄8 in. [244 mm] OD and 40 lbm/ft [59.53 kg/m], and occur at the rock /cement or cement/casing interface
the Young’s modulus and Poisson’s ratio of the rock were depends on numerous parameters, such as the presence
580,000 psi [4,000 MPa] and 0.15, respectively. The

284 Well Cementing

of a thick mud cake at the cement/rock interface. As bonded to the casing has sufficient strength to support
demonstrated by bond logs, debonding often occurs at the the weight of the casing plus the weight of the drillstring
casing/cement interface, and a microannulus is created. during shoe drilling or other events. The strength of
The following field case is an example of the problem, interest is the shear bond strength, often empirically
although more severe cases have been encountered. An associated with the tensile strength of the cement.
intermediate section of a gas well is analyzed. The open- Casing support is usually calculated using the follow-
hole diameter is 12 1⁄4 in. [311.1 mm], with a 9 5⁄8-in. ing equation.
[244-mm], 66.8-lbm/ft [99.41-kg/m] casing. During
completion the heavy mud used to drill the deeper sec- F = 0.969 S × d × h, (8-33)
tion was replaced by a completion fluid with a density where F is the load to break the cement bond in lbf, S is
close to water. This resulted in a 3,290-psi [23-MPa] the compressive strength in pounds per square inch
pressure drop in front of the casing. However, during (assuming it is 10 times the shear-bond strength), d is
production, the temperature of the completion fluid the outer diameter of the casing string in inches, and h is
increased by 36°F [20°C]. The combined effect of the the height of the cement column in feet.
pressure decrease and temperature increase is shown in Assuming that only 5% of the cement is supporting the
Fig. 8-21. First, a 58-μm microannulus was created. load (which can be viewed as a very strong safety factor),
However, as the casing was being heated, it dilated and it can be shown (di Lullo and Rae, 2000) that, even with
reduced the size of the microannulus to 28 μm, which the heaviest casings, the required compressive strength
is still sufficient to create a gas leak. The only solution is around 143 psi [1 MPa]. This is well below the
would be to use an expanding cement (Seidel and accepted 500-psi [3.5045-MPa] standard. In practice,
Greene, 1985; Moran et al., 1991; Baumgarte et al., 1999, with normal casings, 50-psi [0.35-MPa] compressive
Le Roy-Delage et al., 2000; Chapter 7). However, as strength would be sufficient, especially if one considers
mentioned earlier, the cement must be more flexible than that the confining pressure might even increase the
the surrounding rock (Baumgarte et al., 1999, Le Roy- strength.
Delage et al., 2000) to ensure good bonding. For perforating and fracturing, a required compres-
sive strength of around 1,000–1,400 psi [7–10 MPa] is
often mentioned. The earlier discussion shows that this is
60 not necessary, because the pressures generated during
50 perforating and hydraulic fracturing result in a wellbore-
pressure increase. In many cases, a compliant or highly
40 flexible cement with low compressive strength will pro-
Micro- vide better support than a strong stiff cement. The
annulus 30 cement does not need to have high compressive strength
size (µm) to fulfill its role in casing support or zonal isolation.
20
Field observations support this statement (Deeg et
10 al., 1999). However, the flexible cement must be of good
quality and proven resistant to shock. For example,
0 highly permeable cements saturated with water might
0 10 20 30 40 50
suffer excessive shock because of the generation of
Time (min) excessive pore pressure in the cement matrix.
Fig. 8-21. Microannulus as a function of time.
8-4.7 Tectonics, subsidence, and formation creep
Tectonic stresses, subsidence, and formation creep can
8-4.6 Casing support, perforation and fracturing lead to loading of the cement by the formation and even-
tually the collapse of the casing. The loads involved are
The compressive strength required to support casing and often so large (except in the case of formation creep)
survive perforation and hydraulic fracturing treatments that it is practically impossible to prevent failure alto-
has been a subject of debate for many years. Traditionally gether. However, by using a careful completion design,
it was believed that the higher the compressive strength one can at least retard the collapse.
of the cement, the better the result would be. Recent When rock deformation or rock loading occurs, the
developments allow a clarification of the subject. cement can be strongly confined and compressed.
The situation concerning casing support is straight- Plastic flow of the cement leading to shear ductile failure
forward. Casing support means that the set cement would be the main deformation and failure mechanisms.

Chapter 8 Mechanical Properties of Well Cements 285

For these situations, it is necessary to know the cement
behavior as a function of the confining pressure.
The influence of rock stresses or rock deformation on
cement and casing failures can be grouped in three main
loadings.
1. The first includes a more or less continuous loading of
the cement bond and casing. It can be very isotropic if
rock creep is involved, or strongly directional if tec-
tonic stresses are involved (Fig. 8-22). The most
remarkable example is the influence of tectonic
stresses on cased wellbores in the foothills of
Colombia, South America (Last et al., 2002), which
leads to the ovalization of the casing string. Field data
show that the presence of the cement sheath
improves the stability of the casing owing to better
load distribution on the casing, and that a ductile
cement that deforms before failure also helps. Their
analysis is supported by a numerical study based on
the finite element method, in which casing and
cement elastoplastic behavior are taken into account.
Overall, the main recommendation is to use a dual
cemented string. The cement between the two strings
has the goal to strengthen and stiffen the dual string,
so it must be as strong and stiff as possible. Fig. 8-23. Casing shearing owing to shear band activation.

3. The last loading is axial compression of the casing

linked with the compaction of the formation adjacent
to it. This leads to casing buckling (Fig. 8-24), essen-
tially in poorly cemented zones. Heating of the casing
will have the same consequences because of thermal
vertical expansion. Experimental studies (Veeken
et al., 1994) again show that the presence of the
cement is essential to provide casing support and
avoid buckling.
Fig. 8-22. Casing ovalization caused by tectonic loads.

2. The second loading is linked to the activation of

natural faults (Fig. 8-23) or the creation of new
ones. This is typical of the overburden response to Zone with
strong compaction and leads to severe shearing of the no cement
casing. Recommendations (Dusseault et al., 1998) are
essentially based on either a strengthening of the
casing (dual cemented string or thicker casing) or an
increase of system compliance such as the under-
reaming of the zone, the use of ductile cements, and Fig. 8-24. Casing buckling owing to compaction.
even the avoidance of cementing. Pattillo and
Kristiansen (2002) used the finite element method to
analyze the stress field in the casing and cement and
showed that ductile cement aids stability by avoid-
ing extreme casing deformation.

286 Well Cementing

 The overall recommendation to maintain wellbore sta-
bility as long as possible is to strengthen the casing as
(2.109 − 1.174 ) × 914.4 × 9.81 = 8.39 MPa
much as possible and to ensure that the annulus is fully 1, 000
cemented. The cement must have good plastic/ductile or (0.914 − 0.509) × 3, 000 = 1, 215 psi
behavior and flow easily under stress to avoid casing
■ The required cement mechanical properties to sur-
pitching and to provide casing support.
Poor cement or microannulus formation can also lead vive such a change of downhole conditions are then
to a gas or oil leak between the reservoir and the over- determined using a stress model of the cemented
burden. In overpressured reservoirs, such fluid leaks can cased wellbore. The input data to the model include
lead to the pressurization of natural faults, which in turn the well geometry; steel, cement, and rock properties;
become unstable. Shear reactivation and shear collapse and the expected variation of downhole conditions.
of the casing can therefore be either promoted or The cement properties are obtained from the data-
induced by these leaks. Good bonding must be ensured. base mentioned above.
■ Cement selection using such a model is performed
with an iterative process, though one can imagine
8-5 Guidelines for cement design that, in the future, the model will be able to deter-
mine the best configuration automatically. The best
To design a cement system that resists the downhole solution is the most cost-effective cement that does
conditions, the following methodology is proposed. not fail under the input loading condition. The model
■ It is first essential to have a database of cements can be used to provide recommendations on cement
or alternate sealants with the desired mechanical design by tuning the properties of the cement to
properties (elastic parameters, compressive strength, minimize the expected problem. Unfortunately, in
tensile strength, shrinkage/expansion) at various complex situations, such as in tectonically active
temperatures and confining pressures. For example, areas, the modeling can be very involved.
having a variety of cement systems with low, medium, ■ The model must also be able to predict whether the
and high Young’s moduli is recommended. The data- loading can induce the formation of a microannulus,
base allows us to know what can be achieved under and in that case it is necessary to select the cement
various constraints such as slurry density, slurry accordingly (for example expanding cement with a
rheology, and cost. Rock properties are also needed Young’s modulus lower than the rock properties).
and can be obtained from sonic logs.
■ Finally, the integrity of the cement sheath should be
■ The second step is to properly understand and
monitored over time to validate the selection.
describe the downhole situation to identify the cause
of potential problems. This requires a review of all the To help the field engineers design cements based
events that may occur after the cement sets, such on mechanical considerations, software tools that follow
as a change of wellbore fluid, change of temperature, the above guidelines have been developed. An applica-
formation creep, subsidence, or other tectonic event. tion of such software tools for the cementation of steam-
In situations in which the casing can be damaged assisted gravity drainage wells is given in Stiles and
by the external loading, determination of the casing Hollies (2002). This stress analysis model software con-
failure mode (Section 8-4.7) from downhole informa- tains databases on cement, rock, and steel mechanical
tion is necessary. properties; a calculator to determine downhole stress
Example: A casing has been cemented with a conditions; a stress model to predict whether the
0.821-psi/ft or 15.8-lbm/gal [1,900-kg/m3] cement cement will fail (as described in Section 8-4); and a tool
displaced with a 0.509-psi/ft or 9.8-lbm/gal to help in designing the cement.
[1,174-kg/m3] mud. The reference pressure is the This software tool greatly facilitates the design of spe-
pressure applied inside the casing by the mud during cial cement for long-term zonal isolation, and this type of
cement setting, not the pressure applied by the software tool is recommended to design critical cases.
cement in the annulus. The drilling of the next sec-
tion, in which overpressured shales are encoun-
tered, requires a mud-weight increase to 0.914 psi/ft
or 17.6 lbm/gal [2,109 kg/m3]. The maximum pres-
sure increase is at the shoe, located at 3,000 ft
[914.4 m]. The pressure increase is

Chapter 8 Mechanical Properties of Well Cements 287

8-6 Conclusion in preventing microannulus formation. Other configura-
The determination of optimal long-term mechanical tions require high-strength, high-stiffness cements, such
properties that cements must have to ensure zonal as the cement used in a double-string casing and in
isolation is a relatively new approach in the oil and kickoff plugs. In a high-stress environment, it seems
gas industry. It is therefore an area in which new that the main requirement is that the cement must
development occurs at a high pace. The determination ensure isolation and keep the rock away from the casing
of cement mechanical properties under downhole as much as possible. To achieve its goal, it must be
conditions and the modeling of cased wellbore deforma- able to flow without cracking. Soft cements might help
tions are current topics of research. Numerous aspects reduce the load to the casing when shear fault reacti-
are still poorly known, such as the influence of the vation is a problem, in cases in which the slip of the fault
slurry-to-solid transition on the initial state of stress, is limited.
the role of cement pore pressure, and the modeling Cements and alternate sealants are being developed
of cement expansion or shrinkage in a downhole with properties that are more appropriate for downhole
environment. Nevertheless it is clear that cement conditions than conventional oilwell cements. There is
strength is not the only key mechanical property, and no doubt that new products with enhanced properties
that the cement’s elastic properties play a significant will continue to appear in the coming years.
role in maintaining the integrity of the cased wellbore.
In most configurations, especially when the casing
expands because of pressure increase or temperature 8-7 Acronym list
increase, soft elastic cements are the best solution. OD Outside diameter
Expanding cements must also have low Young’s moduli UCA Ultrasonic cement analyzer
to be properly confined by the rock and to play their role

288 Well Cementing

9-1 Introduction
Annular Formation Fluid Migration

Annular fluid migration during drilling or well-comple-

tion procedures has long been recognized as one of the
David Stiles—Schlumberger

Well 1
Interzonal communication
9
Well 2
Gas to surface

most troublesome problems of the petroleum industry. It

is defined as the invasion of formation fluids into the
annulus owing to a pressure imbalance at the formation
face. The fluids may migrate to a lower pressure zone or Low-pressure
possibly to the surface (Fig. 9-1). Gas migration is the zone
most common form of annular fluid migration and no
doubt the most dangerous (Carter and Slagle, 1970;
Sutton and Faul, 1984). Gas migration is the primary
focus of this chapter; however, most of the concepts pre-
sented here also apply to other formation fluids, such as High-pressure
shallow water (Stiles, 1997; Watters and Beirute, 1998). gas zone
Gas migration—also called gas communication or gas
leakage (Carter and Slagle, 1970), annular gas flow
(Garcia and Clark, 1976), gas channeling (Parcevaux et
al., 1983), flow after cementing (Webster and Eikerts,
1979), or gas invasion (Bannister et al., 1983)—is a
problem for both gas-producing and gas-storage wells.
The severity of the problem ranges from the most haz-
ardous (e.g., a blowout situation arising from a severe
pressure imbalance during drilling or cementing) to the Fig. 9-1. Two scenarios of annular gas migration.
most marginal (e.g., a residual gas pressure of a few psi
at the wellhead). In addition to surface-related difficul-
ties, communication between two or more subsurface
zones can occur. Such problems are more difficult to sure or gas flow at the wellhead, may dictate well aban-
detect. donment. More frequently, remedial cementing is per-
This chapter concentrates on gas migration after pri- formed until gas flow is stopped and gas pressure is
mary cementing. However, many of the elements of gas reduced to a level compatible with the operator’s safety
migration are similar to well control during drilling policy and local regulations. However, the efficiency of
(Moore, 1974). Although drilling is beyond the scope of squeeze cementing in such circumstances is very poor
this chapter, some similarities between gas migration for three essential reasons: (1) gas channels are diffi-
during cementing and gas migration during drilling are cult to locate, especially if they are less than 1 mm in
highlighted. size; (2) gas channels may be too small to be effectively
filled by cement; and (3) the pressure exerted during the
squeeze job is sometimes sufficient to break cement
bonds or even to initiate formation fracturing, further
9-2 Practical consequences of gas worsening downhole communication problems.
migration Furthermore, cement repair operations are expensive,
The potential consequences of gas migration following especially offshore or at remote locations (Cooke et al.,
primary cementing are numerous, but not always imme- 1982). Therefore, preventing a gas migration problem is
diately detectable. At the extreme, those that manifest definitely preferable to attempts to repair it. A thorough
themselves at the surface, e.g., sustained casing pres- discussion of remedial cementing appears in Chapter 14.

Well Cementing ■ Chapter 9 Annular Formation Fluid Migration 289

 Gas migration between two or more subsurface zones, ■ Applications of new products and techniques in the
with no surface manifestations, is very difficult to detect field (Kucyn et al., 1977; Watters and Sabins, 1980;
(Fig. 9-1). In such cases, gas production may be Cheung and Myrick, 1983; Seidel and Greene, 1985;
impaired, gas may be diverted to an upper depleted zone Sepos and Cart, 1985; Matthews and Copeland, 1986)
(possibly followed by gas migration to the surface ■ Empirical qualitative prediction techniques (Sutton
through another well), or the efficiency of stimulation et al., 1984; Rae et al., 1989).
treatments may be reduced (Cooke et al., 1982). Such
downhole channeling can sometimes be evaluated by Successful numerical simulations of the process or
special methods such as noise logs (Garcia and Clark, scaled laboratory experiments that could allow a gener-
1976) or acoustic logs (Catala et al., 1984; Rang, 1987) alized and quantitative prediction of gas migration have
(Chapter 15). Hydraulic communication testing is not not been reported.
recommended. If not properly designed and controlled,
these tests may induce communication across properly 9-3.1 Root causes for gas migration
cemented zones or aggravate minor cement-job defects.
A complete discussion of cement-job evaluation is pre- Annular gas migration has three distinct root causes
sented in Chapter 15. (Fig. 9-2).
1. The hydrostatic pressure in the annulus falls to a level
that is less than or equal to the pore pressure of a gas-
9-3 Physical process of gas migration bearing zone.
2. Space in the annulus allows gas entry.
Gas migration is a complex process that is influenced by 3. A path is present in the annulus through which the
many factors: fluid density control, mud removal, gas can migrate.
cement-slurry properties, cement hydration, and inter-
actions between the cement, casing, and formation. The All three root causes must be satisfied for annular gas
cementing industry first recognized gas migration in the migration to take place. The root causes involve various
early 1960s, when major gas-communication problems factors inherent to the cementing process. These factors
occurred in gas storage wells in the United States (Stone are described in this chapter.
and Christian, 1974). Since then, the industry has
worked diligently to understand the problem and find
solutions.
Extensive research has been performed to under-
stand the fundamentals of the physical process, and a Path to
vast quantity of literature describes various aspects of migrate
gas migration.
■ Field case study analyses and experiments for making
practical recommendations (Vidovskii et al., 1971; Gas
Stone and Christian, 1974; Garcia and Clark, 1976; migration
Cooke et al., 1982; Lukkien, 1982; Al-Buraik et al.,
1998)
■ Laboratory investigations to improve the understand-
Annular ≤ Formation
Space for entry
pressure pressure
ing of gas-migration fundamentals (Guyvoronsky and
Farukshin, 1963; Carter and Slagle, 1970; Carter et Fig. 9-2. Root causes of annular gas migration.
al., 1973; Webster and Eikerts, 1979; Sabins et al.,
1982; Bannister et al., 1983; Parcevaux, 1984b;
Beirute and Cheung, 1990; Moroni et al., 1997; Calloni
et al., 1999; Barlet-Gouédard et al., 2001) 9-3.2 Gas migration categories
■ Development of technical solutions (Levine et al., Understanding and modeling the gas migration phenom-
1979; Tinsley et al., 1979; Cheung and Beirute, 1982; enon is difficult because the conduit through which the
Parcevaux et al., 1983; Stewart and Schouten, 1986; gas can channel (i.e., an annular column full of cement,
Sykes and Logan, 1987; Dean and Brennen, 1992) with possibly some spacer and drilling fluid left in the

290 Well Cementing

hole) evolves with time. The physical state of the cement the annulus falls below the formation gas pressure at any
progresses from a liquid slurry during placement to a time, a gas release may occur that, by further relieving
permeable gel during a limited static period, to a per- the hydrostatic pressure, may lead to an irreversible gas-
meable, weak solid when setting, and finally to an imper- entry process. Consequently, the cement-job design
meable solid after hardening. Thus, the physical process should be performed with a free-fall computer simulator
of gas migration is convenient to categorize according to to ensure that the hydrostatic pressure against critical
when it occurs during the cementing operation zones is maintained between the pore and the fracturing
(Fig. 9-3). Three major categories can be defined: pressure at all times during the cement job. An example
1. immediate gas migration (during placement) is shown in Fig. 9-4 (Drecq and Parcevaux, 1988).
It should also be noted that casing reciprocation may
2. short-term gas migration (postplacement)
cause the hydrostatic pressure in the annulus to fall
3. long-term gas migration (postsetting). because of a swabbing effect. This is especially probable

9-3.2.1 Immediate gas migration Placement Pressure Limits

Immediate gas migration, also referred to as gas migra- Fracture pressure
4,000
tion during placement, occurs between the start of the Maximum dynamic
cementing operation and the end of cement placement, 4,500 pressure
which is normally marked when the top wiper plug Minimum hydrostatic
5,000 pressure
lands. Pore pressure
During this time period, gas migration results from 5,500
the loss of hydrostatic-pressure overbalance against gas- 6,000
bearing formations. Preventing gas migration during this 6,500
stage is a relatively straightforward well-control matter, Depth
similar to that practiced during drilling. One of the first (ft) 7,000
approaches to solving immediate gas migration was to 7,500
simply increase the density of the fluid in the annulus. 8,000
Such an approach can be dangerous, because the result-
ing increase in hydrostatic pressure can lead to lost cir- 8,500
culation or formation fracturing. In 1970, Carter and 9,000
Slagle recommended circulating the well before cement- 9,500
Formation
fractured
ing to help remove any trapped gas bubbles that, if not
removed before cement placement, would reduce the 10,000
hydrostatic head of the fluid column. 2,000 4,000 6,000 8,000 10,000
The principal difference between well control during Annular pressure (psi)
drilling and that during cementing is the free-fall or U-
Fig. 9-4. Computer-aided program output (from Drecq and
tubing phenomenon that occurs during the cement job Parcevaux, 1988).
(Chapter 12). Because of the density differences
between the mud, preflushes, spacer, and cement slurry
(or slurries), the hydrostatic pressure exerted at the for-
mation face is not constant during the job (Beirute,
1984; Smith et al., 1987). If the hydrostatic pressure in
et
ug

ts
pl

en
p
m

m
Bu

Ce

Placement Postplacement Postsetting

Minutes Hours Days Months/years

Immediate Short-term Long-term

gas migration gas migration gas migration

Fig. 9-3. Categories of gas migration.

Chapter 9 Annular Formation Fluid Migration 291

during periods when the fluids in the wellbore are in a has increased in recent years because of environmental
static state owing to free fall. As long as the hydrostatic concerns about abandoned gas wells leaking gas into the
pressure in the annulus remains greater than the forma- atmosphere.
tion gas pressure, no gas migration (other than that The primary driver for gas to migrate in the long term
which occurs though dissolution and diffusion processes is the formation of a pathway for the gas to travel after
at the molecular level) should occur during the place- the cement has set. After setting, a normal-density
ment operation. cement system becomes a solid with microdarcy perme-
One final point should be made concerning density ability. As a result, gas cannot migrate at any detectable
control during the cementing operation. Many large rate within the partially water-saturated pores of the
cement jobs are performed on a continuous-mix basis cement matrix. It should be noted that low-density
(i.e., “on the fly”). Density fluctuations may occur during cement systems with high water-to-cement ratios might
the course of the job, resulting in the placement of a exhibit fairly high permeabilities (0.5 to 5.0 mD).
nonuniform column of cement in the annulus Therefore, it is possible for gas to flow, albeit at low
(Granberry et al., 1989). Such a condition may cause rates, within the matrix of such cements and to eventu-
solids settling, free-water development, or premature ally reach the surface. Such events may take weeks or
bridging in some parts of the annulus. When the poten- months to appear as measurable phenomena at the sur-
tial for annular gas migration exists, either batch mixing face, where they usually appear as slow pressure
or process-controlled continuous mixing is recom- buildups in the shut-in annulus.
mended to ensure a homogeneous cement column. The path for long-term gas migration is more likely to
be through a microannulus, a mud channel, a channel of
bypassed lead cement slurry, a free-water channel, a dehy-
9-3.2.2 Short-term gas migration drated filtercake, or any mechanical failure of the cement
Short-term gas migration, also called postplacement gas sheath caused by imposed stresses. Space for entry can
migration, occurs between the end of the primary come from chemical shrinkage of the cement, bulk
cementing operation (normally marked by the landing of shrinkage of the cement, and dehydration of mud chan-
the top wiper plug) and the setting of the cement. Short- nels, free-fluid channels, and filtercakes. Because the
term gas migration may occur anytime between a few cement has already set during this phase, it is no longer
minutes to a few days after the end of the cementing transmitting hydrostatic pressure across the gas zone.
operation. Gas migration that occurs during this time
period is perhaps the most complex to understand, diffi-
cult to predict, and problematic to prevent. For this 9-3.3 Factors affecting gas migration
reason, most industry work concerning gas migration Numerous factors, many of which have been mentioned
has focused on this category. previously, may contribute to gas migration. A synopsis
The primary driver for the occurrence of gas migra- of these factors and how they relate to both the root
tion during this stage is believed to be the decay of the causes and the categories of gas migration is presented
annular pressure. This pressure decay can be attributed in Table 9-1. It is important to note that one single factor
to a combination of several factors, including fluid loss, does not cause gas migration, but rather a combination
gel strength development, and chemical shrinkage of the of several factors, depending upon the unique conditions
cement during hydration. It can also result from the for- in each well.
mation of annular bridges or the setting of mechanical
devices such as packers that isolate hydrostatic-pressure
transmission. Space for gas to enter the annulus during 9-3.3.1 Fluid loss
this stage can be created by fluid loss, free fluid, chemi- Fluid loss from the cement slurry into the formation
cal shrinkage, and the inherent porosity of the slurry. directly affects all three of the root causes of gas migra-
The gas-migration path during the short-term stage is tion and therefore must be considered one of the main
first through the cement-filtercake permeability and contributing factors. First, it may be responsible for a
then through the permeability of the cement matrix. decrease in annular pressure owing to
■ annular bridging

■ increased slurry gelation effects caused by a reduced

9-3.2.3 Long-term gas migration
water content in the slurry
Long-term gas migration occurs after the cement has
■ a decrease in the height of the hydrostatic column
set, which may occur within a few hours after the end of
the cement job. However, days, months, or even years are owing to a slurry-volume decrease
the usual time frame. Industry interest in understand- ■ friction-pressure losses during the compaction
ing, predicting, and preventing long-term gas migration resulting from a slurry volume decrease.

292 Well Cementing

Table 9-1. Factors Responsible for Gas Migration
Annular Pressure –< Pore Pressure Space for Entry Path for Migration
Immediate Hydrostatic underbalance Fluid displaced from wellbore Fluid displaced from wellbore

Short term Fluid loss Fluid loss Slurry permeability

Gel strength development Free fluid Slurry permeability

Chemical shrinkage of cement Chemical shrinkage of cement Filtercake permeability

Annular bridging Slurry porosity Filtercake permeability

Annular packers Slurry porosity Filtercake permeability

Long term Chemical shrinkage of cement Chemical shrinkage of cement Microannulus

Mud channel Mud channel

Free fluid Free-fluid channel

Strength development of cement Dehydrated filtercake Dehydrated filtercake

Bulk shrinkage of cement Bulk shrinkage of cement

Low cement tops

Cement sheath mechanical failure

Second, owing to a volume decrease, fluid loss may into the formation had been neglected. Baret (1988)
create space within the cement matrix that gas can confirmed the critical importance of fluid loss by more
occupy. Finally, fluid loss may be responsible for control- precise direct computations based on Darcy flow. He
ling the filtercake permeability, which ultimately influ- determined that, even in the presence of drilling mud
ences the migration path. Fluid-loss additives may also filtercake, API fluid-loss rates as low as 10 mL/30 min
act indirectly to reduce the permeability of the cement would sometimes be required to prevent annular bridging.
slurry (Chapter 6). It is important to mention that poor fluid-loss control
The importance of fluid loss as a contributing factor across permeable formations further up the hole can
to gas migration was first recognized by Carter and also impair full transmission of the hydrostatic pressure
Slagle (1970). At that time, the respective influences of to a gas zone. In 1976, Garcia and Clark observed gas
fluid-loss control and cement-slurry gelation were not migration when fluid loss occurred high in the hole,
fully understood. It was, however, pointed out that bridg- and hydrostatic pressure was no longer transmitted
ing or gelation owing to fluid loss could restrict the from the column above the bridging point to the bottom
transmission of hydrostatic pressure. In 1975, Christian of the hole.
et al. derived a method for calculating the fluid-loss con- Parcevaux (1987) discussed how fluid loss causes a
trol required to prevent bridging of the cement across pore-pressure decline and the formation of a void space
permeable formations during and after cement place- in the cement. The interstitial water in cement slurry is
ment. Christian et al. concluded that reducing the mobile; therefore, some degree of fluid loss always
American Petroleum Institute (API) fluid-loss rate to occurs when the annular hydrostatic pressure exceeds
less than 50 mL/30 min would reduce gas invasion and that of the formation. The process slows when a low-per-
lessen cement permeability. In 1977, Cook and meability filtercake forms against the formation wall
Cunningham described a procedure to analyze the gas and can stop altogether when the annular and formation
leakage potential based on a similar fluid-loss-rate com- pressures equilibrate. Once pressure equilibrium is
putation. However, Webster and Eikerts (1979) pointed attained, any volume change within the cement will pro-
out that, because earlier work was not based upon flow voke a sharp pore-pressure decline; consequently,
equations, the relative importance of fluid loss may because of the low cement compressibility, a void space
have been overemphasized. The positive influences of forms within the cement matrix, potentially inducing gas
the drilling mud filtercake and mud-particle invasion influx into the cement.

Chapter 9 Annular Formation Fluid Migration 293

 In 2002, Nishikawa and Wojtanowicz discounted the 8,000
validity of previously published pressure-decline theo- 7,000
ries because the correlation with field data was poor. 6,000
They proposed a different model with the following
assumptions. Once the cement is in place in the annulus, 5,000
fluid is initially lost by filtration driven by a hydrostatic Pressure 4,000
(psi)
overbalance. The resulting volume change causes pres- 3,000
sure reduction owing to the compressibility of the 2,000
system. As the volume reduction continues, the cement
1,000
slurry moves downward in plug flow, causing friction
drag at the wellbore walls and an annular pressure 0
decrease. This pressure transient effect is transmitted 0 100 200 300 400
upwards from the zone of fluid loss to the surface as a Time (min)
function of depth and time. A mathematical model was Recorded Predicted
derived for the pressure unloading. at 8,754 ft at 8,754 ft
at 6,909 ft at 6,909 ft
∞ at 5,488 ft at 5,488 ft
0.052
p( h, t ) = 0.052ρ eq h + 2∑ (ρ slurry − ρ eq ) at 4,787 ft
at 4,632 ft
at 4,787 ft
at 4,632 ft
n=1 Dboc at 3,636 ft at 3,636 ft
⎡ (−1) n+1 ⎤
⎢ 2
⎥ sin ( αh e)c2α 2 t
, (9-1) Fig. 9-5. Predicted and recorded loss of hydrostatic pressure at
⎢⎣ α ⎥⎦ depth (from Nishikawa and Wojtanowicz, 2002). Reprinted with
permission of SPE.
where
c = compressibility (psi–1)
thin, low-permeability filtercakes are more effective at
Dboc = Depth at bottom of cement controlling gas migration.
h = depth (cm)
t = time after pumping slurry (min) 9-3.3.2 Gel strength development
α = constant defined as As early as 1970, Carter and Slagle noticed that
thixotropy or gelation of wellbore fluids was relevant to
⎛ π⎞ 1
α = ⎜ nπ − ⎟ the reduction of hydrostatic pressure but provided no
⎝ 2 ⎠ Dboc explanation. Experiments to quantify the effect of gela-
tion on hydrostatic pressure transmission gave inconclu-
ρeq = equivalent density of normal formation sive results (Carter et al., 1973). Some pressure restric-
pressure gradient (lbm/gal) tion was observed at low curing pressures, but
ρslurry = density of the cement slurry (lbm/gal). experiments at greater pressures (500 to 1,000 psi or 3.5
This model was validated in the field when six pres- to 7 MPa) indicated no pressure change. This was prob-
sure gauges were distributed throughout the interval of ably related to deficiencies in the experimental design.
interest in the annulus of a well. The bottomhole pres- It is interesting to note that, much earlier, hydrosta-
sures calculated from the model closely matched the tic pressure reduction during cement hydration had
pressure reduction in the well (Fig. 9-5). been demonstrated in the laboratory and confirmed by
As shown in Chapter 6, cement filtercakes generally field measurements (Guyvoronsky and Farukshin, 1963;
have very low permeability. Bannister et al. (1983) con- Vidovskii et al., 1971) in the former Soviet Union. Similar
cluded that low-permeability cement filtercake deposi- field measurements were performed later by Cooke et al.
tion at the point of gas invasion could hinder gas flow. It (1982), in which the use of external casing sensors per-
is important to note that API fluid-loss tests are per- mitted the observation of downhole temperature and
formed at a fixed differential pressure across a perm- pressure fluctuations as well as the transmission of
eable medium. Thus, according to Darcy’s law, the same applied surface pressure (Fig. 9-6). From this informa-
fluid-loss rate could be measured for two very different tion, it was possible to derive the extent of vertical fluid
slurries: one with a relatively thin, low-permeability movement into the wellbore, locate the top of the
filtercake, and another with a relatively thick, high- cement column, and measure the cement setting time at
permeability filtercake. From the preceding discussion, different depths.
it is clear that low-fluid-loss cement systems that form

294 Well Cementing

 5,000 process involves the breakage of the slurry’s gel strength.
4,000
Sensor depths (ft)
Perforated
However, gas may also flow at the microscopic level within
6,885 ft
6,585 ft the pores of the gelled cement structure (Section 9-3.3.4)
Pressure
3,000 6,659 and 6,585 ft or directly along the cement/pipe and cement/formation
(psi) 5,969 ft

2,000 4,779 ft 6,659 ft interfaces (Section 9-3.3.7). Any or all of these processes
3,459 ft may contribute to the overall phenomenon of gas migra-
1,000 tion, and this limits the applicability of Eq. 9-2.
225 Geothermal Grachyov and Leonov (1969), Parcevaux (1987), and
Temperature 205 6,659 and 6,585 ft • 6,885 ft

(°F) 175 6,885 ft • 6,659 and 6,585 ft

• 4,779 ft
Drecq and Parcevaux (1988) further formalized the pres-
145 3,459 ft 5,969 ft • 3,459 ft sure-reduction process by taking advantage of the simi-
1 2 3 0 0.4 larities between a gelling cement column and a layer of
Time (thousands of minutes) soil undergoing consolidation. Using the theory of soil
mechanics and assuming that the cement slurry behaves
Fig. 9-6. Annular pressure and temperature measurements from as a virgin sedimentary soil before significant hydration
external casing sensors (from Cooke et al., 1982). Reprinted with occurs, the state of stress in the slurry can be described by
permission of SPE.
Terzaghi’s law (Vyalov, 1986). Either U.S. or SI units may
be used in the following two equations, as long as the unit
system is consistent.
In 1982, Sabins et al. related the kinetics of hydrosta-
tic pressure reduction to the cement-slurry gel strength σ = σ ′ + p, (9-3)
development, fluid-loss volume, volume reduction owing
where
to hydration, and slurry compressibility factor. This work
resulted in an empirical method to predict gas migra- p = interstitial (pore) or hydrostatic pressure
tion, and the following equation was derived. σ = total stress exerted at a given linear depth, z
σ′= intergranular or effective stress related to
⎛ 4 × S gel × L ⎞ ⎛ ΔV fl + ΔVhyd ⎞ gel-strength development.
Δp = ⎜ ⎟ or ⎜ ⎟, (9-2)
⎝ dhole − d pipe ⎠ ⎝ fc ⎠ σ is constant and equal to the full overburden pres-
sure exerted by the fluid column.
where
D
dhole = hole diameter (m)
σ = g ∫ ρ slurry ( z )cos θ( z )d x, (9-4)
dpipe = pipe diameter (m) 0
fc = slurry compressibility factor (dimensionless)
where
L = cement column length (m)
D = total linear depth
Δp = hydrostatic pressure change (kPa)
g = gravitational coefficient
Sgel = static gel strength (Pa)
θ = angular deviation
ΔVfl = fluid-loss volume reduction (L)
ρslurry = specific gravity of the slurry at depth z.
ΔVhyd = hydration volume reduction (L).
The effective shear stress, σ′, is related to the static
In 1979, Tinsley et al. introduced the concept of transi-
gel strength determined in the laboratory, using the
tion state, an intermediate period during which the
method described by Sabins et al. (1982) or by Hannant
cement behaves neither as a fluid nor as a solid and the
and Keating (1985), which incorporates the classic
slurry loses its ability to transmit hydrostatic pressure.
shear-stress equation.
The concept of transition state was quantified by a tran-
sition time, beginning with the first measurable gel 4 × L × S gel
strength (about 21 lbf/100 ft2 [10 Pa]), and ending when σ′ = , (9-5)
dhole − d pipe
gas could no longer percolate within the gelled cement.
They showed that a gel strength range from 250 to
500 lbf/100 ft2 [120 to 240 Pa] was sufficient to prohibit where
gas percolation. Gas percolation can be considered a par- dhole – dpipe = width of the annular gap (m)
ticular type of gas migration in which gas in the form of L = length (m)
macroscopic bubbles invades the slurry and rises owing to
Sgel = static gel strength (Pa)
buoyancy effects in accordance with Stokes law. Cement
slurries behave as non-Newtonian fluids; therefore, this σ′ = shear stress (Pa).

Chapter 9 Annular Formation Fluid Migration 295

 Equations 9-4 and 9-5 can thus be combined to obtain Powers studied the shrinkage of pure cement phases
as early as 1935 and found it to increase along the series
4 × L × S gel C2S-C3S-C4AF-C3A from 1% for C2S up to 16% for C3A. He
p = σ − σ ′ = ρ slurry gD cos θ − . (9-6)
dhole − d pipe found that the absolute shrinkage of Portland cement
pastes varies between 2.3% and 5.1%, according to the
following equation.
The hydrostatic pressure, p, exerted by the slurry in
front of the formation varies as a function of the static ΔVab = K hyd1[C3 S] + K hyd 2[C2S] +
gel strength, σ′. However, the exact value of p at time t
may be different from that given by Eq. 9-6 because of K hyd 3[C3 A] + K hyd 4[C 4 AF] (9-9)
kinetic effects.
When gelation occurs during the induction or dor- Powers assumed that, for each type of cement, the
mant period, there is no significant hydration of the shrinkage is a linear function of the percentages of the
cement grains but essentially a buildup of intergranular four major clinker phases. The values Khyd1, Khyd2, Khyd3,
forces owing to interparticle electrostatic forces and the and Khyd4 are coefficients with values varying with the
precipitation of hydrates (Chapter 2). In a first approxi- age (degree of hydration) of the specimen.
mation, the total stress, σ, remains the same, but a In 1982, Geiker and Knudsen found the rate and mag-
transfer from p to σ′ occurs. Eventually, σ′ increases to nitude of the chemical shrinkage to increase slightly
a point at which the cement becomes self-supporting. At with the water-to-cement ratio but the ultimate degree
this time, the interstitial pressure drops to that of a of shrinkage to decrease with increasing curing temper-
column of water, as shown by Eqs. 9-7 and 9-8. ature. The total chemical contraction is split between a
bulk or external volumetric shrinkage, less than 1%, and
p = ρ slurry gD cos θ (9-7) a matrix internal contraction representing from 4% to 6%
by volume of cement slurry, depending upon the cement
σ ′ = (ρ slurry − ρ w ) gD cos θ (9-8) composition (Parcevaux and Sault, 1984). Thus, when
considering cement shrinkage, a distinction should
where always be made between the two types. In most cases,
data reported in the literature refer to total chemical
ρw = the specific gravity of the interstitial water. contraction.
Shrinkage values of less than 4% were reported by
9-3.3.3 Cement shrinkage Chenevert and Shreshtha (1987); however, their experi-
mental design suggests that the phenomenon measured
Cement shrinkage contributes to gas migration by caus- was not the total chemical contraction, but a combina-
ing an annular-pressure reduction and by providing tion of bulk shrinkage and reabsorption of cement free
space for gas to enter the wellbore. When cement enters water. Chemical contraction is a time-dependent phe-
the setting period and hydration accelerates, intergran- nomenon (Fig. 9-7) that begins during the initial setting
ular stresses increase above the value given in Eq. 9-8 and levels off after the final set (Stewart and Schouten,
because of the intergrowth of calcium silicate hydrates. 1986).
Were no volume change to occur at this stage, the pore In 1979, Levine et al. made a significant contribution
pressure would remain at the level given by Eq. 9-7, and by relating shrinkage to pressure reduction. They mea-
the cement would behave as a porous formation. sured the hydrostatic pressure transmission of cement
However, this is not the case. Cement hydration is slurries in a 47-ft long cell with no external pressure
responsible for an absolute volume reduction of the source (Fig. 9-8). They demonstrated that the hydrosta-
cement matrix, also called cement chemical contrac- tic pressure gradient gradually decreases to that of the
tion, which was first identified by LeChâtelier in 1887 mix water. Later, when the cement slurry begins to set,
(Chapter 2; Appendix B). He showed, for normal the hydrostatic pressure quickly approaches zero
Portland cement, a volumetric shrinkage of 4.6%. This (Fig. 9-9). The hydrostatic pressure reduction is the
shrinkage is well documented in the civil engineering lit- result of shrinkage within the cement matrix caused by
erature (Setter and Roy, 1978) and occurs because the hydration and fluid loss; at this point, the fluid column
volume of the hydrated phases is less than that of the ini- above cannot reestablish the pore pressure.
tial reactants.

296 Well Cementing

 7 30
Pressure data
25 150°F laboratory test
6
20 16.4 lbm/gal Class H (neat)
5 Temperature Hydrostatic 15
pressure Static
4 10
(psi)
Shrinkage Total contraction 5 Circulating
(%) 3 0
2 100
1 80
Bulk shrinkage Slurry 60
0 consistency
40
1 5 10 50 100 (Bc) API thickening time
20 150°F atmospheric pressure
Time (hr) 16.4 lbm/gal Class H (neat)
0
Fig. 9-7. Typical contraction and shrinkage (from Parcevaux, 1987). 0 1 2 3 4 5 6
Temperature is 10°C. Time (hr)

Fig. 9-9. Annular gas flow test results (from Levine et al., 1979).
Reprinted with permission of SPE.

Fresh water
10

150°F bath

20
.. p1
. . Pressure
8.34-lbm/gal water
.. . .
. . . transducers and Depth
. . . .
thermocouples (ft)
Tubing

. .. . 30
. .. . .
. . .
. . ..
p2 16.4-lbm/gal cement
Porous plate 40

Meter
4.0 hr

3.0 hr

2.5 hr
2.0 hr
1.5 hr
1.0 hr

0.5 hr
0.2 hr
0 hr

50
Regulator

0 4 8 12 16 20 24 28
Nitrogen Pressure (psi)

Fig. 9-8. Schematic diagram of apparatus to measure hydrostatic pressure transmission of cement
slurries (from Levine et al., 1979). Reprinted with permission of SPE.

Chapter 9 Annular Formation Fluid Migration 297

 Chemical contraction is also responsible for a sec- the existence of free porosity composed of well-con-
ondary porosity, mainly composed of free and conductive nected pores that begin to appear upon the initiation of
pores (Parcevaux, 1984b). At the same time, interstitial the setting period. The same author went on to confirm
water is trapped within the pores through physicochem- that gas migration is driven by an unsteady permeability
ical and capillary forces and can no longer move when effect through the cement pores. After an initial enlarge-
submitted only to its own hydrostatic pressure gradient. ment of the cement pores, a pseudo-steady state is
The combination of chemical shrinkage and secondary achieved when communication has been established
porosity is responsible for the sharp cement pore-pres- throughout the cement column and gas channels have
sure decrease from the water gradient to the formation reached a stable size.
pressure, or even to less than the atmospheric pressure In 1986, Stewart and Schouten confirmed and
if the system is isolated. This was observed by Levine et expanded upon the earlier results of Levine et al. (1979).
al. (1979) and described by Stewart and Schouten They concluded that, when a stable cement slurry (i.e.,
(1986). featuring negligible particle settling) enters the transi-
Contrary to earlier work, Nishikawa and Wojtanowicz tion state, it begins to gel, and the hydrostatic pressure
(2002) reasoned that chemical shrinkage was not decreases ultimately to that of its water phase. When ini-
responsible for annular pressure reduction. Their con- tial setting commences, this pressure (now a pore pres-
clusion was based on experiments involving a solid steel sure) decreases further. In the same paper, Stewart and
rod suspended in a cylindrical container filled with Schouten questioned the validity of static gel strength
cement. The weight of the rod was continuously mea- for describing the potential pressure restriction in
sured as the cement was setting. The theory behind the Eq. 9-1, arguing that this equation assumes the slurry
experiment was that cement shrinkage would instigate a acts as a coherent “one-phase body.” Such an assump-
downward movement of the cement column while gela- tion is valid for pumping applications but not for cases in
tion would increase the friction at the surface of the rod. which the slurry is depressurized internally by fluid loss
Therefore, if gelation and shrinkage occurred concur- or hydration.
rently, the rod would be pulled down by friction, causing
the measured weight of the rod to increase over time. In
the experiments, they did observe a drop in the level of 9-3.3.5 Free fluid (free water)
the cement column; however, they recorded no weight The effect of cement free-water separation was studied
change, leading them to the conclusion that chemical and discussed by Tinsley et al. (1979) and Webster and
shrinkage does not contribute to annular pressure Eikerts (1979). The former concluded through pilot-
decline. This experiment did not take into account the scale experiments that, although undesirable, free water
reduction in pore pressure that would have occurred in is not an influential factor with respect to annular gas
the cement slurry, which would have had an offsetting flow. The latter group studied the problem by construct-
effect on the frictional forces. The experimental setup ing a 9-ft long acrylic model, inclined up to 70° and con-
may also not have had sufficient hydrostatic overburden nected to a gas entry source and several pressure sensors
to cause a significant degree of compaction. (Fig. 9-10). They observed that, in deviated holes, the
free water could coalesce to form a continuous channel
on the upper side of the hole, forming a privileged path
9-3.3.4 Permeability by which the gas may migrate. Thus, they recommended
The concept of gas migration through the pore structure cement slurries that develop essentially no free water.
of a very permeable gelled or set cement, as well as the Despite their observations in the laboratory-scale
potential gas percolation within the gelling slurry that model, Webster and Eikerts were unable to establish a
can occur beforehand, was first introduced by clear relationship between the angle of deviation and
Guyvoronsky and Farukshin (1963). During the period of the importance of the water channel. The nomenclature
hydrostatic pressure reduction, the cement matrix per- for this laboratory test was changed from free water to
meability was measured to be as high as 300 mD. In free fluid in the 1997 edition of API Recommended
1982, Cheung and Beirute proposed a gas migration Practice (RP) 10B, Recommended Practice for Testing
mechanism, based on laboratory experiments, by which Well Cements (Appendix B). The new procedure pre-
the gas first invades cement pore spaces and eventually scribes cement-slurry conditioning in a pressurized con-
permeates the entire cement matrix; consequently, the sistometer at bottomhole circulating temperature
hydration process is prevented from closing the pore before placement in a tube tilted at the same angle as
spaces. Parcevaux (1984b) further refined this mecha- the hole. Webster and Eikerts (1979) and Bergeron and
nism by studying the pore-size distribution of cement Grant (1989) recommended that testing be performed at
slurries during thickening and setting. He demonstrated a 45° angle, the most severe test condition.

298 Well Cementing

 Free water expansion of the casing. Later, when the casing is
opened after the cement has set, the pressure inside the
Channeling effect casing will drop. The heat generated by cement hydra-
tion will also eventually dissipate. Consequently, the
Pressure
measurement
casing diameter will return to its original size.
valve Finally, a bulk volume reduction of the cement sheath
Angle of owing to chemical shrinkage may result in a microannu-
inclination Nitrogen source lus. However, laboratory measurements have shown this
to be negligible under conditions normally encountered
in a cemented annulus (Baumgarte et al., 1999).

9-3.3.8 Cement sheath mechanical failure

Fig. 9-10. Schematic diagram of a model showing fully developed Cracking of the cement sheath caused by imposed ten-
water channeling (from Webster and Eikerts, 1979). Reprinted with
permission of SPE.
sile stresses, compressional stresses, or both (Fig. 9-11)
can form a gas migration path in the annulus. Wellbore
stresses that may be responsible for cement sheath fail-
ure can result from changes in wellbore temperature
9-3.3.6 Mud removal and pressure, tectonic stresses, subsidence, and forma-
tion creep. The relative strength of the formation behind
Early work attributed gas migration problems to poor
the cement sheath is known to have an impact on deter-
mud removal, poor cement/casing or cement/formation
mining whether a given cement sheath will fail when
bonding, or both (Carter and Evans, 1964; Carter and
exposed to wellbore stresses. Formations that are hard,
Slagle, 1970). While several other important contribut-
as characterized by a high Young’s modulus, will confine
ing factors have since been identified, effective mud
the cement sheath and make it less susceptible to crack-
removal is still recognized as imperative for controlling
ing. Formations that are relatively soft, as characterized
annular gas migration. The importance of effective mud
by a low Young’s modulus, will not provide sufficient con-
removal in controlling gas migration cannot be underes-
finement to prevent cracking. A more detailed discus-
timated because, regardless of the properties of the
sion on cement sheath mechanical failure is presented
cement placed in the annulus, a continuous mud chan-
in Chapter 8.
nel between two permeable zones will favor fluid migra-
tion. A detailed discussion on the fundamentals of mud
displacement mechanics and guidelines can be found in
Chapter 5.
Tangential stress Microannulus

9-3.3.7 Microannulus
Tensile
Another path for gas to migrate is through a microan- ΔT failure
nulus—a gap that may form between the cement sheath
and the casing or the formation after the cement has set.
Compressional
A microannulus is a common occurrence that can result Radial stress failure
from any number of events during the life of a well. It has
long been known that a pressure decrease inside the
wellbore after the cement has set will result in a casing- Fig. 9-11. Tensile and compressional failure of the cement sheath
from wellbore stresses.
diameter reduction, leading to the formation of a
microannulus. This commonly occurs when the density
of fluid inside of the casing is reduced after the cement
job. A wellbore-temperature decrease will also reduce
the casing diameter. 9-4 Predicting short-term gas migration
An example of a situation in which both pressure and Gas migration is a complex physical phenomenon that
temperature reduction can occur is when the casing is comprises many facets; as a result, physical modeling of
closed in at the end of the cement job. The exotherm this phenomenon is a formidable problem. It is a non-
generated by cement hydration will cause thermal steady-state process involving changing pressure fields
expansion of the steel casing. In addition, fluids trapped and fluid saturations, and an evolving matrix structure.
inside the casing will heat up, causing further thermal

Chapter 9 Annular Formation Fluid Migration 299

Heterogeneities within the cement slurry, or boundary 9-4.2 Expanding the factors for gas flow potential
effects at the casing or formation, can induce singular Later, a more detailed method developed by Rae et al.
events (such as nonuniform gas breakthrough) that are, (1989) focused on four factors whose components are
by definition, unpredictable. Therefore, it is impossible considered fundamental to the occurrence of gas migra-
to predict the occurrence of gas migration, nor its defin- tion.
itive solution, on an absolute basis. Various prediction
approaches have been developed, but none is universally 1. Formation
applicable. The most common approach to prediction is 2. Hydrostatic pressures in the wellbore
a systematic well analysis, resulting in a purely qualita- 3. Mud removal efficiency
tive risk assignment. 4. Slurry performance
The calculations are based on reservoir deliverability,
9-4.1 Gas flow potential annular geometry, gas-zone pore pressure, hydrostatic
Sutton et al. (1984) described one of the first systematic pressure, mud removal efficiency, cement hydration
approaches to gas migration prediction by calculating a kinetics, and fluid-loss control.
relative gas flow potential (GFP). The GFP is the ratio The Rae et al. (1989) method was not based on a
of the maximum pressure restriction (MPR) to the single experimental investigation or on numerical simu-
hydrostatic overbalance pressure of the well. lations; instead, it was a pragmatic compilation of the
state of the art at the time it was developed. A statistical
K mpr analysis of data from a wide variety of gas wells in the
K gfp = (9-10) United States, Canada, Latin America, Europe, Africa,
pob
the Middle East, and the Far East allowed the calcula-
MPR is further defined as tion of semiempirical relationships between the four fac-
tors. Rae et al. claimed that the wide range of field con-
⎛ ⎞ ditions covered by their method justified its use in most
L
K mpr = 1.67 ⎜ ⎟, (9-11) real cases.
⎝ dhole − d pipe ⎠

where
9-4.2.1 Formation factor
The first factor, the formation factor, is a dimensionless
dhole = diameter of the open hole (in.)
ratio of the reservoir’s productive capacity, kh, to the
dpipe = diameter of the pipe (in.) critical volume, Vcrit (Eq. 9-12). The critical volume is the
Kgfp = gas flow potential additional cement porosity that is created during setting
Kmpr = maximum pressure restriction by chemical shrinkage between the top of the gas zone
and the pressure balance point. A slurry porosity of 2% is
L = cement column length (ft)
assumed at this stage of hydration, and gas is assumed to
pob = overbalance pressure (psi). permeate the annulus in a uniform fashion over the
The GFP is a dimensionless number that can vary defined length. Assuming all other factors remain con-
between 0 and infinity, and the severity of the potential stant, as the value of the formation factor increases, the
gas migration problem is rated according to Table 9-2. risk of gas migration increases.

kh 467.7 khρslurry
fform = = , (9-12)
⎡
) − ( d pipe ) ⎥⎦
2⎤
Table 9-2. GFP Severity Ratings Vcrit
(
pob ⎢ dhole
⎣
2

GFP Severity Rating

<4 Minor where
4–8 Moderate dhole = hole diameter (in.)
>8 Severe dpipe = pipe diameter (in.)
h = zone height (ft)
k = zone permeability (mD)
pob = overbalance pressure (psi)
ρslurry = cement-slurry density (lbm/gal).

300 Well Cementing

9-4.2.2 Hydrostatic factor Table 9-3. Mud Removal Guidelines
The hydrostatic factor (HF) stems from the work of Excellent Moderate
Levine et al. (1979), who observed that the hydrostatic
Hole in excellent condition Hole in excellent condition
pressure exerted by cement slurries decreases, before cementing before cementing
approaching that of the interstitial water as gel strength
increases. They assumed that further pressure decay Circulate one hole volume Circulate one hole volume
occurs only after the cement has set. In theory, when No gas Condition mud
cementing to the surface, gas zones with pore pressures
greater than the hydrostatic pressure of water can flow Condition mud
as soon as the cement gels. When a mud column remains
Greater than 67% standoff Greater than 50% standoff
above the cement, this must be taken into account as an
additional hydrostatic pressure source that is summed Minimal U-tubing Minimal U-tubing
with that of the cement interstitial water. Thus, the
Compatible fluids Compatible fluids
hydrostatic factor is represented by the ratio of the gas-
zone pore pressure to that of the annular pressure at the Use of washes/spacers Use of washes/spacers
commencement of the initial set (Eq. 9-13). Again,
Engineered displacement Engineered displacement
greater values of the hydrostatic factor indicate a
increased risk of gas migration in a given well situation. 10-min spacer contact time 10-min spacer contact time
pgas Use of two bottom plugs Rotation and reciprocation
fhyd = 19.281 ,
) ( )(
of pipe
(
⎡ρ ×h
⎣ mud mud
+ ρ sp × hsp + 8.32 × hcem ⎤
⎦ ) Rotation and reciprocation
of pipe
(9-13)

where calculated from the results of standard API laboratory

fhyd = hydrostatic factor fluid-loss and thickening time tests.
hcem = height of cement in the annulus (ft)
hmud = height of mud in the annulus (ft)
N sp =
q API ( t100 Bc − t30 Bc ), (9-14)
hsp = height of spacer in the annulus (ft) 30
pgas = gas zone pressure (psi)
ρmud = mud density (lbm/gal) where
ρsp = spacer density (lbm/gal). Nsp = slurry performance number
qAPI = API fluid-loss value of slurry (mL/30 min)
t30Bc = time to 30 Bearden units of consistency (min)
9-4.2.3 Mud removal factor
t100Bc = time to 100 Bearden units of consistency (min).
The mud removal factor (MRF) described by Rae et al.
(1989) was a subjective scale from 1 to 10, based upon how The term
closely a defined set of industry standards for mud removal
(Table 9-3) were followed. More quantifiable means of pre- t100 Bc − t30 Bc
dicting mud removal have since been developed
(Chapter 5) that supersede these qualitative standards. is known as the transition time. It must be emphasized
Consequently, the MRF has been largely superseded by that this equation does not represent the actual perfor-
more sophisticated methods that are described later. mance of the slurries under static downhole conditions
in which the mud filtercake influences the leakoff.
Instead, the SPN provides a method of comparing slurry
9-4.2.4 Slurry performance number performance on a relative basis and provides a useful
The final factor considered by Rae et al. (1989) was the tool in both the design and evaluation of cement pro-
slurry performance number (SPN), which was devel- grams for gas wells. Slurries with high SPNs are very
oped to rank cement systems according to their hydra- poor candidates for gas migration control. Those with
tion kinetics and fluid-loss control. It is based on an low API fluid-loss rates and short critical hydration peri-
approximation of the interstitial water lost as the ods offer a much greater probability of success.
cement undergoes the initial hydration process, which is

Chapter 9 Annular Formation Fluid Migration 301

9-4.3 Recent developments in gas migration where
prediction p = pressure (psi)
More recently, improvements have been made to the Rae T = temperature (°F).
et al. (1989) gas migration prediction approach. These
changes involve a more detailed definition of the forma- One can assume that there is a critical volume of
tion parameter (FP), a more quantitative determination space available for the gas to occupy in the annulus
of the mud removal parameter (MRP), and the calcula- before the gas migration problem becomes serious. This
tion of a pressure decay limit parameter (PDLP), which is called the critical annular volume. As previously dis-
replaces the previous HF. Additionally, the new approach cussed, this space is created by fluid-loss, shrinkage, and
does not use an SPN as part of the risk assessment slurry permeability. A conservative assumption is that
process. Instead, it recommends slurry types that have the available space for gas to occupy is 5% of the total
been qualified through laboratory testing based upon slurry volume above the gas zone. Therefore, the critical
the predicted level of risk from the other three parame- annular volume is calculated as 5% of the total slurry
ters (N. Quisel, unpublished results). volume from the top of the gas-bearing zone to the top of
the caprock for that zone.

⎡
)2 − ( d pipe ) ⎤⎥⎦ L
9-4.3.1 Formation parameter
(
2
0.05 π ⎢ dhole
The revised FP is based upon the reservoir’s ability to ⎣
deliver gas. To estimate the produced gas volume, Vgas, Vcrit = , (9-17)
4
across a unit section, the basic steady-state relation for
gases is used. where
dhole = hole diameter (m)
⎡
( ) − ( p ) ⎤⎥⎦
2 2
πkht ⎢ p pore ann
dpipe = pipe diameter (m)
Vgas = ⎣ , (9-15) L = length from top of gas zone to top of caprock (m)
⎛ r ⎞ Vcrit = critical annular volume
μ gas × pann ⎜ ln res + s⎟
⎝ rhole ⎠ 0.05 = the coefficient for space available gas in
the slurry.
where The FP can then be calculated from the ratio of the
h = gas zone length (m) produced gas volume to this critical annular volume by
k = formation permeability (m2) combining Eq. 9-15 and Eq. 9-17.
pann = annular hydrostatic pressure at top of gas Vgas
zone (Pa) K fp =
Vcrit
ppore = pore pressure at top of gas zone (Pa)
(
⎡
) ( ) 2⎤
2
rhole = hole radius (m) 80 kht ⎢ p pore − pann ⎥
rres = reservoir radius (m) = ⎣ ⎦
⎛ ⎞
s = dimensionless skin factor (s = 0 by default). ⎡
( ) ( ) ⎤ r
2 2
⎢ dhole − d pipe ⎥ Lμ gas × pann ⎜ ln res + s⎟
t = production time estimated from setting time (sec) ⎣ ⎦ ⎝ rhole ⎠
Vgas = gas volume, downhole conditions (m3)
μ gas = gas viscosity (Pa-s). (9-18)
The skin factor is highly dependent upon the base where
fluid of the drilling mud (oil or water). A typical skin Kfp = formation parameter.
value is equal to 20.
The gas viscosity, μgas, depends on pressure and tem- Various levels of severity may then be empirically
perature conditions. It is given by Eq. 9-16. This particu- assigned to the value of the FP (Table 9-4).
lar equation is for methane.

( )(
μ gas = 9.76 × 10 −6 + 0.0126 × 10 −6 T + )
{ ( )( )}
p ⎡⎢ 3.22 × 10 −9 − 4.84 × 10 −12 T ⎤⎥ ,
⎣ ⎦
(9-16)

302 Well Cementing

Table 9–4. Severity Ratings Versus FP Cement concentration
FP Severity Rating i = nZ 0% 100%

>1 Very critical

>0.75 to 1 Critical

>0.5 to 0.75 High

>0.25 to 0.5 Moderate

0 to 0.25 Low
i=0
j=0 j = nφ MD
2D Wellbore Concentration
Mesh Grid Versus Depth
9-4.3.2 Mud removal parameter
The second parameter is the MRP. With the advent of Fig. 9-12. Mathematical simulator for the MRP determination.
improved two-dimensional mathematical simulators for
fluid displacement (Chapter 5), the ability to quantify
the effectiveness of mud removal has advanced signifi- causing the annular pressure to drop below the hydro-
cantly in recent years. As shown in Fig. 9-12, such a static pressure. However, the fluid column has lower
mathematical simulator can be used to calculate cement pressure limits at all depths along the wellbore:
concentration on a mesh grid representing a wellbore. ■ pressure applied at the top of the annulus (normally
Perfect displacement would be predicted if a cement atmospheric pressure)
concentration value of 100% were indicated across the ■ pore pressure in front of permeable zones
entire length.
■ vapor pressure of water in front of impermeable zones
Using industry-accepted practices that have been
developed from field experience, positive zonal isolation (including casing-in-casing annuli).
will usually be achieved with 500 ft of cement coverage The pressure cannot drop below these limits at any
above the top of the gas-bearing zone. An MRP can thus depth. Gas migration can occur only when the annular
be calculated over this zone. pressure at a given depth drops to a value equal to or less
Dtog +500 ft
than the pore pressure of a gas-bearing zone at that
depth. The shear stress at the wellbore wall that causes
∫ ( Ycc ) dz ,
1
K mrp = (9-19) the pressure to reach this critical value for gas entry is
h Dbog the PDLP. The following equation uses oilfield units.

where
Dbog = Depth to bottom of gas zone K pdlp =
(
pob dhole − d pipe ), (9-20)
4L
Dtog = Depth to top of gas zone
h = length from bottom of gas zone to 500 ft above where
top of gas zone Kpdlp = pressure decay limit parameter
Kmrp = mud removal parameter L = the length of the cement column above the gas-bear-
Ycc = cement concentration value (calculated by ing formation
mathematical simulator). pob = overbalance pressure at the end of cement
placement, further defined as
Although the MRP is a useful design tool, one should
not forget the importance of the good cementing prac- pob = pmud + psp + pcem + pback − p pore , (9-21)
tices outlined in Chapter 5.
where
9-4.3.3 Pressure decay limit parameter pback = backpressure (i.e., atmospheric pressure + any
applied backpressure)
The third parameter is the PDLP, which is based upon
the concept of critical wall shear stress described by pcem = hydrostatic pressure from the cement column
Stiles (1997). When a fluid is placed inside a pipe or pmud = hydrostatic pressure from the mud column
annulus, a shear stress may be created along the wall,

Chapter 9 Annular Formation Fluid Migration 303

ppore = pore pressure of the gas-bearing formation ■ Reduce the matrix permeability of the cement.
psp = hydrostatic pressure from the spacer/wash column. ■ Increase the rate of gel strength development.
■ Increase the PDLP.
Various levels of severity may be empirically assigned to
the value of the PDLP (Table 9-5). Decreasing the upper boundary limit can be achieved
It is very important to note that, by definition, the by reducing the matrix permeability of the cement
PDLP is a function of density only. Other slurry proper- (Fig. 9-14). This is accomplished by altering the cement-
ties do not apply. The PDLP is entirely defined by the size slurry design, for example by adding small-particle con-
of the annular gap, the hydrostatic contributions of the stituents. However, because the y-axis is logarithmic, it is
fluids in the well, the length of the fluid column, and the apparent that a relatively large decrease in matrix per-
relevant pressure boundaries. Therefore, the only change meability will result in a relatively small decrease in the
in slurry design that will affect the PDLP is density. CHP.
Increasing the slope of the gel strength–develop-
ment curve has a major effect on the CHP (Fig. 9-15).
Table 9-5. Severity Rating Versus PDLP This can also be accomplished by modifying the cement-
slurry design.
PDLP Severity Rating
0–25 Pa (0–50 lbf/100 ft2) Very critical
10,000
25–75 Pa (50–150 lbf/100 ft2) Critical
Impermeable matrix
75–150 Pa (150–300 lbf/100 ft2) High
1,000
150–250 Pa (300–500 lbf/100 ft2) Moderate
Gel strength PDLP Gel strength
>250 Pa (>500 lbf/100 ft2) Low (lbf/100 ft2) 100

10
CHP
9-5 Theoretical strategies for combating
short-term gas migration 1
Critical Upper time
Theoretically, the root causes for gas migration time boundary
(described in Section 9-3.1) can be combated by manag- Time
ing the annular pressure decline, reducing the space for
entry, and minimizing the path for migration. For short- Fig. 9-13. Plot of gel strength development versus time to define the
CHP.
term gas migration, these strategies must be addressed
during the postplacement period.
One concept for evaluating the postplacement period
is to plot the evolution of gel strength over time (Stiles, 10,000
1997). Because gel strength development tends to be
logarithmic, the log of gel strength can be plotted as a 1,000
straight line versus time (Fig. 9-13). The PDLP can be Impermeable matrix
calculated from Eq. 9-20 and plotted on the gel strength
graph so that the intersection of this value with the gel Gel strength 100
(lbf/100 ft2) PDLP
strength curve will define a critical time, tc, when gas
can first enter the annulus. An upper time boundary, tf,
10
represents the time beyond which the gel strength is too CHP
high to allow gas migration. This occurs after the cement
begins to set and become an impermeable matrix. tf is 1
plotted in the same manner as tc, by drawing a vertical Critical Upper time
line from the x-axis to the gel strength–development/ time boundary
impermeable-matrix gel strength intersection. The time Time
between tc and tf is the critical hydration period (CHP).
Fig. 9-14. Strategy for shortening the CHP by reducing the matrix
Stiles recognized three distinct strategies for permeability of the cement.
shortening the CHP and thus reducing the risk of gas
migration.

304 Well Cementing

 10,000 10,000
Impermeable Impermeable matrix
matrix
1,000 1,000
PDLP

Gel strength 100 Gel strength 100

(lbf/100 ft2) PDLP (lbf/100 ft2)

10 10
CHP CHP

1 1
Critical Upper time Critical Upper time
time boundary time boundary
Time Time

Fig. 9-15. Strategy for shortening the CHP by increasing the rate of Fig. 9-16. Strategy for shortening the CHP by increasing the PDLP.
gel strength development.

Finally, the CHP can be shortened significantly by 9-6 Practical solutions for combating
increasing the PDLP (Fig. 9-16). By reviewing Eq. 9-20, gas migration
which defines the PDLP, it can be seen that, unlike the
previous strategies, the PDLP cannot be affected by Practical strategies for combating gas migration can be
modifying any of the cement-slurry properties (with the classified according to the factors previously outlined in
exception of the density). Instead, the size of the annu- Table 9-1. Table 9-6 presents the strategies as a function
lar gap, the hydrostatic contributions of the fluids in the of the three critical root causes and the three time-
well, the length of the fluid column, and the relevant based categories for gas migration.
pressure boundaries control this parameter. Increasing
the size of the annular gap by decreasing the pipe size 9-6.1 Physical techniques
or increasing the openhole diameter is one way to
increase the PDLP. Incidentally, this would also result It has long been known that a number of physical tech-
in a reduction of the FP (Eq. 9-18), further reducing the niques can, under certain circumstances, help control
potential for gas migration. In most cases, however, nei- gas migration. These include applying annular backpres-
ther decreasing the casing diameter nor increasing the sure or small pressure pulses to the annulus, using exter-
openhole diameter would be considered economically nal casing packers (ECPs) and liner-top packers, and
viable options for managing gas migration. reducing the cement column height (including multi-
Increasing the overbalance pressure (Eq. 9-21) stage cementing). Such techniques are certainly valid
would also increase the PDLP. This could be achieved by under a variety of conditions, but well conditions often
increasing the density of any of the wellbore fluids or by limit their application.
increasing the length of a relatively higher-density fluid Application of annular backpressure after the cement
in the annulus. The overbalance pressure is also is in place increases the overbalance pressure exerted on
increased if backpressure is applied. It is important to gas zones, thus delaying the time when gas can enter the
remember that setting an annular packer will have the annulus. However, the presence of weak zones may
opposite effect. The packer will reduce the overbalance restrict this technique because of the risk of inducing lost
pressure and drastically reduce the PDLP. The final way circulation (Levine et al., 1979).
to improve the PDLP is to decrease the length of the Another technique for delaying gas entry, first
cement column; however, this would not be desirable described by Haberman and Wolhart in 1997, involves
because the effective length of the annular seal above applying pressure pulses to the annulus after the cement
the gas zone would be shorter. is in place. The pressure pulses are applied with com-
pressed air or water at approximately 100 psi at a fre-
quency of 30 to 60 sec/pulse. The concept behind this
technique is that the pressure pulses will disrupt gel
strength development in the cement and therefore main-
tain a hydrostatic overbalance for a longer period of time.

Chapter 9 Annular Formation Fluid Migration 305

Table 9-6. Solutions for Prevention of Gas Migration
Root Causes of Gas Migration Pressure applied
Annular Pressure Space for Path for to slurry displacement fluid
–< Pore Pressure Entry Migration
Primary
Immediate Fluid density na† na cement
Short term Right-angle-set Low-porosity Packers
cements cements

Sandwich Low-porosity Sandwich

squeeze cements squeeze

Compressible Compressible Low- High radial effective

Inflation stress applied to
cement cement permeability cement
cements element/formation contact

Fluid density Compressible Surfactants

cement

Thixotropic Low-fluid-loss Thixotropic Inflatable formation

cements cements cements packers, 20- and 30-ft
lengths
Low-fluid-loss Low-fluid-loss Thixotropic
cements cements cements

Back pressure Zero-free- Low-

water permeability
cements filtercake

Annular pressure Zero-free- Low-

pulses water permeability
cements filtercake Bottom plug and
solid wiper plug
Long Term na na Packers

na na Compressible
cements Fig. 9-17. Use of ECPs (from Suman, 1984).

na na Expansive
cements

na na Flexible
annulus shortly after cement placement, thus reducing
cements the hydrostatic overbalance across gas zones below the
ECP. Slurry volume reduction below the packer, from
na na Mud removal fluid loss or chemical contraction, can then result in gas
† na = not applicable invasion of the cement in this interval at an even earlier
time. This could permit undesirable crossflow between
zones located below the packer.
ECPs (Fig. 9-17), inflated by mud or cement slurry, The technique of reducing the cement column height
control gas migration by forming a positive barrier in the stems originally from the work of Levine et al. (1979).
annulus (Suman, 1984). However, ECPs require a com- Viewing the mix-water hydrostatic-pressure gradient as
petent formation to seal against, and they complicate a natural step in the pressure reduction, they proposed
the job execution. Because of the small clearance minimizing the cement column height above the gas
between the uninflated ECP element and the borehole, zone using a very simple graphical method (Fig. 9-18).
such tools have been known to suffer mechanical The job would be designed such that the pressure sum of
damage while casing is being run or during circulation at an equivalent height of water plus the hydrostatic pres-
high rates. Also, it is not uncommon for the packers to sure above the cement would always exceed the forma-
set prematurely because of unexpected pressure fluctu- tion pressure. There is little doubt that this approach
ations during the course of the job. Parcevaux (1984a) can help the design process in a gross sense; e.g., severe
pointed out that ECPs can exacerbate some problems, risks of underbalance may be avoided. It has indeed
because they effectively isolate the lower portion of the been applied with success across some depleted sands,

306 Well Cementing

 0 (Fig. 9-19). Therefore, their ability to compensate for
volume reduction during the transition state is probably
1,000 more effective in situations close to the surface, where
11 lbm/gal mud gas expansion is more significant. In most cases it is
Casing important to maintain the foam quality, or volume of gas
2,000
at 2,000 ft contained, below about 30%; otherwise, the permeability
of the set cement may be sufficiently high to allow gas
3,000 Case 1: Top of migration.
cement at 3,000 ft
The in situ gas generators are designed to maintain
16
Depth 4,000 .6 cement pore pressure by virtue of chemical reactions
lbm
(ft) /ga that produce gas downhole. The produced gases may be
lc
5,000 Formation
em
en hydrogen (Bulatov, 1970; Sutton, 1982) or nitrogen
pore pressure t
Case 2: Top of (Richardson, 1982; Burkhalter et al., 1984). To the
cement at 6,000 ft author’s knowledge, the field application of nitrogen to
6,000
control gas migration has not been reported. Hydrogen-
8.3

generating agents such as aluminum powder have been

4l
bm

7,000 used in the former Soviet Union (Kucyn et al., 1977) and
/ga
lw

elsewhere (Tinsley et al., 1979; Watters and Sabins,

ate
r

8,000 1980). It is important to note that the gas-generating

417 psi Casing at 9,000 ft agents alone cannot prevent gas migration. Fluid-loss
0 1,000 2,000 3,000 4,000 5,000 6,000 control agents and dispersants are necessary to mini-
Pressure (psi) mize interstitial water leakoff.
The principal drawback of these systems, other than
Fig. 9-18. Comparison of cement column height adjustments (from the potential safety hazard from those that generate
Levine et al., 1979). Reprinted with permission of SPE. hydrogen, is the inability of a gas at typical downhole
pressures to achieve the 4% to 6% volumetric expansion
necessary to maintain pore pressure. Strictly applying
but it is clearly not robust enough. As noted in the same Boyle’s law, the volume of gas required to offset just the
paper, as cement changes from liquid to solid, the hydro- chemical contraction would be excessive at high pres-
static pressure falls to values far below that of water sure. Gas-generating systems must also be carefully sta-
because of fluid loss and chemical contraction. bilized; otherwise, gas bubbles may coalesce and create
An elastomeric seal ring, described by Bol et al. channels for formation gas to follow. These limitations
(1986), presents an additional line of defense for inter- notwithstanding, it is clear that this technology has been
facial migration. The success rate may be improved in used with success.
wells in which downhole stresses, such as density
changes or thermal cycling, induce casing deformation.
However, it is important to note that this device cannot 1
solve the problem of gas flow through the cement matrix; 0.9 fgas = 0
thus, it should be used in concert with other techniques. 0.8 fgas = 0.292
0.7
0.6 fgas = 0.488
9-6.2 Compressible cements Volume 0.5 fgas = 0.6
Compressible cement slurries have been developed to (dimensionless) 0.4
maintain the cement pore pressure above the formation 0.3
gas pressure. In theory, this should prevent any move- 0.2
ment of gas from the formation into the cemented annu- 0.1 fgas = 1
lus. Compressible cements fall into two main categories: 0
foamed cements and in situ gas generators. It is impor- 0 100 200 300 400 500 600
tant to draw a clear distinction between the two. Pressure (psi)
Foamed cements (Chapter 7) become nearly incom-
pressible at high pressures because of the relative Fig. 9-19. Compression of foamed cement slurries. fgas is the volu-
incompressibility of gases under such conditions metric fraction of gas in the slurry at surface conditions.

Chapter 9 Annular Formation Fluid Migration 307

9-6.3 Low-permeability cement slurries
Reducing the matrix permeability of a cement system
during the critical liquid-to-solid transition time
described earlier can prevent gas migration. Several
methods have been developed.
The first approach involved the use of water-soluble
polymers to viscosify the interstitial water of the cement
slurry. Because at least a part of gas migration involves
the displacement of cement pore water, viscosification
of the water tends to limit gas mobility. This approach
is also appropriate for fluid-loss control; unfortunately,
viscosification of the cement slurry is a major side effect
of this technique, with resultant mixing difficulties,
greater displacement pressures, and increased risk to
weak formations.
Cheung and Beirute (1982) described the use of low-
permeability cement that operates by immobilizing the Fig. 9-20. Latex film in cement after coalescence.
fluids within the pore spaces of the cement. Because the
cement mix water cannot be displaced, gas cannot move
within the pore spaces of the cement slurry. According to Blomberg et al. (1986) described another technique
Williams et al. (1983), the system is composed of a com- that uses fine mineral particulates to prepare low-den-
bination of bridging agents and polymers. Such systems sity, low-permeability cements. The preferred particu-
have been applied throughout the 140° to 350°F [60° to late in this application is silica fume (also called
180°C] bottomhole static temperature range (Cheung microsilica), a byproduct of the production of silicon and
and Myrick, 1983). ferrosilicon. The average particle size of this material is
Latex additives to prevent gas migration were first 1 μm; consequently, it is able to fill pore spaces and plug
described in a 1982 patent application by Parcevaux et pore throats. Field success has been reported (Grinrod
al. (issued 1985). Subsequent refinements of this tech- et al., 1988) for shallow, low-pressure gas.
nology (Bannister et al., 1983; Parcevaux and Sault, In the late 1990s (Al-Buraik et al., 1998; Moulin,
1984) have extended its applicability to a wide range of 2001), new polymeric microgels were developed to con-
well conditions, and its field application is well estab- trol gas migration at temperatures less than 160°F
lished (Evans, 1984; Peralta, 1984; Matthews and [71°C], at which latexes were not totally effective. The
Copeland, 1986; Rae, 1987; Drecq and Parcevaux, 1988). gas migration control is primarily attributed to the abil-
Latexes are aqueous dispersions of solid polymer par- ity of the microgels to rapidly plug pore throats in the
ticles, including surfactants and protective colloids, cement filtercake, forming an impermeable barrier
which impart stability to the dispersion. Most latexes across the gas zone. Gas migration control is also attrib-
have film-forming capabilities; thus, when contacted by uted to the fact that the microgels are smaller and less
a gas, or when the particle concentration exceeds a dense than cement particles, and they help to reduce
given threshold value, latex particles coalesce to form an pore-throat continuity (and therefore permeability) in
impermeable polymer barrier. In a wellbore, the gas first the cement matrix. As a solid structure develops in the
invades the portion of the cemented annulus across the setting cement, the smaller pore throats reduce the size
gas zone and contacts the dispersed latex particles in of the gas bubbles that can enter the cement, slowing
the slurry. As shown in Fig. 9-20, the latex coalesces their subsequent migration.
within the pore spaces, blocking further progress up the Designing cement systems with a multimodal parti-
annulus. cle-size distribution can result in cement slurries with
Latexes have a number of other beneficial properties considerably less porosity and permeability than a con-
when used in cement slurries (Parcevaux, 1987). The ventional system. The concept, first described by Moulin
small, spherical latex particles act as lubricants, impart- et al. (1997), is based upon maximizing the particle
ing excellent rheological properties. Fluid-loss control is volume fraction in the dry blend by using three or more
provided by a mechanical plugging mechanism. Civil engi- distinct particle sizes. This leads to an increased solids
neers have long recognized the shrinkage-compensating content in the cement slurry and thus a reduced poros-
and bonding actions of latexes. Such attributes translate ity and permeability. More information on cements with
to improved shear-bond strength and elastic deformabil- an engineered particle size distribution can be found in
ity in well cements (Parcevaux and Sault, 1984). Chapter 7.

308 Well Cementing

9-6.4 Fluid loss and free-fluid control additives (Chapter 7). Some degree of fluid-loss control
Fluid loss and free fluid, discussed previously in Sections for thixotropic slurries can also be obtained by the use of
9-3.3.1 and 9-3.3.5, promote the occurrence of gas migra- low-fluid-loss spacer fluids (Bannister, 1978).
tion. To minimize the effect of these parameters, both Successful field results have been obtained in shallow,
must be reduced to fairly low levels. Fluid-loss rates of 50 low-temperature applications (Sepos and Cart, 1985).
mL/30 min or less, and free-water values of 0.25% or less, Stehle et al. (1985) reported good results at greater tem-
have been reported in the literature as requirements for peratures (250° to 280°F [120° to 140°C]) when cement-
slurries used in gas migration situations (Baret, 1988; ing liners and long strings.
Webster and Eikerts, 1979).
Latexes, anionic synthetic polymers, and some cellu- 9-6.6 Surfactants
losic derivatives (at low temperature) are able to pro-
The use of surfactants in cement slurries and preflushes
vide low fluid-loss rates without inducing free-water sep-
was first described by Marrast et al. (1975). Surfactants
aration. However, many of them may affect other
may, under the right circumstances, entrain invading gas
cement-slurry properties, including gel strength devel-
downhole to create stable foam. This foam presents sig-
opment and thickening time, in a deleterious fashion.
nificant resistance to flow of discrete gas bubbles,
Defossé (1985) described a series of metallic salts that
thereby limiting their upward migration through the
depress free-water development, yet are not antagonis-
cement column. Stewart and Schouten (1986) reported
tic to other aspects of slurry performance. A detailed dis-
the technique to be effective, particularly when com-
cussion of fluid-loss additives is presented in Chapter 7.
bined with the use of elastomeric seal rings, described
earlier.
9-6.5 Thixotropic cement slurries
Carter and Slagle (1970) identified slurry gelation as 9-6.7 Right-angle-set cements
a major potential cause of gas migration. However, the
Right-angle set (RAS) cement slurries can be defined as
work of Sabins et al. (1982) indicated that high gel
well-dispersed systems that show no progressive gelation
strength may help resist gas percolation; for this reason,
tendency yet set almost instantaneously because of
they proposed thixotropic and high-gel strength cements.
rapid hydration kinetics. Such systems are capable of
As discussed in Chapter 7, thixotropic cements may
maintaining a full hydrostatic load on the gas zone up to
be prepared by a number of methods, including the addi-
the commencement of cement setting, at which time
tion of bentonite, certain sulfate salts, or crosslinkable
they develop a low-permeability matrix with sufficient
polymers to a Portland cement slurry. In all cases, the
speed to prevent significant gas intrusion. It should be
transmitted hydrostatic pressure of a thixotropic system
pointed out that the transition time involved here is not
should revert to that of its interstitial water and remain
the same as that described by Sabins et al. (1982), nor is
as such until the setting period begins. Therefore,
the mechanism similar to that of high–gel strength sys-
thixotropic systems are unlikely to be effective in situa-
tems (Kieffer and Rae, 1987). Unlike high–gel strength
tions in which the gas-zone pressure exceeds the water
systems, RAS systems undergo a true set involving the
gradient, unless additional backpressure is held on the
deposition and recrystallization of mineral hydrates.
annulus.
RAS slurries are sometimes characterized as such by
The high gel strength of thixotropic cements can offer
standard high-pressure, high-temperature (HPHT)
considerable resistance to physical deformation and per-
thickening time tests, as shown by Drecq and Parcevaux
colation by a large gas bubble. However, as discussed
(1988). A RAS slurry maintains a low consistency until
earlier, the bubbles may often be smaller than the pore
setting, when the slurry viscosity increases to more than
spaces in the setting cement. Under such circumstances,
100 Bc within a few minutes. The increase in consistency
gas migration may occur without slurry deformation, and
is accompanied by a temperature increase resulting
gel strength is no longer an effective barrier.
from the exothermic cement hydration reactions taking
Thixotropic cement slurries tend to have high fluid-
place (Fig. 9-21).
loss rates; therefore, the risk of dehydration and bridg-
It is difficult to design RAS systems for temperatures
ing must be considered. Sykes and Logan (1987) found
below 250°F [120°C] because of the slower cement
the influence of fluid loss to be preponderant to that of
hydration kinetics at lower temperatures. It must also be
gel strength immediately after placement, and they rec-
noted that the shear imparted to the slurry during the
ommended designing the slurry to be well dispersed
API thickening time test varies significantly from that
until after the bulk of fluid-loss volume reduction has
which occurs downhole during a cementing operation,
occurred. The fluid-loss problem has been reduced
making the set profile on a thickening time test merely
somewhat by the development of improved fluid-loss
a qualitative measure.

Chapter 9 Annular Formation Fluid Migration 309

 9-6.9 Flexible cements
Exothermic
308°F reaction
Cements with improved flexibility have been shown to
100 reduce the potential for stress-induced cement-sheath
Viscosity cracking that may lead to long-term gas migration.
(Bc) Thiercelin et al. (1997) showed that the risk of failure is
50 related to the ratio of the cement tensile strength to the
Right-angle
set property Young’s modulus. The larger this ratio (i.e., cement with
relatively high tensile strength and low Young’s modu-
lus), the more resistant the cement is to failure.
1 2 3 4 5
Onan et al. (1993) reported the use of elastomeric
Time (hr) polymers to improve set-cement flexibility. Le Roy-
Fig. 9-21. Pressurized consistometer output from an RAS cement
Delage et al. (2000) incorporated flexible particles such
system (from Drecq and Parcevaux, 1988). as rubber, thermoplastics, or latexes in cement systems
as a method for improving flexibility. Foamed cement
has also been shown to improve flexibility (Deeg et al.,
1999). More detailed information about these methods
9-6.8 Expanding cements can be found in Chapter 7.
Cements that exhibit bulk expansion have been advo-
cated in places in which a microannulus is the gas
migration pathway, and successful field results have 9-7 Laboratory testing
been reported (Seidel and Greene, 1985). As discussed At this writing, neither API nor the International
in Chapter 7, there are two principal techniques for Organization for Standardization have published a stan-
inducing bulk expansion in Portland cement: crystal dard method or procedure for testing gas migration con-
growth and gas generation. The latter operates on the trol in cement slurries; therefore, there is no recognized
same principle as the compressible cements mentioned standard laboratory procedure to characterize the ability
above, with the exception that the concentration of gas- of a cement system to prevent or reduce gas migration.
generating material (typically aluminum) is reduced A significant number of nonstandard laboratory tests
(Sutton and Prather, 1986). The former relies upon the for gas migration has been developed by the industry. A
nucleation and growth of certain mineral species within wide variety of experimental prototypes that attempt to
the set-cement matrix. The bulk volumetric expansion simulate the gas migration process are described in the
that can be attained through crystal growth is usually literature. Two main types of experimental simulators
controlled to be less than 1% (Griffin et al., 1997). The exist: large-scale pilot devices, which attempt to repro-
bulk expansion provided by either of these mechanisms duce the process as it occurs in the wellbore, and small-
does not eliminate internal chemical shrinkage that has scale, benchtop models, which can be used to derive the
previously been discussed as a cause for gas migration. It fundamental laws of a particular physical process under
only acts to increase the bulk or dimensional volume of investigation. To date, none of the simulators described
the cement. in the literature permits the derivation of a physical
Experimental work performed by Beirute et al. model that quantitatively describes gas migration over a
(1992) found that expanding cements are not effective wide range of conditions. Recently, many investigators
at sealing microannuli along the casing/cement inter- have begun to rely upon measuring individual slurry
face if the adjacent formation does not impose an ade- properties such as gel strength development and shrink-
quate confining stress. Their laboratory studies indi- age. These data are then used either as input into math-
cated that the cement sheath can actually expand away ematical simulators or for comparing the performance
from the casing when confined by soft formations. differences between various slurries.
Further experimental work by Baumgarte et al. (1999)
and Le Roy-Delage et al. (2000) confirmed this phenom-
enon, leading to recommendations to use expanding
cements only across relatively hard formations that can
resist the expansion and impose stresses in the cement
sheath that force it to expand towards the casing.

310 Well Cementing

9-7.1 Large-scale testing Computer Gas-pressure Gas
Typically, large-scale models have been constructed in an system flowmeter inlet
effort to better understand gas migration phenomena
and are not practical for conducting routine experiments Gas Heated water out
for individual well designs. The earliest large-scale simu- Cement out
Top
lation was performed by Carter and Slagle (1970) and Flexible
later upgraded by Carter et al. (1973). In 1976, Garcia polymer plug
and Clark constructed a large-scale device specifically to Pressure and (optional)
temperature
study the influence of uneven cement setting. Levine et transducers Cement
al. (1979) described a simulator for studying hydrostatic
pressure profiles within a cement column at rest
Middle 10 ft
(Fig. 9-8). The apparatus built by Tinsley et al. (1979)
investigated the influence of fluid loss and compared dif-
ferent cement systems (Fig. 9-22). Finally, the equip- Water jacket
ment described by Bannister et al. (1983) evaluated the
influence of filtercake formation from cement fluid loss
and the conductivity to gas of setting cement (Fig. 9-23).
Cement in
Bottom
Heated water in
To constant-pressure 200-mesh screen
water source 300-mL void
48.3 cm Gas-pressure
Chamber charged with Gas
water to 3,447.5 kPa or flowmeter inlet
6,895 kPa
Fig. 9-23. Annular gas flow laboratory testing apparatus (from
Slurry-fill 2.1 m
Fluid-loss vent Bannister et al., 1983). Reprinted with permission of SPE.
vent valve
holes when using
permeable section

Permeable or 45.7 cm
9-7.2 Small-scale gas migration testers
nonpermeable
section Various bench-scale or benchtop devices for characteriz-
ing gas migration have been described in the literature.
The first, described by Cheung and Beirute (1982), used
Check valve 1.0 m a modified API fluid-loss cell to investigate the hydro-
Hot-water jacket static pressure decrease and subsequent gas migration
Gas-entry line
from volume in a setting cement column (Fig. 9-24). Adaptations of
325-mesh screen measurement this model have recently been made commercially avail-
device
able and are known as gas migration analyzers or gas
Rubber diaphragm
Slurry 49.5 cm flow analyzers (Fig. 9-25). The test cell is a modified API
fill line HPHT fluid-loss cell with a hollow hydraulic piston at the
To pressure
recorder Thermocouple top of the cell. The piston is pressurized with mineral oil
To temperature recorder to simulate the hydrostatic overbalance. Fluid loss can
occur at both the top and the bottom of the cell, either
Fig. 9-22. Schematic diagram of test fixture used to study gas through standard 325-mesh fluid-loss screens or through
leakage (from Tinsley et al., 1979). Reprinted with permission of SPE.
actual formation cores. Once the cement slurry has
developed a predetermined gel strength, gas is injected
into the bottom of the cell at an appropriate differential
pressure. Gas flowmeters and pressure transducers mea-
sure any gas migration through the slurry. A separate gel
strength–development test (described in Section 9-7.3)
must be performed on the slurry before performing this
gas migration test.

Chapter 9 Annular Formation Fluid Migration 311

 While this device has been adapted for routine use, it Stewart and Schouten (1986) investigated gas migra-
is important to note that, at the scale of this test, three tion in set and hard cement using a U-tube apparatus
factors can unduly affect the gas migration process. like that shown in Fig. 9-26 (Richardson, 1982).
Fluid loss could result in the formation of an impenetra- Parcevaux (1984a) and Drecq and Parcevaux (1988)
ble filtercake at the gas inlet or outlet. Free-water devel- described a small-scale simulator that eliminated some
opment could artificially delay the pore pressure of the limitations of earlier devices. As illustrated in
decrease by reabsorption during hydration. Finally, in Fig. 9-27, the artificial effects of fluid loss and free water
view of the length of the cement column versus the were eliminated, and the external curing pressure was
external applied pressure, such an experiment can only computer-controlled to maintain a differential pressure
consider gas migration across a short interval. close to zero between the top and bottom of the cell.

Nitrogen gas
Top valve Backpressure receiver
Nitrogen gas

Pressure To temperature controller

Mineral
regulator
oil tank Piston
O-ring Thermocouple

Pressure transducer

325-mesh screen
Slurry
To recorder

Nitrogen gas
Bottom valve Gas Pressure
flowmeter regulator

Figure 9-24. Gas flow simulator (from Cheung and Beirute, 1982). Reprinted with permission of SPE.

Fig. 9-25. Gas flow apparatus (photo courtesy Chandler

Engineering LLC).

312 Well Cementing

 Gas Another bench-scale device called the cement hydra-
source tion analyzer (CHA) was first described in a 1997 API
Pressure
gauges Technical Report (TR) 10TR2, Technical Report on
V Shrinkage and Expansion in Oilwell Cements. This
device was developed to measure the evolution of criti-
V Oil cal parameters during cement hydration and to better
discharge
Standpipe understand the mechanisms of gas entry downhole.
U-tube Under fixed-pressure conditions, the CHA measures
Gas hydration rate, shrinkage (with regard to chemical con-
column Oil column traction or development of porosity as the slurry
Water hydrates), and mechanical properties.
column The gas migration cell is an 8-in. long, 2-in. diameter
Cement slurry disposable cylinder (Fig. 9-28). The top of the cell is
closed and contains a pressure transducer and a tem-
V V perature probe. At the bottom of the cell, a sliding piston
(or diaphragm), connected to a gas source, is used to
Water p pressurize the slurry. The rate of the gas entering the
discharge cell and the rate of the water pushing the piston as
shrinkage develops are carefully monitored. The cell is
placed in a thermostatic oven until thermal equilibrium
is reached. As long as the cement slurry remains liquid,
Water the pressure measured at the top of the cell stays con-
reservoir stant. The upward movement of the piston compensates
Fig. 9-26. U-tube gas migration tester (from Richardson, 1982). for any slurry shrinkage. As the slurry hydrates, the tem-
perature in the cell increases and shrinkage continues.
The piston movement is eventually stopped by the
Valve Bleedoff Thermocouple increasing wall shear stress of the slurry, and the pres-
N2 pressure
source sure at the top of the cell decreases. When the pressure
at the top of the cell falls to a predetermined value, the
gas valve connected to the piston is opened. The gas is
Valve
allowed to enter the cell, driven by the pressure
Head- decrease and the continuing shrinkage of the cement.
pressure The permeability of the slurry to gas can be deter-
regulator
Hydrostatic mined by analyzing the pressure and gas flow rate
Cement pressure (Fig. 9-29). If the cement is permeable to gas, the pres-
columns cell sure in the cell will reach equilibrium with the gas-inlet
Dynamic
permeability
pressure once the gas valve is opened, and gas flow can
cell Differential be measured by the gas flowmeter. The amount of gas
pressure that enters the cell gives a good indication of the amount
transducer of shrinkage. A visual observation of the cement after
the test can indicate the manner by which gas migrated
Backpressure Thermostatic oven (bubbles, micropercolation, or fracture). If the cement
regulator Pressure
transducer is impermeable to gas, the pressure in the cell will con-
tinue to decrease and no gas flow will be observed.
Flowmeter N2 pressure source

Fig. 9-27. Dynamic permeability apparatus (from Parcevaux, 1984a).

Reprinted with permission from Elsevier.

Chapter 9 Annular Formation Fluid Migration 313

 Thermocouple
Pressure
gauge

Cement
slurry

Water flow

Gas and water

pressure

Fig. 9-28. Cement hydration analyzer. Reprinted with permission from Elsevier.

100 100

80 95
Pressure

60 90
Pressure Temperature
(bar) Temperature (C°)
40 85
Gas valve
opened
20 80
Shrinkage
Gas flow rate
0 75
0 4 8 12 16 20 24
Time (hr)
100 100

80 95

60 90
Gas rate Temperature
Temperature
(scf/min) (C°)
40 85

Gas valve
20 80
opened
Shrinkage
0 75
0 4 8 12 16 20 24
Time (hr)

Fig. 9-29. Output from the CHA. Cement that is permeable to gas (top) will exhibit a stable pressure
in the cell equal to the gas-inlet pressure, and gas flow will be measured with the gas flowmeter,
while cement that is impermeable to gas (bottom) will exhibit a continued pressure decrease
caused by shrinkage and no gas flow will be measured.

314 Well Cementing

9-7.3 Gel strength testing A similar modification of this method is known as the
Numerous methods for measuring the static gel strength shearometer tube. This device consists of a thin-walled
(SGS) of cement slurries have been developed. Today tube that is placed in a sample of cement slurry and
there are four common methods: using pressure-drop allowed to rest statically for a given period of time.
tubes, using vane rheometers, using rotating-paddle Weights are then applied to the top of the tube until it
devices, and employing acoustic techniques. It is impor- begins to move, and the SGS is calculated based upon
tant to note that comparative testing between these the force imparted by the weights and the surface area
devices has shown significant variations in the measured of the tube. This test can only be carried out at atmos-
SGS. As noted by Prohaska et al. (1993), gel strength pheric pressure.
development can be significantly impacted by the
amount of shear that is applied to the slurry before com- 9-7.3.2 Vane rheometers
mencing measurement and the pressure and tempera-
Early SGS testing was attempted using standard rota-
ture that the slurry encounters. These factors, along with
tional viscometers by measuring the maximum deflec-
the inherent differences in the devices, account for the
tion of the bob while rotating at 3 rpm after the slurry
variability in results between the various methods.
had been static for a designated period of time (normally
1 min or 10 min) (Appendix B). Slippage at the interface
9-7.3.1 Pressure drop tubes between the bob and the cement slurry, coupled with the
One of the simplest means of measuring SGS is by mea- relatively high minimum shear rates (5.1 sec–1), gave
suring the pressure drop across a length of tubing unreliable results. Furthermore, this test could only be
(Fig. 9-30). Cement slurry is placed in a small-diameter carried out under ambient conditions. In an effort to
tube and pressurized with water. A sensitive pressure reduce interfacial slippage, Haimoni and Hannant
transducer measures the pressure drop in the cement (1988) replaced the cylindrical bob with a vane. In the
column as it undergoes gelation. Equation 9-20 can then late 1990s, the vane geometry was further advanced. The
be used to calculate the SGS. One limitation of this test improved device allowed testing at elevated tempera-
is that it is not able to differentiate between pressure ture and pressure. In addition, the device allowed one to
drop caused by gelation and pressure drop caused by impart stress into the static slurry without shearing it
chemical shrinkage. and disturbing the evolving gel structure (Fig. 9-31).

9-7.3.3 Rotating-paddle devices

Constant Sabins et al. (1980) developed a device designed specif-
pressure ically to test SGS at elevated temperature and pressure.
source
The device, similar to a HPHT consistometer, used a spe-
cially designed paddle placed inside a pressurized slurry
container. The slurry could be stirred with a low-friction
Cement-filled tube
magnetic drive to simulate placement. After stirring, a
cord-and-weight bucket was attached to the magnetic
Δp drive, and the weight required to rotate the paddle 10°
Temperature bath
was measured at specific time intervals (normally 5 to
15 min). The SGS could then be calculated from the
geometry of the paddle and the weight required to move
it. Further improvements to this device, which included
direct torque measurement, were patented by Moon et
al. (Fig. 9-32) (1986).
Screw pump

Fig. 9-30. Schematic diagram of pressure-drop tube apparatus for

measuring SGS.

Chapter 9 Annular Formation Fluid Migration 315

Fig. 9-31. Pressurized vane rheometer to measure SGS.

Pulley rotates 0.50–2°/min

paddle rotational speed
Transducer Acoustic
pulsing waveform
Force electronics capture
transducer
Cement
slurry
Digital
Stepping signal
Sample motor processing
shearing
Torque
Measurement
Calculation
of signal
of SGS value
amplitude
Time
Fig. 9-32. Rotating-paddle SGS measuring device (from Moon and Fig. 9-33. Static gel strength analyzer block diagram (from Sabins
Wang, 1999). Reprinted with permission of SPE. and Maki, 1999). Reprinted with permission of SPE.

9-7.3.4 Acoustic devices 9-8 Conclusions

As a cement sample develops SGS, the amplitude of an Gas migration is a complex phenomenon involving fluid-
acoustic signal passing through the sample increases. density control, mud removal, cement-slurry properties,
Correlations can be developed between the change in cement hydration, and interactions between the
amplitude of the acoustic signal and the SGS of the cement, casing, and formation. It has been the subject of
slurry. Using this concept, Sabins and Maki (1999) pro- a substantial amount of research over the past 50 years,
duced a device similar to the ultrasonic cement analyzer yet it remains one of the most serious cementing prob-
(Appendix B) that could measure the SGS (Fig. 9-33). lems the industry faces. Uncontrolled, gas migration can
An apparent advantage of this method is that the lead to loss of production, environmental and regulatory
slurry is not mechanically sheared; therefore, the SGS is problems, and ultimately risk to equipment and person-
measured at zero shear rate. However, it is important to nel. The cost of remediating a gas migration problem can
note that, similar to that of the ultrasonic cement ana- far exceed the costs of active prevention.
lyzer, the gel strength is not a direct measurement, but Gas migration can be categorized according to when
only a correlation that has been established by compari- it occurs in relationship to the cement job. Gas migra-
son of acoustic signal attenuation and testing with pres- tion that occurs during cementing can be referred to as
sure-drop tubes and shearometers. immediate gas migration; gas migration that occurs after

316 Well Cementing

the cement is in place but before it has set is referred to 9-9 Acronym list
as short-term gas migration; and gas migration that API American Petroleum Institute
occurs after the cement has set is referred to as long-
term gas migration. Regardless of when it occurs, there CHA Cement hydration analyzer
are three distinct root causes that must occur for annu- CHP Critical hydration period
lar gas to migrate. ECP External casing packer
1. The pressure in the annulus must fall to a value that FP Formation parameter
is less than or equal to the pore pressure of a gas- GFP Gas flow potential
bearing zone.
HF Hydrostatic factor
2. Space must be available in the annulus for gas to
occupy in order for it to enter the annular column. HPHT High pressure, high temperature
3. A path must be present in the annulus through which MRF Mud removal factor
the gas can migrate. MPR Maximum pressure restriction
Successful strategies, both theoretical and practical, MRP Mud removal parameter
for controlling gas migration rely on minimizing the PDLP Pressure decay limit parameter
impact of one or more of these root causes. Particular RAS Right-angle set
factors that should be addressed include fluid loss, gel RP Recommended practice
strength development, cement shrinkage, cement per-
meability, free fluid, cement hydration kinetics, mechan- SGS Static gel strength
ical properties of the set cement, and last, but not least, SPN Slurry performance number
proper mud removal. TR Technical report
Because the factors surrounding gas migration are so
complex, it must be understood that there is no single
solution for all cases. Numerous predictive models
designed to understand and assess the risk of gas migra-
tion for a given set of conditions have been developed.
These models, coupled with the appropriate laboratory
techniques, aid the engineering and design of cementing
programs that will significantly minimize the likelihood
of annular gas migration.

Chapter 9 Annular Formation Fluid Migration 317

 Erik B. Nelson and Véronique Barlet-Gouédard—Schlumberger

10-1 Introduction
High-temperature wells present special cement system
design challenges. The physical and chemical behavior
Thermal Cements

10-2 High-temperature chemistry

of Portland cement
10
As discussed in Chapter 2, Portland cement is essentially
of well cements changes significantly at elevated tem- a calcium silicate material, the most abundant compo-
peratures and pressures. One must also pay close atten- nents being tricalcium silicate (C3S) and dicalcium sili-
tion to the chemical and physical properties of the for- cate (C2S). Upon addition of water, both hydrate to form
mations in contact with the cement. Corrosive water a gelatinous calcium silicate hydrate called “C-S-H
zones and very weak formations are not uncommon in phase,” which is responsible for the strength and dimen-
thermal wells. Without careful slurry design, the sional stability of the set cement at ordinary tempera-
integrity of the set cement may deteriorate, potentially tures. The reaction also liberates a substantial amount
resulting in the loss of zonal isolation. of calcium hydroxide (CH).
Thermal cementing encompasses three principal C-S-H phase is the early hydration product even at ele-
types of wells: deep oil and gas wells, geothermal wells, vated temperatures and pressures, and it is an excellent
and thermal recovery wells. In this chapter, each sce- binding material at well temperatures less than 230°F
nario is discussed separately, because the cement [110°C]. At higher temperatures, C-S-H phase is subject
system design parameters can differ significantly. to metamorphism, which usually results in decreased
During the last 50 years the most commonly used compressive strength and increased permeability of the
cements to complete thermal wells have been Portland set cement. In the petroleum literature, Swayze (1954)
cement, silica-lime systems, and high-alumina cement. described this phenomenon as “strength retrogression.”
More recently, other cement systems have been devel- At temperatures above 230°F [110°C], C-S-H phase
oped specifically for thermal-well environments. Before often converts to a phase called alpha dicalcium silicate
discussing the cement system design for the various hydrate (α-C2SH). α-C2SH is highly crystalline and much
types of thermal wells, the hydrothermal chemistry of more dense than C-S-H phase. As a result, matrix shrink-
cements used to complete thermal wells is presented. In age occurs that can be deleterious to the set-cement
this chapter, the special chemical notation for cement integrity. This effect is illustrated in Fig. 10-1, which
compounds is used. The reader is referred to Chapter 2 depicts the compressive-strength and water-permeabil-
for an explanation of the customary abbreviations. ity behavior of conventional Portland cement systems

10
40
4
1
30 3 1

Compressive 2 Water 2
strength 20 permeability 0.1
(MPa) 1 (mD) 3
10 0.01

4
0 0.001
0 1 0 1
Curing time (months) Curing time (months)

Fig. 10-1. Compressive strength and water permeability behavior of Portland cement at elevated
temperatures (from Nelson and Eilers, 1985).

Well Cementing ■ Chapter 10 Thermal Cements 319

cured at 446°F [230°C]. Significant loss of compressive the conditions for the formation of various calcium sili-
strength occurred within 1 month; nevertheless, the cate compounds, many of which occur geologically
levels to which the compressive strength fell are suffi- (Taylor, 1964). The C/S ratio is plotted versus the curing
cient to support casing in a well (Suman and Ellis, 1977). temperature. C-S-H phase has a variable C/S ratio, aver-
The real problem lies in the severe permeability aging about 1.5. The conversion to α-C2SH at 230°F
increases. To prevent interzonal communication, the [110°C] can be prevented by the addition of 35% to 40%
permeability of well cements to water should be no more silica (by weight of cement [BWOC]), reducing the C/S
than 0.1 mD. Within 1 month, the water permeabilities ratio to about 1.0. At this level, a mineral known as tober-
of the normal-density Class G systems portrayed in Fig. morite (C5S6H5) is formed; fortunately, this mineral pre-
10-1 (Curves 1 and 2) were 10 to 100 times higher than serves high compressive strength and low permeability.
the recommended limit. The permeability of the high- As the curing temperature increases to about 300°F
density Class H system (Curve 3) was barely acceptable. [150°C], tobermorite normally converts to xonotlite
The deterioration of the lower-density extended cement (C6S6H) and a smaller amount of gyrolite (C6S3H2) with
(Curve 4) was much more severe. minimal deterioration of cement performance.
The strength retrogression problem can be prevented Tobermorite sometimes persists to 480°F [250°C] in
by reducing the bulk lime-to-silica ratio (C/S ratio) in Portland cement systems because of aluminum substitu-
the cement (Menzel, 1935; Kalousek, 1952; Carter and tion in the lattice structure (Kalousek and Chow, 1976).
Smith, 1958). To accomplish this, the Portland cement is The improved performance of “silica-stabilized”
partially replaced by ground quartz, usually as fine silica Portland cements at elevated temperatures is illustrated
sand or silica flour. In some regions, special cements are in Fig. 10-3. Normal-density Class G cements, stabilized
available in which quartz has been ground with Portland with silica sand or silica flour, were cured at 446° and
cement clinker (Italcementi, 1977). Figure 10-2 depicts 608°F [230° and 320°C].

Mole fraction CaO/CaO + SiO2 of starting material

0 0.3 0.4 0.5 0.6 0.7 0.8 1.0
1,000 Rankinite α-CaSiO4
α-CaSiO3
Quartz

800 β-CaSiO3 Kilchoanite

(+ silica)
600 Calcio-
chondrodite
400 Foshagite Phase Y
Cristobalite

300 Tricalcium
Truscottite Xonolite
Calcium hydroxide
Temperature silicate
(°C, log scale) γ-C2S hydrate hydrate
200
Gyrolite Hillebrandite
150
Z-phase 11 Å Tobermorite α-C2S hydrate
100 Afwillite
Hydrous

14 Å Tobermorite
silica

75
Ca(H3SiO4)2 C-S-H(I) C-S-H(II)
50
0 0.5 0.6 0.7 0.8 1.0 1.3 1.5 2.0 2.5 3.0
CaO/SiO2 mole ratio of starting material

Fig. 10-2. Formation conditions for various calcium silicates (from Taylor, 1964).
Reprinted with permission from Elsevier.

320 Well Cementing

 Silica sand (230°C)
Silica flour (230°C)
Silica flour (320°C)
60 1

40 0.1
Compressive Water
strength permeability
(MPa) (mD)
20 00.01

0 0.001
0 1 3 6 12 24 0 1 3 6 12 24

Fig. 10-3. Compressive strength and permeability behavior of 16.0-lbm/gal Class G systems stabilized
with 35% silica (from Nelson and Eilers, 1985).

At 480°F [250°C] the phase truscottite (C7S12H3) The discussion so far has been limited to the behavior
begins to appear. As the curing temperature approaches of the silicate hydration products. To the authors’ knowl-
750°F [400°C], both xonotlite and truscottite are near edge, the hydrothermal behavior of the aluminates and
their maximum stable temperatures, and dehydration of aluminoferrites has not been specifically described in
the residual CH to C occurs. At higher temperatures, the the literature. The common hydrated aluminate and alu-
xonotlite and truscottite dehydrate, resulting in the dis- minoferrite hydrates (Chapter 2) are not typically
integration of the set cement. observed when Portland cements are cured hydrother-
In addition to the compounds cited above, other mally. Ettringite is not stable in hydrothermal condi-
–
phases such as pectolite (NC4S6H), scawtite (C7S6CH2), tions, and is not normally detected. Some of the Al3+,
reyerite (KC14S24H5), kilchoanite (C3S2H approxi- Fe3+, and SO42– ions from ettringite are incorporated
mately), and calciochondrodite (C5S2H approximately) into the silicate phases.
may appear in Portland cement systems cured at ele- The preceding discussion illustrates the complexity of
vated temperatures. These phases can affect the perfor- the hydrothermal behavior of calcium silicate hydrates.
mance of the set cement, even when present in small The performance of the set cement depends not only on
quantities. the downhole temperature, but also on the presence of
Cements containing significant amounts of truscot- subterranean brines and other minerals. As a result, the
tite are usually characterized by low permeability standard conditions for equilibrium transformations
(Gallus et al., 1978). The formation of pectolite, a that are reported in the literature are not always
sodium calcium silicate hydrate, is accompanied by observed downhole (Langton et al., 1980). Therefore,
cement expansion (Nelson and Eilers, 1982); in addition, the set cement must be considered to be metastable,
pectolite appears to render cements more resistant to because its composition can evolve as downhole condi-
corrosion by highly saline brines (Nelson and Kalousek, tions change.
1977; Nelson et al., 1981). Scawtite has been shown to
enhance cement compressive strength when present in
minor amounts (Eilers et al., 1983; Bell et al., 1989). In
general, set cements that consist predominantly of cal-
cium silicate hydrates with C/S ratios less than or equal
to 1.0 tend to have higher compressive strengths and
lower water permeabilities.

Chapter 10 Thermal Cements 321

10-3 High-alumina cement 50
High-alumina cement is a special material manufac-
tured primarily for applications in which a refractory 40
binder is required (Robson, 1962; Scrivener and Compressive
30
Capmas, 2001). In wells, it is used where the in situ strength
combustion process is employed for fireflooding (MPa)
20
(Section 10-8) and is also useful for cementing across
permafrost zones (Chapter 7). The primary cementitious 10
constituent is monocalcium aluminate (CA). As illus-
0
trated in Fig. 10-4, three initial metastable calcium 200 400 600 800 1,000 1,200
aluminate hydrates occur when water is added to CA:
Temperature (°C)
CAH10, C2AH8, and C4AH13. They ultimately convert to
C3AH6 (Quon and Malhotra, 1979). In addition to C3AH6, Fig. 10-5. Compressive strength of high-alumina cement/crushed
aluminum hydroxide (AH3) is often present as a binder firebrick concrete after 4 months’ exposure from 68° to 2,190°F [20°
phase. Unlike Portland cement, set calcium aluminate to 1,200°C] (from Heindl and Post, 1954).
cement does not contain calcium hydroxide.

C12A7 crystals intergrow and form a tightly bonded

27°C ceramic network. In thermal wells, temperatures above
CA + H20 CAH10, C2AH8, C4AH13 C3AH6 1,830°F [1,000°C] are not generally attained; thus, it is
(metastable) important to ensure that the minimum compressive
225°C
550°C to strength is sufficient for maintenance of well integrity.
to 275°C The strength and durability of high-alumina cements
950°C between 440° and 1,830°F [225° and 1,000°C] are pri-
C +C12 A7 C3 AH1.5 marily controlled by the initial water-to-cement ratio.
Depending upon the application, the amount of added
Fig. 10-4. Sequence of reactions of high-alumina cement at various
curing temperatures (from Nelson and Eilers, 1985). Reprinted with water should be the minimum required to prepare a
permission from the Petroleum Society of CIM. pumpable slurry. The use of dispersants is particularly
helpful for pumpability. A higher proportion of cement
relative to aggregate extender is also necessary. For
C3AH6 is probably the only stable hydrated calcium most applications, at least 50% of the solids should be
aluminate at temperatures below 437°F [225°C]. At cement.
higher temperatures, the water content begins to drop, A variety of materials may be used to extend calcium
and at 527°F [275°C], C3AH1.5 is found. As the temper- aluminate cement slurries, provided they have suitable
ature continues to increase, decomposition of C3AH1.5 stability at high temperatures and do not decompose or
occurs with the liberation of C. Between 1,022°F show anomalous thermal expansions or inversions. Silica
[550°C] and 1,742°F [950°C], a recrystallization occurs, sand should not be used if temperatures exceeding
ultimately resulting in C and C12A7. 572°F [300°C] are anticipated. Because of changes in
It is important to realize that, in ultrahigh-tempera- the crystalline structure, thermal expansion of quartz is
ture wells with temperatures up to 930°F [500°C], the relatively high at these temperatures, and thermal
proportional strength loss of high-alumina cement sys- cycling could eventually disrupt the cement. The most
tems is often greater than that experienced by unstabi- commonly used extender for high-alumina cements is
lized Portland cements. High-alumina cement is used crushed aluminosilicate firebrick. Other suitable materi-
because it can withstand wide-ranging temperature fluc- als include calcined bauxite, certain fly ashes, diatoma-
tuations, owing mainly to the absence of calcium hydrox- ceous earth, and perlite.
ide. Figure 10-5 illustrates the effect of curing tempera-
ture upon a high-alumina cement extended with 70%
crushed firebrick (Heindl and Post, 1954). The initial 10-4 Class J cement
strength loss between room temperature and 212°F Class J cement was developed in the early 1970s for
[100°C] is primarily caused by the conversion of the ini- cementing wells with static temperatures exceeding
tial hexagonal calcium aluminate hydrates to the cubic 260°F [126°C] (Maravilla, 1974; Degouy and Martin,
C3AH6. With further heating, the strength continues to 1993; Bensted, 1995). It was recently dropped from the
drop because of dehydration and the formation of C and list of American Petroleum Institute (API) cements
C12A7. Strength improves above 1,830°F [1,000°C] as the

322 Well Cementing

owing to low usage; however, it is still manufactured in environment, the stable minerals would be anorthite,
the Far East, mainly for geothermal well applications. A grossular, prehnite, epidote, and zeolite, possibly associ-
similar cement, belite-silica cement (BSC) has been ated with quartz (Frey et al., 1991). Roy et al. experi-
used in the CIS for high-temperature well cementing mented with compositions that would lead to
(Bulatov, 1985). These cements are advantageous from a the formation of the mineral anorthite. The raw materi-
logistical point of view, because the addition of silica is als were Class J cement, calcium aluminate cement, and
not required. β-C2S. The results were encouraging in terms of
Like Portland cement, Class J cement is a calcium sil- strength; however, the preparation of mixable and
icate material; however, no aluminate phases or C3S is pumpable slurries, with controllable set times, was a
present. The composition is essentially β-C2S, α-quartz, serious problem.
and CH. As discussed in Chapter 2, the hydration rate of More recently, other approaches to preparing anor-
β-C2S is relatively slow; consequently, retarders are thite cements were investigated (Barlet-Gouédard and
rarely necessary at circulating temperatures less than Goffé, 2002). To promote the formation of anorthite, var-
300°F [149°C]. The C/S ratio of Class J cement is ious alumina-rich materials were evaluated in combina-
adjusted such that tobermorite and xonotlite are tion with Class G cement and silica flour. Calcined kaoli-
obtained upon curing (Sasaki et al., 1986; Kalousek and nite was determined to be an ideal alumina source;
Nelson, 1978). Scawtite also occurs frequently. In addi- however, it must be very finely divided (≈2 μm) to pro-
tion, the sulfate resistance of Class J cement is very high vide sufficient reactivity to readily form anorthite. As a
owing to the absence of C3A. result, similar to the experience of Roy et al., Barlet-
Gouédard and Goffe initially encountered difficulties
preparing a mixable and pumpable system. Their solu-
tion involved controlling the particle sizes of the ingre-
10-5 Calcium aluminosilicate systems dients and optimizing the packing between them. Such
In the late 1970s, Roy et al. (1979; 1980) reported work controlled-granulometry techniques are described in
to develop cement systems that resemble the composi- detail in Chapter 7. The anorthite system is a trimodal
tion of the formation in geothermal wells. Such cements mixture: calcined kaolinite and silica flour (fine parti-
should exhibit better chemical and thermal durability cles), Class G cement (medium-size particles), and
than the conventional calcium-silicate or calcium-alumi- ceramic microspheres (coarse particles). The micros-
nate systems. Of particular interest was the pheres are also useful for lowering slurry density.
CaO • Al2O3 • SiO2 system (Fig. 10-6). In a hydrothermal

Si
100 Anorthite system (1.49 g/cm3)
10 Class G cement
90 Anorthite
20 Xonotlite
80 Garnet
30 Kilchoanite/afwillite
70 Pectolite
40 Prehnite
60 Wairakite
50 Epidote
50 Quartz
60 Diopside
40 Hauyne
70 Margarite
30
80
20
90
10
100

100 90 80 70 60 50 40 30 20 10
Al Ca

Fig. 10-6. Ternary diagram showing compositions of minerals in the CaO • Al2O3 • SiO2 system.

Chapter 10 Thermal Cements 323

 A complete solid-solution series extends from anor- calcium aluminate cement as the base reactant. The
thite, Ca(Al2Si2O8), to albite, Na(AlSi3O8). As a result, principal cementitious phase is hydroxyapatite
anorthite can accommodate the substitution of various [Ca5OH(PO4)]. Such systems normally set and harden
ions into its crystal lattice. This property makes anor- within a few minutes at ambient temperature. Citric
thite adaptable to aggressive chemical environments. acid is an effective set retarder at circulating tempera-
The typical properties and performance of an anorthite tures up to around 180°F [82°C] (Brothers et al., 1999).
cement exposed to a carbon dioxide–laden synthetic Calcium phosphate cement normally has a density in
geothermal brine are given in Table 10-1. the range of 15–17 lbm/gal [1,800–2,040 kg/m3]. To reduce
the cost or the density of the slurry, extenders are usually
added. The most common extenders include ASTM Class
F fly ash or nitrogen, which is used to prepare a foamed
10-6 Calcium phosphate systems cement system (Brothers et al., 2001). The composition
Calcium phosphate cement systems were originally and compressive-strength performance of foamed cal-
developed for applications in dentistry. Recently such cium phosphate systems are shown in Table 10-2.
cements have found application as thermal cements The carbon dioxide resistance of calcium phosphate
(Sugama and Carciello, 1992; 1993; 1995), particularly cement, compared to conventional silica-stabilized
for geothermal wells in which high concentrations of Portland cement, is shown in Table 10-3. The cement sys-
carbon dioxide are encountered (Weber et al., 1998; tems were cured in 1-wt% Na2CO3 solution at two tem-
Brothers et al., 2001). Portland cement systems are sub- peratures: 200° and 500°F [93° and 260°C]. The solution
ject to attack by CO2, resulting in the loss of cementi- pH was 2.0–2.7. The weight loss of the cement specimens
tious material from the cement matrix (Chapter 7). The was measured at curing times of up to 32 days. The
consequences of CO2 attack are loss of compressive results clearly show that, within the 32-day test period,
strength and an increase of permeability. the calcium phosphate binder demonstrated excellent
Calcium phosphate cements can be synthesized by resistance to carbonation.
the acid-base reaction between a soluble phosphate
[usually NaH2PO4 or (NaPO3)n] as the acid solution and

Table 10–1. Typical Performance of Anorthite Cement Systems

Slurry Density Plastic Yield Point Thickening 28-D Compressive Water Permeability
(lbm/gal [g/cm3]) Viscosity (cp) (lbm/100 ft2) Time at 300°F Strength in Brine at at 28 D in Brine at
[150°C] BHCT † BHST ‡ of 572°F [300°C] BHST of 572°F [300°C]
(hr:min) (psi [MPa]) (mD)
10.9 [1.31] 75 0.6 5:00 1,400 [9.8] 0.2

12.4 [1.49] 130 6 3:00 2,320 [16.2] 0.05

13.5 [1.62] 144 6 5:40 2,200 [15.4] 0.03

14.0 [1.68] 110 4 3:30 2,570 [18.0] –

16.2 [1.95] 55 10 7:40 2,290 [16.0] –

† BHCT = bottomhole circulating temperature
‡ BHST = bottomhole static temperature

Table 10–2. Composition and Performance of Calcium Phosphate Cement Systems Cured at 600°F †
Water Calcium Sodium Fly Ash Foaming Foam Slurry Density 28-Day
(wt%) Aluminate Polyphosphate (wt%) Agent Stabilizer (lbm/gal Compressive
(wt%) (wt%) (wt%) (wt%) [kg/cm3]) Strength
(psi [MPa])
23.3 17.5 15.6 40.8 1.9 0.9 12.1 [1,450] 570 [4.0]

22.3 21.0 14.9 39.2 1.8 0.8 15.1 [1,810] 1,060 [7.4]
† from Brothers et al. (1999).

324 Well Cementing

Table 10–3. Corrosion-Test Results of Thermal Cements in 1% Na2CO3 Solution; pH = 2.0–2.7†
Slurry Slurry Test Percent Weight Loss After:
Density Temperature
(lbm/gal) (°F)
2D 9D 13 D 20 D
Calcium phosphate 15.7 200 0.1 0.1 0.1 0.1

Class G + 40% silica 16.2 200 1.6 5.7 11 21

18 D 25 D 32 D
Calcium phosphate 16.0 500 0 0 0

Class G + 40% silica 16.2 500 9.5 15 22

† from Weber et al. (1988).

10-7 Deep oil and gas wells 10-7.1 Thickening time and initial compressive
Wells with depths exceeding 15,000 ft [4,570 m], with strength development
bottomhole temperatures above 230°F [110°C], are Cement slurries for deep wells are usually designed to
common throughout the world. Since the 1970s, hun- have at least 3 to 4 hr of pumping time. However, there
dreds of wells with depths exceeding 25,000 ft [7,600 m] are several complicating factors that must be men-
have been completed (Arnold, 1980; Wooley et al., 1984). tioned.
Such wells represent a large investment of time and As the length of the casing string or liner increases,
money; therefore, obtaining a successful well completion the problem of achieving a cement seal becomes more
is of paramount importance. difficult (Suman and Ellis, 1977). In many cases, the
The procedures for cementing deep wells are basi- static-temperature differential between the top and
cally the same as those for shallower wells; however, bottom of the cement column can exceed 100°F [38°C].
because of the severe well conditions and more complex Sufficient retarder must be added to the cement slurry
well architecture, such wells are usually considered to to allow adequate placement time at the BHCT; conse-
be critical (Smith, 1987). Higher temperatures, nar- quently, such a slurry may be over-retarded at the top of
rower annuli, overpressured zones, and corrosive fluids the cement column, resulting in a very long waiting-on-
are commonly encountered. Consequently, the cement cement (WOC) time. If high-pressure gas exists behind
system design can also be complex, involving an elabo- the casing string or liner, the risk of gas invasion into the
rate array of retarders, fluid-loss additives, dispersants, cement is high (Chapter 9). In recent years, advances
silica, and weighting materials. One must be certain that have been made in retarder chemistry and cement-
the cement system can be properly placed and will main- system design that have helped to mitigate such prob-
tain zonal isolation throughout the life of the well. At lems (Chapter 3).
present, Portland cement is used in virtually all deep oil When designing cement slurries for deep, hot wells, it
and gas well completions. is very important to use accurate static and circulating
Typical casing programs and cementing procedures temperature information. Such data may be obtained
for deep wells are given in Chapter 13. Detailed informa- from drillstem tests, logs, special temperature recording
tion regarding cement additives is found in Chapter 3. In subs, or circulating temperature probes run during hole
this section, the design of appropriate cement systems conditioning (Jones, 1986). Computer simulators have
for deep high-temperature wells is presented. also been developed to better predict well temperatures
(Chapter 12). If fluids are circulated in the well for sev-
eral hours before cementing, the well temperature may
be lowered significantly. In such cases, one must be
careful not to overestimate circulating temperature and
over-retard the cement slurry.

Chapter 10 Thermal Cements 325

 The cement slurry is exposed to high pressures in 10-7.3 Cement-slurry density
deep wells and, as shown in Fig. 10-7, a significant accel- Deep wells often involve cementing across high-pressure
erating effect is observed (Bearden, 1959). Earlier com- formations. To maintain control of the well, the hydro-
pressive strength development and higher ultimate static pressure of the wellbore fluids must meet or
compressive strength are also observed as curing pres- exceed the formation pressure at all times. Con-
sure increases (Handin, 1965; Metcalf and Dresher, sequently, cement slurries with densities as high as
1978). Therefore, when designing a proper cement- 27 lbm/gal [3.24 g/cm3] are often placed. When large
slurry composition in the laboratory, performing the quantities of weighting materials are present in the
tests at the anticipated pressure is recommended slurry, slurry sedimentation can be a major concern.
(Appendix B). Maximizing the packing volume fraction of the slurry
In general, as the circulating temperature increases, solids, using a multimodal particle size distribution, can
the sensitivity of Portland cement systems to subtle mitigate such problems (Chapter 7). Silica can be incor-
chemical and physical differences between the slurry porated in one of the granulometric fractions to provide
ingredients also increases. Therefore, all laboratory tests stabilization.
should be performed with samples of the water, cement,
and additives that will be used during the job.
10-7.4 Fluid-loss control
As discussed in Chapters 3 and 6, fluid-loss control is
600 necessary to preserve the chemical and physical charac-
teristics of the cement slurry and to prevent the devel-
500 opment of a cement filtercake that could cause bridging
in the annulus. For most primary cementing operations,
400
an API/ISO fluid-loss rate of between 50 and 100 mL/30
Thickening
time 300 min is generally considered to be adequate.
(min)
200
10-7.5 Long-term performance of cements
100 for deep wells
Once the cement system is successfully placed in the
annulus, it is important to ensure that it will provide
0 5,000 15,000 25,000 35,000 adequate casing support and zonal isolation throughout
Pressure (psi) the life of the well. As discussed earlier in this chapter,
Fig. 10-7. Effect of pressure on pumpability of cement (cement: API
the most common thermal cement systems are based on
Class H with 0.3% retarder; BHCT: 200°F [93°C]) (after Smith, 1976). silica-stabilized Portland cement.
Reprinted with permission of SPE. A typical slurry composition for a deep, hot well would
consist of Class H or Class G cement, 35% to 40% silica
(BWOC), a dispersant, a fluid-loss additive, a retarder,
and a weighting agent. The long-term performance of
10-7.2 Cement-slurry rheology such cement systems would be very similar to that
The narrow annuli often associated with deep-well com- shown in Fig. 10-3.
pletions increase the difficulty of achieving a good bond When high-density slurries are unnecessary, or if
between the cement, the pipe, and the formation. The lower-density slurries are required to prevent lost circu-
small clearance between the casing and the formation lation or formation breakdown, extenders such as fly
increases the risk of cement contamination by drilling ash, diatomaceous earth, bentonite, or perlite are com-
fluid that was not displaced by chemical washes or spac- monly used. The long-term performance of typical sys-
ers. Proper casing centralization can be difficult to tems in laboratory tests is illustrated in Figs. 10-8 and
achieve. As discussed in Chapter 5, the rheology of the 10-9. All systems contained 35% silica flour (BWOC). In
fluids in the wellbore should be carefully controlled. In Fig. 10-8, the systems have been cured at 450°F [232°C]
some cases, the cement slurry is designed for turbulent- under saturated steam pressure for up to 2 years, and
flow displacement, requiring the use of dispersants. compressive strength and permeability measurements
When designing highly dispersed slurries, one must be have been performed at periods ranging from 1 day to
careful to avoid slurry sedimentation or free-water devel- 24 months. Figure 10-9 presents data for systems cured
opment. This is especially important when the borehole at 600°F [315°C]. It is important to note the nonlinear
is highly deviated (Chapter 13). time scale and the logarithmic permeability scale.

326 Well Cementing

 Fly ash: 15.6 lbm/gal [1.87 g/cm3] Fly ash: 15.6 lbm/gal [1.87 g/cm3]
Perlite/bentonite: 12.9 lbm/gal [1.55 g/cm3] Perlite/bentonite: 12.9 lbm/gal [1.55 g/cm3]
Perlite/bentonite: 11.9 lbm/gal [1.43 g/cm3] Perlite/bentonite: 11.9 lbm/gal [1.43 g/cm3]
Diatomaceous earth: 13.8 lbm/gal [1.66 g/cm3] Diatomaceous earth: 13.8 lbm/gal [1.66 g/cm3]

50 50

40 40
1
1
30 Compressive 30
Compressive
4 strength
strength 2
(MPa) 20 (MPa)
20
2
3
10 10
3 4
0 0
0.03 1 3 6 12 24 0.03 1 3 6 12 24
Curing time (months) Curing time (months)
100 100

10 10
1

1 1
Water Water
permeability 3 permeability 3
(mD) 2 (mD) 0.1
0.1
1 4 2
0.01 0.01
4
0.001 0.001
0.03 1 3 6 12 24 0.03 1 3 6 12 24
Curing time (months) Curing time (months)

Fig. 10-8. Compressive strength and permeability performance of Fig. 10-9. Compressive strength and permeability performance of
conventionally extended Portland cement slurries—450°F [232°C] conventionally extended Portland cement slurries—600°F [315°C]
after Nelson and Eilers, 1985). Reprinted with permission from the (after Nelson and Eilers, 1985). Reprinted with permission from the
Petroleum Society of CIM. Petroleum Society of CIM.

System 1 contained Type F fly ash as an extender and Systems 2 and 3 were extended with perlite and ben-
was the heaviest of the four. Despite the density advan- tonite. System 2 performed well at both 450° and 600°F
tage and the highest initial compressive strength, the [232° and 315°C] with regard to compressive strength.
performance of System 1 over a 2-year period was The permeability of System 2 varied back and forth
no better than that of lower-density systems at 450°F across the 0.1-mD line. System 3 was the least dense of
[232°C], and was the poorest of the four at 600°F the four. The compressive strength performance was
[315°C]. This delayed degradation of fly-ash-containing adequate at both curing temperatures, but the perme-
systems was probably the result of alkali contaminants abilities were too high. It is important to point out that
in the fly ash. Such contaminants can slowly react and perlite is compressible, and its extending effect
form substituted calcium silicate hydrates, notably decreases as the hydrostatic pressure in the well
reyerite, with deleterious effects (Eilers and Root, 1974). increases (Chapter 3). For this reason, perlite is rarely
It is important to mention that cement degradation asso- used today. System 4, containing diatomaceous earth,
ciated with fly ash has not been observed at curing tem- was a rather poor performer in the strength category, yet
peratures below 450°F [232°C]. had low permeability.

Chapter 10 Thermal Cements 327

 Figure 10-10 shows the typical performance of a If competent cement systems with densities less than
normal-density neat Class J system. Its behavior is simi- 12.5 lbm/gal [1.5 g/cm3] are necessary, microsphere-
lar to that observed with normal-density silica-stabilized extended, multimodal-particle-size, or foamed cements
Portland cement systems. (Chapters 3 and 7) may be appropriate. However, when
The behavior of these systems illustrates that high contemplating the use of ceramic or glass microspheres,
compressive strength and low water permeability are not one must be certain that they can withstand the hydro-
necessarily linked. Although water permeability is not as static pressure. Ceramic microspheres and most grades
convenient to determine as compressive strength of glass microspheres can withstand no more than
(Appendix B), one should do so before the application of 3,000 psi [20.7 MPa], which eliminates them from con-
a cement system in severe downhole conditions. In addi- sideration in most deep well completions. However, glass
tion, the data suggest that conventionally extended microspheres with hydrostatic crush strengths as high
Portland cement systems with densities below about 12.5 as 10,000 psi [69.0 MPa] are available. Foamed cement,
lbm/gal [1.5 g/cm3] may not be able to perform suitably occasionally used in deep high-temperature wells, is
in high-temperature wells, except perhaps as “filler” sys- more common in geothermal and steamflood wells.
tems that are not placed across producing zones.

10-8 Geothermal well cementing

6,000
Projects to extract geothermal energy exist throughout
the world. Virtually any location with thermal anomalies
is a potential site for geothermal well drilling. Some of
the more notable geothermal projects are located in
4,000 California, Utah, and New Mexico, United States;
Mexico; Central America; The Philippines; Indonesia;
Compressive Slurry density New Zealand; Kenya; Iceland; and Italy (Geothermal
strength 16 lbm/gal Education Office, 2001).
(psi) [1.92 g/cm3] Most geothermal plants are configured as shown in
2,000 Fig. 10-11. Superheated formation water that lies above
the geothermal formation is produced to the surface,
whereupon it is “flashed” into steam. The steam is used
to power turbines that generate electrical power. The
spent water is injected back into the reservoir, not for
0 replenishment, but for environmental reasons. The for-
0 1 3 6 12 24
mation waters are often highly saline and corrosive and
Curing time (months) contain toxic heavy metals. Another plant design, called
1
“hot dry rock,” is employed in regions where no geother-
mal formation waters exist. Two intersecting wells are
drilled into the hot formation. Water from the surface is
pumped down one well and becomes superheated. The
superheated water is then produced out of the other well
0.1 and flashed into steam.
Geothermal wells are usually completed in much the
Water same manner as conventional oil and gas wells; however,
permeability the environment with which the cements must contend
(mD)
is frequently much more severe. The bottomhole tem-
0.01 perature in a geothermal well can be as high as 700°F
[370°C]. The failure of wells in several geothermal fields
has been directly attributed to cement failure (Radenti
and Ghiringelli, 1972; Berra et al., 1998); as a result,
research has been conducted to identify cement formu-
0.001
0 1 3 6 12 24
lations that perform suitably under such conditions.
Curing time (months)

Fig. 10-10. Compressive strength and permeability behavior of Class

J cement at 450°F [232°C].

328 Well Cementing

 Injection well

Production well

Fig. 10-11. Geothermal power plant (from Geothermal Education Office, 2001). Drawing courtesy
of Geothermal Education Office, Tiburon, California, USA.

10-8.1 Well conditions associated with An economical geothermal reservoir requires that
geothermal wells large quantities of hot water or steam must be produced
With the exception of hot, dry rock completions with cir- from each well. Therefore, the reservoirs are usually nat-
culating temperatures as high as 500°F [260°C] (Carden urally fractured and have effective permeabilities that
et al., 1983, Duchane, 1994), most geothermal wells are are probably greater than 1 D. The integrity of the for-
not cemented under “geothermal” conditions, because mations ranges from poorly consolidated to highly frac-
the fluids circulated during drilling cool the formation. tured, and the fracture gradients tend to be low.
The maximum circulating temperatures during the Consequently, lost circulation is the most serious obsta-
cement job seldom exceed 240°F [116°C]; therefore, the cle to successfully cementing geothermal wells (per-
design of cement systems with adequate thickening sonal communication, Weber, 2003). It is not uncommon
times is usually not a problem. Most geothermal wells to have losses in the casing strings set above the target
are less than 10,000 ft [3,050 m] in depth. Downhole reservoir, and in many cases total losses occur before the
pressures are seldom above the water gradient. intended setting point for the intermediate string. For
The drilling programs for geothermal wells usually these reasons, low-density cement systems are required
call for setting surface and production casing above the by most geothermal operators (Nelson et al., 1981).
reservoir. In some cases, a slotted liner is hung through Lost circulation also hampers the determination of
the producing zone, but cementing the liner is not con- cement placement temperatures. Placement tempera-
sidered critical. It is very important to cement the cas- ture simulation and modeling is essential to formulate
ings to the surface; otherwise, creep or elongation will the appropriate cement system (Chapter 12).
occur because of thermal expansion when the well is
brought into production (Shryock, 1984).

Chapter 10 Thermal Cements 329

 The chemistry of the reservoir fluids varies from fresh 8,000 Mesh 325 140 50
water to brines with greater than 200,000 mg/L total dis-
solved solids. The fluids extracted from dry steam fields 6,000 50,000
contain relatively few salts and low concentrations of Compressive
noncondensible gases, the most noticeable being H2S. strength 4,000 30,000 (kPa)
The saline brines often contain significant quantities of (psi)
2,000 10,000
carbonate and sulfate.

1 4 10 40 100 400
10-8.2 Performance requirements and design
Average silica size (μm)
considerations
Geothermal wells arguably present the most severe con- 300°F [150°C]
ditions to which well cements are exposed. As a result, 450°F [232°C]
617°F [325°C]
the performance requirements are among the most
stringent. Geothermal well cements are usually designed
to provide at least 1,000 psi [7.0 MPa] compressive
strength, and no more than 0.1-mD water permeability Mesh 325 140 70
100
(API Task Group on Cements for Geothermal Wells, 80
1985). In addition, the set cement often must be resis-
Crystalline 60
tant to degradation by saline brines. composition
Silica-stabilized Portland cement compositions are 40
(%)
almost exclusively used to complete geothermal wells; 20
however, their dominance is being challenged by sys- 0
tems that offer better resistance to the severe chemical 20 40 60 80 100 120 140 160 175
environments. Each is described in this section. Average silica size (μm)
Scawtite Kilchoanite
10-8.2.1 Portland cement–based geothermal well Xonotlite Quartz
cement compositions
When Portland cement–based cement systems are
Mesh 325 140 50
expected to contact highly saline and corrosive geother- 10.0
mal brines, the particle size of the added silica is an 4.0
important consideration. As explained in Chapter 3, 1.0
there are two forms of silica commonly used in well Water 0.4
permeability 0.1
cementing: silica sand, with a particle size of approxi- 0.04
(mD)
mately 175–200 μm, and silica flour, with an average par- 0.01
ticle size of approximately 15 μm. Field personnel usu- 0.004
ally prefer silica sand, because its lower surface area 0.001
facilitates easier slurry mixing. However, in certain geot- 1 4 10 40 100 400
hermal environments, silica sand cannot be relied upon Average silica size (μm)
to provide adequate stabilization.
Eilers and Nelson (1979) investigated the effect of Fig. 10-12. Effect of silica particle size on the performance of Class
G cement cured in geothermal brine (from Eilers and Nelson, 1979).
silica particle size on the performance of Class G cement Reprinted with permission of SPE.
formulations cured at various temperatures in a geo-
thermal brine. The salinity of the brine was 25,000 mg/L
total dissolved solids. Figure 10-12 shows the relation-
ships between the silica particle size and several para- Figure 10-13 shows that the silica particle-size effect is
meters—compressive strength, water permeability, and significantly more pronounced with lower-density
cement phase composition. The slurry density was cement compositions.
15.8 lbm/gal [1.90 g/cm3]. A decrease in compressive High concentrations of sodium chloride depress the
strength and an increase in water permeability occurred rate at which silica enters solution (personal communi-
when the average particle size of the added silica cation, R. Fournier, 1979); as a result, when the silica
exceeded about 15 μm. Xonotlite was also replaced particle size is large, the rate of dissolution of silica is
by kilchoanite as the predominant cement phase. insufficient to allow the formation of the desired calcium

330 Well Cementing

 Compressive strength Permeability
Mesh 325 140 50
4,000 10 Mesh 325 140 50
300°F [150°C]
450°F [232°C] 4
617°F [325°C] 1
3,000
0.4
Compressive
2,000 Permeability 0.1
strength
(mD)
(psi) 0.04
1,000 0.01
0.004
0.001
1 4 10 40 100 400 1 4 10 40 100 400
Average silica size (μm) Average silica size (μm)

Fig. 10-13. Effect of silica particle size on the performance of a 13.5-lbm/gal Class G-perlite-bentonite
system cured in geothermal brine (from Eilers and Nelson, 1979). Reprinted with permission of SPE.

silicate hydrates (C/S ratio < 1). The kinetics of dissolu- 40% SiO2
tion can be affected by the particle size of the solute. 70 100% silica fume
Reducing the particle size of the silica increases its sur- 100% silica flour
60 50% silica fume, 50% silica flour
face area; consequently, a sufficient supply of silica is 33% silica flour, 67% silica fume
available. 50
25% silica flour, 75% silica fume
Grabowski and Gillott (1989) and Dillenbeck et al. Compressive
(1990) studied the effects of silica “fume,” with an aver- strength 40
age particle size of approximately 0.1 μm (Chapter 3), (MPa)
upon Portland cement systems at elevated temperatures 30
and pressures. With a constant SiO2 concentration (40%
20
BWOC) and water-to-solids ratio (0.5), samples were
prepared containing silica fume, combinations of silica 10
fume and silica flour, and silica flour. Curing was per- 7 28 56 90 210 270
formed at 450°F [230°C] and 400 psi [2.75 MPa] for
Total age (days)
7 days, using samples aged under ambient conditions for
periods up to 270 days. The systems containing silica 10–1

fume developed less compressive strength, but lower

permeability, than equivalent systems containing only
silica flour (Fig. 10-14). The major phase found in all of
the samples was xonotlite (scawtite was detected in the
samples containing only silica flour); however, the Permeability
(mD)
microstructures were different. The samples containing
silica flour exhibited short parallel needles of xonotlite.
10–2
As the quantity of silica fume increased, the texture of
the xonotlite was granular. In general, the samples with
needle-shaped xonotlite crystals exhibited higher per- 5 × 10–3
meabilities.
The presence of carbonate in certain geothermal 7 14 28 270
brines presents a serious difficulty for Portland cement Total age (days)
systems (Milestone et al., 1986 and 1987). Calcium sili- Fig. 10-14. Compressive strength and permeability behavior of
cate hydrates are not stable in such a chemical environ- silica-stabilized Portland cements containing various amounts of
ment, even at ordinary temperatures (Taylor, 1964). silica fume (after Grabowski and Gillot, 1989). Ambient curing for 7
Upon exposure to carbonate solutions, calcium silicate days at 230°C, 100% real-time humidity, and 2.75 MPa compressive
hydrates are eventually converted to a mixture of cal- strength. Reprinted with permission from Elsevier.

Chapter 10 Thermal Cements 331

cium carbonate and amorphous silica. This phenomenon Table 10-4. Compositions of Typical Geothernal Cement
has been observed in well cements by numerous Systems†, ‡
researchers (Onan, 1984; Bruckdorfer, 1986; Shen and Sample Parts by Components Slurry
Pyle, 1989). High-alumina cements are also known to Code Weight Weight
degrade in the presence of carbonate (Crammond and
1 100 API Class G cement 15.1 lbm/gal
Currie, 1993). At present there appear to be no pub- (64, 2C, 21.5S, 3.9A, 3.8F) [1.81 g/cm3]
lished data regarding the behavior of high-alumina 35 Silica flour
cements in a carbonaceous environment at elevated 1 Lignin-sugar
temperatures. 54 Water
The principal, and generally successful, defense 2 100 API Class J cement 15.4 lbm/gal
against such degradation has traditionally been the (37.3C, 54.2S, 1.1A, 1.0F) [1.85 g/cm3]
placement of low-C/S-ratio cement systems with very low 0.4 Lignin-sugar
permeability. However, such systems have been shown 44 Water
to be inadequate for geothermal wells with formations 3 100 API Class F cement 15.1 lbm/gal
containing very high concentrations of CO2 (Hedenquest (63.9C, 21.1S, 3.1A, 54.F) [1.81 g/cm3]
and Stewart, 1985). Milestone et al. (1986 and 1987) 40 Silica flour
demonstrated that tobermorite and xonotlite are among 0.7 Lignin-sugar
the cement phases least resistant to carbonation, and 63 Water
their deterioration is accelerated when bentonite is 4 30 API Class J cement 13.7 lbm/gal
present in the cement. Milestone et al. discovered that 40 Pozzolan [1.65 g/cm3]
reducing the silica flour concentration from 35% to 20% 30 Blast furnace slag
(BWOC) improves the cement’s resistance to CO2. When 0.5 Carboxymethylcellulose
60 Water
less silica is present, weaker and more permeable cal-
cium silicate hydrates form; however, a substantial quan- 5 100 API Class G cement 13.5 lbm/gal
tity of calcium hydroxide also remains in the system. (64.2C, 21.5S, 2.9A, 3.8F) [1.62 g/cm3]
Upon substantial carbonation, the calcium hydroxide 35 Silica flour
8.5 Perlite
reacts to form a protective layer of calcite, the perme- 2 Bentonite
ability decreases, and further attack is inhibited. 1 Lignin-sugar
Because of the presence of weak formations and low 116 Water
fracture gradients, lower-density cements are often
required in geothermal wells. Therefore, research has 6 100 API Class G cement 14.0 lbm/gal
(64.2C, 21.5S, 3.9A, 3.8F) [1.68 g/cm3]
been performed to develop low-density systems that will 35 Silica flour
perform adequately. The typical extenders used to pre- 10 Diatomaceous earth
pare low-density geothermal cements are bentonite 1 Lignin-sugar
and diatomaceous earth. Additional silica flour, up to 91 Water
100% by weight of cement, is sometimes added in lower- 7 100 API Class G cement 15.5 lbm/gal
density systems to ensure proper stabilization (Gallus et 40 Silica flour [1.86 g/cm3]
al., 1979). 0.8 Dispersant
Ultralow-density foamed cements (Rickard, 1985; 0.8 Fluid-loss agent
Sugama et al., 1986) and microsphere-extended systems 0.4 Retarder
60.3 Water
have been used to cement geothermal wells. Such sys-
tems have been used successfully in thermal recovery 8 100 API Class G cement 13.6 lbm/gal
wells (Section 10-9.1); however, information is sparse 100 Silica flour [1.63 g/cm3]
regarding the long-term stability of these systems to cor- 0.3 Retarder
85.1 Water
rosive brines. Until sufficient data are available, it would
be prudent to restrict the use of ultralow-density sys- 9 100 API Class G cement 15.4 lbm/gal
tems to applications in which the formations fluids are 80 Silica flour [1.85 g/cm3]
not aggressive. 0.5 Fluid-loss agent
0.3 Retarder
Table 10-4 lists the compositions of both normal and 76.8 Water
low-density systems that are often used as geothermal
cements. The compressive strength and water perme- 10 100 API Class G cement 15.7 lbm/gal)
ability upon long-term exposure to actual geothermal 40 Silica flour [1.89 g/cm3]
1 Retarder
conditions are shown in Figs. 10-15 and 10-16, respec- 59.2 Water
tively. † fromAPI Task Group on Geothermal Well Cements, 1985. Reprinted with permission from
Oil & Gas Journal.
‡A = AI2O3, C = CaO, F = Fe2O3, M = MgO, S = SiO2.

332 Well Cementing

 Cement Sample Designation
1 2 3 4 5 6 7 8 9 10
1 day
12,000
3 months 80
Cubes 6 months
12 months
10,000
Cups 60
8,000
Compressive Compressive
strength strength
(psi) 6,000 40 (MPa)

4,000
20
2,000
1,000 psi

0 0
Compressive strength of cement cube and sandstone cup samples after aging periods of 1 day and 3, 6,
and 12 months. Cup samples were cured and aged downhole. Cubes were laboratory cured under water
at 392°F [200°C] for 1 day, then exposed downhole for 3, 6, and 12 months in the Cerro Prieto geothermal
field, Mexico. The downhole temperature was 417°F [214°C].

Fig. 10-15. Compressive strength performance of typical geothermal well cements under
actual conditions (from API Task Force on Geothermal Well Cements, 1985). Reprinted
with permission from Oil & Gas Journal.

Cement Sample Designation

1 2 3 4 5 6 7 8 9 10

2 5
1 day
3 months
1 4
6 months
12 months
0 3
0.1 mD
Log10 –1 2 Log10
permeability permeability
(log mD) –2 1 (log nm2)

–3 0

–4 –1

–5 –2

–6 –3
Water permeabilities of cement samples taken from slurry-filled sandstone cup holders after
curing 1 day and 3, 6, and 12 months downhole in the Cerro Prieto geothermal field, Mexico.
The downhole temperature was 417°F [214°C].

Fig. 10-16. Permeability performance of typical geothermal well cements under actual
conditions (from API Task Force on Geothermal Well Cements, 1985). Reprinted with
permission from Oil & Gas Journal.

Chapter 10 Thermal Cements 333

10-8.2.2 Alternate geothermal well cement common methods practiced today. These techniques
compositions have been the salvation of many oil fields with high-vis-
More recently, calcium phosphate and calcium alumi- cosity crudes, and essentially involve the trading of heat
nosilicate systems were developed specifically for use in for viscosity reduction (Kastrop, 1965; Butler, 1991).
geothermal wells, and both show superior resistance to As in geothermal wells, the formations associated
CO2. The chemistries of each system are described ear- with steam recovery and fireflood wells are frequently
lier in this chapter. problematic. Weak and unconsolidated zones with low
Calcium phosphate systems (Section 10-6) have been fracture pressures and high permeability are often pre-
used since 1997 to successfully cement geothermal wells sent; as a result, severe lost circulation and fluid-loss
in Japan and Indonesia, and service life is estimated problems are common.
to be about 20 years (Weber et al., 1998; Brookhaven Thermal recovery wells are usually less than 3,000 ft
National Laboratory, 2000). The cement blends consist [915 m] in depth and are frequently deviated (30° to
of fly ash, calcium aluminate cement, sodium polyphos- horizontal). The circulating temperatures during pri-
phate, and water in compositions that vary with the mary cementing operations are often less than 104°F
depth at which the cement will be used. [40°C], and accelerators such as calcium chloride or
Calcium aluminosilicate systems (Section 10-5) were sodium chloride are often added to promote early
developed more recently (Barlet-Gouédard and Vidick, cement strength development.
2001; Barlet-Gouédard and Goffé, 2002). At this writing, Thermal recovery wells are always cemented to sur-
the performance of these systems in actual wells has not face. When heat is initially supplied, the temperature
been published. rise should be controlled to prevent undue thermal
Placing cements that are chemically inert to corro- shock to the casing and cement. Nevertheless, because
sive geothermal brines would be an attractive strategy. of thermal expansion, high levels of stress are built up in
Such systems, commonly referred to as “synthetic the pipe and the cement sheath (Carter et al., 1966);
cements,” are used routinely to complete wells for therefore, the strongest possible cement-to-pipe and
CO2-flooding projects or chemical waste disposal cement-to-formation bonds are necessary. Failure of the
(Chapter 7). Epoxy-based polymer systems are most bonds could allow interzonal communication and pipe
commonly used for such applications; unfortunately, expansion. The ultimate result would be casing failure
they would suffer thermal degradation at the tempera- by buckling or telescoping (Humphrey, 1960). A sub-
tures encountered in geothermal reservoirs. However, stantial amount of work has been performed to devise
epoxy cements are often used near the surface, because cementing techniques that minimize the effects of ther-
they can withstand the temperature cycling associated mal expansion. Such methods include the placement of
with geothermal well production. thermal packers (Smith, 1966) and the inclusion of a
Polymers that are stable to high temperatures have sliding sleeve in the casing string that can move freely in
been examined by various researchers. Zeldin and response to thermal stress (Greer and Shryock, 1967). A
Kukacka (1980) developed an organosiloxane polymer third procedure involves holding the casing in tension
cement that was proven suitable as a geothermal during the cement job to minimize the expansion when
cement in an API study. A coal-filled furfuryl alcohol- thermal stress is eventually applied (Farouq Ali and
base cement system for geothermal wells was proposed Meldau, 1979).
by Pettit (1979) and Eilers (1985). Degouy and Martin The cement must also be able to withstand the ele-
(1993) demonstrated that phenolic resins, with fillers vated temperature exposure and thermal cycling associ-
such as calcium carbonate and sand, provided accept- ated with steamflood and fireflood wells. To maximize
able performance at curing temperatures up to 330°F the delivery of heat to the pay zones, an insulating
[200°C]. No commercial use of these technologies has cement is desirable in thermal recovery wells; however,
been reported. the presence of such cements places additional thermal
stress on the casing (Leutwyler, 1966). Thermal conduc-
tivity is more dependent upon the cement density than
cement composition (Nelson, 1986). Typical laboratory
10-9 Thermal recovery wells data are shown in Fig. 10-17. At equivalent density, the
The application of heat to stimulate heavy oil production thermal conductivity of foamed cement is only margin-
has been practiced for more than 50 years. Methods such ally different from that of conventionally extended
as in situ combustion (fireflood), downhole heaters, hot cement.
fluid injection, and steam stimulation have been used. In
situ combustion and steam injection are the most

334 Well Cementing

 Cement density (g/cm3)

0.9 0.7 0.9 1.1 1.3 1.5 1.7 1.9

0.8

Steam
0.7

Oil
0.6
Thermal
0.5
conductivity
BTU/hr ft °F 0.4
0.3
0.2
0.1
5.8 7.5 9.1 10.8 12.5 14.1 15.8
Cement density (lbm/gal)

Fig. 10-17. Typical cement density/thermal conductivity relation-

ship (from Nelson, 1986). Reprinted with permision from Oil & Gas
Journal. Caprock
Sand

10-9.1 Steam recovery wells

Enhanced oil recovery may be accomplished by steam-
flooding or cyclic steam stimulation (Gates and Holmes,
1967; Curtis et al., 2002). Steamflooding consists of intro- Steam injection
ducing steam into an injection well and sending the Heated heavy oil
steam through the formation to a production well. Cyclic Shale
flows to well
steam stimulation of production wells involves the injec-
tion of steam into the production well for a short period Fig. 10-18. SAGD process (from Curtis et al., 2002).
of time and returning the well to production
(Earlougher, 1968). Steam recovery techniques are prac-
ticed extensively throughout the world (Chu, 1983).
During the last decade, the steam-assisted gravity permeability is required, and normal- to high-density
drainage (SAGD) method for recovering heavy crudes slurries are best at providing these qualities.
has been extensively developed (Butler, 1998 and 2001). Unfortunately, because of the lost circulation and ther-
The process uses twin horizontal wells drilled and mal conductivity considerations, such slurries are gen-
extended into the base of a reservoir with the horizontal erally unsuitable. Therefore, much research has been
steam injector placed directly above the horizontal pro- performed to devise low-density slurries with high com-
duction well (Fig. 10-18). In an ideal SAGD process, a pressive strength and low permeability.
growing steam chamber forms around the horizontal Conventionally extended Portland cement systems,
injector, and steam flows continuously to the perimeter containing perlite, bentonite, diatomaceous earth, etc.,
of the chamber, where it condenses and heats the sur- generally perform adequately in steamflood wells, pro-
rounding oil. As the viscosity of the oil decreases, it vided the slurry density is above 12.5 lbm/gal [1.5 g/cm3].
drains to the horizontal production well underneath. Their long-term performance is very similar to that
Thus, the use of gravity increases the efficiency of oil exhibited by such systems in deep wells (Fig. 10-19).
production. The formations in steamflood wells are often so
The most important steamflood fields are located in incompetent that cement systems with densities less
central and southern California, United States; Alberta than 12.5 lbm/gal [1.5 g/cm3] are required to avoid lost
and Saskatchewan, Canada; Venezuela; The Netherlands; circulation or formation damage. Thus, silica-stabilized
West Germany; and Indonesia. Reservoir temperatures foamed cements (Smith, 1983) and microsphere-
seldom exceed 600°F [315°C]; therefore, Portland cement extended systems (Ripley et al., 1980) are very common
is used in virtually all steamflood well completions. in steamflood well completions today. Previously, multi-
The characteristics of steamflood wells and the asso- stage cementing was necessary to successfully complete
ciated performance requirements of cementing materi- these wells.
als are often at cross-purposes. A strong cement with low

Chapter 10 Thermal Cements 335

 12 lbm/gal [1.44 g/cm3]
11 lbm/gal [1.32 g/cm3]
10 lbm/gal [1.20 g/cm3]

Cured at 450°F [232°C] Cured at 600°F [315°C]

6

4
Compressive
strength 3
(thousand
psi)
2

100

10

Water
permeability 1
(mD)

0.1

0.01
1 3 6 12 24 1 3 6 12 24
Time (months) Time (months)
Fig. 10-19. Long-term performance of glass microsphere systems cured at elevated temperatures.

Typical slurries using glass or ceramic microspheres peratures (unpublished data, Nelson, 1987). X-ray dif-
are prepared with a silica-stabilized Portland cement– fraction analysis of the systems revealed the coincident
based slurry. The long-term performance of glass micro- appearance of reyerite and certain aluminosilicate
sphere systems cured at 450° and 600°F [232° and 315°C] hydrate phases. Ceramic microspheres are derived from
is shown in Fig. 10-19. The slurry densities vary from 10.0 fly ashes, and the delayed (reyerite-related) deteriora-
to 12.0 lbm/gal [1.20 to 1.45 g/cm3]. tion of normal-density fly ash cement systems has been
The performance of silica-stabilized ceramic micros- discussed earlier in this chapter.
phere systems at 450° and 600°F [232° and 315°C] is Typical foamed cement systems for thermal wells are
shown in Fig. 10-20. Initially, these systems were gener- prepared from a normal-density base slurry of Portland
ally stronger and less permeable than their glass micros- cement, at least 35% silica flour, a surfactant, and a
phere counterparts. However, between 1 and 2 years of foam stabilizer. The long-term performance at 450° and
curing, significant deterioration was noted at both tem- 600°F [232° and 315°C] of three foamed cement systems

336 Well Cementing

 12 lbm/gal [1.44 g/cm3]
11 lbm/gal [1.32 g/cm3]
10 lbm/gal [1.20 g/cm3]

Cured at 450°F [232°C] Cured at 600°F [315°C]

6

4
Compressive
strength 3
(thousand
psi)
2

100

10

Water
permeability 1
(mD)

0.1

0.01
1 3 6 12 24 1 3 6 12 24
Time (months) Time (months)

Fig. 10-20. Long-term performance of ceramic microsphere systems cured at elevated temperatures.

with densities ranging from 9.0 to 12.0 lbm/gal [1.08 to Febus, 1984). Compressive strength and permeability
1.44 g/cm3] is shown in Fig. 10-21. Comparison of the data for systems cycled between 550° and 100°F [288°
foamed cement data with those of equal-density micros- and 94°C] are shown in Table 10-5. More recently, ther-
phere systems reveals the foams to have significantly mal cements containing additives that impart flexibility
higher compressive strength. The water permeabilities (Chapter 7) have been successfully introduced for
of the foamed cements are also higher (>0.1 mD), and steamflood applications (Stiles and Hollies, 2002; Stiles,
more variable with curing time. 2006).
Foamed cements have also been shown to resist
repetitive thermal cycling, which occurs when the cyclic
steam stimulation technique is applied (Harms and

Chapter 10 Thermal Cements 337

 12 lbm/gal [1.44 g/cm3]
11 lbm/gal [1.32 g/cm3]
9 lbm/gal [1.08 g/cm3]

Cured at 450°F [232°C] Cured at 600°F [315°C]

6

4
Compressive
strength 3
(thousand
psi)
2

100

10

Water
permeability 1
(mD)

0.1

0.01
1 3 6 12 24 1 3 6 12 24
Time (months) Time (months)

Fig. 10-21. Long-term performance of foamed cement systems cured at elevated temperatures.

Table 10-5. Effect of Thermal Cycling on Performance of 10-9.2 In situ combustion wells
Foamed Cements for Steamflood Conditions† In situ combustion recovery, or fireflood, consists of ini-
Properties of Foamed Cement Density tiating combustion in an injection well and then propa-
Foamed Cement gating the combustion front by the injection of air or
10 lbm/gal 11.5 lbm/gal 13 lbm/gal
oxygen through the reservoir to the production wells
Compressive strength 1,210 psi 1,680 psi 2,260 psi (Chu, 1981; Petit et al., 1992). In such wells, the cement
after 20 days at 550°F
is exposed to maximum temperatures between 700° and
Compressive strength 1,630 psi 1,550 psi 2,440 psi 1,700°F [371°and 926°C] near the burning zone. Such
after 100 days at 550°F‡ temperatures exceed the stable range of Portland
Compressive strength 1,240 psi 2,020 psi 2,430 psi
cement; therefore, high-alumina cement is necessary.
after 160 days at 550°F§ Fireflood wells are physically similar to and are usu-
ally found in the same locations as steam injection wells.
Air permeability after 2.4 mD 1.0 mD 0.9 mD Thus, the formation conditions and cement performance
100 days requirements are basically the same. Usually, most of the
† Surface slurry: 15.4 lbm/gal Class G, 40% silica flour, 3% lime (from Harms and Febus, 1984).
Reprinted with permission of SPE. casing is cemented with Portland cement systems, with
‡ Cycled to 100°F twice.
calcium aluminate cement placed opposite and about
§ Cycled to 100°F three times.

338 Well Cementing

100 ft [31 m] above the pay zone as a tail slurry. However, because of the previously described conversion of the
this practice is not without risk. Calcium aluminate initial aluminate hydrates to C3AH6. The water-perme-
cement is a strong accelerator of Portland cement, and ability values are extremely low as well.
setting can occur within a few minutes after the two The performance of foamed calcium aluminate
slurries commingle in the annulus. For complete safety, cements has also been investigated (Nelson and Eilers,
calcium aluminate cement should be the only system 1985). Figure 10-23 shows the compressive strength and
pumped. water permeability of three systems cured for 7 and
The performance of two normal-density calcium alu- 28 days at 1,250°F [677°C] in a refractory furnace. Two
minate cement systems is depicted in Fig. 10-22. Data foams, with densities of 11.0 and 9.0 lbm/gal [1.32 and
are given for systems cured at 100° and 220°F [38° and 1.08 g/cm3] were prepared from a neat calcium alumi-
93°C] and heated in a refractory furnace at 600°, 1,000°, nate cement-based slurry. Another foam, with a density
and 1,500°F [315°, 538°, and 815°C]. The compressive of 11.0 lbm/gal [1.32 g/cm3], contained fly ash. The com-
strengths of the aluminate systems at the lower temper- pressive strength was adequate; however, the water
atures are adequate, yet considerably lower than similar- permeabilities were excessive.
density Portland cement systems. This is primarily

538°C 538°C
38°C 93°C 315°C 815°C 38°C 93°C 315°C 815°C
15 100

10

10
1
Compressive Water
strength permeability
(MPa) (mD)
0.1
5

0.01

0 0.001
1 3 7 1 7 3 1 7
Curing time (days) Curing time (days)

100% ground firebrick (BWOC)–15.6 lbm/gal [1.87 g/cm3]

200% ground firebrick (BWOC)–15.7 lbm/gal [1.87 g/cm3]

Fig. 10-22. Compressive strength and permeability performance of calcium aluminate cement systems
at various temperatures (from Nelson and Eilers, 1985). Reprinted with permission from the Petroleum
Society of CIM.

Chapter 10 Thermal Cements 339

 15 10
Neat: 11.0 lbm/gal [1.32 g/cm3]
Fly ash: 11.0 lbm/gal [1.23 g/cm3]
Neat: 9 lbm/gal [1.08 g/cm3]

10 1

Compressive Water
strength permeability
(MPa) (mD)

5 0.1

0 0.01
7 28 7 28
Curing time (days) Curing time (days)

Fig. 10-23. Performance of foamed calcium aluminate cement systems at 1,250°F [677°C] (from Nelson
and Eilers, 1985). Reprinted with permission from the Petroleum Society of CIM.

■
10-10 Conclusion Microsphere cement systems can be used in thermal
The preceding discussion has demonstrated that ther- wells, provided the base slurry is stabilized to high
mal cements encompass a wide variety of wellbore con- temperatures, and the collapse pressure (usually
ditions and complex chemical processes. Many factors 3,000 psi or 20.7 MPa) is not exceeded.
■ Foamed cement, made from a stabilized base slurry,
must be considered to determine the optimum cement
composition for a particular situation. Nevertheless, can be used with confidence in most thermal wells. In
there are several basic points that the engineer should geothermal wells, in which corrosive fluids are pro-
remember when contemplating this problem. duced, the long-term stability of foamed cements has
■ When static temperatures exceed 230°F [110°C], 35%
not been proven.
■ If the cement will be exposed to temperatures
to 40% silica BWOC must be added to Portland
cements; otherwise, strength retrogression will occur. exceeding 750°F [400°C], Portland cement should
■ If saline geothermal brines are present, fine silica
not be used. High-alumina cement is suitable.
■ Silica is deleterious to the stability of high-alumina
flour (less than 15-μm particle size) should be added
to Portland cement as a stabilizer. Silica sand does cements at temperatures exceeding 572°F [300°C].
not reliably provide adequate protection. Crushed aluminosilicate firebrick or fly ash is suit-
■ If high concentrations of CO2 are present, using cal-
able.
■ During laboratory testing, accurate static and circu-
cium aluminosilicate or calcium phosphate cements
is recommended. If Portland cement is used, degra- lating temperatures must be used to obtain an opti-
dation can be inhibited by reducing the silica con- mal thickening time and compressive strength at the
centration to 20% BWOC. wellsite.
■ Most common cement extenders are compatible with
thermal cements; however, if the static temperature
exceeds 450°F [232°C], fly ash should not be used in
Portland or Class J cement systems. Bentonite, per-
lite, and diatomaceous earth are suitable.

340 Well Cementing

■ High pressure strongly affects the behavior of thermal 10-11 Acronym list
cement systems; therefore, laboratory testing must be API American Petroleum Institute
performed at the anticipated bottomhole pressure.
■
ASTM ASTM International, formerly the American
Thermal cements are sensitive to subtle chemical Society for Testing and Materials
changes; therefore, laboratory testing should always
be performed with samples of the cement, additives, BHCT Bottomhole circulating temperature
and location water that will be used during the job. BHST Bottomhole static temperature
■ The common assumption that high compressive BSC Belite-silicate cement
strength is automatically linked with low permeabil- BWOC By weight of cement
ity is false. Permeability should be measured in the C/S Bulk lime-to-silica (ratio)
laboratory before a cement system is placed in a ther-
mal well. CA Monocalcium aluminate
CH Calcium hydroxide
C-S-H Calcium silicate hydrate
SAGD Steam-assisted gravity drainage
WOC waiting on cement

Chapter 10 Thermal Cements 341

 Cementing Equipment and Casing Hardware
Ed Leugemors, James Metson, and Jean-Louis Pessin—Schlumberger

11-1 Cementing materials

R.L. Colvard—Weatherford International
C. Dennis Krauss and Mark Plante—Baker Oil Tools

Before describing the design and function of cementing

equipment, one must be familiar with the physical and
11
(94 lbm) or SI (50 kg) sacks, so-called “big bags” (1 to
1.5 SI tons generally), or larger quantities (truck, rail-
way car, or ship).
chemical properties of the various cementing materials.
A thorough discussion is presented in Chapters 2 and 3; 11-1.2 Water
however, a rapid review of the principal points is useful Fresh water is normally used for cementing onshore
for this discussion. wells and seawater for offshore locations. However, one
must be aware that fresh waters are often not very
11-1.1 Cement “fresh.” Inorganic salts and organic residues from vege-
As discussed in Chapter 2, Portland cement is used tation are frequently present in significant quantities.
during almost all well cementing operations. It is a Such materials are known to affect the performance of
finely divided and highly reactive powder. The vigorous Portland cement systems.
hydration of Portland cement during initial mixing, as
well as the changeable slurry properties during place- 11-1.3 Dry cement additives
ment, complicate the design of cement mixing and Table 11-1 summarizes the relevant properties of dry
pumping equipment. additives with regard to cementing equipment and
Portland cement is usually stored in silos at a central logistics.
storage location. Alternatively, it may be packaged in U.S.

Table 11-1. Important Properties of Dry Additives with Respect to Cementing Logistics
Material Chemically Inert Chemically Active
Form Insoluble powder or finely cut material Soluble powder or finely cut material

Examples (function) Powdered coal (extender) Lignosulfonate (retarder)

Hematite (weighting agent) Sodium silicate (extender)
Barite (weighting agent) Calcium chloride (accelerator)

Maximum concentration 100% (BWOC) 10% (BWOC)

(order of magnitude)

Influence of accuracy in Concentration acts directly on system Materials may have secondary effects.
concentration on slurry quality density; no unexpected effect.

Handling The additives are blended with the dry The additives are normally dry-blended with the
cement in a special blender at the central cement. They are sometimes added to mix water
storage location up to several days before on location in an open horizontal tank, just prior
the pumping job. The blended material is to the job.
then transported to the wellsite. (If the
amount required is small and the material
easily scattered in the water, it is treated
as a soluble material.)

Well Cementing ■ Chapter 11 Cementing Equipment and Casing Hardware 343

11-1.4 Liquid cement additives Table 11-2. Important Properties of Liquid Additives
Offshore, liquid additives are usually preferred. Such with Respect to Cementing Logistics
materials are more compatible logistically, because their Type of material All liquid additives are chemically active.
mixing requires less space. Because liquid additives are Frequently, they are water solutions of
preblended with the mix water, the resultant slurry the corresponding dry additive.
tends to be more homogeneous than one mixed from a Examples Lignosulfonate solution (retarder)
dry-blended powder mixture. Table 11-2 is a summary of (function) Naphthalene sulfonate solution (dispersant)
the relevant points. Latex (gas migration prevention)

Concentration (order Up to 25 L/100 kg cement

of magnitude) (3.0 gal/U.S. sack)
11-2 Basic equipment
Influence of Materials may have detrimental
Figure 11-1 is a schematic flow diagram of cement-slurry accuracy in secondary effects.
preparation that indicates the steps performed at the concentration
central storage location and at the wellsite. Each func- on slurry quality
tion in the flow diagram also represents a major piece Handling Additives are blended with the mix water
of equipment. Some functions may be combined into a on location in a horizontal open tank,
multipurpose “cementing unit.” In the following discus- shortly before the job. They also can
sion, the design and operation of each class of equip- be blended on the fly during the job.
ment are presented in detail. The second method is preferred.

Localization Mainly offshore

11-2.1 Cement and dry additive blending
Figure 11-2 is an overview of the delivery and blending
processes that occur at the central storage location. pneumatic loading bottles (Fig. 11-3), mechanical screw
Cement is delivered to the central storage location. elevators, or combined systems (for unloading from
Upon delivery in bags, the cement is usually combined. ships). These transfer systems may also be used to load
Equipment that may be used in the transfer includes dry additives.

Central storage location Transfer Wellsite

Storage Storage

Dry
cement (1)
Dry (3) Blend Surge
additive (3) tank
blending
HP steel
Dry
flow hoses
additives (2)
Water
Liquid Cement
Cement
additive slurry
mixer
mixing pumper
Liquid (2) Liquid
additives (2) additives (4)

Well casing
(1): Usually bulk, possibly U.S. (94-lbm) or metric (50-kg) sacks, or “big bags” (0.5 to 1.5 metric tons) topped with
(2): Manufacturer‘s packaging the cement
(3): Bulk (for very small and/or unplanned jobs, cement is sometimes stocked in paper sacks) head
(4): Bulk in special containers, except if mixing is done in an open tank

Fig. 11-1. Typical cementing process.

344 Well Cementing

 Cement from cement mill delivered
to the central bulk station

1.

2.

3.

4. 6.
1. Standard sacks or large bags
2. Bulk (notice the absence 5. Cement from the bulk
of surge tank on trailer) station sent to the
3. Sacks, large bags, or bulk rig location
4. Pneumatic silo 7.
5. Pneumatic loading bottle
6. Dry-additive blender 8.
7. Air-compressor plant
8. Horizontal tank trailer with
9.
merge tank
9. Twin vertical tank trailer
with merge tank
10. Supply boat, cementing 10.
barge, or vessel

Fig. 11-2. Delivery, storage, and distribution of cement and dry additives.

Loading The bulk cement is stored in pneumatic or atmos-

(unsacked dry pheric silos. Transfer systems are available to move the
materials)
cement from one silo to another or to a blender, road
transport unit, or supply boat. When the transfer system
is pneumatic, several silos are connected permanently to
Hopper save time and labor. In humid climates, an air dryer may
be installed in the system.
The cement and dry additives are usually combined in
Transfer to Bottle a pneumatic blending tank (10- to 20-ton capacity) at
dry-additive the central bulk material plant (Fig. 11-4). Bulk materi-
blender, bulk als are usually air-blown, and sacked materials are
storage silo, or
bulk transport Pressurized poured into the tank through a hopper located on top.
air The sacked additives may also be poured into pneumatic
additive bottles (1- or 2-ton capacity) and then blown
into the blending tank. Pressurized air is supplied by one
Fig. 11-3. Pneumatic dry-material loading bottle.
or more air compressor units.
The bulk materials are loaded into the blending tank
first for weighing purposes. A weigh cell is permanently
integrated into the tank frame. The sacked additives are
loaded last. Air, pressurized to about 35 psi [2.5 bar], is
injected through nozzles into the mass of the materials
until thorough blending is accomplished. Then the blend
is transferred pneumatically to a bulk material transport
container for delivery to the wellsite. To obtain the
required amount of cementing materials, more than one
batch must frequently be prepared.

Chapter 11 Cementing Equipment and Casing Hardware 345

 rough (e.g., corrugated desert track). This problem is
Dust magnified if the blended materials have significantly
collector different particle sizes or densities. If small quanti-
ties of soluble or easily dispersible additives are
needed, mixing the dry additives with water at the rig
Hopper
Transportation site is often preferred.
Vent skid ■ If the cement job is canceled or postponed for a long
Pressure
period after the blend has been prepared, a storage or
filling lines reuse problem may arise. Adding liquid additives only
during the job prevents such situations because shelf
Air-jetting Vent from other life problems are eliminated.
system tank (if needed†)

Delivery Delivery 11-2.2 Transportation of bulk materials or blends

†Used
to the wellsite
if the other pressurized tanks
in the bulk station vent out only The equipment used to deliver the cementing materials
through this dust collector to the wellsite varies according to the location. The var-
Fig. 11-4. Dry-additive blender. ious types of transports are discussed below.
Land rigs: Trucks or semitrailer transports are gen-
erally used for land operations. As shown in Fig. 11-5,
truck-mounted vertical tanks or semitrailer-mounted
Although very popular, this method has two draw- vertical or horizontal tanks are the most common.
backs. Limited access locations: Cementing materials in
■ Some particle segregation (caused by shaking) can sacks or big bags can be transported to remote locations
take place during transportation to the rig site, espe- by truck, cargo plane, barge, ship, or helicopter. The con-
cially if the distance is long and the road surface is

Air compressor
Surge
tank

Fill Vent

High-pressure Low-pressure
delivery delivery

Air compressor Surge

tank
High-
pressure
delivery

Low-
pressure
Vent delivery
Fill

High-pressure delivery: Material transferred to storage tank or to mixing unit surge tank
Low-pressure delivery: Material air-blown to surge tank on rear of unit and dumped into cement mixer hopper

Fig. 11-5. Bulk material transports.

346 Well Cementing

tainers are made of durable, water-resistant materials. Side view Back view
Sometimes, the containers are placed in stackable Vent
crates for added protection. Cement
Offshore rigs: Supply boats or cementing vessels are
used for offshore locations. They normally have built-in
tanks, pneumatic unloading equipment, and a supply of
hoses. Sometimes, mobile skid-mounted tanks with a low
center of gravity and mobile unloading equipment are

Back
Face
used. It is important to note that the pneumatic equip-
ment must be sufficiently powerful to blow heavy mate-
rials such as barite (specific gravity of 4.33) up to the
drilling rig tanks, a vertical distance of 130 to 200 ft [40 Porous
to 60 m]. The air compressors used for this task typically material
deliver 250 to 350 ft3/min [7,100 to 9,900 L/min] with a Air box
Valve Fill Vent
pressure rating of 28 to 44 psi [2 to 3 bar]. Sock
Air box
Delivery into
11-2.3 Wellsite storage of cement or blends Aeration and cement mixer
fluidization air hopper
As discussed above, pneumatic bulk trucks or trailers
transport neat or preblended dry cement to the wellsite Fig. 11-6. Atmospheric transportable bulk tank (typical piping
from the central storage and blending plant. Neat arrangement).
cement can also arrive directly from the cement mill.
The material is then transferred pneumatically to trans-
portable tanks that are either brought to the rig site for Safety valve
the cement job or are a permanent part of the drilling rig
equipment. Such tanks are similar to those used at cen-
tral storage locations, but their dimensions allow trans-
port on standard or specially designed (with a built-in
hydraulic laying/raising system) trailers. When empty,
the tanks must not exceed the weight limits specified by
various countries. A large variety of storage tanks for
road travel exists within two principal categories—
atmospheric and pressurized. Both are equipped with a
set of skids for proper installation on imperfectly leveled
ground and for easy winching onto trailers.
The atmospheric tank is always operated in a vertical
position. Air at low pressure (about 3 psi [0.2 bar]) is
blown into a gutter fixed to the slanted bottom of the 8 jets (1-in.)
2-in. pressurizing line
tank. The roof of the gutter is made of a porous material. 5-in. bleedoff/vent line
5-in. fill
The air passes through the porous partition and fluidizes 5-in. material delivery
Quick manhole
the cement blend. The cement blend glides along the 50-in. ID 6 jets (1-in.)
slanted bottom to a chute gate and then to the hopper of 2-in. air (letting) Bronze porous floor
a slurry mixing system. As illustrated in Fig. 11-6, atmos- 2-in. air (aeration)
pheric tanks are made in the shape of a parallelepiped.
Pressurized tanks use air at about 44-psi [3-bar] pres- Fig. 11-7. Pressurized bulk tank (typical piping arrangement).
sure and can operate horizontally or vertically.
Figure 11-7 is a schematic diagram of a typical unit. As
shown in Fig. 11-8, the vertical tanks are generally cylin- mixer. For versatility, some vertical pressurized tanks
droconical in shape, while horizontal models are more are also equipped to release the cement directly to a
complex. In the first stage, pressure-reduced air is blown hopper at atmospheric pressure (Fig. 11-9).
from the bottom through the mass of cement for aera- The bulk trailers are sometimes used for additional
tion and fluidization. Then air at 44 psi [3 bar] is storage. Indeed, they can serve all storage needs on the
injected into the tank, and the cement flows out through rig site, provided they are equipped with their own surge
a discharge line to a surge tank, which feeds the cement tanks, described later.

Chapter 11 Cementing Equipment and Casing Hardware 347

 Horizontal (supply boat)

4-in. vent line

5-in. delivery line
Fill line
3-in. air line
Vertical, single Vertical, twin

Fig. 11-8. Pressurized bulk tanks (various configurations).

Delivery to surge tank

(in high-pressure mode)

Bulk cement
(or blend)

Valve‡
Porous
Enlarged material Annular
to show air box
detail Valve† Valve†
Sock
Aeration and
fluidization air
Delivery to mixer hopper
Open in atmospheric mode, closed in high-pressure mode
†

Closed in atmospheric mode, open in high-pressure mode

‡ (in atmospheric mode)

Fig. 11-9. Pressurized bulk tank (typical piping arrangement).

11-2.4 Metering of water freshwater or seawater distribution system is used. Each

Contrary to what one might think, the simple method of batch of mix water is then used successively to feed the
employing a flowmeter is not used for water metering. A cement mixer.
set of twin 10-bbl (or sometimes 20-bbl) tanks is pre- The additives may be preblended with the water in
ferred. A “displacement tank” (Fig. 11-10), which is the storage tanks or may be blended while the water is
divided equally by a partition, is also used. passing through the displacement tank. In the second
Both sides of the partitioned tank are filled with mix case, a liquid-additive metering system (described later)
water from the rig storage. If the job is offshore, the is required.

348 Well Cementing

 Additives (from exist on the rig site, especially offshore. If the job is can-
liquid additives celed or postponed, the costly solution may have to be
metering system) thrown away. Also, if a larger-than-expected volume of
slurry becomes necessary during the job, the volume of
Water the premixed additive solution may be inadequate. Thus,
methods that allow continuous (“on-the-fly”) mixing are
often preferred. On-the-fly methods employ a semiman-
ual or automatic metering system that delivers the cor-
Closed
Open rect amount of additives to each side of the displace-
Partition ment tank.
Water
pipe

Level scale
11-2.5.1 Liquid additive metering system with
metering tanks
All liquid-additive metering systems consist of two prin-
cipal parts—a storage and transfer unit and a metering
unit.
The storage/transfer unit generally includes four stor-
age tanks of various capacities (usually between 6.2 and
Drain
Open
25 bbl [1,000 and 4,000 L]). The storage and transfer
pipe Closed unit allows the independent metering of additives
according to the requirements of a particular job. This is
Mixing water
(to the mixing
convenient because well cement slurries typically con-
pump) tain two or three additives.
Each storage tank is equipped with its own air-oper-
Fig. 11-10. Displacement tank system. ated diaphragm pump and agitation system (recircula-
tion, as illustrated in Fig. 11-11, or air-operated stirrer)
to avoid segregation of the additive components.
For precise placement of the slurry in the wellbore, Therefore the operation of the unit requires a source of
the volume of the displacement fluid must be accurately clean and dry air at 120 to 145 psi [8 to 10 bar]. The con-
measured. After the cement slurry has passed through figuration of the unit varies depending on whether it is
the mixing system, the displacement fluid usually passes designed for use on land (skid or trailer-mounted) or off-
through the displacement tanks for volume measure- shore (containerized).
ment and is pumped by the cementing unit instead of The metering unit generally consists of a set of three
the rig’s mud pumps. (Fig. 11-11) or four 25-gal or 10-L tanks, with visible level
scales. To prepare a batch (10 bbl or 20 bbl [1.6 m3 or
3.18 m3] according to the displacement tank), the
11-2.5 Liquid additive storage and mixing
proper amounts of the selected additives are introduced
The simplest method of mixing liquid additives (and dry into the metering tanks. The additives are then released
additives at less than 3% by weight of cement) with into one of the two displacement-tank sections that is
water consists of pouring the required amount of each being filled with water. Finally, the mixture is agitated to
additive into a tank of water. One should measure the obtain a homogeneous solution. The same operation is
additives and water accurately to obtain the correct con- repeated for the following batch in the other half of the
centration; the preparation of a slight excess of solution displacement tank and so on. The repetitions of the
is also advisable. The mixing can be achieved with a operation may be automatically or semiautomatically
paddle mixer, circulation pump, jetting system, or a com- controlled.
bination of these.
The premix method has several disadvantages.
Premixing requires an extra tank, which must be clean
and sufficiently large. Extra tanks are not always avail-
able, and sufficient space to accommodate them may not

Chapter 11 Cementing Equipment and Casing Hardware 349

 A1 A2 A3
Additive
metering
tank(s)

Open Closed Air-powered

Closed Open stirrer

Water
Additive
storage
tank(s)

Air-powered
diaphragm
pump(s)

Mixing water

Fig. 11-11. Liquid-additive metering system (with metering tanks).

11-2.5.2 Liquid-additive metering system without 11-2.6 Surge tanks

metering tanks For smooth cement mixer operation, the supply of
The liquid-additive-system metering rack (Fig. 11-12) is cement (or blend) should be steady, and the pressure at
used to provide accurate (±2%) delivery of up to four the mixer bowl should remain constant. The bulk
additives into displacement tanks. The operator cement is moved from the storage tank toward the
depresses a button to initiate the delivery. The metering cement mixer, driven by the differential pressure cre-
rack behaves as four “smart valves,” installed between ated between the tank and the end of the line. If the line
the additive pumps and the displacement tanks. The is longer than approximately 23 ft [7 m], the cement
valves are controlled by a microcomputer using data tends to separate from the conveying air into slugs, cre-
from electromagnetic flowmeters. ating pulsating flow. To smooth the flow and allow for
operational requirements, such as changing from one
storage tank to another, a surge tank is used.
Window As shown in Fig. 11-13, the surge tank has a cylindro-
Control cabinet conical shape. It has a capacity of about 12.5 bbl
[2,000 L] and connects the end of the transfer line to the
top of the mixer bowl. This device maintains the pres-
A4 sure above the mixer bowl at atmospheric level plus the
To cementing unit A3 hydrostatic head of the material. A dust separator is
displacement tank From also installed where the conveying air is vented to the
A2 additive atmosphere.
A1 storage In some cases, the surge tank is maintained at higher
tanks
than atmospheric pressure. The advantage is a higher
On/off box delivery rate to the cement mixer, but the drawback is a
higher required pressure from the bulk system to feed
the surge tank. A pressurized surge tank is also used
Fig. 11-12. Metering rack hardware.

350 Well Cementing

with certain types of recirculating cement mixers to 11-2.7 Cement mixing
maintain a steady delivery rate. The tank is pressurized The cement mixer is a device in which a flow of pressur-
by controlling the flow or air in the vent line. This type ized water (possibly containing additives) meets a flow
of surge tank is used almost exclusively offshore, where of cement (possibly containing additives), and a cement
the bulk system is required to transport the cement long slurry is formed at a prescribed rate. Several types of
distances. mixing systems exist; they are described individually
below.

Vent
Dust separator 11-2.7.1 Conventional jet mixer
Air-forced The conventional jet mixer consists of a hopper, a mixing
cement bowl, a discharge gooseneck, and a slurry tub. The max-
Windows imum slurry-generating capacity of the conventional jet
(for watching mixer, evaluated in rate of dry material, is slightly higher
Air-jetting cement level) than 2,200 lbm/min (1 SI ton/min). Figure 11-14 shows a
system
(not shown) configuration for sacked cement, and a system for pneu-
matically delivered cement is illustrated in Fig. 11-15.
The cement is delivered to the hopper. The water is
Valve injected into the bowl through jets for mixing with the
cement and into the gooseneck for adjusting the slurry
Sock density. The jets are chosen according to the operating
pressure, slurry fabrication rate, and type of dry materi-
Delivery als. The movement of cement down through the hopper
(into mixer
hopper) is assisted by the high-pressure flow of water through the
jets. The resulting pressure drop pulls the dry cement
into the stream of water. To reinforce this effect, the
Fig. 11-13. Surge tank. gooseneck can be given a venturi tube profile. Further
along at the gooseneck, turbulent flow mixes the cement
particles with the water, and the result is a cement
slurry.

Mixing water
(from the mixing Slurry (to the
manifold) displacement
pump(s)

Additional
Knife water
Sack
Suction
Gooseneck pipe
Hopper Dry Grating
cement
Two or three jets
y
Slurr
Cutting Bowl
table

Mixing water Slurry tub (200 L)

Fig. 11-14. Conventional jet mixer (sacked cement).

Chapter 11 Cementing Equipment and Casing Hardware 351

 Mixing water
Surge tank or (from the mixing Slurry [to the
atmospheric manifold) displacement
tank pump(s)]

Additional
water
Butterfly valve
Suction
Sock pipe
Dry cement Gooseneck
Hopper Grating
Two or three jets

Bowl

Mixing water Slurry tub (200 L)

Fig. 11-15. Conventional jet mixer (bulk cement).

The slurry density is adjusted by using the bypass ■ The slurry density is adjusted by operating the sliding
system to change the water-to-cement ratio. As the gate.
bypass is opened, the suction effect decreases and ■ The slurry is removed from the slurry tub by a recir-
reduces the amount of cement drawn out of the hopper. culation jet, fed by a centrifugal pump. The centrifu-
At the same time, the water bypassing the jets enters the gal pump force feeds the displacement pumps and
slurry. The combined effect is a decrease in slurry den- recirculates some slurry through the mixing system.
sity. Conversely, if the bypass is closed, the density ■ Water is always injected ahead of the recirculation
increases.
jet.
The conventional jet mixer can be operated at low
(175 to 200 psi [12 to 14 bar]) or high (880 to 1,180 psi Recirculation through the mixer heart and the tub
[60 to 80 bar]) water pressure. In the first case, the mix- improves the homogeneity and rheology of the slurry.
water pump is a centrifugal pump. In the latter case, it Adjustment of the slurry density is also easier.
is a reciprocating pump, usually identical (except per-
haps in plunger size) to the displacement pump. The
“double high-pressure pump cementing units,” which 11-2.7.3 Recirculation mixer without conventional
are the most widely used throughout the world, are jets
equipped to mix at either low or high pressure. The low- Available equipment includes a variety of mixers without
pressure method is preferred for two main reasons. Less conventional jets (Fig. 11-17). The maximum capacity of
horsepower is required and, because both high-pressure most mixers, evaluated by rate of dry material, is close to
pumps are available to displace the slurry, higher mixing 4,400 lbm/min [2 SI tons/min]. They all consist of the fol-
and displacement rates are possible. With the high-pres- lowing.
sure method, the jets and the bowl-and-gooseneck ■ A sophisticated metering system to mix cement with
assembly are less apt to become plugged with dirty water and a device to mix the resulting slurry with
mixing water or poor-quality cement. previously mixed slurry from the mixing tub
■ A centrifugal pump or similar device (located at the

11-2.7.2 Recirculation jet mixer bottom of the tub) to improve the initial mixing by
The maximum capacity of the recirculation jet mixer shearing, ensure recirculation through the mixer, and
(Fig. 11-16) is slightly more than 4,400 lbm/min feed pressurized slurry to the downhole pump
(2 SI ton/min). The recirculation jet mixer differs from ■ A mixing tub that can be divided into two sections,

the conventional type in several ways. each of which can be equipped with a stirrer to
■ A remotely controlled sliding gate is present between
improve mixing, allowing a film-like flow over the
the hopper and the bowl. common partition that assists the release of
entrapped air

352 Well Cementing

 Depending on the model, the density of the slurry is cement is delivered to the mixer. Normally, the cement
remotely controlled by metering the cement and/or is transferred directly from a pressurized tank without
water. The water rate is usually kept constant, and the passing through a surge tank.
slurry density is controlled by altering the rate at which

Mixing water
(from the mixing
manifold) Slurry [to
Centrifugal the displacement
pump pumps(s)]

Dry cement

Suction
Remotely controlled pipe
sliding gate Grating
Tub
Jet(s) recirculating
line

Mixer
Slurry tub recirculating
line

Fig. 11-16. Recirculation jet mixer.

Bulk cement

Cement
mixing valve
Mixing water
Water metering
valve (annular) Slurry

Recirculation
line

Centrifugal
pump

Slurry (to the

displacement Deaeration Mixing tub
pumps) partition
(optional)

Fig. 11-17. Recirculation jet mixer (without jets).

Chapter 11 Cementing Equipment and Casing Hardware 353

11-2.7.4 Cement mixing units Front view
As discussed above, the batch-blending system and the Vent
liquid-additive metering system have been designed to Cement Missing water
solve the proportioning problems encountered with Surge (from the mixing
or blend tank
cementing materials. However, the slurry properties are manifold on the
affected not only by the proportions of cement, water, cementing unit
and additives, but also by the shearing that occurs
during mixing.
New
Proper operation of a mixing unit should help blend slurry
the appropriate quantities of the cement blend and the
mix water. The correct ratio will provide the expected Cement
mixer
slurry density and other properties. Continual verifica- Holding
tion of slurry density is essential; however, some density tank
fluctuations during slurry mixing are unavoidable. The
longer the mixing time and the larger the slurry volume, Recirculating
the better the homogeneity of the resulting slurry. pump
Finally, the slurry should be given the proper amount
of shearing, which is a function of mixing energy and
mixing time. Because a centrifugal pump is an ideal Top view
shearing device, it is advisable to increase the volume of
slurry being recirculated so that such a pump can be Slurry [to the
downhole Recirculation
used. pump(s)] slurry
Recirculating mixers are available in a variety of con-
figurations (skid-, truck-, or trailer-mounted, diesel- or
electric-powered, sometimes soundproofed) and sizes.
They all have certain common features (Figs. 11-18 and Fig. 11-18. Single 6-bbl tank mixing unit (with conventional jet mixer).
11-19).

Cement
From mixing or blend
manifold Surge
Mixing tank
water

Holding Recirculation Holding

Tank No. 1 jet mixer Tank No. 2
Stirrer Stirrer

Densitometer

Slurry Slurry
To downhole To downhole
pump Centrifugal Centrifugal pump
Pump No. 1 Pump No. 2
Drain Drain

Available Available
Fig. 11-19. Twin-tank mixing unit (with recirculation jet mixer).

354 Well Cementing

■ A surge tank with a capacity of 9.4 to 25.2 bbl [1,500
to 4,000 L] Discharge
■ A conventional or recirculation jet mixer port
■ One or two holding tanks with a capacity ranging Discharge
from 6.3 bbl [1 × 1,000 L] to 50 bbl [2 × 8,000 L] (size valve
is limited by transportability concerns)
■ Two recirculating centrifugal pumps (only one on the
smallest units), with a maximum displacement rate Plunger
of 25 bbl/min [4,000 L/min], to either circulate the
slurry in the holding tank(s) to improve shearing and
homogeneity or to feed the slurry to the downhole Suction
pump stroke
Suction
■ A pair of paddle stirrers, hydraulically or electrically valve
driven, to maintain homogeneity
■ A manifold, sufficiently versatile to be used in a vari-
ety of combinations
In particular cases such as very small jobs, or when
the proportions of additives and the slurry density are
very critical, the total volume of slurry needed to com-
plete the job (including the usual excess) is prepared
before pumping downhole. The liquid additives are not
metered as described earlier; instead, they are released Plunger
directly into the tank or added through a jet mixer. No Discharge
special practice exists regarding dry additives. The vol- stroke
umes or weights of dry additives are usually measured in
the conventional manner.
Suction manifold
(not shown)
11-2.8 High-pressure pumps
All current cementing pumps are of the reciprocating Fig. 11-20. Reciprocating plunger pump fluid end (section).
type, mostly with three plungers (triplex) and spring-
loaded suction and discharge valves (Fig. 11-20). The
transformation of the rotating motion of the input shaft pressure and flow ratings without modifying the maxi-
into the reciprocating motion of the plungers is gener- mum available horsepower. The plungers used in
ally accomplished either by a system of a crankshaft and cementing usually have a diameter between 3 and 6 in.
connecting rods or by a swash plate and connecting rods [7.6 and 15.2 cm].
system. These pumps include an internal fixed-ratio
speed reducer. Depending on the make and the model,
the plunger stroke varies from 5 to 10 in. [12.5 to 25 cm]. 11-2.8.2 Hydraulic horsepower
The global efficiency of triplex pumps is 85% to 90%. Depending on the make and the model of the pump, the
If adequately pressurized, the volumetric efficiency can maximum horsepower varies between 200 and 500 hhp
reach 98% with water at 80% of the maximum speed. The [150 kW to 370 kW].
construction is particularly rugged, allowing the pumps
to handle the heaviest and most abrasive slurries.
11-2.8.3 Versatility
Heavy-duty triplex pumps, which can handle gravel-
11-2.8.1 Convertibility laden fluids, can also perform hydraulic fracturing
Depending upon the manufacturer, the “size” of a pump treatments. In the 200-hhp to 500-hhp [150-kW to
can be altered by changing either the fluid-end assembly 370-kW] range, the same pumps are used for cementing
or the plungers and packing system, using adapters to and stimulation.
convert the fluid-end body. Size alteration changes the

Chapter 11 Cementing Equipment and Casing Hardware 355

11-2.8.4 Maximum flow rate and pressure Slurry density: The slurry density is traditionally
Available makes, models, and sizes offer a large variety measured manually by a pressurized mud balance
of specifications. One should bear in mind that, during (Fig. 11-21). More sophisticated systems determine the
most cement jobs, only one pump is used to pump slurry density with a mass flowmeter that employs the
downhole. Usually the maximum rate is 8 bbl/min Coriolis effect (Fig. 11-22) or with a radioactive densito-
[1.3 m3/min]. This limitation is based upon the maxi- meter connected to a central computer-based recording
mum allowable rate for the 2-in. treating line most com- unit (Fig. 11-23).
monly used for cementing.
The pump pressure usually does not exceed 1,030 psi
[70 bar] (cement squeezes excepted). As a matter of
fact, if the density of the cement slurry is equal to that
of the drilling mud, the pumping pressure is simply a
consequence of the friction losses in the surface equip-
ment (steel flow hoses and cement head) and below the
surface.
Fig. 11-21. Pressurized mud balance (photo courtesy of Chandler
11-2.8.5 Drive Engineering, LLC).

The pumps on mobile units are driven by a diesel engine

and may have either automatic or manual transmissions.
Those permanently installed on an offshore rig are fre-
quently driven electrically (usually with a directly cou-
pled DC motor).

11-2.9 Controls and instruments

At or before the start of a job, some control devices on
the mixer are selected (e.g., chokes on a jet mixer) or set
in position (e.g., a mix-water valve on most recirculating
mixers) according to the composition and density of the
slurry and the desired injection rate. During the job, the
final adjustment is made with either the cement (blend)
or the mix-water metering valve, depending upon the
type of equipment. Adjustment of the downhole pump
rate may also be necessary to maintain a constant level
in the slurry tank and maintain the pumping pressure
within fixed limits (e.g., squeeze jobs).
Cement jobs require the measurement of many para-
meters. A discussion of the various types of measuring
equipment is given below.
Mix water: The volumes of water are measured in the
displacement tanks.
Cement (blend) and slurry: The volumes of mixed
slurry and dry cement are calculated from measure-
ments of the mix-water volume and the slurry density. Fig. 11-22. Coriolis effect mass flowmeter for slurry-density
Flow rate: The slurry flow rate is observed at the measurement.
downhole pump-stroke counter. A flowmeter is used if
job parameters are being recorded continuously.
Pressure: The pumping pressure is read at a gauge or
mechanical recorder. An electronic pressure transducer
is used if the various parameters are recorded by a cen-
tral unit.

356 Well Cementing

 Pipe Three key parameters are monitored: water flow rate,
Radioactive
mixing-tub level, and slurry-flow rate. The water- and
source slurry-flow rates are measured with electromagnetic
Amplifier Panel
To power flowmeters. The mixing-tub level is monitored by an
supply and ultrasonic-guided wave sensor. The solids fraction can be
Detecting processor calculated by computing the difference between the
cell slurry-flow rate and the water flow rate. The mixing-tub
level is important because the volume of slurry in the tub
Gamma can fluctuate, and the system must be able to instantly
rays compensate for such fluctuations as it calculates the
solids volume fraction.
Slurry rheology: Rheology measurements are not
Fig. 11-23. Schematic diagram of radioactive densitometer. routinely performed at the wellsite. However, a couette
viscometer (Appendix B) can be used at the wellsite for
ambient-temperature rheology measurements.
Compressive strength: This measurement is rarely
Solids fraction: With the advent of engineered parti- made at the wellsite because of the long curing times.
cle size cement systems, it has become possible to Slurry samples are normally taken to a central labora-
design high-performance cement systems at slurry den- tory for postjob tests.
sities approaching that of water (Chapter 7). Under Computer-based central recording units are available
these circumstances, the densities of the mix water, the to continuously record vital pumping parameters
cement blend, and the final slurry are essentially identi- (Fig. 11-25). The recorders significantly improve onsite
cal. Therefore, the mixing process cannot be monitored job monitoring while simultaneously storing data for
or controlled by conventional slurry-density measure- postjob evaluation. The stored data can be used to pro-
ments. To overcome this problem, a different quality- duce job-data sheets and enhanced graphics for in-depth
control principle based on solids fraction measurement analysis. Slurry density, flow rate, and pressure can be
was developed (Vigneaux et al., 2003). monitored for various configurations of pumping equip-
The solids fraction of a slurry is the percentage of ment. The microprocessor calculates and displays cumu-
solids contained in slurry sample. It can be expressed in lative volumes, and the total volume is printed on the log
terms of volume or mass. Expressed in terms of volume, in increments as short as one second.
the solids fraction is 1 minus the inverse of slurry poros- One can imagine that the continuous monitoring of
ity (Chapter 7). key pumping parameters could be used to automatically
A schematic diagram of equipment to monitor the control slurry mixing. As of this writing, equipment
solids volume fraction is shown in Fig. 11-24. applying such a principle is beginning to appear in the

Surge
can

Tub level

Water
flowmeter
Slurry
flowmeter
Mixing
tub
Mixer Mixing water

To triplex
pump

Fig. 11-24. Schematic diagram of solids fraction monitoring equipment (Vigneaux et al., 2003). Reprinted with
permission of SPE.

Chapter 11 Cementing Equipment and Casing Hardware 357

Fig. 11-25. Computer-based monitoring system for cement jobs.

field. In the near future, automatic cement mixing will Articulated (loop) section
undoubtedly become a routine procedure.

Three Two
11-2.10 Steel flowhoses and cement head swivels swivels
A “cement head” (Section 11-5.14) is screwed into the
top casing collar or landing joint, depending on the type
of cement job. The discharge side of the downhole pump Half union Half union
and the cement head are connected by a series of artic- (male) (female)
ulated or straight sections of high-pressure steel pipe,
also known as “treating iron” (Fig. 11-26).

Straight section

Fig. 11-26. Steel flow hoses.

358 Well Cementing

11-3 Cementing units
The various components of cementing units, which fab-
ricate and inject the cement slurry, have been described
individually in Section 11-2. Figure 11-27 illustrates the
combination of the components to assemble a basic
cementing unit. A variety of configurations and composi-
tions exists, tailored to the type of rig to be serviced and
the redundancy, versatility, and mobility required.
The various configurations are described below,
according to the type of rig to be serviced.
Skid-mounted units: Illustrated in Fig. 11-28, skid-
mounted units are most applicable to isolated land rigs,
offshore rigs, cementing barges (lakes and rivers), and
open-sea cementing vessels.
Truck-mounted units: Shown in Fig. 11-29, such units
are suitable for almost any land rig. However, the chassis
must be adapted to the type of surface upon which the
unit will travel. The “standard” unit is designed to travel
on roads and must conform to local road regulations. The
“off-road” unit is built for more difficult terrain. The
“desert” unit can be driven over soft surfaces, even sand
dunes.
Fig. 11-28. Typical cementing skid and control console.

Liquid additives from Cement and

the liquid additive Basic cementing unit dry additives
metering system
3 6

2
From rig
4 storage

Slurry
Water from to well
rig storage 9
1 5
10

7 8
1. Centrifugal water supply pump 6. Mixing water manifold
2. Water distributor 7. Cement mixer (conventional jet mixer shown)
3. Additive distributor 8. Slurry tub
4. Displacement tank system 9. Centrifugal pressurizing pump
5. Mixing water pump (centrifugal—low-pressure 10. Reciprocating displacement (downhole) pump(s)
mixing; reciprocating—high-pressure mixing)

Fig. 11-27. Mixing and pumping equipment on rig site (typical setup).

Chapter 11 Cementing Equipment and Casing Hardware 359

 However, single-pump units now exist. One such unit
is shown in Fig 11-31. This unit provides a cost-efficient
solution for less-critical, lower-pressure well casing jobs.
High-pressure treating iron is replaced with a flexible
hose for faster wellsite rig-up.

Fig. 11-29. Typical cementing truck.

Semitrailer-mounted units: Like the truck-mounted

units, semitrailer-mounted units as shown in Fig. 11-30
are appropriate for almost any land rig. They can be
drawn by many types of tractors, providing a logistical
advantage. A heavy tractor-drawn unit with five axles has Fig. 11-31. Single-pump cementing unit.
a better weight distribution ability than the correspond-
ing truck with only three. The maximum authorized pay-
load is greater than that of the truck, which allows the For economic reasons, a cementing unit is designed
loading of more equipment on the same chassis. to meet the requirements (including road regulations)
Helicopter units: Helicopter units are intended for of as many locations as possible. In Europe, for example,
rigs totally inaccessible by land or water. The units, the the required specifications vary from one country to
mixing equipment, and the cement silos are designed to another, and the unit must conform to the most stringent
be transported by helicopter. They can be dismantled regulations. In addition, soundproofing is more fre-
into smaller components, incorporating lifting frames, quently demanded because of the proximity of wells to
and are often made of lighter materials to reduce weight. residential areas (Fig. 11-32).
Traditionally, a cementing unit contains two of each Cementing-unit designs are developed with special
vital item. This redundancy is necessary because a well attention to safety and environmental requirements.
can be severely damaged or lost if it becomes impossible Environmental release of wastes (liquid or dry chemi-
to complete a job after it has commenced. The extra cals, cement slurries, or pump or engine oil spillages) is
equipment serves as an “insurance policy” to protect the prevented by better equipment design and the use of
operator’s investment. recovery receptacles.

Fig. 11-30. Semitrailer-mounted cementing unit.

360 Well Cementing

 Surge tank Control cabin Roof Soundproof power
unit (diesel)

Cement

Recirculation
jet mixer
(not shown)

Pressurizing
pump Slurry tub Low-pressure Two high-
mixing pump pressure pumps

Water Water supply pump Double displacement tank

Water distributor

Mix water manifold

(inside cabin/not shown) Slurry

Fig. 11-32. Semitrailer-mounted cementing unit (Europe).

The safety requirements with which the equipment 1. Special water-cooled manifold rated to cool exhaust
should comply depend upon the location and are espe- gas to 200°C (392°F) maximum, and with a surface
cially dependent upon possible sources of flammable or temperature not exceeding 200°C at any point.
explosive gases. Whenever the unit can be placed more 2. Oversized radiator.
than 98 ft [30 m] away from the well (as on most land rig
3. Inlet air combustion, slam-shut valve.
sites) there are no special requirements. Standard
equipment can often be used without modification. This 4. Inlet air flame trap.
distance condition is often difficult to satisfy on offshore 5. Exhaust gas spark arrestor, DNV type approved.
rigs, where every compartment or deck location is clas- 6. Overspeed valve, which closes the engine blower
sified according to the potential risk of explosion or fire. flapper valve when speed exceeds the normal maxi-
The classification is made by official regulatory bodies mum by 10%.
according to standards that may vary slightly from one
7. High-water-coolant temperature valve, which shuts
country to another; however, operators usually adhere to
down the engine when water temperature exceeds
the most stringent regulations.
95°C (204°F). Fuel rack actuated.
For example, the following is a summary of the Det
Norske Veritas (DNV) requirements for diesel engines to 8. Low-water-coolant level valve, which shuts down the
be located in a hazardous area, classified as “Zone 2,” in engine.
which an explosive gas mixture may exist for a short time 9. High-exhaust-gas temperature valve, which shuts
only under abnormal conditions. Diesel engines are ban- down the engine when the gas temperature exceeds
ished from Zones 0 and 1, which are more sensitive 200°C.
areas. The DNV is the Norwegian certification body, and 10. Special control panel.
its standards serve as a reference in the North Sea.

Chapter 11 Cementing Equipment and Casing Hardware 361

 Diesel engines must often be adapted further to meet Years ago, most casing hardware was manufactured
fire-protection standards. The equipment required to and sold by the cementing company, and the operator
adapt an engine is sophisticated, entirely made of high- depended upon the service companies to develop new
quality stainless steel, and bulky. It is extremely expen- tools. Today, a multitude of engineering groups and man-
sive. Electric motors for Zone 2 areas are typically con- ufacturers supports and supplies the cementing compa-
fined in a closed shelter, which is pressurized with air nies with most casing hardware.
taken from a safe area. An overpressure is maintained so Some cementing tools are retrievable devices that
that no gas from the hazardous area can enter the shel- may require some form of operation from the surface.
ter. When a cementing unit is to be operated offshore in These tools are usually operated by specialists or tool
a nonhazardous area, the drilling and service companies operators within the cementing companies and are typi-
often opt for “protected” diesel engines, which provide cally associated with remedial cementing operations
increased security at a more reasonable cost. Following such as squeeze jobs or pressure testing. Packers, bridge
is a list of the devices that should be installed on a stan- plugs, and retainers are examples of cementing tools
dard diesel engine to ensure this protection. that are commonly used for squeeze or plugback
1. Overspeed valve, which closes the engine blower flap- cementing.
per valve when speed exceeds the normal maximum
by 10%.
2. High-water-coolant temperature valve, which shuts 11-5 Casing hardware
down the engine when water temperature exceeds A typical application of casing hardware for a primary
95°C. Fuel rack actuated. cement job of moderate depth is shown in Fig. 11-33. The
3. Low-oil-pressure valve, which shuts down the engine lower end of the casing is protected by a guide shoe or
when engine oil pressure is below a value to be set- float shoe. A float collar or autofill collar is placed one
tled with the manufacturer. Fuel rack actuated. or two joints above the shoe to provide a seat to land
cement plugs and to halt cement displacement. The
4. High-oil-temperature valve, which shuts down the
short section of casing between the shoe and float collar
engine when oil temperature exceeds 130°C (266°F).
is called the shoe joint and is provided as a buffer within
Fuel rack actuated.
the casing to contain fluid contamination that may build
5. Special control panel. up ahead of the top wiper plug. The length of the shoe
joint may be as few as 1 joint (40 ft [12 m]) or as many
as 10 joints, but most operators use 2 joints. The length
11-4 Introduction to casing hardware of the shoe track is sometimes adjusted on the rig to
accommodate a lack of confidence in the accuracy of the
This section focuses on equipment used on or within the
rig pump efficiency. Rig pumps are often used to displace
casing string to enhance casing placement and cement-
the cement and plug(s). Plugs act as physical barriers to
ing operations. Within the petroleum industry, this
separate the cement slurry from the drilling mud and
equipment is commonly referred to as casing and liner
displacement fluids.
hardware and cementing tools. These tools fall into five
Centralizers are placed in critical sections to prevent
basic categories.
sticking while the casing is lowered into the well. Then
■ Float equipment
they keep the casing in the center of the borehole to aid
■ Centralization and flow enhancement tools mud removal and help ensure placement of a uniform
■ Elastomers—casing or liner wiper tools and cement sheath around the casing in both the openhole
inflatables and cased hole sections.
■ Stage tools

■ Torque and drag reduction tools 11-5.1 Guide shoes and float shoes
An entire textbook could be written about these tools; Guide shoes and float shoes are tapered, commonly
for this textbook, the discussion will be limited to the bullet-nosed devices that are installed at the bottom of
most common or basic types, with the emphasis placed the casing string. They guide the casing toward the
on application, principles of operation, and basic design center of the hole to minimize hitting rock ledges or
characteristics. washouts as the casing is run into the well. The outer
portions of these shoes are usually made from steel, gen-
erally matching the casing in size and threads. The

362 Well Cementing

 Rigid centralizer

Liner hanger

Casing packer

Spring-bow centralizer

Tandem-rise centralizer Low-drag roller centralizer

Stage tool

Centralizer sub
Semirigid centralizer
Reamer shoe

Fig. 11-33. Typical application of casing hardware for a primary cement job (drawing courtesy of
Weatherford International).

inside (including the taper) is generally made of con-

crete or thermoplastic, because this material must be
drilled out if the well is to be deepened beyond the
casing point. Guide shoes and float shoes are made in a
wide size range, from 31⁄2 to 24 in. [8.9 to 61 cm] in diam-
eter, and they are generally used in surface-to-interme-
diate casing strings at shallow depths.
The guide shoe differs from the float shoe in that it
lacks a check valve. The check valve in a float shoe can
prevent reverse flow, or U-tubing, of cement slurry from
the annulus into the casing. The float shoe also reduces
hook weight, because the check valve increases the
buoyancy of the casing string by preventing backflow of
fluid as the casing is lowered into the well.
A basic guide shoe is shown in Fig. 11-34. The steel
Fig. 11-34. Basic guide shoe (drawing courtesy of Weatherford
shell is molded into a rounded nose with a central orifice. International).

Chapter 11 Cementing Equipment and Casing Hardware 363

 A variation is the “down jet” guide shoe. This modifi-
cation incorporates some side jets or ports to divert all
or part of the fluid (Fig. 11-35). The purpose of the jets
or ports is to promote high displacement efficiency by
distributing the fluids evenly around the annulus. The
jets come in a wide variety of sizes and shapes. The con-
crete nose is fairly strong, yet can be easily drilled with
polycrystalline diamond compact (PDC), insert, and
roller cone bits.

Fig. 11-36. Aluminum nose float shoe (drawing courtesy of

Weatherford International).

Fig. 11-35. Cement nose guide shoe with down jet option.

11-5.1.1 Nose materials and shapes

Guide shoes and float shoes are available with various
nose designs, each built for a specific purpose. Most
common among these is the cement nose guide shoe.
The slightly rounded nose is usually made with high-
strength concrete that has a compressive strength of
8,000 and 10,000 psi [5.5 and 6.9 MPa]. The rounded
nose allows the casing to be deflected off of ledges and
obstructions as the casing is lowered. The shells of guide
shoes and most float equipment have grooves into which
the cement flows when poured. As a result, the cement
is not held in place by shear bond alone, and it has the
strength to withstand the stresses associated with casing
placement and the subsequent cement job.
Aluminum nose shoes are another option; however, Fig. 11-37. Eccentric composite-nose float shoe (drawing courtesy
their popularity has been diminished by the introduction of Weatherford International).
of concrete and composite materials. Aluminum noses
can be cast into a wide variety of shapes. Lengthened
noses can ease movement past bridges and obstructions
typical 9 5⁄8-in. eccentric nose float shoe can withstand
downhole (Fig. 11-36).
end loadings greater than 100,000 lbm [45 SI ton] at
Composite nose guide shoes are the newest option,
temperatures exceeding 250°F [121°C]. Their perfor-
made possible by the advent of composite materials that
mance exceeds the American Petroleum Institute (API)
can withstand high load weights and resist abrasion, ero-
specifications, and they can withstand running into well-
sion, and high temperatures (Fig. 11-37). For example, a
bores up to and beyond 90° inclination.

364 Well Cementing

 Other types of guide shoes include the Texas pattern 11-5.2 Float equipment
(beveled nose) and the sawtooth pattern (Fig. 11-38). The primary purpose of float equipment is to be able to
The noses of these guide shoes are formed or cut out of pump cement slurries into the well that are heavier then
the steel shell. The beveled nose shoe has the same out- the mud. Check valves prevent the cement from U-
side diameter (OD) as the couplings and an inside diam- tubing back into the casing or liner string (Fig. 11-39).
eter (ID) that is the same as the casing or drillout diam- Float equipment consists of specialized casing shoes
eter. The bevel is advantageous in situations in which it and collars with check valves to prevent wellbore fluids
is necessary to pull back the bottomhole assembly from entering. As the casing is lowered, the hook load or
through the previous casing. The beveled guide reduces hanging weight is reduced by the weight of fluid dis-
the probability of sticking. placed. The casing is filled from the surface, and the
The beveled nose shoe is most often run in shallow, hook load or amount of buoyancy is controlled by moni-
vertical wells. Such bevels are also incorporated into toring the weight indicator. The frequency of filling is
most float shoes. generally once every 5 to 10 joints; however, some large-
In recent years, the price of float shoes has diameter or thin-wall casings may require more frequent
decreased, and many operators use them routinely to filling to prevent casing collapse. In addition to proper
have the option of an additional valve in the casing filling, the casing should be lowered at a slow, steady
string. However, in deepwater applications, guide shoes rate to prevent pressure-surge damage.
are making a comeback. The guide shoes leave an unre-
stricted path for tripping devices to pass through.

Texas pattern

Fig. 11-39. Typical float shoe (drawing courtesy of Davis-

Sawtooth pattern Lynch, Inc.).
Fig. 11-38. Texas pattern and sawtooth pattern guide shoes
(drawing courtesy of Weatherford International).

Chapter 11 Cementing Equipment and Casing Hardware 365

 As the demand for larger and heavier casings has Circulating at least one hole volume of mud is typi-
increased, so has the concern for derrick stress and cally required before performing the primary cement
fatigue. Float equipment can be used to reduce derrick job; however, to optimize hole and mud conditions for
stress by inducing flotation or increased casing buoyancy cementing, some drilling programs call for many hours
by simply not filling the casing or liner while running into of circulating (Chapter 5). Most float equipment valves
the wellbore. After landing, the casing is filled and cir- have been rated (after testing according to Recom-
culation is slowly established to begin hole conditioning. mended Practice for Performance Testing of Cementing
Flotation devices must be carefully deactivated, slowly Float Equipment [API RP 10F]) to withstand more than
releasing trapped air inside the casing; otherwise, casing 24 hr of solids-laden fluid flow at 10 bbl/min and to hold
collapse may occur. A specialist from the vendor supply- 5,000 psi of backpressure after circulation is complete.
ing the flotation device is usually present to monitor its After the cement is displaced, the float valve must
use. A typical casing flotation system is shown in prevent backflow into the casing. Should the float valve
Fig. 11-40. The operating sequence is shown in Fig. 11-41. fail, surface pressure must be applied until the cement
Owing to the risk of casing collapse during the flota- hardens to prevent U-tubing. Applying surface pressure
tion of larger strings, some operators and consulting is undesirable because it expands the casing while the
groups are studying casing drag more closely. In shal- cement hardens. When the pressure is released, the
lower, high-angle wellbores, casing drag and differential casing relaxes, potentially creating a microannulus
sticking frequently impede pipe placement to total between the casing and cement.
depth (TD). Roller centralizers are being proposed as a
safer method to reduce friction (Section 11-5.11).

Mud- Special Special

filled Flotation application application
casing collar float collar float shoe

Air-filled Semirigid
casing centralizer
Mud

Fig. 11-40. Typical casing flotation system (drawing courtesy of Davis-Lynch, Inc.)

366 Well Cementing

 Bottom
Opening plug
sleeve
Shear
pins

Seals

Circulating ports
Bottom sleeve

1. Bottom portion of casing is run dry 2. Casing pressure is increased until 3. Bottom cementing plug is launched
(not filled with fluid), with flotation collar the opening sleeve shifts down to ahead of cement. After landing on the
installed at desired depth. Casing above permit fluid and air to swap. After a bottom sleeve, it pushes both sleeves
the collar is filled with drilling fluid as fluid stabilization period, the casing ahead of the cement to the float collar
casing run continues to desired depth. is filled with drilling fluid. below.

Top plug

Casing
Casing
Bottom
plug
Bottom
plug

Float collar

Float collar

4. Bottom cementing plug and sleeves land and seal on 5. Top cementing plug seals and locks on bottom
the float collar. Bottom cementing plug ruptures, and cementing plug/collar assembly at the float collar.
cement is pumped through and out of the float equipment.

Fig. 11-41. Operating sequence of casing flotation system (drawing courtesy of Davis-Lynch, Inc.).

Chapter 11 Cementing Equipment and Casing Hardware 367

11-5.3 Valve materials and types
The three most common types of check valves used
in casing hardware are the ball, flapper, and poppet
valves. They have been adapted for floating and autofill
applications and are commercially available from sev-
eral suppliers. There are many specialized variations,
the latest of which is the large-bore double-flapper valve,
used for surge reduction in deepwater applications. The
large bores offer less resistance during the autofill
process.
API RP 10F provides standard methods for equipment
testing and a classification system based on perfor-
mance capabilities. On scales of I, II, and III and A, B,
and C, the best equipment rating is III-C.
Materials used in valves vary depending on require-
ments for drillability, compatibility with bits, and tem-
perature stability. The major valve components are gen-
erally made of phenolic resin, composites, or aluminum. Ball valves are still used Float collar sized for
Nitrile rubber is used for valve seals and coatings rated primarily in smaller float ball valve tubing.
for service at temperatures up to 300°F [149°C]. Beyond equipment, like this tubing
float shoe.
that temperature, viton rubber is used. Most large
springs are made from a phosphorus bronze, beryllium, Fig. 11-42. Tubing-size ball valves (drawings courtesy of
or aluminum alloy. Steel parts such as springs and pins Weatherford International).
are generally too small to pose a problem; nevertheless,
they are always tested to verify that they are drillable
with PDC bits. Plastic parts are typically made from als used and the construction of the valve. The use of ball
thermoset materials such as phenolic resins that are valves as the primary valve for cement placement is
stable to temperatures as high as 450°F [232°C] and waning. Poppet valves (described later) are more popu-
inert to most oils and synthetic mud components. They lar today because of their superior erosion resistance.
are easily drilled and are compatible with most bits. Flapper valves are composed of a spring-loaded flap-
Aluminum is easily drilled, but it has a tendency to foul per hinged to a plate with an integral seat (Fig. 11-43).
or smear over some diamond or fine-cutter bits when The flapper and plate are generally made of aluminum or
found in large masses unless the PDC bit is modified to composite materials. The spring-loaded seating action is
drill soft metals. virtually unaffected by hole deviation. The flapper opens
Ball valves consist of the typical ball-and-cage proportionally to the rate at which fluid is being pumped
arrangement (Fig. 11-42). The ball is generally made and the strength of the spring in the flapper. Because of
from weighted phenolic resin and may be rubber-coated. the irregular surfaces behind the flapper and spring,
The cage may be made of plastic or aluminum. The ball smaller valves tend to induce a large amount of turbu-
is buoyant in many drilling fluids and cement slurries; lence during flow. This increased turbulence can easily
however, when the hole is deviated, reverse flow may be cause erosion and seal damage when the flapper is not
required to return the ball to its seat. Ball check valves completely open, leading to a short valve life.
may not be effective in preventing a slight reverse flow A variety of flapper valves is available to fit within
or gas migration in highly deviated holes. most API threads such as 8-round or buttress as well as
The minimum clearance between the ball and the cemented versions of conventional shoes and collars.
cage may be one-half inch or smaller. Solids such as mud The new generation of composite large-bore flappers
scale or cuttings should be avoided to prevent wedging of (diameters greater than 3 in. [7.6 cm]) opens to provide
the ball or plugging the valve. Likewise, ball valves are a large, unobstructed flow path that passes tripping balls
not recommended for use in muds containing heavy con- and extraneous materials without blockage or harm to
centrations of lost circulation materials (LCM). The the valve. The temperature rating of composite flapper
backpressure resistance of the ball valves is more than valves is less than 300°F [149°C]. Aluminum valves are
adequate for most applications. The temperature stabil- available for service at higher temperatures.
ity and wear resistance depend greatly upon the materi-

368 Well Cementing

 Flow path during normal Backflow path blocked by
circulation, in which spring-loaded poppet (cone).
hydraulic force against
the poppet pushes it
down and compresses
the spring.
Aluminum nose flapper Flapper valve float collar with
float shoe with autofill autofill tube installed and Fig. 11-44. Poppet valve (drawing courtesy of Weatherford
tube installed and conversion ball shown above. International).
conversion ball shown
above.
the poppet should be avoided. The spring-loaded seating
action is virtually unaffected by hole deviation. The tem-
perature resistance depends on the type of materials
used and should be specified by the supplier. The pres-
sure resistance typically exceeds that of the flapper and
ball valves. Wear resistance is also generally good but
should be verified before using poppet valves in
Insert flapper equipment as used with extended-circulating-time applications.
8-round and buttress casing. The poppet spring is usually a nonferrous material
Fig. 11-43. Flapper valves (drawings courtesy of Weatherford such as phosphorus bronze or beryllium copper.
International). Generally, softer alloys are used for the spring materials
and are chosen according to the temperatures they will
experience. It is widely assumed that the springs in
Bump pressure ratings are based on the way the valve poppet valves are very strong, but this is far from true.
is cemented in place or the number of threads in the The purpose of the spring is to simply lift the poppet into
valve body. Backpressure resistance is highly dependent its seat. The differential pressure of the fluid trying to
on the valve body and is generally more than adequate for reenter the pipe holds the poppet in place.
most conditions. The wear resistance varies with size and There are two types of poppets on the market today.
manufacturer; it should be specified for circulating appli- The more common type, shown in Fig. 11-44, has the
cations longer than 8 hr. head on the top of the stem. Another version has the
Poppet valves are composed of a spring-loaded head or sealing area on the bottom of the stem. This gen-
poppet (also referred to as a plunger or cone) housed in erally results in a shorter piece of equipment for the
a cage much like that of a ball valve (Fig. 11-44). float collar, but may pose a problem on the float shoe (for
Phenolic resin or aluminum is used for the cage, and the clearance).
poppet is often made from nitrile rubber or is rubber- Table 11-3 is a summary of the advantages and disad-
coated. Like the ball valve, the poppet valve has a vantages associated with the various types of valves.
restricted flow path. Solid debris that may plug or bind

Chapter 11 Cementing Equipment and Casing Hardware 369

Table 11-3. Advantages and Disadvantages of Various Types of Valves
Types of Valves Flapper Ball Phenolic Poppet or Cone
Advantages Inexpensive Simple Self-closing
Least restrictive Easy to use Erosion-resistant
Self-closing Easy to drill out 0–90° inclination
Easy to drill out

Disadvantages High erosion susceptibility Dependent on flow to close Low tolerance to some LCM (e.g., fibers)
Easily blocked Poor performance in inclined wellbores
Difficult to drill out High erosion susceptibility
Easily damaged

11-5.4 Surge pressure and float equipment

Pressure surges are generated each time the casing is
raised and lowered; they are the product of inertia and
flow resistance of the displaced fluid. Pressure surges
combined with hydrostatic-pressure differentials may
exceed the casing-collapse or the formation-fracture
pressures, causing loss of mud or permanent formation
damage. As the trend continues towards reducing the
number of casing strings in a well, many operators are
running reduced-clearance combinations such as
■ 17 7⁄8 -in. to 18-in. casing through 20-in. hole

■ 16-in. to 17 7⁄8-in. casing through 18-in. hole

■ 13 3⁄8-in. to 13 5⁄8-in. casing through 16-in. hole

■ 113⁄4-in. to 11 7⁄8-in. through 13 3⁄8-in. to 135⁄8-in. hole

Closed Open Autofill Autofill with
■ 9 5⁄8-in. to 9 7⁄8-in. casing through 113⁄4-in. to 117⁄8-in. hole diverter
■ 7-in. to 7 5⁄8-in. casing through 9 5⁄8-in. to 9 7⁄8-in. hole. Fig. 11-45. Different types of surge models (drawing courtesy of
Pegasus Vertex).
External attachments such as centralizers and recip-
rocating scratchers may increase the flow resistance and
should be considered when determining a safe lowering
speed. Fig. 11-45 shows four ways to run casing or liners
into a well. Pipe ending conditions include
■ closed pipe

■ fully open pipe

■ pipe with autofill restriction v = v–

■ pipe with circulating sub (flow diverter). v=0

In all of these scenarios, the surge-pressure calcula- v = –vpipe

tions are based on many assumptions. Many drilling
fluids are compressible; therefore, surge pressures
should be reduced compared to conventional steady-flow
models. Unsteady flow conditions are reduced to a sta-
tionary flow regime model by considering an improved
mud clinging constant (Fig. 11-46). Clinging constants vpipe
were addressed in earlier works by Burkhardt (1961)
and Fontenot and Clark (1974) and are incorporated
Fig. 11-46. Graphic modeling of clinging characteristics of fluids on
into much of the mud displacement software used today. moving pipe (drawing courtesy of Pegasus Software).

370 Well Cementing

 Calculation of the pipe speed during the running velocity is 1.5 times the average velocity. Controlling the
process or reciprocation is based on creating an accept- lowering speed of 90-ft stands of drillpipe is extremely
able annular flow velocity, often equivalent to that used difficult without compensating for acceleration and
during drilling. Unfortunately such a velocity is not deceleration. As a safety precaution, if 1.5 ft/sec is
always possible, because the fluid viscosity varies with desired, one should request 1.0 ft/sec.
temperature and time. In openhole sections, the forma- The primary method of lowering the surge pressure
tion may react with the drilling fluid, resulting in while running casing or liner pipe is to allow fluids to
swelling (reducing the annular volume) or caving as enter the casing from the bottom. This lowers the
loose material falls into the annulus. volume of fluid passing up through the annulus. This led
In the absence of surge pressure calculation software, early float equipment developers to expand on the use of
Eq. 11-1 may be used. The equation is derived from the automatic fill float equipment many years ago. Today, it
Bingham plastic model and can be used to estimate a is a major component of most deepwater casing strings.
safe maximum lowering speed at a particular depth. The
effects of hole abnormalities and external attachments
are neglected. Turbulent flow is assumed, and a worst- 11-5.5 Automatic fill-up equipment
case friction factor of 0.016 should be used. Automatic fill-up shoes and collars contain check valves
similar to those used in float equipment (Fig. 11-47).
( ) ⎤⎥ ⎡ ⎤
0.5 However, the check valves are modified to remain in the
⎡
v pipe = ⎢25.6 Dρ ps
dhole − d pipe (
⎢ d
× ⎢ hole
)2 − 1⎥, open position to allow filling or even reverse circulating.
⎥
⎢ ⎥
( ) The casing fills continuously as it is run downhole, saving
f frict 2
⎣ ⎦ ⎢ d pipe ⎥
⎣ ⎦ rig time and reducing the pressure surges associated
with float equipment. The valves are usually designed to
(11-1) reduce casing overflow but, as the clearance between
where
dhole = hole diameter (in.)
dpipe = pipe diameter (in.)
D = depth (ft)
ffrict = mud friction factor (use of a worst-case 0.016 is
suggested)
gfrac = fracture gradient
ps = the lesser of psf or psc, where
psf = 0.5 L ( gfrac – 0.052ρ), for formation
protection
psc = 0.5 ( pmin – 0.052ρ), for casing protection
pmin = minimum casing collapse resistance (psi)
vpipe = maximum casing velocity to prevent damage to
casing or formation
ρ = fluid density (lbm/gal).
Some apply general safety guidelines such as limiting
the pipe speed to 2 ft/sec [0.6 m/s]. Such rules are not
practical. For example, running 13 3⁄8-in. casing inside a
16-in. casing and a 171⁄2 in. open hole certainly would not
be the same as running a 5-in. liner inside a 6-in. open
hole. Another concern with pipe velocity analysis is
making sure the drill crew is measuring the maximum
velocity. As a rule, most programs assume the maximum

Running position Flapper valve closed

Fig. 11-47. Large-bore double-flapper type autofill float equipment

(drawings courtesy of Davis-Lynch, Inc.).

Chapter 11 Cementing Equipment and Casing Hardware 371

casings becomes smaller, this becomes challenging.
Stopping overflow while reducing surge pressure
requires one to reduce the pipe running speed. Running
the casing quickly while reducing the surge pressure
requires one to eliminate restrictions on the float equip-
ment and contend with mud returns on the rig floor or
downhole. This can be accomplished with a diverter tool
that is placed directly above the restrictive section, usu-
ally the first joint of drillpipe above a liner or casing
hanger.
Autofill equipment must be activated, or converted,
to begin functioning as a one-direction check or float
valve. Conversion is generally performed after the casing
is in place; however, it can also occur while running Fig. 11-48. Orifice fill flapper valve and conversion ball with autofill
casing to prevent overflow, control the hook load, or to tube in place (drawing courtesy of Weatherford International).
control the well. Sudden stops while lowering the casing
should be avoided to prevent premature valve conver-
sion. The maximum flow rate of fluids down the casing rate is estimated at 200 ft/min [61 m/min]. The ball may
may also be limited to prevent the valve from being unin- be pumped down; however, should it set while pumping,
tentionally activated. the conversion may occur without an indication that is
Small-orifice and poppet valves are not recommended easily recognized. With the ball properly seated, the ori-
for use with drilling fluids containing high concentra- fice tube may be discharged by applying 300 to 800 psi
tions of LCM, or large-particle-size LCM. Using a large pressure, depending on the manufacturer.
number of reciprocating scratchers and other external
attachments may increase the annular flow resistance
and cause overflowing. Large-bore autofill float equipment
A new family of large-bore autofill float equipment was
developed to meet the need for increased surge reduc-
Casing insert equipment tion. Most of the newer types of large-bore autofill equip-
Insert equipment offers an economical means of provid- ment use a double-valve design flapper held open by a
ing floating and autofill valves for shallow-to-moderate single 24- to 36-in. [0.6- to 0.9-m] long aluminum or com-
depth and pressure applications. Insert equipment is posite tube. Composite tubes are preferred because they
generally available for 103⁄4-in. and smaller casings, sup- are easier to drill out with PDC bits. The openings are
plied with API 8-round long or short couplings. They are typically 3.5 in. [8.9 cm] across the ID of the tube and
made of aluminum and composite materials and contain have a 3.25-in. seat for the ball (Fig. 11-49). As a result,
ball, flapper, or latch-in valve mechanisms. the path of least resistance for any fluid is up the casing
Insert valves are installed within a coupling and are string, offering the maximum surge reduction.
trapped between the above and below joints of casing. There are two operating modes in autofill flapper
Their strength is limited to the material trapped with valve equipment—filling and nonfilling. In the filling
the coupling, which is usually aluminum. The insert mode, it is possible to circulate the fluid through the
valve must be fully made up to prevent damage and casing before landing without losing the autofill capabil-
interference with proper casing engagement. When ity. This capability is required to circulate past an
landing a plug, the rate should be reduced to prevent obstruction or a sand bridge. The nonfilling mode can
pressure spikes that may exceed the strength of the be used when one is absolutely sure that autofilling is
insert. not required or that filling can be performed from the
Autofill or orifice fill (flapper) valves are converted surface.
by ejecting the orifice tube, allowing the spring-loaded This next option is dictated by the operator’s desire
flapper to close (Fig. 11-48). for fast response to well control issues or hole deviation.
The conversion operation usually requires the use of If the conversion method requires dropping a ball, it can
a weighted phenolic ball or, in some cases, a small zinc be dropped from the surface for near-vertical wells, usu-
or brass tripping ball. To save time in vertical wells, the ally up to 30° deviation. If the ball is not pumped, one
ball is generally dropped into the casing and allowed to should allow 3–5 min per 1,000 ft for the ball to reach its
free fall while running the last 5 to 10 joints. The free-fall target.

372 Well Cementing

 A B C D

Fig. 11-49. Operational sequence of a typical large-bore autofill collar (drawings courtesy of Weatherford International).
A: Run-in position
B: Circulation position after ball is dropped into place
C: Circulation pressure increased until the pressure drop across the small ports at the bottom of the inner sleeve
exceeds the shear pin ratings holding the sleeve in place
D: Converted position, now standard double-flapper float collar

The other option is to have the ball trapped inside the

autofill valve in a cage mechanism above the valve,
where it will remain until downward circulation begins.
Circulation flow will force the ball into the seat, build up
backpressure, and deactivate the autofill mechanism.
The limiting factor is that the ball can restrict the flow
of autofilling fluids and the solids carried therein.
If the 31⁄2-in. ball cannot pass through the drillpipe,
liner equipment, or subsea plugs caused by ID restric-
tions, then the ball can be inserted inside the casing
before the hanger or running tools are threaded
together. This is commonly called “floating the ball.”
Another option is to locate the ball inside a launching
device (Fig. 11-50) below the restrictions (e.g, a subsea
or liner plug).

Fig. 11-50. Mechanism that allows a ball to be launched from

the end of a subsea or liner plug set (photograph courtesy of
Weatherford International).

Chapter 11 Cementing Equipment and Casing Hardware 373

Other autofill float equipment valves 11-5.6 Differential fill equipment
The poppet autofill valve (Fig. 11-51) contains a spring- Differential fill shoes and collars combine the benefits of
loaded plunger that is held open to allow filling. There floating and autofill equipment (Fig. 11-52). They are
are several methods available for holding the plunger designed to automatically fill and regulate the fluid level
open, and most valves require a preset downward flow within the casing. Most differential fill units (shoe or
rate for release. They include the older ball and wedge collar) will keep the casing approximately 90% full with
design mechanisms. The minimum flow rate is generally respect to the annular fluid level. Therefore, in theory,
between 4 and 8 bbl/min [0.6 and 1.3 m3/min]. However, the casing should remain about 81% full when both a
some devices are designed to release at a predetermined shoe and collar are used. As with all equipment, pipe
hydrostatic pressure. Poppet collars are usually speed, fluid viscosity, gel strengths, and other factors
designed to retain the tripping mechanism, so two tend to reduce this accuracy, and the utmost care should
poppet units (shoe and collar) may be used to provide be taken to verify the fluid height inside the casing to
added seal insurance. While the autofill valve can be run prevent pipe collapse.
with a guide shoe, it is important to not combine autofill Differential fill and autofill equipment is often used
float valves with standard poppet valves. on long strings to reduce surge pressures and the possi-
New autofill poppet valve designs use a spinning bility of formation damage, which is normally associated
impeller mechanism to release the poppet stem. After a with float equipment (Fig. 11-53). This equipment also
certain amount of pumped fluid at a given rate is circu- saves time, which reduces the probability of sticking.
lated past the impeller, the impeller threads itself off a The fluid-level regulating feature reduces the hook load
brass stem to convert the valve to a standard poppet and, at one time, was thought to prevent overflow pro-
valve. vided the annulus is not restricted. Experience has
There are also hydrostatically sensitive mechanisms shown that this is not always true. Circulation may be
that allow the poppet valve stem to be released after established without harming the valve as long as the
reaching a predetermined hydrostatic pressure. These valve does not have a trapped conversion ball. The valve
devices are accurate to ±100 psi. A thorough knowledge will resume operation when the fluids in the casing and
of the true vertical depth is required at all stages when annulus reach the designed level.
using this valve.

Impeller type Hydrostatic pressure type

Running Converted
Fig. 11-51. Poppet autofill valves (drawings courtesy of
Weatherford International). Fig. 11-52. Differential fill float collar (drawing courtesy of
Davis-Lynch, Inc.).

374 Well Cementing

 verify that the drop ball to convert the collar is compati-
ble with the guide shoe.
The following are some additional precautions.
■ To reduce float valve wear during long circulating and
conditioning periods, the tripping operation may be
delayed until just before pumping cement.
■ Because of restrictions in the fill-up path, the casing
should be lowered at a moderate rate (generally 1
ft/sec) to reduce pressure surging. The availability of
surge prediction software has made this a much safer
operation today, and this software should be used
before running casing. For optimal protection, it is
best to enter the mud properties measured during the
wiper trip.
Run-in Autofill Ball in Converted ■ LCM may tend to slow or prevent filling, which may
position position place position
increase surging or lead to collapse. Periodic circu-
lation and monitoring the weight indicator may be
Fig. 11-53. Differential fill valve (drawings courtesy of Weatherford necessary.
International).
■ Hole deviation and casing size may prohibit the use
of a weighted tripping ball. Some manufacturers
offer ball guides for deviations that exceed 20 to 30°.
The typical differential valve regulates fill-up through
Others trap or preload the ball, which should allow
the action of a floating differential piston. The piston
use at any deviation; however, circulation is also pre-
slides up to open and down to close and is designed so
vented before tripping. The supplier should provide
that the upper pressurized surface area is approximately
information about the maximum allowable deviation.
10% larger than the lower. The forces acting to shift the
piston are produced by hydrostatic pressures acting on
the upper and lower surfaces. Because the upper area is
larger, less pressure above is required to balance the 11-5.7 Inner string cementing equipment
forces across the piston. When the pressure above Inner-string cementing is a technique typically used
(casing hydrostatic) exceeds 90% of the pressure below with large-diameter casings in which drillpipe is placed
(annular hydrostatic), the piston will slide down to halt inside the casing as the conduit for pumping fluids from
the filling. Likewise, when the pressure below exceeds the surface to the casing annulus (Fig. 11-54). The inner
90% of the pressure above, filling will resume by moving strings can be run in one of three ways.
the sleeve upwards. This cycle is continuously repeated
as the casing is lowered. However, cycling may not begin
until the hydrostatic pressure is sufficient to overcome Stab-in
frictional forces associated with the O-rings. With the stab-in method, a simple stinger is fitted
The inoperative flapper valves may be converted to to the bottom of the drillpipe and then lowered into
function as float valves at any time. Converting most the receptacle (usually plastic) where a set of O-rings or
valves requires a tripping ball (usually in the 1.5 to special seals are engaged (Fig. 11-55). Once the casing
2.0 in. size) operation like that described for orifice fill reaches the desired depth, the stab-in stinger and cen-
equipment. Circulating before dropping the ball may tralizer are connected to the drillpipe and run into the
help clear the ball seat of debris. To verify proper trip- casing. The drillpipe is lowered until the stinger engages
ping pressure, the ball should be allowed to fall to the the receptacle. For stingers without a latch, additional
seat before pumping. The pressure required to trip most weight must be applied to counteract the lifting force
valves is generally between 500 and 800 psi [3.4 and 5.5 (backpressure) created while cementing. The maximum
MPa]. Because only the ball is discharged, a shoe and lifting force may be estimated by multiplying the maxi-
collar may be used, and both may be tripped with a mum expected pumping pressure by the end area of
single ball. A differential fill or guide shoe may also be the stinger, including the seals. A simple guideline is
used below a differential collar, provided the tripping to apply 10,000 to 20,000 lbf [44.5 to 89.0 kN] to pro-
ball is compatible with both units. One should always vide adequate pipe weight. One should also check for

Chapter 11 Cementing Equipment and Casing Hardware 375

 Stop collar
Drill pipe
centralizer

Casing

Fig. 11-55. Stab-in stinger (drawing courtesy of Weatherford

Tag-in adapter International).

for 95⁄8-in. and larger casings, but are most often used in
Tag-in float collar 18-in. and larger sizes. In some floating-rig offshore loca-
tions, the drillpipe stinger is actually suspended above a
standard float collar or shoe, and the cement is pumped
into the float equipment to compensate for wave motion,
because it represents the path of least resistance.
Stab-in receptacles on the float collar are often fitted
with a latch-in plug receptacle, which differs from a
latch-in stinger. Using drillpipe gives the operator a pos-
itive indication that the displacement process is com-
plete because there is direct observation from the sur-
face. It also leaves behind one more seal to prevent any
Down-jet float shoe U-tubing of the cement back inside the casing.

Screw-in
The screw-in stinger (Fig. 11-56) is also mounted on the
end of the drillpipe, but the stinger has a set of coarse
Fig. 11-54. Inner string cementing (drawing courtesy of threads located above the seal assembly. The threads are
Davis-Lynch, Inc.). mated to an aluminum or cast-iron receptacle on top of
the float collar and can be released with about 11 turns
to the left. This feature is sometimes used to run the
fluid coming up the annulus. If it is not sealed, then add casing like a scab liner (Chapter 14), but this operation
an additional 10,000 to 20,000 lbf [44.5 to 89.0 kN]. is dangerous because the casing is in compression when
Achieving this weight may require the use of drill collars picked up from the bottom by the drillpipe, thus reduc-
or heavy-weight drillpipe, so this weight availability ing the collapse pressure. Most float equipment of this
should be considered ahead of time. type is rated to about 100,000 lbf [445 kN] of pickup
The shoes and collars are basically larger versions of force, but the rating must be verified by the vendor. This
the types previously discussed, with the addition of a version of float equipment is seldom made to be PDC-
seal receptacle and beveled surface. They are available drillable.

376 Well Cementing

Fig. 11-56. Screw-in stinger (drawing courtesy of
Davis-Lynch, Inc.).

Latch-in Fig. 11-57. Latch-in stinger (drawing courtesy of Davis-Lynch, Inc.).

The latch-in system is similar to the screw-in stinger, but
the stinger is fitted with a split threaded ring that allows
the stinger to be stabbed in (Fig. 11-57). The stinger ■ Is more environmentally friendly because it reduces
must be rotated out or unscrewed to be released. Today, the amount of built-up materials around the wellhead
very few operators use the latch-in stinger because of before abandonment
potential release problems in the field.
Casing running operations are designed according to
the type of valve used. The lowering speed should be 11-5.8 Shoe and collar shell design
slow enough to prevent surging. Float equipment may The shell is the outer steel portion of a cementing shoe
require more frequent filling to prevent collapse. and collar (Fig. 11-58). It becomes an integral part of the
The inner string isolates the casing interior from the casing and must be capable of meeting chemical, dimen-
pumping and hydrostatic pressures created while sional, and strength requirements. Aside from an API
cementing. Care should be taken to prevent creating threading standard, Specification for Threading,
pressure differentials that may exceed the collapse Gauging, and Thread Inspection of Casing, Tubing, and
resistance of the casing. To help prevent collapse, pres- Line Pipe Threads (API Spec 5B), there are no specific
sure may be applied to the inside of the casing with the industry standards covering shoe or collar shells.
use of a packoff head. Most float shoe and collar shells are made from collar
The principal advantages of using stab-in equipment stock, which is made from the same alloy composition as
are listed below. casing. The OD of this equipment generally conforms to
■ Greatly reduces displacement volumes and time API collar OD standards.
■ Wastes less overdisplaced cement when cementing Shoe and collar cases are usually made of low-alloy or
to surface carbon steels such as API K55 or N80 and are suitable
for H2S applications. To prevent interference and allow
■ Decreases the need for extended-working-time
use over a broad range of casing weights, shoe and
cements collar internal and external diameters generally con-
■ Contributes less contamination because of reduced form to the lightest-weight API casing drift and cou-
area and turbulent velocities in the drillpipe pling external diameters, respectively. This also pro-

Chapter 11 Cementing Equipment and Casing Hardware 377

 Some manufacturers offer a heavyweight range to more
closely match internal diameters and to improve exter-
nal thread strengths. Table 11-4 shows the pressure rat-
ings of various float collars.
General-purpose equipment is most readily available
with API 8-round and buttress connections. To allow
makeup to either API long or short connections, 8-round
equipment is usually provided with API long internal
threads and short external threads where applicable.
External thread joint strengths may be as low as that of
lightweight K55 casing but are generally adequate for
bending and tensile stresses at the lower end of the
casing. Some manufacturers offer double-box (internal
thread) collars, which eliminate the weak part, the
external thread. Premium threads are not commonly
stocked and may require an extended delivery time.
Typical float equipment Groove pattern
shell with internal grooves float equipment
placed above the valve shell 11-5.9 Cementing plugs
Cementing plugs are semirigid barriers used to separate
Fig. 11-58. Float equipment shells (drawings courtesy of
Weatherford International). cement slurry from drilling fluids, to wipe the casing,
and to indicate when cement placement is complete.
Plugs were once made of gunnysacks, wood, and leather.
vides a heavier than usual wall, which allows use of low- Present designs include top and bottom plugs con-
strength steels, such as K55, to meet N80 burst and col- structed of nitrile or polyurethane molded over PDC-
lapse pressures. Because of the lack of pressure differ- drillable high-density plastic cores (Fig. 11-59). Most
ential across the shoe, general-purpose shoes are plugs are designed to be nonrotating; as a result, they
considered suitable for use with high-strength casings. are easier to drill out (Fig. 11-60).

Table 11-4. Nonrotating Standard Float Equipment Ratings†, ‡

Float Collar Model 1 Model 2 Model 3 Three Nonrotating Float Collar
Size (in.) Pressure Ratings
Plug Bump Back Plug Bump Back Plug Bump Back
Pressure Pressure Pressure Pressure Pressure Pressure
Rating Rating Rating Rating Rating Rating
(psi) (psi) (psi) (psi) (psi) (psi)
7 5,000 5,000 na§ na 8,000 5,000

75⁄8 5,000 5,000 na na 8,000 5,000

85⁄8 5,000 5,000 na na 8,000 5,000

95⁄8 5,000 5,000 na na 5,000/8,000 5,000 Model 1 Model 2

103⁄4 4,000 4,000 na na 6,000 5,000

113⁄4 3,300 3,600 na na 5,000 5,000

133⁄8 3,200 3,600 na na 5,000 5,000

16 na na 2,000 2,000 3,000 2,000

185⁄8 na na 1,500 1,700 3,000 1,700

20 na na 1,500 1,700 3,000 1,700 Model 3

† To meet API RP 10F - Classification IIIC requirements.
‡ Courtesy of Weatherford International
§ Not available

378 Well Cementing

Fig. 11-59. Cementing plugs are available in all shapes, sizes, and
materials (photograph courtesy of Weatherford International, Inc.).

Fig. 11-61. Typical top and bottom nonrotating cement plugs. Color
differences help differentiate the top from the bottom plug. The core
materials for both plugs are the same (drawing courtesy of
Weatherford International, Inc.).

Fig. 11-60. Plugs with nonrotation features (drawings courtesy of

Davis-Lynch, Inc.). preflushes clean before they enter the annulus. A second
use of the third plug is to measure displacement effi-
ciency of the mud pumps. Because of casing-ID vari-
Although similar in external appearance, top and ances and pump-liner inefficiencies, the displacement
bottom plugs differ considerably in internal design and volume sometimes varies from job to job. With a three-
operation (Fig. 11-61). Bottom plugs were developed to plug system, the exact number of pump strokes neces-
precede the cement slurry, requiring an internal bypass sary to bump the plug can be measured before pumping
or flow-through feature. The bottom plug features a thin the cement slurry. This allows operators to pressure test
membrane that is designed to rupture and permit flow the casing while the cement slurry is still fluid without
once the bottom plug has seated (usually on the float damaging the cement-to-pipe bond.
collar). Bottom plugs also provide a seat for landing top Top plugs are occasionally used alone. They are
plugs and sealing off displacement. To ensure compati- designed to withstand the pressures and forces gener-
bility, top and bottom plugs and float equipment should ated when landed abruptly. When both top and bottom
be from a common manufacturer. The use of bottom plugs are used, it is vital that they not be launched out of
plugs with high LCM concentrations in the slurry can be sequence. Because of the exterior similarity, top and
risky, because the LCM may tend to ball up ahead of the bottom plugs are generally color coded. The conse-
wiper plug and bridge the float valve. quence of pumping the top plug first is that it will land
More recently, using a third wiper plug has become and not rupture, leaving the casing full of cement. If this
more common. The extra plug separates chemical happens, the only recourse is to drill out the casing.
washes or spacers from the drilling mud, keeping the

Chapter 11 Cementing Equipment and Casing Hardware 379

 Other common plugs include the following:
■ tapered plugs, used in multiple-ID strings
■ subsea plugs, used with subsea completions
■ latch-in plugs, used with latch-in equipment
■ flexible-fin plugs, used when passing through stage
equipment.
Figure 11-62 shows one of the newest large-bore
subsea wiper-plug-set designs. The new sets come with
built-in swivel equalizers and are released with darts
instead of balls. The reason for dart release is to provide
wiping of the drillpipe strings on deep casings or liners
and to avoid having to wait for the gravity settling of
balls. The operational sequence is illustrated in
Fig. 11-63.

Fig. 11-62. Large-bore subsea wiper plug system, with top and
bottom plugs and darts (courtesy of Weatherford International, Inc.).

1 2 3 4 5 6
Fig. 11-63. Operational sequence on typical top and bottom subsea release wiper plug system. Note that the nonro-
tational profile of the plug perfectly matches up with the nonrotating plate on top of the float collar (courtesy of
Weatherford International, Inc.).

380 Well Cementing

11-5.10 Stage equipment
Stage equipment, consisting of stage collars and port
collars, is placed within the casing string to provide an
intermediate passage to the annulus (Fig. 11-64). The
collars are generally made of match-grade steel. They
may be available in special weight ranges to optimize
strength and internal dimensions. Stage equipment is
generally used to protect weak formations from exces-
sive hydrostatic pressure, to cement widely separated
zones, and to reduce mud contamination.

Fig. 11-65. Diagram of mechanical stage tool (drawings courtesy

of Weatherford International).

pumped and the ports are closed again by seating and

applying pressure to the larger closing plug. Once closed,
the stage collar cannot be reopened. The pressure
required to open and close varies with manufacturers,
but it is generally between 800 and 1,400 psi [5.5 to
9.7 MPa]. When two stage collars are used, a special
Fig. 11-64. Diagram of typical stage tool placement just inside the
cased hole (drawing courtesy of Weatherford International).
upper stage collar is required, and care should be taken
to release the correct plugs in the proper sequence. The
ID of the upper stage collar seats must be larger than the
lower collar seats. For highly deviated holes, the free-fall
Stage collars are available in both mechanical and
dart should be replaced with a pumpdown plug. Drilling
hydraulic versions. However, even the mechanical
is required to remove the plugs and aluminum seats.
method of opening and closing requires hydraulic force.
Figure 11-66 shows the mechanical stage tool in the
Mechanical tools are opened and closed using free-
run-in position, the opened position, and the closed posi-
fall plugs or pumpdown-closing plugs to select and shift
tion. Figure 11-67 shows the plugs and opening devices
the appropriate internal sleeve(s) (Fig. 11-65). The
used to operate the stage tool. As a rule, all stage tools
lower sleeve initially covers the ports. Once the first
should be opened as soon as the primary job is com-
stage is complete, the lower sleeve is pumped down to
pleted, because any cement slurry above the tool can be
uncover the ports by seating the free-fall (or pumpdown)
circulated out.
opening plug and applying pressure. The second stage is

Chapter 11 Cementing Equipment and Casing Hardware 381

 Closing
plug

Closing sleeve

Opening sleeve Fluid ports Double

seals
Brass shear
mechanism
Antirotation Broken shear Lock
feature mechanism ring

Body connection Free-fall

with metal-to- opening device
metal seal and
elastomer
backup

Running position Opened position Closed position

Pin and box threads are identical to Opening device has landed and, after The closing plug has landed and, after
the casing threads. Stage collar integral pressure is applied, the lower set of shear pressure is applied, the upper set of shear
connection is designed for gas tightness. mechanisms is broken and the sleeve mechanisms is broken and the sleeve
Seals on opening sleeve provide internal shifts downward to uncover the fluid shifts downward, shutting off the fluid
and external pressure integrity across ports. Pumping operations can now ports. Double seals above and below the
the fluid ports. be conducted through the stage collar. ports provide pressure integrity.

Fig. 11-66. Mechanical stage tool shown in run-in, open, and closed positions (drawings courtesy of
Davis-Lynch, Inc.).

382 Well Cementing

 Two-stage cementing Two-stage cementing using a bypass
using a first-stage sealing plug, shutoff plug and baffle, free-fall
plug, free-fall opening opening device, and closing plug
device, and closing plug
1. A float shoe, float collar, and stage
1. A float shoe, float collar, and collar are installed in the casing string
stage collar are installed in and the casing is run to bottom.
the casing string and the The shutoff baffle is installed in the
casing is run to bottom. casing string at least one joint
above the float collar. If API threads
2. Circulation is established and are run (8-rd or buttress) the baffle
first-stage cement is mixed and can be installed in the “J” section of
Closing plug pumped. Closing plug a coupling. If premium threads are run,
a separate baffle collar must be run.
3. The first-stage sealing plug is
launched and the cement is displaced. 2. After the hole is conditioned, the
At the conclusion of displacement, bypass plug with the nose piece
the first-stage sealing plug lands is launched ahead of first-stage
and effects a seal against the float cement. This plug will pass through
collar. No baffle is required. the shutoff baffle and land on any
manual- or self-fill float collar. Once
4. The free-fall opening device is landed, approximately 50 psi will
dropped and allowed to gravitate invert the wipers on the bypass plug
to position. Pressure is applied to and allow cement to pass.
Free-fall the casing and the stage collar is Free-fall opening device
opening device opened. 3. After cement is mixed and pumped,
the shutoff plug is launched and
5. Circulation is established and cement is displaced. At the conclusion
second-stage cement is mixed of displacement, the shutoff plug lands
and pumped. and effects a seal in the shutoff baffle.

6. The closing plug is launched and 4. The opening of the stage collar and
cement is displaced. At the conclusion the ensuing second-stage cementing
of displacement, the closing plug and closing of the stage collar are
lands and effects a seal in the stage carried out identically to that described
collar. Pressure is applied to the casing for two-stage cementing with first-stage
and the stage collar is closed. sealing plug.

First stage shut-off plug

Shut-off baffle
First stage
sealing plug

Bypass plug

Fig. 11-67. Plugs and opening devices used to operate stage tools (drawings courtesy of Davis-Lynch, Inc.).

Chapter 11 Cementing Equipment and Casing Hardware 383

 One of the most common causes of stage tool failure
is the inability to close the tool. The correct pressure to
close most tools is related to the lift pressure of the
second-stage cement column plus the shear strength of
the pins holding the sleeve open. For example, suppose
an operator is displacing second-stage cement before
the closing plug bumps at 600 psi [4.1 MPa]. If the tool
is set to close at 1,200 psi [8.3 MPa], then the pump pres-
sure should be raised to at least 1,800 psi (1,200 + 600 =
1,800 psi [12.4 MPa]).
Hydraulic stage tools do not require a free-fall plug
or pumpdown plug to open the tool. The tool opens when
the shutoff plug bumps against either a float collar or
landing collar. Then, the pressure can be increased
slowly until it reaches a preset shear-pin rating in the
stage-tool opening sleeve. The hydraulic stage tool has
an internal sleeve much like the differential sleeve,
meaning it has a larger area inside the tool on the upper
end than the lower end (visible in Fig. 11-68 in orange
color). This differential area, combined with the increas-
ing pressure, keeps the tool closed until the internal
area of the tool is pressurized, usually around 700
to 1,000 psi [4.8 to 6.9 MPa]. During the cementing
process, the cement slurry takes the path of least resis-
tance, following the mud out of the stage tool ports and
up the annulus. The operating sequence is shown in
Fig. 11-69.
There are optional methods for opening a hydraulic Fig. 11-68. Hydraulic stage tool with accessories (drawings
tool with a free-fall plug, and there are special plugs that courtesy of Weatherford).
can be used to open or close the tool if required. These
options are available because some operators run stage
tools only as a contingency, and if they do not want to use
them, they must be able to close them after the job.

384 Well Cementing

 1 2 3

4 5 6

Fig. 11-69. Hydraulic stage tool operational sequence (drawings courtesy of Weatherford International).

Chapter 11 Cementing Equipment and Casing Hardware 385

Liner stage tool options
When running stage tools or annular casing packers
(ACPs) on liners, it is difficult to reliably open and
inflate the tool with standard liner wiper plugs. Figure
11-70 shows a special plug set that works with most liner
systems in conjunction with hydraulic stage tools and
ACPs. This system is sometimes referred to as a four-plug
system, but it is actually two darts and two plugs. The
lower dart launches the lower plug and displaces the pri-
mary cement, landing on a special plate. Then pressure
builds up, inflating and locking the ACP. Further pres-
sure build-up opens the hydraulic stage tool.

Fig. 11-70. Liner plug set built specifically for hydraulic stage tools
and ACPs (courtesy of Weatherford International, Inc.).

Fig. 11-71. Diagram of port collar with running tool (drawing

courtesy of Weatherford International).
Port collars are another type of stage tool, mechani-
cally operated from the surface by a tool connected to an
inner string of drillpipe (Fig. 11-71). They are available
with sliding or rotational valve mechanisms and may be ment should not be placed in the midbody of a stage
opened or closed as often as necessary. Lock-down collar but rather on the coupling ends, where there are
options are also available. The sliding valves are gener- no thin-walled sleeves.
ally opened with an upward motion and closed with a Centralizers and baskets, or annular casing packers
downward motion; they require a minimum of 10,000 lbf (ACPs), are often used with stage equipment. Cement
[44 kN] to stroke. Port collars may be placed as often as baskets or ACPs are placed below the stage collar to help
necessary in the casing string and selected in any support the hydrostatic pressure of the next stage and to
sequence. There are no plugs to use or drillout required. prevent cement from falling through lower-density fluids
Some shifting tools may be fitted with cup-type seals to below.
form a conduit from the inner string to the ports.
Stage collars and port collars must be handled with
care because of the close tolerances between the outer
body and the internal sleeves. Tongs or makeup equip-

386 Well Cementing

11-5.11 Annular packoff equipment
ACPs (also known as external casing packers) are popu-
lar because they provide improved sealing and central-
ization. They are used to prevent gas migration, reduce
the gas-to-oil production ratio, reduce or prevent water
production, isolate production or injection zones, and
avoid squeeze cementing.
ACPs vary in size and valve type depending their pur-
pose. Most can be filled with mud or cement, but the
amount of pressure they can withstand is limited.
Overfilling is a major cause of casing packer failures. Two
types of ACPs are shown in Fig. 11-72.
Annular packoff equipment (packer shoes, collars,
and ACPs) is used to protect areas of the formation from
excessive hydrostatic pressure, contaminating fluids, or
both. The equipment has expanding rubber elements
that pack off against the formation to create an imper-
meable annular barrier. The rubber elements also cen-
tralize the casing when expanded.
ACPs are generally used below stage collars or port col-
lars to protect the formation below from excessive hydro-
static pressure or contamination (Fig. 11-73). They are
also used in an attempt to block gas and fluid migration.
ACPs can be installed on either side of a weak formation.

Continuous Rib ACP Discontinuous Rib ACP

Fig. 11-72. ACP (drawings courtesy of Weatherford International,

Inc.).

Chapter 11 Cementing Equipment and Casing Hardware 387

 Inflate-limit
Drillable valve
closing seat
Double-lock rings Inflate passage
Closing sleeve
Shear pin
Free-fall
Protected seals opening device
Shear pin
Opening sleeve
Drillable split-type Optional
opening seat pumpdown
opening plug
Shear pin
Antirotation lug

Steel-reinforced
inflatable packer

Pumpdown Opening Plug

A pumpdown opening plug can
be used in lieu of the free-fall
Ratchet-type opening device. It is applied in
bottom sub horizontal wells and directional
wells where deviation at the
packer stage collar exceeds 30º.
Customer-specified
casing mandrel

Running in Hole Inflating Element

Shows packer stage cementing collar in running The free-fall opening device enters split-type opening seat,
position with opening and closing sleeves pinned shearing the pins in the lower section. This allows lower section
in place. Lower section of split-type opening seat to move down exposing the inflatable packer element to the
isolates inflate passage, preventing premature fluid and pressure inside the casing. Fluid enters the packer
inflation of the packer. element through the double-seal in the free-fall opening device
and the split-type opening seat and inflation passage in the
tool body.

Fig. 11-73. Typical stage cementing collar (drawings courtesy of Davis-Lynch, Inc.).

388 Well Cementing

 ACPs can be packed off by either inflating or com- Basket shoes use a basket instead of an expanding
pressing the rubber element (Fig. 11-74). The inflatable rubber element to pack off the annulus. They provide the
type is generally larger and more capable of packing off same function as a packoff shoe but are limited to low
oversized or irregular holes. The inflation process gener- differential pressure applications and do not create a
ally begins at a predetermined setting pressure. The set- true hydraulic seal.
ting pressure should be sufficiently high to prevent pre-
mature packoff while conditioning or cementing.
An optional break-off rod may be used to prevent pre- 11-5.12 Cement displacement enhancement tools
mature setting by blocking the inflation port until External casing attachments increase the chances of
broken free by a wiper plug. Once the element is achieving a successful primary cement job. The most
inflated, an internal valve mechanism will hold the common devices are centralizers, baskets, and scratch-
inflated position, and the surface pressure may be ers, described in this section.
released. When possible, the element should be inflated Stop collars are used to hold or limit the travel of the
with cement. external attachments. (Fig. 11-75). They should be used
Packer shoes and collars are hydraulically set by a instead of welding the attachments, which may be harm-
tripping ball (Fig. 11-74). With the casing in place, the ful to the casing. Stop collars are available in solid or
tripping ball is dropped and allowed to fall to a seat on hinged designs and may be fastened with bolts, set
the piston. As pressure is applied, the load is transferred screws, or hammer-locking mechanisms. Stop rings for
to an external sleeve that compresses and expands the high-strength casing applications should contain hard-
element. At a pressure of approximately 800 psi ened inserts to better grip the casing. Care should be
[5.5 MPa], the piston shears free, uncovering the ports. taken to prevent dropping loose articles into the hole,
The external sleeve contains a ratchet mechanism that and all tools should be installed on the pipe rack before
permanently holds the set position. The cement is the pipe is shipped to the location.
pumped and, unlike stage equipment, the ports are not On critical wells, time should be taken to make sure
closed. the stop collar matches the well conditions. It is critical
in most cases that the centralizers be supported and
pulled into the wellbore. Typical stop collars such as the

13⁄4 in.

Set-screw stop collar, Friction clamp stop collar,

hinged hinged

11⁄4 in.

Spiral nail stop collar, Automatic slip-on

hinged stop collar

Fig. 11-75. Stop collars (photographs courtesy of Weatherford

International).

Fig. 11-74. Typical packoff float shoe (drawing courtesy of

Weatherford International).

Chapter 11 Cementing Equipment and Casing Hardware 389

older friction clamps (nut and bolt types) have a grip- including casing size, casing weight, casing seat, hole
ping force of slightly more than 10,000 lbf [44 kN], size, and mud weight, and, when deviation is present,
whereas the set-screw stops have gripping forces up to full survey data, including kickoff point, rate of build,
35,000 lbf [156 kN] with proper installation. The grip- and final deviation. Because centralization is most criti-
ping force of insert-type stop collars can exceed cal through the cemented interval, the anticipated top
60,000 lbf [267 kN] and are easier to install than most of cement must also be provided. With these data, the
set-screw stop collars. Of course, the price is directly computer can run centralizer spacing and standoff pro-
proportional to the gripping force. Purchasing high-qual- grams in two modes. The first and more effective mode
ity centralizers and then installing them with inferior is variable spacing, in which the relevant well data are
stop collars is a waste of money. entered and the computer calculates both the number of
Casing centralizers (Fig. 11-76) are one of the sim- centralizers to run and how to space them to achieve the
plest yet most beneficial mechanical devices used in pri- desired standoff along the cemented interval. The
mary cementing. They are designed to position the second mode is constant spacing, in which the program
casing more centrally in the hole and provide the follow- calculates the standoff that can be expected for a given
ing benefits. number of centralizers across the cemented interval. A
■ Reduce drag and differential sticking while running typical computer output is shown in Fig. 11-77.
casing The type, fit, and strength of a centralizer are impor-
■ Improve
tant factors to consider. There are three basic types of
centralizers—rigid, semirigid, and spring bow.
– mud removal Rigid centralizers are built with a fixed bow height
– cement placement, creating a more uniform wall and are sized to fit a specific casing or hole size. The
thickness rigid centralizer bows can be solid cast metal, such as
– performance of other external devices, such as aluminum (Fig. 11-78) or iron, molded plastics and phe-
scratchers and baskets nolic resins, or even machined composites. They also can
be made from solid, tubular, or T-bar steel members that
are welded together (Fig. 11-79). The side loading on a
given wellbore dictates to what extreme the operator
must go to centralize the pipe. For example, on a riser
tieback string, one would not use a 133⁄8-in. rigid central-
izer that may be exposed to helical buckling loads of
50,000 to 100,000 lbf [222 to 445 kN], well beyond the
capabilities of any spring-bow centralizer. This situation
could also overload the compressive strength of most
plastic, aluminum, or tubular rigid centralizers. A cen-
tralizer built out of solid steel bar or cast iron would be
able to meet the requirements. Centralizer choice
involves matching loading with restoring force or resis-
tance.

Fig. 11-76. Collection of casing centralizers (photograph courtesy

of Weatherford International, Inc.).

To be effective, centralizers must be properly placed

with respect to spacing and location within the string.
Spacing is a function of several parameters, such as
casing size, hole size, deviation, and all factors relating
to the side forces and drag. Most cementing and service
companies use computer software to determine optimal
centralizer spacing and placement (Chapter 12). The
software requires one to input pipe and well data,

390 Well Cementing

 Typical 9 5⁄8-in. Centralizer Application Analysis
Well Data:
Casing size: 9.625 in. Casing weight: 43.50 lbm/ft Casing seat: 6,300 ft Hole size: 12.250 in.
Mud density: 9.200 lbm/gal Kickoff point: 5,300 ft Rate of build: 10°/100 ft Final deviation: 89°
Constant Spacing Summation Variable Spacing Summation Variable Spacing Summation
with Spring Bow Centralizers with Spring Bow Centralizers with New Semirigid Bow
Spacing: Constant Spacing: Constant Centralizers
Number of Centralizers: 39 Number of Centralizers: 53 Number of Centralizers: 35
0% 100% 67% 70% 67% 70%
0 1.31 0.879 0.919 0.879 0.919
4,400 4,400 4,400

4,600 4,600 4,600

4,800 4,800 4,800

5,000 5,000 5,000

5,200 5,200 5,200

Depth Depth Depth
(ft) 5,400 (ft) 5,400 (ft) 5,400

5,600 5,600 5,600

5,800 5,800 5,800

6,000 6,000 6,000

6,200 6,200 6,200

6,400 6,400 6,400

0 1 2 0.87 0.88 0.89 0.90 0.91 0.92 0.87 0.88 0.89 0.90 0.91 0.92
Standoff (in.) Standoff (in.) Standoff (in.)
Conclusion: Projected standoff Projected standoff before cement Fewer units project comparable standoff
inappropriate for cement emplace- emplacement compares favorably to that of conventional spring bows
ment. Note no standoff at all with API specification for centralizer while effecting reduced casing drag
through build section performance. and installation time.

Fig. 11-77. Casing centralizer computer analysis (courtesy of Davis-Lynch, Inc.).

Fig. 11-78. Spiral-bladed rigid centralizer (photograph courtesy of Fig. 11-79. Rigid centralizer (photograph courtesy of Weatherford
Ray Oil Tools). International).

Chapter 11 Cementing Equipment and Casing Hardware 391

 A problem with rigid centralizers is that they must be
smaller than any restrictions they must pass through;
consequently, the amount of standoff is limited.
Figure 11-80 is a plot showing the amount of standoff
achieved by rigid versus spring-bow centralizers (dis-
cussed later). At low loading forces, the spring-bow cen-
tralizer will give a better standoff. In high-loading situa-
tions, such as a kickoff section of a deviated wellbore,
the rigid centralizer may give a better standoff.
Semirigid centralizers are relatively new and have
most of the attributes of the rigid centralizer in that they
can withstand high side loads. However, when forced
into a loading situation greater than their capability, the
bows collapse without destroying the centralizer
integrity and without leaving loose parts in the hole.
An example would be a semirigid centralizer built for Fig. 11-81. Semirigid centralizers (courtesy of Weatherford
International).
95⁄8-in. casing. The centralizer has an OD of 12 in. and is
run into a 121⁄4-in. open hole. If the centralizer passes a
swollen shale section during the running process, the
bows can collapse to a degree sufficient to allow the cen- Spring-bow centralizers are more commonly used.
tralizer to pass through. Further movement would have They are constructed of spring steel bows with an OD
been prohibited if a rigid centralizer had been run. greater than the hole in which they are run. These bows
Semirigid centralizers are most often a type of must be flexible, yet sufficiently stiff to provide adequate
stamped steel-alloy bow that is part of the solid body or standoff in various hole shapes and diameters. They are
welded on to end rings (Fig. 11-81). Most other materi- available with low or tall bow heights for slim or oversize
als are not acceptable because of potential cracking of holes, respectively. The most common bows are those
the materials under high stress. Examples of semirigid that are straight or aligned with the long axis of the
centralizers are shown below. casing on which they are mounted (Fig. 11-82).

Centralizer Stand Off Diagram

Rigid bow
Decision point
Spring bow
100

80

60
Standoff
(%)
40

20

0
0 500 1,000 1,500 2,000 2,500 3,000 3,500 4,000 4,500
Restoring force (lbm)

Fig. 11–80. Standoff versus loading force for rigid and spring-bow centralizers
(courtesy of Weatherford International).

392 Well Cementing

 Today, many engineers specify the tandem rise or
double bow centralizer because of its superior ratio of
starting force to restoring force (Fig. 11-84). Generally
speaking, one should select a centralizer that provides
the maximum restoring force with the least possible
amount of drag.

Nonweld straight bow Welded straight bow

centralizer with hinge centralizer, hinged version

Fig. 11-82. Spring-bow centralizers (photographs courtesy

of Weatherford International).

Other bows may be placed in a helical pattern (spiral-

bow centralizers) to aid mud removal while running.
Some have thin angled blades beneath the bows and are
able to induce turbulence and aid mud removal
(Fig. 11-83). Such devices are known as centralizers with
fins or turbolizers.
Although most spring-bow centralizers are similar in
Fig. 11-84. Tandem rise centralizers. The unusual double-hump
appearance, subtle differences in bow or hinge thick- structure of the spring bow acts like two short bows instead of one
ness or the lack of proper heat treatment may consider- longer one (photograph courtesy of Weatherford International).
ably reduce their performance or strength. There are
dozens of primary bows readily available from various
suppliers and hundreds of other specialty bows built for
specific requirements. To establish minimum strength requirements and
a basis to determine the centralizer spacing, the API
created the Specification for Bow-Spring Casing
Centralizers (API Spec 10D). This specification defines
strength requirements and methods of testing for stan-
dard and close-tolerance centralizers.
Close-tolerance centralizers are designed for casing-
hole clearances less than 0.75 in. per side. Spring-bow
centralizers can be made to fit tighter casing/hole com-
binations, but it can be difficult to keep the centralizer
in place while running casing. If the centralizer slips, it
will move up to a coupling or some other obstruction on
the casing string. Then, the longitudinal compression
forces will exceed the centralizer’s strength and the cen-
tralizer will deform and generate more drag. In some
cases, the centralizer may fall apart. Either situation
may prevent further downward movement of casing. The
slippage of centralizers occurs most frequently in tight
annular conditions or where the centralizer is not pro-
viding adequate standoff to prevent the stop collar from
scraping along the annulus.
Fig. 11-83. Turbulence-inducing centralizers, or turbolizers
(photographs courtesy of Weatherford International).

Chapter 11 Cementing Equipment and Casing Hardware 393

 To overcome this problem, several companies have When installed over stop collars, casing couplings,
developed a new type of centralizer called the central- and some centralizer subs, the centralizer is designed to
izer sub (Fig. 11-85). The purpose of the centralizer sub be pulled into the wellbore. This is the optimal way to
is to allow the spring-bow centralizer to compress while move it downhole with the least drag or damage to any
running through hole sections that are only large enough parts. One reason that stop collars occasionally fail on
for the casing connectors. After passing through the close-tolerance centralizers is that the dimensions
restrictions, the spring-loaded centralizer bows pop (thickness) do not allow the stops to fit under the cen-
back out and continue centralizing the pipe. tralizer bows at full compression. If this restriction is
The standard centralizer sub is built from float equip- known before the job, the centralizers may be run with
ment collar stock, much like the float collar shell. The stop collars on either end of the travel area; however,
entire OD is the same size as the connection couplings. this can increase the drag as the centralizer is pushed
Where the centralizer is installed on the sub, the OD is into the wellbore.
again reduced to produce a reduced slot for the end col- Pushing a spring-bow centralizer compresses it
lars and the bows. These slots are produced by milling lengthwise. This tends to cause the bows to expand out-
the OD out instead of welding on any rings, which would ward, increasing drag. When possible, centralizers should
compromise the shell body properties. The OD of the be installed so that they are pulled by the casing rather
compressed centralizer should be no larger than the than pushed. This reduces the chance of centralizer
casing coupling. Centralizer subs with rigid blades are damage caused by hanging. Pulling is accomplished by
available but have a limited benefit when dealing with placing a stop ring or lug within the centralizer collars.
under-reamed wellbores. In large or oversize hole conditions, centralizers may
be placed over couplings (but not couplings that have
an OD greater than 1 in. more than the pipe size), pro-
vided the couplings do not interfere with bow deflection
(compression). Most close-tolerance centralizers are
designed with low profiles and should not be placed over
stop rings or couplings that may contact or interfere
with the bows. The most common gripping method is to
install set screws into the collars of close-tolerance cen-
tralizers.
Standard spring-bow centralizers are available with
solid or hinged collars. The hinged style is the simplest
to install and is most often used. Solid-collar centralizers
are recommended for close-tolerance conditions (typi-
cal of liner applications), because they are generally
stronger and cause less drag resistance.

Torque- and drag-reduction centralizers

As high-angle or horizontal wellbores have become more
common, the need for drag reduction during running
casing has become a major priority in casing accessories.
Special coatings for spring-bow centralizers were origi-
nally used, but this method had limitations (tempera-
ture and abrasion). The next evolutionary step was to
integrate rollers with the solid blade centralizers
(Fig. 11-86).

Fig. 11-85. Centralizer subs (photograph courtesy of Weatherford

International).

394 Well Cementing

 Low-drag centralizers feature Low-torque centralizers feature
tangential rollers and are installed tangential as well as axial rollers
between stop collars or between and are installed between stop
casing couplings. More than 50% collars or between casing
Torque- and drag-reduction tools for drag reduction can be achieved couplings. This centralizer allows
drillpipe, completion strings, and liners with this centralizer. reciprocation and rotation of the
casing or liner.

Fig. 11-86. Roller-equipped centralizers (photographs courtesy of Weatherford International).

The first roller tools had tangential rollers and axles taining longitudinal wires or cables. They are attached
mounted perpendicular to the pipe. The next generation lengthwise to the casing to remove mud while rotating.
of roller tools had axial and tangential rollers. This Scratchers and wellbore wipers are most effective
change was intended to reduce drag and torque while when the casing is centralized and manipulated before
moving pipe. The axial rollers are arranged to contact and during cementing. To prevent solids or mudcake
the pipe inside the centralizer, allowing the centralizer buildup, scratchers should be spaced to ensure overlap-
to remain stationary while the pipe rotates inside. ping of areas worked by adjacent scratchers. It is very
At this writing, the record for the longest rotated important that circulation be established before pipe
liner in a high-angle wellbore was achieved in the Wytch movement. Reciprocating scratchers may be allowed to
Farm field (United Kingdom). The approximate vertical float between the stop collars. Rotating scratchers are
depth was 5,000 ft, and the total measured depth was attached with special clamps that are usually different
slightly greater than 31,000 ft. Overcoming torque and from normal stop collars.
drag to insert the liner to TD was a major accomplish- The use of scratching devices has declined in recent
ment, and rotation was performed for more than 10 hr years. The problem is not related to their effective-
while the hole was circulated and cemented. ness—sometimes they work too well. Mud filtercake
that is removed from the wellbore wall while running the
pipe adds solids to the displaced mud column and
Other casing-mounted cementing devices increases fluid loss to the formation as the cake is
Scratching devices are external devices designed to rebuilt. Both of these occurrences increase the hydro-
remove immobile mud from the wellbore and mudcake static pressure exerted by the fluid column.
from the wellbore wall (Fig. 11-87). These devices are Scratchers are most effective when the engineer
categorized by how aggressive they are. The basic spring- takes their effects into account. About every 10 to 15
loaded bristle scratcher is more aggressive than the joints of pipe, one should stop and circulate the annulus
cable scratcher (also known as a wellbore wiper). to move the displaced solids up into the previous casing
Scratchers are also designed for casing reciprocation or to surface. During the cement displacement process,
and rotation. Reciprocating devices, consisting of a pipe movement is still important because cement filter-
collar with radial wires or cables, are designed to remove cake takes the place of the removed mud filtercake.
mud during pipe reciprocation. The up-and-down move- Cementing baskets are used in low-differential-pres-
ment causes the wires or bristles to disturb the wellbore sure applications to separate fluids and to help support
wall. Rotating devices are straight strips of steel con- the hydrostatic pressure of the cement slurry. Cement

Chapter 11 Cementing Equipment and Casing Hardware 395

 Bristle-type scratcher,
designed for reciprocation.

Cable scratcher, sometimes

called a wellbore wiper because
of its soft wiping action.

Strip wellbore wiper, designed

to scratch during pipe rotation.

Fig. 11-87. Scratching devices (photographs courtesy of Weatherford International).

baskets are not an annular seal and they should never be

used to replace an ACP. They exist simply to offer a
restriction to wellbore fluids that tend to intermix
because of their different densities.
One place to use a cement basket would be 10 ft
directly below a stage tool. After the second-stage
cementation is completed and the stage tool is closed,
the heavier cement slurry will tend to drop into the
lighter fluid below the stage tool. The baskets can help
prevent this occurrence by inhibiting flow. Baskets are
typically constructed of thin steel petals arranged in an
overlapping pattern (Fig. 11-88). Cement baskets are
available in both the half-basket and whole-basket con-
figurations, and both are installed over a stop collar
(usually a set screw).
The packoff or barrier is formed as the petals spread Metal petal cement basket Metal petal half basket
to fit the hole, creating a sort of upside-down umbrella
within the annulus. Cement baskets were not designed Fig. 11-88. Cement baskets (photographs courtesy of Weatherford
International).
to be reciprocated rapidly, and they will not lift the
cement column during reciprocation. Experience has
shown that reciprocation can turn the baskets inside out
and destroy them. Casing rotation is possible with plan-
ning and the right type of stop collar.

396 Well Cementing

 Baskets are most effective when centralized and 11-6.2 Liner types
placed in a gauge section of the hole. Most baskets are There are four basic liner categories (Fig. 11-90).
not hinged and must be slipped onto a joint of casing Drilling or intermediate liner. This type of liner is
before makeup with the stop collar inside. The position normally run to cover a weak or troublesome formation
of the stop collar inside the basket allows the basket to that impedes further drilling operations. After the liner
remain stationary while picking pipe out of the slips is cemented in place, drilling to a deeper zone proceeds.
during running. Production liner. A production liner is run across a
zone of interest or a producing zone. These liners can be
cemented in place like drilling liners or consist of
11-6 Liners predrilled or screen-wrapped tubulars that are not
cemented in place.
11-6.1 Introduction Scab liner. A scab liner is a section of casing or tubing
A liner is a string of pipe made from conventional casing for repairing damaged casing. This liner is normally run
or tubing. Unlike conventional tubulars that are hung in old wells but may be placed in new wells in which tools
from the wellhead, a liner is hung from a previously set or the drillstring has worn through or damaged the pre-
casing string (Fig. 11-89). The distance between the top vious casing string. The scab liner can be run to repair
of the liner and the wellhead can be from less than 50 ft the leak and/or damage. Short scab liners are often not
[15 m] to several thousand feet. cemented in place; instead, they are frequently run in
A liner is lowered into the wellbore using the drill- place with a packoff seal at the top and the bottom.
string. This is normally drillpipe or tubing, but in some Scab tieback liner. A section of casing that extends
cases, the liner is lowered into position using coiled upward from the top of an existing liner but does not
tubing. reach the surface is called a scab tieback liner.

Pay zone

Drilling liner Production Scab Scab tieback

Liner Liner Casing liner liner liner
Fig. 11-89. Differences between a liner and conventional casing Fig. 11-90. Types of liners (drawings courtesy of Baker Oil Tools).
(drawings courtesy of Baker Oil Tools).

Chapter 11 Cementing Equipment and Casing Hardware 397

11-6.3 Reasons for installing a liner string. The original drillstring can still be used above the
There are many reasons why a liner is installed in a liner, saving time and money. Liners allow sidetracking
well. Some are related to the well. Others involve logis- during the drilling process without losing hole size.
tics, rig-handling capabilities, wellhead design, drilling
cost reduction, or contingency plans when drilling 11-6.4 Principal components of a liner
through unknown formations or encountering unex-
Most conventional cemented liners consist of five seg-
pected difficulties.
ments that run from the rig floor to the bottom of the
When a liner is run, the openhole section below the
well (Fig. 11-91).
previous casing string can be cased off more rapidly.
The liner is run on the drilling string, not on individual 1. Surface cementing equipment
tubular sections reaching to the wellhead. This saves rig 2. Liner setting tool assembly
time and allows the string to be quickly placed across 3. Liner setting sleeve/liner top packer assembly
sensitive or weak formations, possibly preventing the
4. Liner hanger
openhole section from caving in. In addition, because
the liner is set on bottom or hung from the previous 5. Shoe track assembly
casing string, the weight of the liner is not borne by the Throughout this discussion about liners, the reader
wellhead. will often be referred to Fig. 11-91 to clarify the location
Liners are usually much shorter than a conventional of the various components. For example, Fig. 11-91[2]
long string; therefore, they are easier to manipulate (i.e., refers to the liner setting tool assembly.
rotation, reciprocation, or both) during the cement job.
Such movement can greatly improve cement bonding
(Chapters 5 and 13).
Normally, cement slurries are more dense than the
drilling fluid. The increased hydrostatic pressure can 1. Surface cementing
lead to lost circulation (Chapter 6). When liners are equipment
used, the height of the cement-slurry column is shorter,
the hydrostatic pressure is lower, and the probability of
lost circulation is reduced.
Because liners are shorter than conventional long
strings, the pressure surge (piston effect) that occurs
while they are run into the hole is less severe. This also
decreases the chances of lost circulation. 2. Liner setting tool assembly
The tubular section that comprises the liner string
normally weighs more than the drillstring. Therefore, 3. Liner setting sleeve/
deeper wells can be drilled with smaller rigs, because Liner top packer assembly
of the decrease in overall string weight. Also, large rigs
can drill wells beyond 30,000 ft [9,140 m]—a feat that 4. Liner hanger
cannot be accomplished with conventional casing
strings run all the way back to the wellhead on conven-
tional drilling rigs.
The section of casing above a liner has a larger ID
than the liner. This allows more flexibility in the type of
completion equipment run in the well. The larger uphole
ID permits larger completion strings and, in most cases,
increased well productivity.
With conventional casing strings, one may eventually
reach a point at which the drillpipe will no longer fit
inside the next casing size. This forces the operator to
switch to a smaller-size drillstring. Installing a liner
5. Shoe track assembly
allows the operator to lay down only a portion of the

Fig. 11-91. Segments of a liner (drawing courtesy of Baker Oil Tools).

398 Well Cementing

11-6.4.1 Surface cementing equipment (Fig.11-91[1])
The surface equipment is located on the drilling rig floor. Topdrive connection
It connects the running string to the block of the rig and and circulation inlet
connects the displacement, circulation, and cementing
unit lines to the drillstring. The surface equipment holds
the displacement plugs and setting balls for hydrauli-
cally operated tools downhole. There are two types of Combination swivel and
cementing/displacement inlets
cementing heads for running liners. The first (Fig. 1-92)
is used for conventional drilling rigs and the other is
used for drilling rigs that operate with a topdrive system
(Fig. 11-93). This equipment allows total control of the
weight set on the liner, because the surface cementing
equipment carries the string weight at all times during
the cementing and liner-installation operations. Plug-retaining flapper

Elevator lifting neck

Plug-retaining flapper

Positive ball dropping device

Plug retaining pin

Cementing and
displacement inlets

Fig. 11-93. Cement head for topdrive drilling system (drawing

courtesy of Baker Oil Tools).

Swivel

Fig. 11-92. Plug-dropping cement head for conventional drilling

rigs (drawing courtesy of Baker Oil Tools).

Chapter 11 Cementing Equipment and Casing Hardware 399

 In some parts of the world, local regulations require ■ Provide a reliable releasing mechanism to detach the
the cementing heads to be operated directly from the rig running string once the liner is at the desired position
floor. For these areas, a remotely controlled pneumatic in the well [this can be done mechanically
version of the cementing head is used. The remote-con- (Fig. 11-95), hydraulically (Fig. 11-96), or both,
trol panel is shown in Fig. 11-94. depending on the type of running tool used]
■ Allow rotation of the liner before and after the
drillpipe has been disconnected from the drillstring
■ Provide the ability to set liner top packers after the
setting tool is released
■ Provide a seal between the drillpipe and the liner
while running the liner in the well and during the
cement job (diverting the fluid to the bottom of the
liner while circulating)

Fig. 11-94. Control panel for pneumatically operated plug-dropping

head (drawing courtesy of Baker Oil Tools).

Pneumatically operated surface equipment is also

used in areas where rotating liners are run on a regular
basis. With conventional systems, it is a safety standard
to stop the rotation when the displacement plugs or set-
ting balls are being launched. Resuming liner rotation
after stopping may be difficult or impossible. The pneu-
matic design permits continuous rotation of the liner,
enhancing the safety and efficiency of the liner-cement-
ing operation.

11-6.4.2 Liner setting tool assembly (Fig. 11-91[2])

The liner setting tool assembly is a retrievable device
that connects the drillstring to the top of the liner. In the
1930s, it was most commonly known as the “letting-in
tool,” and mainly consisted of a modified casing release
spear. Today, the setting tool is run with a liner setting
sleeve or liner top packer (Fig. 11-91[3]) with a corre-
sponding setting tool profile. Together they perform sev-
eral key functions during the liner running and cement-
ing operations.
■ Support the weight of the liner while running in the
hole A B
■ Allow the operator to push, pull, and circulate freely
Fig. 11-95. Mechanical liner running tools (drawings courtesy of
until ready to disconnect Baker Oil Tools).

400 Well Cementing

 Packer setting
dog

A B
Fig. 11-97. Packer-setting dog sub (drawing courtesy of Baker Oil
Hydraulic running Coiled tubing Tools).
tool running tool

Fig. 11-96. Hydraulic and coiled tubing liner running tools (drawings
courtesy of Baker Oil Tools).

When the liner is run with a liner top packer also create a temporary closed chamber at any time
(Fig. 11-91[3]) at the top of the liner string, an addi- during the installation to activate hydraulically operated
tional device is required in the liner setting tool assem- tools in the liner string. Some of these hydraulic oper-
bly. This tool is usually called a packer-setting dog sub ated tools are hangers (Fig. 11-91[4]), running tools
(Fig. 11-97). (Fig. 11-91[2]), hydraulic casing packers, hydraulically
As explained above, the liner setting tool assembly operated openhole packers, and cementing-type valves
not only acts as the quick disconnect between the drill- located in the liner string.
string and the liners, but also creates a seal between the Packoff assemblies come in a variety of models that
two strings. This seal is required to ensure that the fluid can be categorized into four types. Each has its advan-
pumped at surface through the drillstring is pushed out tages and disadvantages according to the type of liner
of the bottom of the liner string. Together with a ball and the well conditions.
seat or temporary plugging device inside the liner, it can

Chapter 11 Cementing Equipment and Casing Hardware 401

 Cup seal system (Fig. 11-98). The cup seal system is large-diameter liners or liners at shallow depths, in which
an economical setting tool packoff for low-pressure and no drillstring weight is available to keep the running
-temperature service. Normally, it only holds pressure string in place while pumping fluids. The insert has bidi-
from one side and is part of the liner setting tool assem- rectional seals, is normally made of aluminum or compos-
bly. Thus, it is retrieved after the job is finished and can ite material, and is drilled out after the liner cementing
be reused after it has been inspected and the cups have operation is completed. Sometimes this liner-top restric-
been replaced. The cup seal system provides an unre- tion can aid in highly deviated wells and where there is a
stricted liner top at the end of the cementing operation. possibility of excess cement settling out inside the liner.
Drillable packoff bushing system. This system com- Polished-bore receptacle (PBR) packoff system. The
prises one tool installed at the top of the liner and another PBR packoff system (Fig. 11-100) has the same advan-
corresponding tool at the bottom of the liner setting tool tages as the cup seal system in that it leaves the liner fully
assembly, usually called a slick stinger. The tool run in the open after the cement job. In addition, it has the high
liner string is normally run between the liner setting pressure and temperature bidirectional pressure capabil-
sleeve or liner top packer (Fig. 11-91[3]) and the liner ities of the drillable packoff bushing system. It consists of
hanger (Fig. 11-91[4]). The drillable packoff bushing two components, one located at the bottom of the liner
system (Fig. 11-99) creates a positive two-way seal at the setting tool assembly and the other placed in between the
liner top. Upward piston forces are greatly reduced com- liner setting sleeve or liner top packer (Fig. 11-91[3]) and
pared to other seal systems, particularly when running the liner hanger (Fig. 11-91[4]).

Drillable packoff bushing

Slick stinger

Fig. 11-99. Drillable packoff bushing system (slick stinger) (drawing

Fig. 11-98. Cup seal system (drawing courtesy of Baker Oil Tools). courtesy of Baker Oil Tools).

402 Well Cementing

 Polished-bore Profile sub
receptacle

Retrievable
seal system

PBR packoff

Machined recess

Fig. 11-100. PBR packoff system (drawing courtesy of Baker Oil Fig. 11-101. Retrievable seal system (drawing courtesy of
Tools). Baker Oil Tools).

Retrievable seal system. The retrievable seal system sections. Normally, most liner setting sleeves are manu-
(Fig. 11-101) combines the advantages of the drillable factured with some type of tieback extension or PBR
packoff bushing system, in which there are very low above (Fig. 11-102) and in some cases below
pumpout forces, and the PBR packoff system with the (Fig 11-103) the liner setting tool profile.
bidirectional high-pressure, high-temperature seal and This tieback extension or PBR is used to stab in a set
the unrestricted liner top at the end of the cementing of seals built on a seal mandrel with the same ID as the
operation. The retrievable section below the running liner string. This seal mandrel can be integral to a
tool is locked into a special profile sub at the liner top by tieback packer and used to repair liner-top leaks, run as
a set of locking dogs. These dogs will transfer most of the a scab tieback liner, or used as a conventional tieback
upward forces back into the liner instead of placing string that extends all the way back to the wellhead. In
them on the drillstring. At the end of the cement job, the some complex liner completions, this tieback extension
dogs are allowed to collapse inwards, owing to a dedi- or PBR is designed as part of the completion.
cated machined recess at the bottom of the slick stinger. In many cases, the overlap between the liner top and
This allows the retrievable seal system to be pulled out the bottom of the previous casing string is not long
of the liner top, leaving the liner top fully open. enough to create a proper cement seal at the top of the
liner. A 500-ft overlap has been the industry standard;
however, some operators and well conditions require the
11-6.4.3 Liner setting sleeve/liner top packer overlaps to be as short as 50 ft. In some wells, in which
assembly (Fig. 11-91[3]) the adjacent formation is very weak, the cement will
The liner setting sleeve is the “crossover” between the never reach the top of the liner; therefore, instead of
liner setting tool assembly (Fig. 11-91[2]) and the liner relying upon cement alone, a supplementary seal should
string itself. The upper profile or thread matches that of be used. For such applications, the liner setting sleeve is
the liner setting tool assembly (Fig. 11-91[2]), and the replaced by a crossover between the liner setting tool
bottom thread and ID matches that of the liner tubular assembly and the liner string itself. The crossover fea-

Chapter 11 Cementing Equipment and Casing Hardware 403

 Setting tool profile

PBR extension

PBR
extension PBR
extension

Setting
sleeve
Setting
sleeve

Fig. 11-103. Setting sleeve with tieback extension below the

setting-tool profile (drawing courtesy of Baker Oil Tools).

Mechanical Hydraulic

Fig. 11-102. Setting sleeve with tieback extension above the profile Much has changed since the first liner top packers
(drawing courtesy of Baker Oil Tools). were introduced in the 1920s. The original devices were
made of lead, canvas, and rubber elements. Even in the
mid-1980s, only about 20% of the cemented liners were
tures a packing element that creates a seal between the run with liner top packers owing to problems with the
liner top and the previous casing string. This liner top seal during the installation process. High circulation
packer is normally set after the cement has been rates, temperature increases, and cuttings in the drilling
pumped in place. Other functions of the liner top packer fluid would damage or swab off the seal. Compression-
include the following. type elastomeric seals and various types of backup rings
■ After the liner has been hung and cemented in place, have greatly improved the reliability of the conventional
the liner top packer insulates the cement from the liner top packer (Fig. 11-104). In the early 1990s, a mile-
hydrostatic pressure of the drilling fluid and aids the stone was reached with the introduction of expanding
backpressure equipment inside the shoe track metal technology to liner top packers (Fig. 11-105).
(Fig 11-91[5]). As a result, the cement slurry is held These expanding seals can be run in the most rigorous
in place until it sets. environments and are typically not the weak link in the
■ When excess cement above the liner top is reversed
liner system. As a result, liner top packers are used in
about 80% of liner installations today.
out, the liner top packer isolates the remaining
Liner top packers are also equipped with the same
cement slurry behind the liner from the pumping
type of tieback extension as the conventional liner set-
pressure, reducing the possibility of formation break-
ting sleeve (Fig. 11-102). However, the tieback extension
down or loss of cement in the liner lap.
on a conventional liner setting sleeve is either threaded
■ If a poor cement job has occurred, the liner top
or part of the tool, while that on a liner top packer moves
packer serves as a seal around the top of the liner independently.
assembly.

404 Well Cementing

 PBR extension

Setting tool profile

PBR extension

Packer

Setting tool profile

Hold-down slips

Hold-down slips

Packer

Fig. 11-105. Liner top packer with expanding metal seal (drawing
courtesy of Baker Oil Tools).

Fig. 11-104. Liner top packer with elastomeric seal (drawing

courtesy of Baker Oil Tools). 11-6.4.4 Liner hanger
Liner installation and cementing techniques have
improved since the early days. Initially, liners were
After the liner has been installed and cemented in installed by drilling the required openhole section below
place, the liner settting tool assembly is removed from the previous casing string, and cement was placed in the
the tieback extension. This exposes and activates the openhole section. Liner joints were made up on the rig
dogs on the packer-setting dog sub (Fig. 11-97). Slacking floor and allowed to fall freely in the well. The drillstring
off to a predetermined amount of set-down weight will was used to push the liner into the cement before it set
place weight on the tieback extension and not on the (Fig. 11-106).
setting sleeve/liner system, because the dogs on the The introduction of the liner hanger (Fig. 11-91[4])
packer-setting dog sub (Fig. 11-97) cannot reenter the in combination with the liner setting sleeve assembly
tieback extension. This will allow the packing element or (Fig. 11-91[3]) and the liner setting tool assembly (Fig.
expanding metal seal to move outwards against the tubu- 11-91[2]) at the top of the liner string allowed liner
lar in which the liner is being run to create a seal. placement in a more controlled manner. The liner

Chapter 11 Cementing Equipment and Casing Hardware 405

 Cone

Slip

Jay cage

Bow spring

Fig. 11-107. Single-cone mechanically set liner hanger (drawing

courtesy of Baker Oil Tools).

Drill well Place cement Push liner into

in open hole place with
drillstring Hanger body
Fig. 11-106. Early liner placement technique (drawings courtesy of
Baker Oil Tools). Cone

hanger is basically a simple wedge or inclined plane and

ensures that the liner will not collapse because of its
own weight, become buckled, or shift position in the
hole after it has been installed. The liner is suspended
and cannot sink into the formation bottom, nor can it fall
over into a cavity and lose alignment with the cased hole
above. Being perfectly aligned with the axis of the hole
and free of bending, the suspended liner is not subjected Slip
to undue wear from tools, rods, or tubing strings run
through or operated inside the liner string. Initially,
liner hangers were mechanically set, much like the
device shown in Fig. 11-107, except that the 360° cone
(Section 11-6.4.4.2) (Fig. 11-108) has been replaced with
cone pads to increase bypass or flow area.
Fig. 11-108. Slip, cone, and liner-hanger body (drawing courtesy of
Baker Oil Tools).

406 Well Cementing

 Hanger designs are influenced by several factors. 11-6.4.4.1 Predicting liner hanger capacities
1. Annular dimensions. The available space between There is no single method to accurately predict the liner
the liner OD and the casing ID in which the liner is load that a hanger will support, because many downhole
to be run. and mechanical variables are not well quantified. Aside
2. Annular area. Backpressure on the formation during from downhole conditions, which naturally vary consid-
circulation and cementing should be kept to a mini- erably, the friction factor for steel on steel will vary with
mum. Multicone designs on conventional hangers the following:
increase the bypass area and provide higher load- ■ machined finish
carrying capabilities. ■ lubrication or films between surfaces
3. Casing grade, size, and wall thickness. The weaker ■ vibration
the casing in which the liner hanger is to be set, the ■ extent of contamination between surfaces
lower the liner weight that can be suspended from
■ yield strength and dimension of tools versus mini-
the hanger itself.
mum yield and nominal dimensions
4. Length of the liner and liner top packer rating. Liner
■ condition of the tubular in which the liner hanger is
length is normally not a significant problem and is
essentially unrestricted in most applications. Also, to be set and the wellbore environment behind the
the rating of the liner top packer (especially in unce- tubular.
mented or partially cemented liner applications) As mentioned previously, the liner hanger mechanism
must be taken into consideration when choosing a is a simple wedge or inclined plane. Formulas for deter-
hanger type. Tubing with higher pressure ratings will mining the magnitude and direction of these forces are
add substantial weight to the hanger. readily available in mechanical engineering handbooks.
5. Setting. Unlike retrievable packers or other retriev- These formulas can be used as a starting point to esti-
able tools, most liner hangers and liner top packers mate the maximum load that a liner hanger can carry.
are designed for a one-time operation. As a result, The most common workable solution relies heavily on
they are less expensive. the following:
6. Pipe movement. Special rotating or reciprocating ■ past successful performance and installations
designs allow movement during the cementing oper- ■ the use of conservative friction factors
ation to aid cement displacement. The first rotating ■ the use of conservative safety factors
liners were designed in the early 1950s but did not
■ testing.
gain wide industry acceptance until the mid-1980s.
7. Hole bridging or collapsing. Where this problem Pull tests are not more reliable than the formulas in
occurs, drill-down liner hangers and systems are engineering handbooks for predicting the actual loads
available. that can be supported. In fact, relying solely on pull-
8. Environment. Corrosive environments dictate the through tests can be misleading, unless a large number
use of special materials or controlled hardness of tests are conducted to establish validity.
designs. This is more typical for production or com- Some liner manufacturers have translated the hand-
pletion rather than drilling liners. book formulas for a simple wedge for use as a starting
point for designing multiple slips and cones to be
9. Lost circulation. This is associated with the bypass
installed on the hangers. The formulas are shown in Eqs.
area requirement. A hanger and liner top packer will
11-2 and 11-3.
help hold the cement until it sets. For conventional
packers, the higher the rating of the liner top packer, 1. Maximum supported load Fmax that the casing in
the larger the OD will be; therefore, the annular which the liner hanger is to be set will withstand:
clearance between the liner OD and the casing ID Fmax = pburst EG, (11-2)
will be smaller. This can cause lost circulation.
10. Collapse, burst, and tensile requirements. These where
requirements may also dictate the type of hanger pburst = burst pressure or internal yield pressure of
and liner top packer that can be used in environ- casing in which liner hanger is to be set
ments in which pressure ratings must exceed 10,000
psi [68.9 MPa] and 500,000 lbm [2,220 kN] of load. ⎡ Lπdslip ⎤
E=⎢ ⎥ × N slips
⎢⎣ N slipseg ⎥⎦

Chapter 11 Cementing Equipment and Casing Hardware 407

 where matter. Factors that have a direct bearing on the down-
dslip = slip OD hole liner capabilities are the following.
■ Is the supporting casing in new or near-new
L = slip length
condition?
Nslips = number of slips
■ Is the supporting casing backed up by hard,
Nslipseg = total slip segments cut
good-quality cement?
tan α + μ ■ What is the actual ID (not nominal ID) of the casing
G= in which it is set?
1 − μ tan α
■ What is the actual friction factor for that specific
where hanger in a particular well?
α = hanger cone angle The first three questions cannot always be answered
μ = coefficient of friction. at the time the hanger is designed. Possibly, a reason-
able estimate of actual existing conditions could be
2. Maximum load that the hanger body (or liner joint
found before running the liner, but even this is doubtful.
on which the hanger is made) will support:
For all of the above reasons, the formulas discussed
Fmax = pcol BG (11-3) above are used when recommending a hanger load of a
slip-and-cone type hanger design. Historically, most liner
where hangers designed using these calculations do not fail or
pcol = collapse pressure of liner hanger body allow the liner to move down the hole.

⎡ Dπdbody ⎤ 11-6.4.4.2 Liner hanger types

B=⎢ ⎥ × N cpads ,
⎢⎣ N coneseg ⎥⎦ Liner hangers can be classified into two types:
■ Mechanically set liner hanger
where ■ Hydraulically set liner hanger.
dbody = OD of body
Both types are available in a variety of designs. The
Nconeseg = total cone segments cut type of setting mechanism required, or the amount of
Ncpads = number of cone pads weight the liner hanger will encounter at its setting
depth, determines the number of slips and cones it will
tan α + μ need. Simple designs have a single row of slips and
G= .
1 − μ tan α cones; other designs may have two or more rows of slips
and cones to increase the carrying capability.
Equation 11-2 provides the maximum hanging load Both designs use the same wedge technology, which
that may be imposed on the supporting casing, assuming: consists of the following components (Fig. 11-108).
■ the casing in place is not supported by cement or Slip. This device is wedged between the cone on the
formation behind the pipe liner hanger and the casing in which it is set. This slip
■ the supporting casing is new pipe that has not been
has hardened sharp teeth on the outside that bite into
the casing, preventing the liner from sliding downhole.
weakened or greatly enlarged by drillpipe wear or
Cone. A ramp on the liner hanger body that the slip
pressure
moves along until the slips bite into the casing wall.
■ the correct friction factor is used.

Equation 11-3 gives the maximum hanging load that Mechanically set cone-and-slip liner hanger
the hanger will support before the hanger body fails. The mechanically set liner hanger (Figs. 11-107, 11-109,
Both calculations are performed, and the lesser of the and 11-110) was the first type designed. It was intro-
two results is used to determine how much liner weight duced in the 1920s. The setting mechanism consists of
can be carried by the hanger and in the casing in which two items: the jay cage and the bow springs. The jay cage
it is placed. is a sleeve attached to the hanger body. The sleeve can
For many years, thousands of pull tests and field runs be released from the body by means of pipe manipula-
were performed, and these have shown that liner capac- tion. This allows the slips to move from the run-in posi-
ities are greater than those calculated by the formulas tion to the set position. The bow springs are made from
above. But translating either the pull tests or the calcu- hardened steel that will not wear out easily while run-
lated results to actual downhole conditions is another ning in the hole. Their main function is to hold backup

408 Well Cementing

 Cone

Slip

Cone
Cone

Slip
Slip

Fig. 11-109. Single-cone mechanically set liner hanger (drawing

Fig. 11-110. Multicone mechanically set liner hanger (drawing
courtesy of Baker Oil Tools).
courtesy of Baker Oil Tools).

friction on the jay cage to facilitate the liner-hanger set- pumped down to temporarily close off the liner
ting operation. system. Using this type of liner hanger will also pre-
The same design principles are still used today. The vent overpressuring the formation.
mechanically set liner hanger is available in a variety of ■ There is a continuous mandrel without any setting
designs with the same features and benefits. ports and seals.
■ There is excellent bypass area, because the hanger
■ The hangers can be set and unset multiple times on a
does not require a setting piston that may restrict the single trip.
bypass and decrease the overall pressure rating. ■ The hangers are not sensitive to temperatures or
■ The hangers can be set and activated with pipe
harsh environments that may compromise the seals.
manipulation. Depending on the model and type of ■
setting tool, setting is performed by picking up the Erratic circulation pressures do not interfere with the
drillstring and the liner in tension at the required set- hanger setting mechanism.
ting depth and either rotating the string to the left or
the right and slacking off. Hydraulically set, slip, and cone liner hanger
■ The hangers match the liner pipe strength in almost
The hydraulically set liner hanger (Figs. 11-111 and
all liner sizes. 11-112) was introduced in the early 1940s. It was
■ The hangers do not require setting pressure to acti- designed to allow running a liner in highly deviated or
vate. This saves rig time because no ball or plug is crooked holes or through casing windows, in which the

Chapter 11 Cementing Equipment and Casing Hardware 409

 Cone
Cone

Slip

Slip

Cone

Slip

Fig. 11-111. Single-cone hydraulic-set liner hanger (drawing

courtesy of Baker Oil Tools).

Fig. 11-112. Multicone hydraulic-set liner hanger (drawing

courtesy of Baker Oil Tools).
activation mechanisms of mechanically set hangers can
become stuck. Hydraulically activated liner hangers
have a smoother outside profile. out a topdrive system. Hydraulically set liner hangers
Two types of hydraulic-set hangers are available— do not require left-hand rotation of the drillstring,
one in which the hydraulic setting mechanism is built which could potentially back off a connection during
into the hanger and one with the hydraulics built into the setting process.
the running tool. ■ The smoother profile on the outside aids the running
The cone and slip design (Fig. 11-108) is identical to
of liner systems into and past existing liner tops.
the design of mechanically set hangers. Thus, both types
■ The hydraulically activated independent setting
have the same load-carrying capabilities. The only dif-
ference is that the hydraulically set hanger is activated mechanism does not rely on friction between the bow
at the required setting depth by applying a predeter- springs and the casing, as mechanically set liner
mined differential pressure at the liner hanger hangers do. Certain drilling fluids decrease the fric-
(Fig. 11-113). tion between the bow springs and can prevent the set-
The hydraulically set liner hanger is available in a ting of the hanger.
variety of designs with the following features and bene- ■ The liner string can be rotated before setting, allow-
fits. ing the liner to more easily pass ledges and bridged-
■ Hydraulically set liner hangers do not require pipe off areas in the well. This is virtually impossible with
manipulation to activate the hanger, as some mechanically set hangers, because rotation of the
mechanically set liner hangers do. Pipe manipulation pipe may activate the setting mechanism.
may be difficult in heavy long liners and highly devi- ■ The shear mechanism is normally preset and pre-
ated wells or on conventional rotary drilling rigs with- tested for downhole conditions and reliability.

410 Well Cementing

 Although this hanger improved rotational perfor-
mance, it provided only small cross-sectional areas
between the liner OD and casing ID, resulting in severe
dynamic load constraints. These liner systems were run
on a J-type setting tool and setting sleeve combination
with inherent load and torque weaknesses. They could
be run only on short, lightweight liners.
In the early 1980s, operators showed more interest in
rotating liners without altering the conventional liner
installations. New custom-designed bearings (Fig. 11-114)
Moving specifically for liner applications were introduced. In
seal most cases, the new bearings did not increase the over-
Pressure all maximum OD of the conventional liner hanger. Newly
port designed liner setting tools for rotating liner hangers
Static were more robust than the conventional running tools.
seal

Upper bearing race

Shear
pin
Tapered
bearing
rollers

Lower
bearing
Run in Apply internal Shear pin race
pressure and set slips

Fig. 11-113. Hydraulic cylinder settting mechanism (drawing Fig. 11-114. Bearing for rotating liner hanger (drawing courtesy of
courtesy of Baker Oil Tools). Baker Oil Tools).

Rotating liner hangers that allow liner rotation during Rotating type hangers are designed for both the
the cement job mechanically set (Fig. 11-115) and hydraulically set
Conventional hangers are normally set in the wellbore, liners (Fig. 11-116), with the same setting mechanisms
and running tools are released before pumping the and operating and running guidelines as described ear-
cement in place. Pipe manipulations during the cement lier (Section 11-6.4.4.2). The running tools used with
displacement greatly increase the overall quality of the these types of rotating type liner hangers can be seen in
cementing job (Chapter 5). Figs. 11-95B (mechanical running tool) and 11-96A
The first rotating liner systems were introduced in (hydraulic running tool).
the early 1960s. A mechanically set rotating liner hanger There are other aspects to consider when comparing
was designed with a special liner setting tool that would rotating liners to conventional ones. First, rotating hang-
allow the liner hanger to be set and the liner setting tool ers use bearings and are rated for a particular load car-
assembly to be released from the liner. One could still rying capacity, depending on the design and size. The
rotate the liner string. The liner hanger was equipped bearings must withstand the total weight of the liner
with a full-circle floating cone at the top of the liner while the hanger is set, plus the drillstring weight placed
hanger. A conventional industry-standard ball bearing on the setting tool. As described in Section 11-6.4.2, the
assembly was located above the cone and below the top drillstring weight prevents the packoff assemblies from
connection. being lifted out of the top of the liner during the pumping

Chapter 11 Cementing Equipment and Casing Hardware 411

 Bearing assembly
Bearing assembly

Fig. 11-115. Mechanically set rotating liner hanger (drawing Fig. 11-116. Hydraulically set rotating liner hanger (drawing
courtesy of Baker Oil Tools). courtesy of Baker Oil Tools).

stages. This total combined weight may exceed the load When running a rotating liner, one must be sure to
rating of a bearing when a cup seal system (Fig. 11-98) or use the proper centralizers. Not all centralizers used on
PBR packoff system (Fig. 11-100) is used in large liner conventional liners are suitable with rotating liners
sizes. Such packoff systems typically are not used with (Section 11-5.11).
liners that can be rotated during the cement job. Last but not least, when a rotating liner is planned,
Second, torque is applied to the drillstring and liner the drilling rig must be considered. The ability to rotate
string to rotate the liner. Therefore, both the drillstring the liner and drillstring while maintaining control of
and the liner-joint threads that attach the liner joints to the amount of weight that is placed on top of the liner
their accessory equipment must be considered. Threads during the job requires special surface equipment
with low torque ratings and the nonshouldered connec- (Section 11-6.4.1). When working from a rig with a top-
tions used on conventional liners should not be used on drive system, one should use the equipment shown in
a rotating liner, because this may prevent liner rotation Fig. 11-93. When working off of a conventional rig with-
during the job. Torque failure owing to excess torque can out a topdrive, one should use the equipment shown in
result in several negative consequences. Leaks may Fig. 11-92 and carefully maintain control of the liner
occur, and the liner ID could be reduced, preventing weight.
proper cleanup. In addition, it may be impossible to seal
additional tools into the liner at a later stage of the well
completion.

412 Well Cementing

Liner hangers for all types of conditions The standard hangers are available in mechanical
Up to this point, only the conventional slip-and-cone (Fig. 11-117) and hydraulic (Fig. 11-118) versions, both
liner hangers have been discussed. In the late 1980s, a with and without the bearing option.
hanger was introduced that does not require any weld- The slips on these hangers are enclosed inside the
ing, which allowed manufacturers to make liners cost- slip seat and are particularly appropriate for liner sys-
effective and comply with any corrosion resistant alloy tems that must be rotated while running in the hole to
(CRA) material an operator specified. The slip-and-cone the desired setting depth. Some running tools (mainly
design has been replaced by the slip-and-slip seat hydraulically operated types) will allow this operation to
design, which has uniform casing contact, and allows the occur during the installation. The slips on conventional
slips to match the contour of the casing in which the hangers are completely open and are held in place with
hanger is set. This even liner-load distribution greatly small screws; such hangers are not appropriate for oper-
increased the load-carrying capabilities versus conven- ations that require rotating while running. Slips could
tional liner hangers. Operators could also run extra-long break from the hanger while running in the hole, and the
liners in combination with the higher loads created by system could become stuck. Expensive fishing and
liner top packers with a high pressure rating. Some milling operations would then be necessary.
hanger sizes can hold 1.8 million lbm [816,470 kg] in Because these hangers can be assembled from a kit,
combined loading, with little or no casing damage. they can be delivered with various special options for
Because these hangers are supplied in a “kit form,” unique liner applications without having to design a
they can be made in various designs without changing completely new tool. These options include the following.
many components and can sometimes be converted in Double grip. Some liners are so short and light that
the field without having to order completely new tools. it is difficult to detect if the setting tool assembly has

Mechanically Mechanically Hydraulically Hydraulically

set set rotating set set rotating

Fig. 11-117. Mechanically set liner hangers (drawings Fig. 11-118. Hydraulically set liner hangers (drawings
courtesy of Baker Oil Tools). courtesy of Baker Oil Tools).

Chapter 11 Cementing Equipment and Casing Hardware 413

released from the liner. The double grip function allows piece of material. These hangers are only made in a
the operator to pull tension into the liner after the hydraulically set version because of the rotational
hanger has been set and the setting tool has been requirements before reaching setting depth. The con-
released. Overpull (pull-up weight greater than that of ventional setting cylinder is replaced by a setting piston
the liner) indicates that the setting tool assembly is still mechanism to increase the bypass area and prevent the
attached to the liner. No overpull indicates the setting cylinder from becoming stuck during run-in. Some of
tool assembly is disconnected from the liner. these liner hangers are designed with a liner top packer
Antileft-hand rotation feature. The antileft-hand to make the system even smoother on the outside and
option is also available for short liners that are run with shorter in length (Fig. 11-120).
hydraulically operated tools and require some type of
mechanical backoff. On conventional short and light
liners with a bearing liner hanger, the liner may rotate
on the bearing after the hanger has been set.
Disconnecting the setting tool assembly from the liner
may be impossible.

Drill-down liner hangers

Drill-down liner systems are often confused with wash-
down liners, which require some rotation for short peri-
ods of time. Some well conditions require treating the
liner like a conventional drilling string to run it to set-
ting depth. In such cases a special drill-down liner
hanger (Fig. 11-119) is run that is machined out of one

Hanger
activation
mechanism

Fig. 11-120. Liner hanger with liner top packer (drawing courtesy
of Baker Oil Tools).

Fig. 11-119. Drill-down liner hanger (drawing courtesy of Baker Oil

Tools).

414 Well Cementing

11-6.4.5 Shoe track assembly
The shoe track assembly (Fig. 11-91[5]) is run at the
bottom of the liner. Its function is basically the same as
a shoe track of a full casing or long string. It consists of Liner wiper plug
a chamber between a number of backpressure valves. latch profile
The assembly prevents contaminated cement from
entering the openhole section at the end of a cement
job. The backpressure valves have the same function as
float valves in a full casing or long string. They prevent
the cement slurry from U-tubing, or flowing back into the
liner after the setting tool has been pulled out of the
liner top. The liner shoe track assembly normally con-
sists of three parts.
■ Shoe or guide at the bottom of the string. As with con-
Fig. 11-121. Seat-and-latch landing collar for liners (drawing
ventional casing tools, the shoe or guide at the bottom courtesy of Baker Oil Tools).
of the string is called a float shoe (Section 11-5.2).
The float shoe guides the liner through the openhole
section to the bottom of the well or the desired set-
ting depth. Side and bottom ports ensure reliable cir-
culation at all times, even if the liner is set on bottom
or is lying on the low side of a highly deviated well-
bore. The float shoe is delivered with either one or Liner wiper plug
two backpressure valves. It is made of easily drillable latch profile
material for removal after the cement has set and
Integral ball seat
before drilling the next hole section.
■ Upper valve system at the top of the shoe track. On
conventional casing strings, this valve is called a float
collar (Section 11-5.2). It is normally placed one or
two joints, and in some cases several joints, above the
float shoe. The spacing between the float shoe and
float collar depends on the liner size, expected cont-
aminated cement after the cement job, and, in some
Fig. 11-122. Seat and latch landing collar for liners with integral
cases, particular field or well cluster characteristics. ball seat (drawing courtesy of Baker Oil Tools).
The inner components of the float collar are made
from the same drillable material as the float shoe.
■ Landing collar. The landing collar is usually run one
ball seat below the displacement plug seat. When acti-
or two joints above the float collar. If the operator vating hydraulic devices in the liner string, such as
decides not to use a float collar, the landing collar can hydraulically set liner hangers (Section 11-6.4.4.2),
be installed above the shoe. Normally, this tool is only ACPs, or cementing valves, the liner string must be tem-
run with liners and not installed in a full-casing-string porarily plugged off by dropping a setting ball to the seat
cementing operation. in the landing collar. This creates a closed chamber that
Landing collars catch displacement plugs and provide allows the string to be pressurized. Pressurizing the
a secondary backup to the float valves in the float shoe drillstring and liner creates a differential pressure
and float collar. When the running tools are pulled out of inside and outside the string. This pressure increase
the liner top, the landing collars prevent backflow or U- activates the hydraulically set devices in the string at
tubing of the cement. There are two types of landing col- various intervals. A further increase in pressure shears
lars. One (Fig. 11-121) provides a seat-and-latch assem- out the setting ball and seat in the landing collar, restor-
bly to latch the displacement plugs at the completion of ing full circulation through the shoe for cementing oper-
cement displacement. Combined, they are designed to ations. When hydraulically operated devices are only run
ensure rapid and complete drillout if needed for addi- at the top of the liner string, the setting ball and seat
tional hole sections below the liner. The second type may be placed closer to the liner top. In such cases, a
(Fig. 11-122) has one additional feature—an integral standard landing collar without the ball seat is used.

Chapter 11 Cementing Equipment and Casing Hardware 415

11-6.5 Displacement plugs for liner applications liner and the ID of the openhole section (Fig. 11-123D).
Displacement plugs perform the same function as those A pump-pressure increase at the surface indicates that
in conventional long string casing cementing operations. the plugs have landed and reached the landing collar.
Fluids such as drilling muds, completion fluids, spacers, The method described above only separates the end
washes, and cement slurries should be separated to pre- of the cement slurry and the beginning of the displace-
vent cross contamination as they are pumped through ment fluid; thus, cement-slurry contamination at the
tubulars and up the annulus. front end is still possible. Some operators use cement
Conventional casing plugs are pumped directly from displacement plugs before and after the cement slurry.
the surface because they pass through only one continu- This system is basically run with two drillpipe darts run
ous pipe ID (Fig. 11-123). Liners are run on the drill- from the surface and two liner wiper plugs run at the
string to the setting depth and typically have a much bottom of the liner setting tool assembly, described in
larger ID than the drillstring above them; as a result, Section 11-6.4.2. Both liner wiper plugs (Fig. 11-126)
a single plug cannot be pumped from the surface. have ports through the middle to accept different sizes
Therefore, the displacement is performed by two dis- of drillpipe darts and are released accordingly. A special
placement plugs. One plug, known as the drillpipe landing collar (Fig. 11-127) is required at the top of the
dart (Fig. 11-124), is located in the surface cementing shoe track to allow cement-slurry bypass after the first
equipment or cementing manifold on the rig floor drillpipe dart and liner wiper plug have reached the
(Section 11-6.4.1). The plug is located between two landing collar.
inlets in the manifold and is kept in position by a retract- This plug system is also used in special two-stage
ing pin inside the plug-dropping head. The second plug cementing operations in which the formation cannot
is either attached to the bottom of the liner setting tool withstand the hydrostatic pressure of the cement
assembly, as described in Section 11-6.4.2, or is attached column in a single stage. In these applications, a
to the top of the liner hanger assembly. This second plug cementing valve (Fig. 11-128) is placed in the liner
is called a liner wiper plug (Fig. 11-125). It has a port string. This allows a short first-stage cement job to be
through the middle ID to allow fluid to be pumped performed. The cementing valve is then opened above
through it (Fig. 11-123A). the first stage of cement. After the first-stage cement
After the cement has been pumped in the liner and has set, the second-stage cement slurry is displaced on
the drillstring, the drillpipe dart is released from the sur- top of the first stage, thus decreasing the total hydrosta-
face cementing equipment by retracting the pin in the tic pressure.
cementing head. The drillpipe dart is pumped as a sepa- Sometimes the liner sizes are so small that the ID
rator behind the cement and in front of the displace- through the liner is equal to or smaller than the ID of the
ment fluid (Fig. 11-123B). drillstring on which the liner is run. In these cases, the
When the drillpipe dart reaches the top of the liner, it plugs behave similarly to long-string plugs in that they
latches into the liner wiper plug. This plugs the center can be dropped from the surface without having a liner
hole (ID) of the liner wiper plug, and both plugs become wiper plug at the top of the liner or the bottom of the
a single divider between the cement and the displace- running string. The cement-slurry volume pumped
ment fluid. (Fig. 11-123C). through such liners is typically small. Any contamination
Both plugs then travel down the liner to the landing of this small volume of cement may compromise well-
collar in the shoe track (Section 1.4.5) and displace the bore integrity, and separating the cement from spacers
cement into the annular space between the OD of the and displacing fluids is especially important. The system

416 Well Cementing

 Drillpipe dart

Drillpipe
dart

Drillpipe
dart

Liner Liner Liner

wiper wiper wiper
plug plug plug

Drillpipe
dart

Liner
wiper
plug

A B C D

Fig. 11-123. Drillpipe dart and liner wiper plug displacement sequence
(drawings courtesy of Baker Oil Tools).

Chapter 11 Cementing Equipment and Casing Hardware 417

Fig. 11-124. Drillpipe dart (drawing courtesy of Baker Oil Tools).

Fig. 11-126. Dual liner wiper plug system (drawing courtesy of Baker
Oil Tools).

Fig. 11-125. Liner wiper plug (drawing courtesy of Baker Oil Tools).
Fig. 11-127. Landing collar for dual liner wiper plug system (drawing
courtesy of Baker Oil Tools).

consists of two drillpipe darts and a special slimhole the well. A pressure increase allows verification of the
landing collar (Fig. 11-129). The lead pumpdown plug or displacement volume. Increasing the pressure to a pre-
drillpipe dart is dropped from the surface in front of the determined level will open the bypass ports in the slim-
cement slurry, and the second pumpdown is dropped at hole landing collar above the pumpdown plug, allowing
the tail of the cement slurry. The first plug separates the the cement slurry to enter the openhole section. Finally,
cement from the wellbore fluid, is pumped through the the second pumpdown plug is dropped behind the
drillstring into the liner, and lands in the slimhole land- cement slurry and lands on top of the slimhole landing
ing collar at the top of the shoe track near the bottom of collar insert, indicating the completion of the cement job.

418 Well Cementing

 11-6.6 Liner selection
For many years, the oil and gas industry has been search-
ing for a method to reliably determine the correct liner
equipment for a particular situation. There are as many
variables that can influence liner equipment selection
as there are to predict liner hanger capacity, as
described earlier. Some of the variables that greatly
influence the liner equipment selection are discussed
below.
Available equipment in a liner versus casing size.
Some liner models and designs are not available in all
sizes. Alternatively, the annular space between the
casing ID and the liner OD may be so small that even
coupling-type liner joints do not fit. In these cases, some
liner equipment designs are not feasible. In addition,
wellhead restrictions at the top of the well can prevent
certain designs from being run.
Available equipment in the area within the
expected installation date. Although a wide variety of
liner equipment is available around the world, the most
suitable equipment may not be available at a particular
location. In these cases, a compromise must be made,
and the best possible alternative will have to be deliv-
ered to the rig site to meet the expected installation
Fig. 11-128. Cementing valve for two-stage liner cement jobs (draw-
ing courtesy of Baker Oil Tools). date.
Circulation rates and temperature and pressure
requirements. In some cases, the liner equipment is
installed in wells in which the expected temperatures
are well beyond 500°F [260°C]. Such well conditions
exceed the temperature ratings of most sealing materi-
als, making it difficult to chose most conventional
hydraulically operated tools. Pressure requirements at
some liner setting depths are beyond the rating of con-
ventional hydraulic setting cylinders. Also, the combina-
tion of high circulation rates, drilling-fluid densities, vis-
cosity, and temperatures prevents some conventional
liner top packers from being installed in the same trip as
the liner and may require a second trip for installation
(tieback packer).
Well conditions and geometry. Some operators
prefer hydraulic equipment over mechanical equipment
because of its ability to rotate in case the liner does not
slide to the bottom on its own. However, in some wells,
the top of the drilling fluid is at an unpredictable depth
or the drilling-fluid density is not homogeneous. These
situations make it difficult to determine the hydrostatic
pressures of these fluid columns and operate hydrauli-
cally activated tools downhole.
Drilling rig capabilities and available running
string. Some liner installations require special opera-
tions to be performed by the drilling rig. Some rigs do not
have these capabilities and make the proper liner selec-
Fig. 11-129. Slimhole landing collar (drawings courtesy of Baker Oil tions impossible. If coiled tubing, small-size drillpipe, or
Tools).

Chapter 11 Cementing Equipment and Casing Hardware 419

threaded tubing is used in large hole sections, insuffi- others choose hydraulic equipment. Such preferences
cient weight may be available to set the liner top pack- may be based upon previous experience. Last but not
ers or allow rotatation to activate mechanically set least, local regulations may prevent some tools from
equipment. being selected or imported.
These are just a few of the parameters that can influ- Despite all of the variables discussed above, liner
ence liner equipment selection. Of course, the prefer- selection flowcharts provide preliminary guidance con-
ences of operators and drilling and completion engi- cerning the proper liner equipment to use. An example
neers also enter into the equation. Some operators may is shown in Fig. 11-130.
want to use only mechanically set equipment, while

Is a liner hanger
required?

Yes No

Liner top packer

required?
Mechanical set Hydraulic set
Yes No

Rotational Rotational
Standard capabilities Standard capabilities High end
applications during applications during performance? Liner setting
cementation cementation sleeve

Single Yes No
cone, Rotational and push/pull capabilities
single jay Expanding before running?
Low- to Single Single Single metal Conventional
elastomer
medium-end cone cone cone seal
applications Single Yes No
cone,
multijay

Hydraulic Right-hand manipulation required

to activate hanger?
release
Medium- to
high-end Multicone Multicone Drill-down
applications
capabilities Yes No

High to High-load High-load High-load High-load Right-hand Left-hand

extreme end cage type cage type cage type cage type Yes No set sleeve set sleeve
applications
Hydraulic release Hydraulic
drill down release
Drill-down
capabilities

Tie-back extension required

Wiper plug (always required for packer)?
systems Drill down
Yes No
Standard
wiper plug PBR
extension
Packoff
Four-plug system Standard
extension
system
Slick PBR RS Special
Cup seal extension
system stinger packoff retrievable
Single-plug assembly system seal system
slimhole

Landing Shoe track

Two-plug collar
slimhole

With ball seat to Single Double

Without ball valve valve
activate hydraulic
seat
tools

Four-plug Wash/
Single Standard system Slimhole For For Four- For slim- Drill
plug with plug system plug with single standard- plug hole Standard ream
with ball down down
ball seat with ball seat seat ball seat plug plug system system plug

Fig. 11–130. Liner selection flow chart (courtesy of Baker Oil Tools).

420 Well Cementing

11-7 Remedial cementing tools
Remedial cementing tools are mechanical or hydraulic Workstring
devices used downhole to assist in the placement of
cement during plugback or squeeze cementing opera-
tions. They generally isolate areas of the casing from
treating pressures and cement. Some are available in
retrievable or drillable designs, each being suited for a
particular set of well conditions. Remedial cementing
Expansion joint
tools are generally provided with service. Details of a
specific tool’s operation or limitations should be
obtained from the service company or manufacturer. Tubing tester

11-7.1 Cased hole squeeze packers

Squeeze packers are primarily used to isolate the upper
portion of the casing and wellhead from cement
and squeeze pressures and to improve control and place- Unloader valve
ment of fluids during squeeze cementing operations. They
are available in either drillable or retrievable types
and differ considerably in appearance and operation.
Retrievable squeeze packers may be set and released
repeatedly; they are most useful in multiple setting opera-
tions, such as selective testing and cementing of multiple
zones. Squeeze packers are run on tubing and are avail-
able in compression-set (Figs. 11-131A and 11-131B) or
tension-set designs (Figs. 11-132A and 11-132B).
Compression-set packers are usually more versatile
and are therefore preferable whenever sufficient tubing
height is available. Generally 10,000 to 20,000 lbf [44.5 to Compression-set
89.0 kN] is the minimum required compressive or tensile packer
force to pack off the elements of either compression or
tension packers. Most compression packers are set by
slightly lifting the tool, rotating to the right (one-quarter
turn at the tool), and setting down to apply the neces-
sary force. This operation frees the lower slip assembly
to slide over the lower cone and engage the casing wall.
As the tubing is lowered, the elements are compressed
until they are packed off against the casing. Compression
packers are released by simply raising the tubing.
Tension packers are usually set at depths less than
4,000 ft [1,219 m] and are set by running the packer 20 ft
A B
[6 m] past the setting depth and then picking back up to
setting depth. This procedure prevents buckling caused Fig. 11-131. Compression-set service packers (drawings courtesy
by running in the hole with the tubing in compression. of Baker Oil Tools).
Next, right-hand rotation (three quarters of a turn) is
applied to the workstring at the tool. To ensure the tool
is set, slack off (release the hanging weight at the well- the operator should slack off 10,000 lbf or the string
head) 10,000 lbf, or the weight of the string if 10,000 lbf weight. This will open the bypass and equalize the pack-
is unattainable. After verification that the tool is set, the ing elements. Once equalized, the string weight is taken
operator can pull 25,000 lbf [111 kN] at the tool to close to neutral and the workstring is rotated three quarters of
the bypass and pack off the elements. The tool is now set a turn. The packer is now unset and free to come out of
and ready to perform its job. To unset the tension packer, the hole.

Chapter 11 Cementing Equipment and Casing Hardware 421

 ting the packer. Also, the provision of a bypass area
Workstring beneath the elements reduces swabbing and piston
effects while running and allows reverse circulation of
excess cement slurry without exerting excessive pres-
sures. Packers with full-opening mandrels permit the
use of through-tubing perforating guns, pressure
recorders, or other wireline accessories.
Drillable squeeze packers, commonly referred to as
Tubing tester cement retainers (Fig. 11-133), may be set on wireline or
tubing. They are generally made of cast iron and are
made compact to minimize drilling time. A sliding sleeve
or poppet valve is provided to control slurry placement
and preserve final squeeze conditions. Sliding sleeve
valves are operated by raising and lowering the tubing.
They prevent flow in either direction. Cement retainers
Unloader are often used instead of retrievable packers to prevent
backflow of cement when dehydration is not expected as
well as to isolate the treated area from pressures
because of the reversing of excess cement from the
tubing. Cement retainers are also better suited to situa-
tions in which potential communication with upper per-
forations or casing problems that may lead to cementing
a retrievable packer in the hole.
Cast-iron cement retainers are set using any of three
methods. When accurate depth control is an issue, a
wireline-set cement retainer (Figs. 11-133A and
11-133B) is deployed with an adapter to connect the
Tension-set cement retainer to the wireline setting tool. The cement
packer retainer is lowered to the proper position and set by
electrically firing a slow-burning charge in the setting
tool. When the cement retainer is completely packed off,
the setting tool shears free and is retrieved with the
wireline. To cement through the retainer, a stinger is run
on tubing, drillpipe, or coiled tubing and inserted into
the cement retainer.
The second method of setting is on drillpipe or
threaded tubing (Fig. 11-133C). Either tubing or
drillpipe is used because of the need to rotate the tubing
to set the cement retainer.
Third is the coiled tubing setting method
(Fig. 11-133D). The cement retainer is connected to a
A B
tubing setting tool, and a valve is opened to allow the
Fig. 11-132. Tension-set service packers (drawings courtesy of tubing to fill as the cement retainer is lowered. The valve
Baker Oil Tools). is pushed open by lowering the tubing and closed by rais-
ing the tubing. Hydraulic pistons push out the top slips,
allowing the retainer to be set by pulling up on the coiled
Retrievable squeeze packers usually include many tubing. The stinger is connected and run in the hole with
design features to improve performance and versatility. tubing to perform the squeeze. Rotating the tubing to
Most compression packers use hydraulic hold-down slips the right releases the upper slips and initiates packoff in
to resist upward forces generated by the treating pres- some models. The tubing is then pulled to complete the
sures below the packer. Some packers provide a bypass packoff. When the proper setting tension is achieved, the
valve that may be opened or closed while in the set posi- setting tool shears free. The setting tension may range
tion. This feature provides more accurate control of from 18,000 lbf [80 kN] for 41⁄2-in. sizes to 48,000 lbf
slurry placement by allowing circulation without unset- [214 kN] for 95⁄8-in. sizes.

422 Well Cementing

 Workstring Workstring
Coiled tubing or
threaded pipe

Crossover

Wireline
setting
tool

Mechanical
setting tool

Stab-in tool
Hydraulic
setting tool

Wireline
adapter
kit

Wireline-set Mechanically
Wireline-set cement set cement
cement retainer Hydraulically
retainer
retainer set cement
retainer

A B C D
Wireline Wireline Drillpipe Coiled tubing
set set set set
Fig. 11-133. Drillable cement retainers (drawings courtesy of Baker Oil Tools).

Chapter 11 Cementing Equipment and Casing Hardware 423

11-7.2 Cased hole bridge plugs However, tension packers are not recommended for
Bridge plugs are normally used to isolate the casing tandem running operations with certain bridge plugs,
below the zone to be treated. When set, bridge plugs act because their combined operations may prevent freeing
as solid barriers to prevent flow and resist pressure from the tension packer, locking the entire assembly down-
above or below. They are available in drillable and hole.
retrievable designs for wireline or tubing-set operations. Retrievable bridge plugs are available in cup or
Retrievable bridge plugs are often used in multiple zone packer configurations. Cup bridge plugs (Fig. 11-134D)
applications, because they can be set and released as are generally used in shallow, moderate pressure appli-
often as necessary. They may be run in tandem with cations. They are simpler and more economical than
retrievable packers for single-trip straddle operations. packer models. However, the cups are in constant con-

Workstring Workstring
Tubing

Retrieving Electric
head line setting Retrieving
tool head

Retrieving
head

Wireline
adapter Latching
kit type
packer
plug

Packing
element
bridge
plug Wireline-set
retrievable
bridge plug
Permanent
packer
Packer cup
bridge plug

A B C D
Packing element Wireline set Latching type Cup style
packer plug

Fig. 11-134. Retrievable bridge plugs (drawings courtesy of Baker Oil Tools).

424 Well Cementing

 packer. Packing element bridge plugs (Figs. 11-134A,
11-134B, and 11-135C) use a packer element design that
is more durable and well suited for deep, high-pressure
applications. They have smaller external diameters that
permit faster running and reduce swabbing and piston
effects.
Retrievable bridge plugs are coupled to the tubing by
a retrieving sleeve and are typically set by rotating to the
right while lowering the tubing to apply force (generally
at least 10,000 lbf [44 kN]). Some models feature a left-
hand rotational set that requires no additional weight.
These may be used in shallow applications or to permit
removal of surface equipment. They are released from
the tubing by pulling up slightly (1,000 lbf [4 kN]) while
rotating to the left (one quarter turn at the tool).
Retrievable bridge plugs are retrieved by lowering the
retrieving sleeve while circulating to remove sand and
debris from above the bridge plug. When solid contact is
made, the tubing is raised to apply a slight upward force
(2,000 to 3,000 lbf [9 to 13 kN]) while rotating to the
right. Once free, the bridge plug may be retrieved or
moved to another position. To prevent cement and per-
foration debris from interfering with the retrieval, sand
should be placed on top of retrievable bridge plugs. Sand
may also be used above drillable bridge plugs to reduce
damaging shock waves caused by closely placed perfo-
rating guns.
Drillable bridge plugs are used to create a temporary
or permanent plug for squeezing or plugback applica-
A B C tions (Figs. 11-136A and 11-136B). They are often used to
Cup style Packing element Wireline-set seal off nonproductive zones or wells to be abandoned.
retrievable style bridge plug retrievable They are made of cast iron and are constructed similarly
bridge plug bridge plug to cement retainers. The basic difference is that the
Fig. 11-135. Retrievable bridge plugs (drawings courtesy of Baker mandrels are plugged and do not contain check valves.
Oil Tools). Drillable bridge plugs may be set on wireline, threaded
tubing, drillpipe, and coiled tubing. Setting on wireline
allows the establishment of zonal isolation without a rig;
tact with the casing while the bridge plug is being run, however, this cannot be done in high-angle wells.
which causes wear and increases swabbing and piston The setting tools and procedures are often the same
effects. Cup bridge plugs may be run on tubing or a cable as those used for cement retainers. Some drillable
(sand line) and are released and retrieved with a retriev- bridge plugs are designed to allow pressures above and
ing sleeve. However, a special retrieving sleeve and pro- below the plug to equalize before drilling through the
cedure are required for sand line operations. When top slips. This feature is most important when gas or
released, the bridge plugs are automatically set by apply- high pressures are expected below the tool.
ing pressure from above or below. Cup bridge plugs
(Fig. 11-134D) may be used in tandem with a tension

Chapter 11 Cementing Equipment and Casing Hardware 425

 Hydraulic
setting tool
Electric line
setting tool

Wireline
Wireline adapter kit
adapter kit

Cast-iron Cast-iron
drillable drillable
bridge plug bridge plug

A B A B
Tubing tester Uploading sub
Fig. 11-136. Cast-iron bridge plugs (drawings courtesy of Baker Oil
Tools). Fig. 11–137. Tubing tester and unloader (drawings courtesy of
Baker Oil Tools).

11-7.3 Tubing testers and unloaders closed by rotating the tubing (one-quarter turn at the
Tubing testers (Fig. 11-137A) are downhole valves used tool) to the right and lifting. They are reopened by simply
to check the tubing for leaks. They are typically used lowering the tubing. A simple ball-seat sub may also be
during squeeze cementing operations, because of the used to test the tubing. However, the ball seat somewhat
potential problems that exist while pumping cement restricts the ID of the tubing, and the ball must be
under moderate-to-high differential pressures. A leaking reverse circulated to the surface before pumping.
connection may permit local cement dehydration, creat- Tubing unloaders (Fig. 11-137B) or tubing bypass
ing false squeeze indications or completely plugging the valves are placed in the tubing string to provide an alter-
tubing. nate passage for circulating or spotting fluids. They are
Tubing testers are typically equipped with a fully open- often used with packers that are not equipped with built-
ing flapper mechanism. The full opening bore permits the in bypass valves and therefore must be unset to permit
use of through-tubing perforating guns or other wireline circulation. Tubing unloaders are operated by raising or
tools. Tubing testers are placed above a packer and are lowering the tubing. They are suitable for tension or
run in the open position to allow filling. They are typically compression packer operations.

426 Well Cementing

11-7.4 Cementing using inflatable service tools
Inflatable service tools have been used in the oil and gas
industry for more than 50 years to perform a variety of
remedial workovers. Inflatable packers, bridge plugs,
and cement retainers are used in open holes, cased Pup joint
holes, slotted liners, and gravel-pack screens, but they
should be used only when conventional tools are not
suitable. Inflatable packers employ a large rubber ele-
ment that is inflated with a fluid to form a seal against
the casing or open hole. They have a much smaller OD
than conventional tools. These packers can be run
through wellbore restrictions and inflated up to 250% of
their run-in OD. J circulating
Inflatable tools are especially useful in open holes of valve
uncertain size. Just like conventional packers and
bridge plugs, inflatable service equipment can be set up
in any array (i.e., retrievable packer, resettable packer,
retrievable bridge plug, and cement retainer), allowing
the same operations to be performed as with conven- Single-set
tional equipment. Inflatable equipment is in neither pressure-
compression nor tension. The pressure inside the inflat- inflatable
packer
able element holds the equipment in place. Inflatable
packers can be run on threaded pipe, drillpipe, coiled
tubing, and wireline, making them just as adaptable as
their conventional counterparts.

11-7.5 Inflatable service packers

Inflatable squeeze packers (Figs. 11-138A, 11-138B, and
11-138C) are used primarily to isolate the upper portion
of the casing and wellhead from cement and squeeze A B C
pressures. They are also used to improve control and Single-seal Double-seal Openhole
placement of fluids during squeeze cementing opera- element element seal
tions. Normally, inflatable squeeze packers are set and
released only once, but they can be configured to allow Fig. 11-138. Inflatable service packers (drawings courtesy of
Baker Oil Tools).
multiple settings. The inflatable service packer is a
retrievable device that may be run on drillpipe, threaded
pipe, or coiled tubing. It is best suited for establishing
zonal isolation in open hole or casing. Because of the
inflatable elements’ large expansion capacities, the
packer can be run through wellbore restrictions and set
in the larger ID below. In such applications, the inflat-
able service packer or resettable service packer is run
through a casing restriction and set above a production
zone to protect the restriction above. The treatment
fluid is spotted to the circulating valve (Fig. 11-139), the
valve is closed, and the treatment fluid is injected into
the formation below.

Chapter 11 Cementing Equipment and Casing Hardware 427

11-7.6 Circulating valve 11-7.7 Inflatable bridge plugs
The circulating valve was designed to open and close the Inflatable bridge plugs are primarily used to establish
tubing immediately above an inflatable packer and to isolation between zones. They can be run on drillpipe,
permit reversing, equalizing, circulating, and spotting of threaded pipe, coiled tubing, and wireline. Because of
acids, cement, or other fluids. This device is similar to the inflatable elements’ large expansion capacities, the
the unloaders discussed earlier. The circulating valve is packer can be run through wellbore restrictions and set
a simple lug and J-slot mechanism with a left-hand-turn in the larger ID below. Inflatable bridge plugs
opening sequence. Shear pins prevent premature open- (Fig. 11-140A) are run with an on-off connector
ing while running in the hole (Fig. 11-139). (Fig. 11-140B). This allows the operator to activate and
deactivate the inflatable bridge plug mechanically with
pipe rotation. If the operator cannot use pipe rotation, a
hydraulic release running tool (HRRT) (Fig. 11-140C)
can be used. The HRRT is used to isolate a lower zone or
for zonal abandonment. This device can be run on wire-
line, coiled tubing, threaded pipe, or drillpipe. A
hydraulic disconnect is ideal for use with the wireline
setting tool (Fig. 11-141) or coiled tubing if rotation is a
problem. It is ideal for openhole abandonments and is
simple to use. A variety of accessories can be used with
these products to aid in setting, retrieving, circulating,
spotting, and treating.

Tubing

Fill-up
sub

On-off tool/
retrieving
head

Inflatable
Fig. 11-139. Circulating valve (drawing courtesy of Baker Oil Tools). bridge
plug

A B C

Fig. 11-140. Inflatable bridge plug and on-off connectors (drawings

courtesy of Baker Oil Tools).

428 Well Cementing

11-7.8 Through-tubing electric wireline setting tool a poppet-style check valve. At a predetermined pressure,
The through-tubing electric wireline setting tool the setting tool hydraulically disconnects from the
(EWST) (Fig. 11-141) is an electric downhole pump pow- inflated tool, allowing retrieval of the EWST and wireline
ered from the surface by means of conventional wireline. to surface.
The EWST uses fluid from either the wellbore or a reser-
voir carrying system to inflate and pressurize inflatable 11-7.9 Inflatable cement retainer
packing elements. The setting tool incorporates a mag-
Combining a flapper-valve assembly (Fig. 11-142A) with
netic casing collar locator to allow accurate depth con-
a permanent inflatable bridge (Fig. 11-142B) creates a
trol. The operation of the system is tracked at surface by
cement retainer. Cement retainers are usually used to
monitoring the amperage draw of the tool on the power
squeeze off unwanted production or gas channels
supply. When used with an inflatable packing element,
between the open hole and casing. The bottom bull plug
fluid is drawn into the setting tool and filtered before
is removed and replaced with a shear-out ball seat. The
entering the pump section. The pump section forces the
lift sub on top is replaced with the valve assembly.
fluid into the inflatable element, where it is trapped by
An inflatable cement retainer allows cement to be
pumped into channels. Once the cement is in place, the
hydrostatic pressure is relieved by pulling out of the
retainer. Once out of the retainer, a valve closes and does
not allow further squeezing.

A B
Inflatable cement Permanent
retainer inflatable
Flapper valve bridge plug
assembly

Fig. 11-141. Electric wireline setting tool (drawing courtesy of Fig. 11-142. Inflatable cement retainer and permanent inflatable
Baker Oil Tools). bridge plug,

Chapter 11 Cementing Equipment and Casing Hardware 429

11-7.10 Software for inflatable tools
Most tool manufacturers provide job-design software for
operations involving inflatable products. An example is
shown in Fig. 11-143.

Prejob
Information

Volume Calculations

Element Differential Pressure

Interpolation Spreadsheet

Element
Hydraulics

Performance Envelope

Fig. 11-143. Software for inflatable tool job design (courtesy of Baker Oil Tools).

430 Well Cementing

11-8 Through-tubing inflatable products
In the early 1980s, coiled tubing began to play an impor-
tant part in the workover industry. Many operators Coiled tubing
sought an efficient way to perform workovers without
killing the well. Coiled tubing was the answer.
In the past, the only way to perform a kill-free Slip-type coiled
workover operation was to use a snubbing unit. This tubing connector
technique was slow, required many safety precautions,
Fluid loading valve
and presented environmental risks in offshore locations.
Another obstacle of snubbing operations was that
they had to be performed without pulling the production Dual flapper type
backpressure valve
tubing and without pipe rotation. Many companies
began developing tools that could be run through the
Ball operated secondary
production tubing and set in the casing below. This hydraulic release
required a unique tool—one that could expand up to
300%, withstand elevated temperatures, and be as
durable as possible. Through-tubing packers were born Tension-operated
from this need, and today any job one can do on a con- equalizing/circulating
ventional rig can be done using coiled tubing (Graves et valve
al., 2004).

11-8.1 Through-tubing inflatable retrievable Retrievable

packer packer
The through-tubing inflatable retrievable packer
(TTIRP) (Figs. 11-144A and 11-144B) provides a means
of performing remedial and stimulation operations with-
out removing the production tubing (Mackenzie and
Plante, 1999). This versatile tool is typically conveyed
into the wellbore with coiled tubing and allows the oper-
ator to selectively perform operations on zones within
the wellbore. Typical applications for the retrievable A B
packer are remedial cementing, acid stimulation, chem-
ical water shutoffs, selective production testing, Fig. 11-144. Through-tubing retrievable packers (drawings
courtesy of Baker Oil Tools).
hydraulic fracturing and pressure testing operations.
Setting of the TTIRP is performed by placing a setting
ball in the coiled tubing at surface and pumping it to the
ball seat at the bottom of the coil. Once the ball is on Pulling the TTIRP through the tubing end is not a
seat, pressure is applied and the element is inflated to problem because the element has been compressed to
the tubing ID and held for 20 to 30 min. Once the maxi- its run-in size. If the packer becomes stuck in the
mum inflation pressure is reached, any additional pres- tailpipe, a hydraulic disconnect can be activated by
sure will shear the ball sub, resulting in a pressure drop. dropping a larger ball to disconnect from the packer and
The pressure drop causes a poppet inside the element to allow the coiled tubing to be pulled from the well. The
close and maintain the pressure inside the element. Now packer left behind has a fishing profile that is compati-
the operations can be performed. When the job is com- ble with through-tubing fishing tools. Pressure inside the
plete and the operator is ready to pull equipment out of element is predetermined using a software application
the hole, a larger ball is pumped to shear out an equal- (Section 11-8.5) that plans the job and tells the operator
izing seat. With the packer equalized, the packer can be how much fluid will be required.
retrieved. A straight pull is all that is needed, and the
release pins will shear and allow a slot in the mandrel to
be pulled across the deflate ports. The packer is then
deflated and pulled from the wellbore.

Chapter 11 Cementing Equipment and Casing Hardware 431

11-8.2 Through-tubing inflatable retrievable 11-8.3 Through-tubing inflatable permanent
bridge plug bridge plug
The through-tubing inflatable retrievable bridge plug The through-tubing permanent bridge plug (TTPBP)
(TTRBP) (Figs. 11-145A and 11-145B) is typically con- (Figs. 11-146A and 11-146B) is a nonretrievable inflat-
veyed into the wellbore with coiled tubing or EWST (Fig. able bridge plug used to permanently plug the casing
11-141), allowing for all the inherent advantages of below the production tubing. Applications include: per-
working over a well with the completion in place and not manent water shutoff, permanent shutoff of a lower
having to kill the well. Applications for the TTRBP are zone, and well abandonment. The TTPBP can be run on
the same as those for any standard retrievable bridge coiled tubing, threaded pipe, or wireline using an EWST
plug. The TTRBP may be retrieved by means of threaded (Fig. 11-141A).
pipe, coiled tubing, or slickline, and equalized, released,
and retrieved in a single run.

Coiled tubing
Electric wireline
Electric cablehead
crossover Slip-type coil tubing connector
Dual flapper-type
Reservoir system backpressure valve
electric wireline
setting tool Ball-operated secondary
hydraulic releaser

Nipple profile locator

Ball operated circulation valve

Test pressure
equalizing valve Inflation valve

Hydraulic disconnect running tool

Pressure extender

Belleville-type
secondary pull release
Hydraulic disconnect
running tool Permanent bridge plug

Retrievable bridge plug

A B
Fig. 11-146. Through-tubing permanent bridge plug and completion
(drawings courtesy of Baker Oil Tools).
A B

Fig. 11-145. Through-tubing retrievable bridge plug (drawings

courtesy of Baker Oil Tools).

432 Well Cementing

11-8.4 Through-tubing inflatable permanent 11-8.5 Through-tubing inflate design software
cement retainer Job-design software is available for through-tubing
The through-tubing cement retainer (TTCR) (Figs. 11- inflate operations (Fig. 11-148). The program was devel-
147A and 11-147B) allows permanent isolation and oped to help in the information gathering, engineering,
cementation of a lower zone without pulling the produc- execution, and postjob reporting stages of an isolation
tion tubing and packer. The built-in retrievable spotting planned with through-tubing inflate technology
valve (Fig. 11-147B) provides a means of spotting (Mackenzie and Patterson, 2004).
cement to the top of the TTCR so that unwanted fluids
are not pumped into the formation. Because the TTCR
and spotting valve operations do not require rotation, 11-9 Acronym list
the tool can be run on coiled tubing or threaded pipe. ACP/ECP Annular casing packers/external casing
The TTCR houses opposing flapper valves, negating the packers
chance of cement contamination after release of the
running string from the retainer. API American Petroleum Institute
BWOC By weight of cement
CRA Corrosion resistant alloy
DNV Det Norske Veritas
EWST Electric wireline setting tool
Coiled tubing
HRRT Hydraulic release running tool
ID Inner diameter
LCM Lost circulation material
Slip-type coiled
tubing connector OD Outer diameter
Dual flapper-type PBR Polished-bore receptacle
backpressure valve
PDC Polycrystalline diamond compact
Ball-operated
secondary TD Total depth
hydraulic release TTCR Through-tubing cement retainer
Nipple profile TTIRP Through-tubing inflatable retrievable
locator packer
TTPBP Through-tubing permanent bridge plug
Flow actuated
inflation valve TTRBP Through-tubing inflatable retrievable
bridge plug

Cement retainer
spotting valve
11-10 Suggested reading
Baker Hughes Inc., http://www.bakerhughes.com/baker-
hughes/ (accessed May 23, 2006).
Permanent
cement
Davis-Lynch Inc., http://www.davis-lynch.com/ (accessed
retainer May 23, 2006).
Halliburton Co., http://www.halliburton.com/ (accessed
May 23, 2006).
Schlumberger Ltd., http://www.slb.com/ (accessed May
23, 2006).
Weatherford International Inc., http://www.weather-
ford.com/ (accessed May 23, 2006).

A B

Fig. 11-147. Through-tubing inflatable cement retainer (drawings

courtesy of Baker Oil Tools)

Chapter 11 Cementing Equipment and Casing Hardware 433

 Inflatable Element
Well Condition 1

Prejob
Information

Volume Calculations

Performance
Envelope

Element Differential Pressure

Interpolation Spreadsheet

Fig. 11-148. Job-design software for through-tubing inflate operations (courtesy of Baker Oil Tools).

434 Well Cementing

12-1 Introduction
The previous chapters have shown that there are many
facets to a cementing operation. This chapter provides
Primary Cement Job Design

12
Bernard Piot—Schlumberger

expected life of the well. A fourth category of concern

must therefore be added: One must carefully consider
any parameter that may affect the integrity of the
guidance on construction of a well-specific cementing cement sheath and the quality of isolation during the
program assessing the key elements of slurry selection well’s productive lifetime and after abandonment. This
and cement placement. The engineer must consider is particularly important in certain environments in
data from a variety of sources and optimize the parame- which the cost of remedial intervention to repair a
ters of the cementing operation according to the well defective primary cementing job or restore zonal isola-
conditions. This chapter describes in detail how to sys- tion may be very high.
tematically analyze the data to align the design of the
cementing operation with its operational and economic
objectives.
12-2.1 Depth and dimensional data
The information that must be gathered when planning a
primary cementing operation includes
■ vertical depth
12-2 Problem analysis
■ measured depth
Every cementing operation has a set of specific objec-
tives combining technical considerations (ranging from ■ angles and azimuths of deviation
simple fill-up criteria to sophisticated isolation require- ■ casing size and weight
ments) with project economics and local regulations. ■ openhole size
For the cementing operation to fulfill its objectives,
■ string type (e.g., full string, liner, tieback, and multi-
the design engineer must first consider three basic types
of well data. stage).
■ Depth and dimensional data Depth data are particularly important because they
■ Wellbore environment, including pressure regime and strongly influence the temperature, fluid volume, and
drilling fluid engineering hydrostatic and dynamic pressures. High-angle wells
■ Temperature regime
are particularly challenging and require strict control
of mud displacement and slurry stability (Keller et al.,
These data dictate the cement properties and the dis- 1983).
placement regime for a given well. The annular configu- No well is exactly vertical unless directional drilling
ration and the pore and fracture gradients play a large techniques are available. Figure 12-1 shows a three-
role in determining the flow regime, because they dimensional (3D) survey of a conventional well that was
restrict the density and rheology of the fluids and conse- considered vertical. Most boreholes twist and turn in
quently influence their pump rates. various directions. Unless the exact hole trajectory is
Wellbore conditions also indicate whether special known, the relatively flexible casing string will not be
materials must be considered owing to the presence of adequately centralized and may touch or lie on the bore-
gas, salt, or other influences. These factors, along with hole wall at various locations.
the pressure and temperature profiles, guide the selec- Chapter 5 illustrates the dramatic adverse influence
tion of additives for controlling the slurry-flow properties of eccentered pipe on mud-removal efficiency.
and setting behavior. Performing a detailed directional survey, or 3D survey,
All the data types listed above are related to the state of the borehole (measured depth, deviation angle, and
of the well at the time of drilling. However, the objective azimuth) over the entire interval provides sufficient
of well cementing is to not only place cement correctly knowledge of the well path to ensure proper centraliza-
in the annulus after a section is drilled but also to pro- tion, mud removal, and zonal isolation. Drilling with bot-
vide zonal isolation between the respective zones for the tomhole assemblies containing measurement while

Well Cementing ■ Chapter 12 Primary Cement Job Design 435

 East

North

Magnetic north

Ring size 0.5 m

Fig. 12-1. 3D survey of a "vertical" well (from Maeso and Tribe 2001; reprinted with permission of SPE).

drilling (MWD) or logging while drilling (LWD) tools is The first calipers had two or three arms. If a true
becoming more common. Thus, such information is often caliper log was not available, an estimated hole size
readily available. could be derived from measurements recorded by a
When planning the centralization of a casing in a single-arm system used to convey logging tools. However,
“vertical well” without 3D survey data available, the hole sizes calculated from such measurements were
design engineer should always use a minimum deviation highly inaccurate if the well was deviated or if the size
of a few degrees (usually 3°) to account for uncertainties and shape of the hole were irregular.
in trajectory. Alternatively, some use randomization Changes in hole size and shape often occur in one
techniques to introduce variations in trajectory between preferred direction, usually perpendicular to the mini-
two survey points. mum stress in the formation. Stress-induced shape
In principle, the openhole size is dictated by the drill- changes can result in highly oval wellbores in which two-
bit size. Along with the casing size and type, the open- and three-arm calipers are inaccurate. A four-arm tool
hole size should be selected according to the expected was introduced to measure these holes accurately
well conditions and completion configuration. In an (Fig. 12-3). Four-arm calipers, also called X-Y calipers,
actual well, the open hole is rarely “gauge” (i.e., per- provide rough hole-shape measurements and more accu-
fectly round and cylindrical). Soft, unconsolidated zones rate hole-volume estimates than single-axis calipers.
or formations containing shales tend to be unstable. If However, real-world boreholes are never truly round or
the formation stresses and the mud density are not bal- truly elliptical. Thus, the X-Y caliper is often coupled
anced, the formation can slough and break into frag- with a pad-mounted tool, such as a dipmeter or one of
ments that can be difficult to remove from the wellbore. the more modern imaging tools.
Variable hole shape affects the required slurry volume Hole size can also be measured ultrasonically. This
and well control. It can also dramatically affect the dis- technique is often used to measure casing corrosion
placement mechanics. Various wireline-caliper tools because of its high resolution. The rotating sensor mea-
have been developed to measure hole size. Drawings and sures the transit time and computes the standoff 140 to
basic characteristics of each are summarized in Fig. 12-2 180 times per revolution. The use of such ultrasonic
and Table 12-1. In general, calipers that record a greater technology is limited by the attenuation of the wellbore
number of independent measurements provide a better fluid.
estimate of hole size and volume (Table 12-2).

436 Well Cementing

 Table 12-2. Caliper Types and Hole Geometry
Round Hole Oval Hole

Two-pad single
measurement

Correct volume Wrong volume

Three-pad single
measurement

Correct volume Volume too small

Four-pad two-axis
measurement

Correct volume Volume OK

Six-pad independent
measurement

Correct volume Best determination

Fig. 12-2. Examples of single-, two-, and three-arm calipers.

Table 12-1. Characteristics of Different Calipers

Number of Arms Phasing of the Arms Max. Diameter Remarks
Density 2 180° 22 in. [560 mm] 2 coupled arms: one reading

Sonic 3 120° 16 in. [406.4 mm] 3 coupled arms: one reading

Borehole 4 90° 40 in. [1,016 mm] 4 arms coupled 2 by 2:

geometry with extension arms; two-axis caliper
30 in. [762 mm] standard

Various imaging 4 90° 17.5 to 21 in. 4 arms coupled 2 by 2:

tools [444.5 to 533.3 mm] two-axis caliper

Environmental 6 60° 40 in. [1,016 mm] 6 independent readings

measurement with extension arms;
sonde 30 in. [762 mm] standard

Ultrasonic imaging Ultrasonic Rotating sensor 12 in. [317.5 mm] 140 to 180 readings per revolution;
tool limited by mud attenuation

Chapter 12 Primary Cement Job Design 437

 fluid, containing an easily identifiable material that is
detected when it returns to the surface. A drawback of
this method is that it requires at least one full circula-
tion cycle. Additionally, it does not account for the dis-
persion of the tracer in the entire mud volume and
assumes that all of the drilling fluid is mobile and being
circulated.
With the advent of MWD and LWD tools, openhole size
can be derived from acoustic measurements (Orban et
al., 1991; Minear et al., 1996; Maranuk, 1997), nuclear
measurements (Paske et al., 1992), or resistivity mea-
surements in conductive drilling fluids. These measure-
ments provide more information about borehole stabil-
ity when made during both drilling and pipe-tripping
operations.
Combinations of LWD measurements are particularly
useful for detecting borehole-stability and trajectory
problems. Such early diagnostic information allows the

Ovalization Depth
Long axis diameter (ft)
10 0 10
(in.)
Short axis diameter
10 0 10
(in.)
Bit size
10 0 10
(in.)

4.59 7.22
3.63 8.52

XO25 5.53 9.67

4.46
4.22
6.13 9.09 9.43

4.93 10.32
Fig. 12-3. Four-arm calipers and ultrasonic tool.
3.82
6.79 13.88 10.07

6.11
XO50 4.86
In recent years, a sonde using six independent arms 7.86 10.37 8.94
has dramatically improved the determination of the pre- 5.18 8.68
cise borehole shape. Such calipers provide extensive 5.20
9.78
drilling and petrophysical information. An ovality algo- 5.59 7.33

rithm helps determine the shape of the borehole for 5.85

5.32 4.58
stress-analysis studies, an important element for pre- 10.27
XO75 5.93 4.63
dicting wellbore stability.
4.94
In many areas, however, wireline caliper tools are not 5.20 5.03
3.69

run, particularly in the large surface-hole sections. It is 4.88 4.17

therefore a common practice to specify a given percent-

age of excess cement slurry to ensure the required annu-
lar fill-up. This widely accepted field practice is not with-
out drawbacks (Section 12-4), because this “excess” is
largely based on local offset analysis, including geology,
mud, and drilling practices.
In the absence of any hole-size measurement, the
“fluid caliper” (Stringer and Barrett, 1990) offers an
alternative method to estimate the volume of the open
hole. A fluid caliper uses a small volume (pill) of tracer
Fig. 12-4. The six-arm caliper tool and log.

438 Well Cementing

mud and drilling engineers to propose remedies and Average Ultrasonic
take immediate action before the problem worsens, Standoff
resulting in a better-quality openhole section. For exam- Bottom Left
ple, the combination of ultrasonic-caliper and density Up Right 3D Textured Image
measurements produces borehole-shape images that 0 (ms) 2.5 0 (ms) 2.5
can be transmitted in real time via telemetry (Maeso
and Tribe, 2001). Figure 12-5 shows the LWD tool, and Overgauge
section
Fig. 12-6 shows an example of the information that drilled by
allows the driller to make sound decisions concerning mud motor
the drilling process.
The integration of the ultrasonic caliper information Pipe worked
and data derived from neutron and density measure- extensively
ments produces a robust and accurate description of the 80 ft
wellbore that can be used by the driller and the petro-
physicist alike (Ortenzi et al., 2002). Given that each of
the caliper estimates derived from LWD measurements
is valid only within a certain operating range and envi- Gauge section
ronment, the integrated caliper uses a simple weighting drilled with
logic based on environmental indicators. rotary steerable
Labat et al. (2002) describe the development of an motor
azimuthal 3D LWD caliper from 16 sector density mea-
surements. This application is useful for drillers, who Fig. 12-6. LWD borehole shape (from Maeso and Tribe, 2001;
can optimize drilling parameters and reduce wellbore reprinted with permission of SPE).
stability problems such as ovalization, washouts, and spi-
raling. In addition, cementers not only have a more
exact estimate of the hole volume from the 16 radial measurements, but also information on the hole shape
and trajectory. This information helps optimize central-
izer placement and drilling-fluid removal, critical ele-
ments of a good cementing operation.
Before selecting a particular casing weight, grade, or
metallurgy, engineers consider the mechanical stresses
Neutron to which the casing will be exposed and take into
detectors account the possible corrosive nature of the formation
Latched coupling fluids. The well-design engineer can use stress-analysis
software to select casing that will withstand the produc-
Azimuthal Neutron source
tion operation conditions. However, the cementing engi-
density
neutron tool
neer must also account for the possible high differential
Electronic pressures across the pipe wall, which may cause the
Pulse carrier casing to collapse or burst. This can occur when there
motor Density source are large fluid-density differences between the cement
slurry and displacement fluid or when there are large
Resistivity Density
tool detectors
final-displacement or casing-test pressures.
Longitudinal forces must also be considered. The dead
Ultrasonic weight of the casing will cause it to stretch. Buoyancy will
detector cause compression. For larger casing sizes, one must take
Steerable measures to ensure that buoyancy will not cause the
motor casing to be pumped uphole during the cementing oper-
Batteries ation. The calculations necessary to ensure this does not
happen are described in Appendix C.
In many primary cementing operations, other factors
prevail. Logistics, availability, and standardization will
Tool bus
often dictate the use of a standard pipe weight and grade
Fig. 12-5. Combined LWD tools. when the well conditions do not impose a much stricter
choice.

Chapter 12 Primary Cement Job Design 439

12-2.2 Wellbore environment Productive formations must be effectively isolated
The specific problems of the openhole interval to be from each other and from nonproducing intervals or
cased require careful evaluation. The presence of pay aquifers to ensure maximum long-term productivity from
zones, overpressured formations, low fracture gradients, the well (Chapter 1). If the formations are known to con-
gas, or massive salt zones presents particular challenges. tain gas, special cement slurries (along with other pre-
Careful job design is required to overcome these well- cautions) may be required to ensure that gas does not
bore-related problems. migrate through the column of setting cement
It is important to know the pressure regime along the (Chapter 9). It is important for the engineer to not only
entire openhole section. Pore pressures are important consider the pay zones, but also any zones that contain
for well control, and such information may be obtained fluids (gas or brine) that may flow through the hydrating
from logs. If logging facilities are not available on loca- cement column.
tion, the mud weight provides a fair indication of the The physical and chemical properties of the mud
maximum pore pressure in any given interval. Even influence cement-job design (Sauer, 1987). Cement slur-
when the mud weight has been chosen for other consid- ries are usually incompatible with drilling fluids. Mixing
erations, such as wellbore stability, it provides an upper these fluids leads to rheological difficulties and poor
bound of the pore pressures in the well. mud removal. To prevent commingling, chemical
If a kick occurs during the course of drilling, it pro- washes, spacers, or other flush fluids are usually pumped
vides important information about pore pressure. At between the drilling fluid and cement slurry
the other extreme, the risk of fracturing the formation (Chapter 5). Such fluids are designed to be chemically
must be carefully assessed, and a mean fracture-pres- compatible with both the drilling fluid and the cement
sure gradient is normally provided for each openhole slurry. OBMs invariably require the use of solvents or
interval. These values are normally based on either surfactants in the spacers or washes to improve compat-
leakoff testing, which is performed upon drilling out the ibility, to remove the oil film from the casing and forma-
shoe of the previous casing string, or a formation tion surfaces, and to leave the surfaces water-wet
integrity test at any particular depth of interest. (Carter and Evans, 1964). Each solvent/surfactant
Other sources of formation-pressure information system is tailored to the particular OBM, particularly its
include the following. base oil.
Complete displacement of the drilling fluid over the
■ Data from formation tests, particularly fluid sampling
interval of interest is a primary objective of any cement-
and pressure readings ing operation. To facilitate removal, the viscosity of the
■ Data from stimulation or squeeze cementing treat- drilling fluid should be as low as feasible. Whenever pos-
ments in offset wells sible, the anticipated downhole properties of the
■ LWD and annular pressure-while-drilling measure- drilling fluid should be used, because these may be sig-
ments [these tools provide an accurate downhole nificantly different from those measured during the
equivalent circulating density (ECD), especially regular mud checks at surface conditions. This is par-
important when drilling high-pressure, high-temper- ticularly true for OBM, for which pressure and tempera-
ature (HPHT) wells with highly compressible oil-base ture play a significant role in its properties.
mud (OBM)] Debris and cuttings should be cleaned out of the
borehole by circulating at an adequate rate for an appro-
Pay zones merit special attention. It is important, for priate amount of time. Numerical simulators are useful
example, to avoid unnecessary damage from excessive to verify that preflushes do not bypass the mud and do
fluid leakoff from the cement slurry. The slurry filtrate is not allow the mud and cement to commingle
ordinarily not a concern because it does not penetrate (Section 12-4). In cases in which 100% mud removal
the formation as deeply as the drilling fluid. In most cannot be assured, a cement slurry may be modified to
cases in which the well is perforated before production, minimize the adverse effects of commingling (Rae and
the charges penetrate the damaged zone. Cement grains Brown, 1988). Data on compatibility are obtained by per-
are too large to penetrate any porous rock matrix. forming laboratory tests in accordance with procedures
Nevertheless, care should be exercised when cementing defined by the International Organization for Standard-
fractured and/or depleted reservoirs to avoid opening ization (ISO) and the American Petroleum Institute
fractures and losing the whole cement slurry deep into (API) (Appendix B), operators, and service companies.
the reservoir.

440 Well Cementing

12-2.3 Temperature regime As part of MWD or LWD technologies, different tem-
The bottomhole circulating temperature (BHCT) and peratures can be measured during the drilling and cir-
the bottomhole static temperature (BHST) must both be culating process. The tool’s temperature probe measures
considered, in addition to the temperature differential the temperature inside the tool to allow necessary cor-
between the bottom and top of the cement column. The rections to the other measurements. Therefore, only
first of these, BHCT, is the temperature that the cement dedicated probes that measure the annular-fluid tem-
slurry will theoretically encounter as it is placed in the perature should be used. It is important to note that this
well. It is the temperature used for HPHT thickening- measurement is not a direct indication of the tempera-
time tests of a proposed cement formulation. The BHCT tures that will be experienced during the cementing
dictates the selection of specific retarders and other operation. The flow contact areas, flow rate, and flow
additives, depending upon their efficacy under such con- regime may be significantly different during the cement
ditions. job. Furthermore, additional heat is generated during
The BHCT is calculated using sets of temperature the drilling process.
schedules last published in ISO 10426-1:2000 or the As for the temperature subs, the main problem with
regional American National Standards Institute LWD/MWD temperatures is that they are obtained from
(ANSI)/API 10A/ISO 10426-1-2001 version (April 2002). different geometries and with different flow rates.
Work performed by the API Task Group on Cementing Unless these conditions can be simulated to calibrate
Temperature Schedules between 1984 and 1991 was the temperature measurements, these temperatures
published in 1999, and resulted in an update of the tem- should not be used as the slurry test temperature.
perature schedules used in Recommended Practice for Computer simulators that model the physics of heat
Testing Well Cements, API Recommended Practice (RP) transfer under dynamic conditions are now widely used
10B (1999). However, the ISO/API standards do not (Guillot et al., 1993). These methods consider all of the
account for unusual well conditions such as the cold well parameters important for heat transfer, including
temperatures found in deepwater wells (Calvert and annular geometry (flow and contact area), fluid rheol-
Griffin, 1998) or the hot temperatures found in some ogy, flow rate, and injection temperature. Therefore,
deviated wells (Wooley et al., 1984). They further assume they can predict temperatures at various points in a well
that the surface-formation temperature is a constant and are especially valuable when cementing deepwater,
standard of 80°F [26.7°C]. A method called the API deviated, or horizontal wells. The temperature simula-
Equivalent Wellbore has been recently proposed to cor- tors have been validated against field data in land and
rect these shortcomings (Kutasov, 2002). This method shallow-water offshore wells (Davies et al., 1994; Merlo
recalculates a pseudolength of wellbore (the equivalent et al., 1994); temperature simulation for deepwater wells
wellbore), and uses a special equation (the Kutasov- (Romero and Touboul, 1998) was validated in 2001
Targhi equation) to calculate the BHCT. The equivalent (Ward et al., 2001). There is a variety of simulators that
wellbore method works best when used for deeper wells consider various degrees of complexity. The cementing
in which the length of the section being investigated is team must agree on the circulation rate and duration
negligible with respect to that of the entire well. The required before cementing, because these two parame-
method does not take into account well geometry, flow ters critically affect the bottomhole cementing circulat-
geometry, or the fluids pumped (rheology, rate, time, ing temperature.
and flow regime for heat exchange). As an example, Fig. 12-7 shows the temperature pro-
Some operators prefer to design cement systems file in the casing and the annulus of a long horizontal
using temperatures actually measured in the well during well (measured depth [MD] = 27,900 ft [8,500 m], true
circulation. One method to obtain such temperatures is vertical depth = 6,900 ft [2,100 m], and BHST = 185°F
to circulate small temperature probes in the mud and [85°C]). The measurements were taken after one com-
retrieve them as they exit the well (Jones, 1986). plete cycle of mud circulation, after which cement was
However, the poor recovery rate of the probes has led to placed. The profile in this example is essentially flat over
abandonment of this technique. Temperature subs a large portion of the open hole, suggesting that the rhe-
(Cloud, 1992) have been used by Honore et al. (1993) in ology of the fluids in the annulus will not change signifi-
conjunction with temperature recorders and by Ward et cantly during placement. On the other hand, the tem-
al. (2001) to record temperatures and temperature pro- perature history curve shows that the cementing
files in wells. However, using simulators requires that circulating temperature is 167°F [72°C]. This tempera-
measurements be correlated with the flow geometry ture is significantly higher than that predicted by the
(drillpipe and hole/casing size) and flow history (mud corresponding API schedules (127°F [53°C]). Such a
rheology, injection rate, and injection periods). discrepancy could lead to significant differences in

Chapter 12 Primary Cement Job Design 441

 Mud Temperature Profile at 510 min
Annulus

100
Geothermal profile
Geothermal profile

Tubular fluids
Tubular fluids
75
Mud
Annulus
temperature
history
(°C) API BHCT

50

25
0 250 500 750 1,000 1,250 1,500 1,750 2,000
Time (min)

10 20 30 40 50 60 70 80 90
Temperature (°C)

Fig. 12-7. Temperature profile and temperature history plots.

cement-slurry behavior. If one chose a retarder concen- ture at the bottom of the well was very low, close to the
tration assuming a circulating temperature of 127°F seabed temperature of 39°F [4°C]. Then the tempera-
[53°C], the actual thickening time in the well could be ture slowly rose and returned to the BHST. The simula-
much shorter than desired and serious consequences tion shows that the temperature would reach 119°F
could result, including loss of the well. [48°C] after 1 day. Such a slow temperature recovery
The BHST is the undisturbed temperature at the could dramatically affect the cement-slurry setting time
bottom of the wellbore. BHST is important for predicting and the development of compressive strength. Using the
the rate of compressive strength development and the normal BHST for strength development tests would lead
long-term stability of a given cement system. It is usually to misleading results. Temperature simulators are thus
calculated from the mean geothermal gradient in the used to derive realistic temperature schedules for per-
area of interest, but it may be estimated from logging forming laboratory tests that most closely simulate the
measurements. Although the ISO/API standards stipulate actual conditions encountered by the cement slurries.
that compressive strength be measured at BHST, com- The temperature differential between the top and
puter simulators can predict the rate at which the well bottom of the cement can also be extremely important
temperature will rise from the BHCT to the BHST. For when embarking upon a cementing design. A cement
critical jobs, following such schedules may be preferable. system that has been retarded for adequate placement
Figure 12-8 shows the temperature recovery at the time at the BHCT may remain liquid or have poor
bottom of a 2,625-ft [800-m] openhole section in a deep- strength development when circulated back to a shal-
water well at 8,200 ft [2,500 m] water depth. The BHST, lower depth in the well. A useful guideline is to ensure
predicted by the geothermal gradient, is 119°F [48°C]. that the static temperature at the top of the cement
The cement slurry was injected through drillpipe as part (TOC) exceeds the BHCT. Sabins et al. (1981) devised
of a stab-in cementing operation without a riser. After similar guidelines based on an experimental study of a
cement placement (429 min in Fig. 12-8), the tempera- number of cement formulations. When it is not possible

442 Well Cementing

 50

45
Geothermal profile
40
API BHCT
35
Tubular fluids
30
Mud
temperature 25
history
(°C)
20

15

10
Annulus
5

0
0 250 500 750 1,000 1,250 1,500 1,750 2,000
Time (min)

Fig. 12-8. Temperature recovery in a deep offshore well.

to meet these criteria, compressive strength tests should ■ cement composition (API class, commercial light-
be performed that simulate the conditions at the TOC. If weight, proprietary blends)
satisfactory compressive strength development cannot ■ project economics
be achieved, it may be necessary to execute the job in ■ well objectives.
more than one stage. Such guidelines provide a simple
means of calculating a suitable depth for the location of
the stage collar.
12-3.2 Compressive strength and mechanical
properties
12-3 Slurry selection Today, compressive strength is not as dependent upon
slurry density. Strong cements can be designed at very
A number of considerations come into play during the low densities by controlling the particle sizes of the
selection of a final slurry design for a specific well appli- slurry ingredients (Chapter 7) (Baret et al., 1996; Moulin
cation. et al., 1997). Also, as discussed in Chapter 8, the practi-
cal importance of compressive strength has diminished
12-3.1 Slurry density (Thiercelin et al., 1997; Thiercelin et al., 1998; Di Lullo
In many cases, the selection of slurry densities is dic- and Rae, 2000). Other mechanical properties such as
tated by factors other than simple pore and fracture Young’s modulus and Poisson’s ratio, previously ignored
pressures. In the past, cement systems were often mixed in the context of well cementing, are now increasingly
at high densities for rapid compressive-strength devel- recognized as important performance parameters.
opment. However, economics may dictate the use of low- However, the industry’s understanding of these para-
density extended or “filler” cements, which provide a meters in the context of well construction is still in its
higher slurry volume per sack of cement. However, the infancy. These criteria address the flexibility of a cement
mechanical properties of the set cements may be infe- system and its ability to withstand temperature, pres-
rior. The following drive slurry density: sure, and tectonic fluctuations during the lifetime of a
well (Baumgarte et al., 1999; Bosma et al., 1999; Le Roy-
■ fracture gradient
Delage et al., 2000; Ravi et al., 2002).
■ pore pressure

Chapter 12 Primary Cement Job Design 443

 Nevertheless, industry and government regulatory persant may be needed to preserve slurry mixability and
bodies have issued guidelines and specifications for pumpability. Some form of cement-particle dispersion or
compressive strengths of cements used for certain appli- deflocculation is always necessary for optimal fluid-loss-
cations. To learn the appropriate rules for the area in additive performance. In addition, fluid-loss additives
which you are working, contact the appropriate regula- may have secondary retarding effects. Certain other
tory agency. Many of these regulations deal specifically additives, notably some of the sugar-base retarders, can
with shallower regions, where concerns center on adversely affect the performance of a fluid-loss additive.
the satisfactory isolation of freshwater supplies. Such considerations are discussed in detail in Chapter 3.
However, guidelines do exist for preferred strengths Even in the absence of fluid-loss additives, disper-
before drilling out (500 psi [3.5 MPa]) and perforating sants may be required for slurry mixability and to reduce
(2,000 psi [14 MPa]). It is therefore important to select friction pressures. Dispersants are particularly valuable
a design that can meet these criteria. Compressive in situations in which annular clearances are small and
strength can also become a critical consideration when high friction pressures may pose some risk to weaker for-
cementing across intervals such as mobile salts or pay mations. When using dispersants, it is important to pay
zones that will require subsequent stimulation close attention to other aspects of slurry performance.
(Goodwin and Phipps, 1984; Rae and Brown, 1988). In Dispersants can act in synergy with cement retarders,
some areas (e.g., deepwater wells with low formation resulting in a longer thickening time. Adding excessive
temperatures) the rate of strength development is often amounts of dispersants can result in slurry instability,
more important than the ultimate strength. leading to high levels of free water and sedimentation. In
some cases, antisettling additives may be required to
counteract this effect. Stiles and Baret (1993) discussed
12-3.3 Well temperature models and a solution to sedimentation and free water
Well temperature is a key consideration in the slurry- observed in cement slurries. They found that the best
selection process. If the temperature at any location in model to describe particle settling in a flocculated
the well exceeds about 230°F [110°C] at any time during cement slurry is based on the consolidation of the solids.
the well’s lifetime, Portland cement systems must be However, the sedimentation of well-dispersed slurries is
modified to maintain adequate cement-sheath stability not well understood.
(Chapter 10). The addition of silica flour is the most Cement retarders are the largest group of additives.
common technique. Alternatively, one can select cement Retarder selection depends upon the BHCT, the type (or
systems that are specifically designed for use in thermal even brand) of cement, and the exact slurry composi-
wells (Brothers et al., 2001; Barlet-Gouédard et al., tion. The reader is referred to Chapter 3 for a complete
2003). At the other extreme, cements used in arctic or discussion.
other low-temperature cementing applications are spe-
cially formulated to generate a low heat of hydration,
thereby minimizing the melting of permafrost (Smith, 12-3.5 Cement-slurry design
1987) (Chapter 7). Slurry design is an iterative process. The “first-guess”
formulation, which is expected to meet the required per-
formance criteria, is initially based upon experience.
12-3.4 Cement additives Local and regional databases and “expert systems”
Fluid-loss additives are generally incorporated in slur- (Kelly et al., 1992) complement this local experience.
ries that traverse pay intervals or in situations in which Knowledge-based systems also facilitate and standardize
the annular gap is small. These materials reduce the rate the slurry-design process.
at which the aqueous phase of a cement slurry leaks into However, the variability of cements demands that
the formation. Loss of water from a cement slurry can actual laboratory testing be performed to verify the pre-
seriously affect its performance, particularly its viscosity dicted results and refine the design. Although this might
during placement and the ability of the slurry column to seem burdensome, experienced engineers and labora-
transmit pressure after placement. It can also damage tory personnel can dramatically reduce the number of
producing intervals (Suman and Ellis, 1977; Bannister, tests required to arrive at the optimal formulation for a
1985), but this is generally considered a minor problem. given set of well conditions. Testing the final design with
Fluid-loss additives should be selected with care, actual field samples is essential to prevent problems
because most of them have secondary effects. For exam- arising from material variability and contamination.
ple, many fluid-loss additives are viscosifiers, so a dis-

444 Well Cementing

12-4 Placement mechanics In addition to chemical compatibility, fluid dynamics
Good mud removal is the single most important require- play an important role in mud removal. Work by Frigaard
ment for a successful primary cement job. This subject et al. (2001) allows the design engineer to set targets for
has been reviewed extensively by Haut and Crook the required rheological behavior of the various fluids in
(1979), Smith (1982), and Sauer (1987) and is discussed the wellbore. Such information can be combined with a
fully in Chapter 5. neural network-based expert system that recommends
the optimal spacer composition (Théron et al., 2002). In
all cases, compatibility testing between the various
12-4.1 Spacers and chemical washes fluids should be performed to prevent unforeseen inter-
Chemical compatibility between fluids pumped in suc- actions that might undermine spacer performance
cession is equally critical to mud displacement. As dis- (Appendix B).
cussed earlier, cement slurries are usually incompatible While precautions should be taken to formulate fluids
with drilling fluids, causing gelation at the mud/cement to efficiently displace the drilling fluid in the annulus, it
interface and reducing the displacement efficiency. To is strongly recommended to use multiple wiper plugs to
prevent commingling, chemical-wash or spacer fluids keep these washes and spacers from intermixing with
are usually pumped between the mud and cement. each other or with the preceding mud, which would
Careful selection of these fluids is mandatory affect their properties and certainly their efficiency.
(Chapter 5). Numerical placement simulators are cru-
cial tools to help determine the best solution. In some 12-4.2 Casing centralization
situations a chemical wash is appropriate. In other situ-
ations, a weighted spacer fluid is necessary to maintain The pipe eccentering problem was effectively addressed
hydrostatic overbalance across active formations by Couturier et al. (1990) and Bittleston and Guillot
throughout the job. (1991). Their work led to the development of modeling
Spacers are normally pumped at densities between tools to optimize mud removal. These tools eventually
those of the mud and cement, because buoyancy forces evolved into a two-dimensional (2D) numerical place-
have been shown to favorably influence the mud- ment simulator that allows precise visualization and
removal process. Brice and Holmes (1963) first identi- quantification of the fluid-displacement process as each
fied the importance of contact time (the time that the fluid moves up and around the annulus. Deviation,
spacer is in contact with a given location in the annulus) eccentering, and caliper measurements are taken into
to mud removal when using the turbulent-flow displace- account to calculate the fluid properties in each cell of
ment process. Their original recommendation of a mini- the simulation grid. Simulators help identify any ten-
mum 10-min contact time for fluids in turbulent flow dency for channels or films to be left on the pipe and
remained an industry standard for many years. However, formation walls.
this is no longer necessarily considered valid. Owing to borehole irregularities, the casing is never
Today there are numerical simulators that account in the center of the open hole. Fluids will naturally flow
for all of the wellbore and fluid parameters. The simula- more readily on the wider side of the annulus. Thus, any
tors ensure that two incompatible fluids will not mix mud-displacement strategy is seriously compromised
during placement in the annulus and that the pumping unless there is adequate casing centralization.
time will be sufficient to remove the film of mud at the Maintaining an API casing standoff above 67% was an
formation wall. As discussed in Chapter 5, these complex early guideline (Chapter 5). The real purpose of this
numerical simulators allow one to assess the impact of guideline was to verify that centralizers conformed to
eccentricity on the efficiency of the planned displace- minimum specifications. It was never intended as a
ment strategy. cement job design target; indeed, channeling occurs at
The spacer composition depends upon the type of 67% standoff. Therefore, it is important to achieve the
mud, the required flow regime (laminar or turbulent), highest possible standoff to facilitate the displacement
the formations that are contacted, and the nature of the process.
cement slurry. For example, freshwater-base spacers are Centralizer friction, casing rigidity, and borehole
used to remove freshwater-base muds, while salt-toler- geometry may limit the number of centralizers that can
ant spacers may be required for salt-saturated muds. be installed. To optimize centralizer placement, com-
OBMs are typically removed with spacers containing sur- puter models (Juvkam-Wold and Wu, 1992) calculate the
factants, organic solvents, or both. Special surfactants position and the standoff of the casing string at each
may be necessary when low-toxicity paraffinic-base point in the wellbore, accounting for well trajectory and
muds are employed. size. A 3D survey and at least a four-arm caliper are rec-
ommended to determine the proper placement of cen-
tralizers in difficult or critical wells.

Chapter 12 Primary Cement Job Design 445

 In situations in which the available data are limited Computer simulators can optimize the centralizer
to the directional survey, the engineer must obtain a distribution for a given required standoff. In the same
best estimate of the washed-out zones. However, new near-vertical well, achieving 75% standoff over the entire
tools are emerging that allow real-time transmission of interval would require increasing the number of central-
drilling data from the rig to an engineering center. The izers to 70 from 51 (Fig. 12-10); however, the centralizers
cementing design engineer can then use simulators to would be distributed in a more effective way. Such an
calibrate hole size, trajectory, and volume for every increase is a minimal expense compared to the cost of
newly drilled section (Queirós et al., 2003). the operation and the risk of not achieving the required
Centralizer calculations must determine the standoff zonal isolation. There are various types of centralizers
not only at the centralizer itself, but at locations available, and their characteristics are discussed in
between centralizers. Casing can bend and sag between Chapter 11.
centralizers, resulting in dramatically lower standoffs.
This problem increases with well deviation. The example
in Fig. 12-9 shows that for a nearly vertical well (3° devi- 12-5 Well control
ation at 6,562 ft [2,000 m]), the standard distribution
Each well offers an “envelope,” or range, of acceptable
of 1 centralizer per 3 joints gives an excellent standoff of
pressures that must be respected to design and execute
90–95% at the centralizer all along the interval. This
a successful cement job. The limiting pressure bound-
requires 51 centralizers if the entire section is to be cen-
aries are normally the openhole pore-pressure and the
tralized. However, the standoff between centralizers
fracture-pressure profiles; however, it is also important
would be as low as 43% near the shoe, compromising
to consider the burst and collapse pressures of the tubu-
mud removal and jeopardizing the objective of achieving
lars.
a good leakoff test at the shoe for drilling the next inter-
Unless sufficient computing power is available, it is
val. The standoff is also as low as 30% in the small
impractical for the engineer to examine the pressures at
washed-out section between 1,100 and 1,200 m, making
each point in the well throughout the entire treatment.
isolation difficult in this zone and in the interval above it.
For this reason, a good approach is to perform a worst-

0 0

At centralizer

500 500

Horizontal
Between departure
centralizer
TVD 1,000 1,000 TVD
(m) (m)

1,500 1,500

2,000 2,000
25 50 75 100 Well 0 30 60
Pipe standoff (%) Horizontal departure (m)

Fig. 12-9. Standoff in a near-vertical well with 51 centralizers (1 per 3 joints).

446 Well Cementing

 0 0

At centralizer
500 500
1⁄ 3

Horizontal
Between departure
centralizer
TVD 1,000 1,000 TVD
(m) (m)
2⁄1

1⁄ 3

1,500 1,500
2 ⁄5

1⁄ 2

2,000 2,000
25 50 75 100 Well 0 30 60
Pipe standoff (%) Horizontal departure (m)

Fig. 12-10. Optimized centralizer distribution with 70 centralizers.

case scenario analysis, which allows the engineer to tion can reduce the impact of a low-density fluid on the
quickly evaluate a particular design. This involves the net hydrostatic pressure below it, while a tight interval
identification of key problem areas in a given well, which can have the opposite effect.
are typically the zone of highest pore-pressure gradient In the event that the hole is gauge and no single zone
and the zone of lowest fracture-pressure gradient. exhibits an abnormally high pore pressure, a good guide-
However, sections of the well in which the annular con- line is to assume that the shallowest active zone poses
figuration changes should also be carefully examined, the greatest risk to well control. Worst-case calculations
because the contributions of friction and hydrostatic should focus on this zone. One may ignore frictional
pressures can be highly variable. components that may be present when the low-density
Normally, the weak zones in a well will encounter the fluid passes the zone (i.e., only hydrostatic pressure
highest pressure just before the completion of the job should be considered). This ensures that, even in the
(i.e., seconds before the top plug “bumps”). At this event of a shutdown, the well will remain secure in the
point, the longest column of high-density fluid will be in absence of friction pressure. However, in such situations,
the annulus, and friction will be at its highest level the fluids may generally continue to flow owing to U-
(ignoring any pump-rate reduction in anticipation of tubing (Section 12-6). Engineers should make U-tube
bumping the plug). This can be considered the worst- assessments available to wellsite operations personnel,
case scenario for zone breakdown. However, it is much because U-tube events are often misinterpreted in terms
more difficult to propagate a fracture with cement than of losses or gains, causing unnecessary changes in the
it is with OBM. Therefore, the ability to circulate mud at proposed pumping schedule.
the required rate is sufficient to raise confidence levels. In the job-design process, friction pressures are most
When performing an ECD simulation, the entire well- significant when small annuli are involved and must be
bore geometry, including restrictions around the liner calculated carefully. On the other hand, in large annuli,
hanger, must be considered. friction is usually negligible. Instead, a safe estimated
Conversely, from a well-control standpoint, the worst value (50 to 100 psi [0.35 to 0.7 MPa] for the entire well)
situation occurs when the fluid of lowest density (typi- can be ascribed to friction pressure.
cally a chemical wash) passes by an active zone. Pumping excess cement to ensure adequate annular
Depending upon the annular configuration or openhole fill is a common practice, particularly when the borehole
diameter, the zone in question may not be that of high- size and annular volume are uncertain. Negative conse-
est pore pressure. For example, a large washed-out sec- quences may occur unless one considers the possible

Chapter 12 Primary Cement Job Design 447

effects of overfill. If the borehole is closer to gauge than ished, calculation of the wellhead pressure until the plug
expected, the excess cement may be circulated to a bumps indicates how quickly the heavier cement column
point higher than originally intended. If weak zones are is rising in the annulus.
present, the resulting elevated hydrostatic pressure may Considering the importance of annular fluid veloci-
induce losses, compromising the results or endangering ties and pressures to the safe and successful execution
the well. of a cement job, it is clear that U-tubing must be consid-
A similar and no less dangerous situation can arise if ered in any job design. Algorithms exist that permit
large volumes of low-density washes or spacers are fairly accurate simulations of these phenomena
pumped and the hole is more in-gauge than thought. If (Beirute, 1984; Wahlmeier and Lam, 1985). These algo-
such fluids reach an unexpected height in the annulus, rithms have been validated against carefully measured
the resulting loss of hydrostatic pressure may compro- field parameters during cementing operations
mise well control. Thus, having an accurate caliper log (Kelessidis et al., 1994).
of the openhole section is not only useful for computing The numerical manipulation needed to accurately
the required slurry volume, but is also necessary from a simulate the physics of well displacement is consider-
well control standpoint. able. Fortunately, computers with sufficient power to
handle the algorithms at practical speeds are readily
available today.
12-6 U-tubing
Pumping dense fluids such as cement slurries down a
casing string can result in a phenomenon known as free- 12-7 Example of job design procedure
fall or U-tubing (Arnold, 1982; Beirute, 1984). The fluids The following example illustrates how the basic job-
inside the casing and in the annulus will naturally tend design concepts discussed above can be combined with
to achieve a hydrostatic-pressure equilibrium. During the power of computer simulators to provide a realistic
the course of the cement job, some interesting effects and technically suitable well program.
may be observed (for example, mixing of heavy and light Whether computers are used or not, the process of
fluids that are not separated by a wiper plug). designing a cement job complies with the following
The density of cement slurries is usually higher than scheme:
those of the drilling fluid, chemical wash, or spacer. ■ clearly defining the objectives
When the cement slurry is introduced inside the casing, ■ collecting the data (offset and planned wells)
a hydrostatic pressure imbalance is created between the
■ identifying the different options
inside of the casing and the annulus. As a result, the
cement slurry has a tendency to free-fall and draws a ■ assessing the risk of each
vacuum inside the upper part of the casing. ■ making the final selection and decision.
In many cementing operations, the pump rate into
the casing is insufficient to keep the casing full during
the early part of the job. This results in a net flow or 12-7.1 Well conditions and cement job
efflux of fluid from the well. The rate of efflux may be considerations
much greater than the inward flow. Eventually, as hydro-
static pressure equilibrium is approached, the outward A summary of the well data is presented in Table 12-3
flow falls below the inward flow and the casing gradually and illustrated in Fig. 12-11.
refills. At some point, the outward flow may reach zero The plan is to cement a 47-lbm/ft [69.9-kg/m], 9 5⁄8-in.
and the fluid column in the annulus may become sta- [244 mm] intermediate casing at a depth of 6,486 ft
tionary. Such events are easily misinterpreted as partial [1,977 m]. The well is almost vertical, having a maximum
or complete loss of circulation. Finally, when the casing deviation of 3° at the shoe. The open hole is reasonably
is again full of fluid, the inward and outward flow will be in gauge, with most of the section diameter ranging from
equal. However, these values may not remain so for the 12.2 to 13 in. [310 to 330 mm]. A few small, washed-out
remainder of the job. If a low-density wash is present, it sections across the Aptian and Barremian formations
may cause an annular-pressure reduction as it flows past extend the 15-in. open hole to about 18 in. [to 458 mm
the casing shoe. This will in turn cause a second period from 381 mm].
of free-fall, accompanied by another surge of high The Albian formation is a powerful aquifer that must
returns. The beginning and end of U-tubing events can be carefully isolated. However, weak formations below
easily be detected by measuring the surface pressure prevent the use of normal-density cements. Loss zones
during the cement job. Similarly, when U-tubing is fin- and salt sections exist above and below the Albian. A
good cement bond across the casing shoe is required as

448 Well Cementing

Table 12-3. Configurational Data for the 9 5⁄8-in. Section of the Example Well
MD Fracture Gradient Pore Equivalent Diameter Formation Name Typical Lithology
(m) (kg/m3) (kg/m3) (in.)
399 – – 12.519 133⁄8-in. casing Previous casing shoe

402 1,558 1,020 15.297 Senonian salt Evaporite

500 1,558 1,020 12.287 Senonian salt Evaporite

635 1,558 1,020 12.287 Turonian Limestone

702 1,558 1,020 12.287 Cenomanian Evaporite

737 1,558 1,020 12.977 Cenomanian Evaporite

747 1,558 1,020 12.179 Cenomanian Evaporite

894 1,498 1,150 12.179 Albian Sandstone

895 1,498 1,020 12.179 Aptian Sandstone

915 1,498 1,020 14.917 Aptian Sandstone

938 1,498 1,020 12.931 Aptian Sandstone

955 1,498 1,020 12.208 Aptian Sandstone

968 1,498 1,020 12.208 Barremian Sandstone

982 1,498 1,020 14.315 Barremian Sandstone

1,042 1,498 1,020 12.208 Barremian Sandstone

1,066 1,498 1,020 12.767 Barremian Sandstone

1,078 1,498 1,020 15.623 Barremian Sandstone

1,139 1,498 1,020 12.480 Barremian Sandstone

1,165 1,498 1,020 18.264 Barremian Sandstone

1,183 1,498 1,020 13.097 Barremian Sandstone

1,203 1,558 1,020 12.199 Barremian Sandstone

1,367 1,558 1,020 12.199 Neocomian Sandstone

1,380 1,558 1,020 13.143 Neocomian Sandstone

1,485 1,558 1,020 12.207 Neocomian Sandstone

1,695 1,558 1,020 12.207 Malm Siltstone

1,850 1,558 1,020 12.207 Dogger shale Shale

1,977 1,558 1,020 12.207 Dogger anhydrite Evaporite

Chapter 12 Primary Cement Job Design 449

 0 0

Senonian
Evaporite
salt Fracturing
pressure
500 500
Limestone Turonian
Evaporite Cenomanian
Sandstone Albian Water
Depth 1,000 1,000 Depth
(m) Sandstone Barremian (m)

Sandstone Neocomian
1,500 1,500
Siltstone Malm

Shale Dogger shale

Pore
Evaporite pressure
2,000 2,000
0 20 40
Well Lithology Formation pressure (MPa)

Fig. 12-11. Well and formation data in job design example.

the mud weight is raised from 9.8-lbm/gal [1,174-kg/m3] [1,000 L/min] for 3 hr lowers the cementing circulating
to 17.1 lbm/gal [2,050 kg/m3] for drilling the lower 81⁄8-in. temperature to 118°F [48°C]. Although this tempera-
hole section. Common practice would dictate installing ture is not very different from the API BHCT, it has
a stage collar just below the Albian. The 95⁄8-in. casing not yet reached the pseudosteady state (Fig. 12-12).
section would be cemented using a two-stage technique, Therefore, any change in circulation time and rate will
allowing the placement of a strong 15.9-lbm/gal significantly influence the resulting BHCT.
[1,905-kg/m3] slurry across the second stage (Albian). The temperature profile after this 3-hr circulating
Placement difficulties and lost circulation below the period is fairly flat, around 122°F [50°C], with the tem-
Albian formation would normally require two-stage perature at the Albian formation being at or near the
cementing, increasing rig time. Remedial operations geothermal temperature (Fig. 12-13). The temperature
may also be required. A better solution would be to per- near the surface will still be close to 104°F [40°C]. Thus,
form a single-stage cement job, employing a high-perfor- compressive-strength development across the Albian
mance, lightweight lead slurry based on engineered-par- formation or at the surface would not be jeopardized.
ticle-size blends (Chapter 7). A short, normal-density The mud in the hole is a saltwater-base polymer
tail slurry would provide high strength at the shoe. system with a density of 9.8 lbm/gal [1,174 kg/m3], ade-
The lead slurry must be brought to surface. To avoid quate to cover the equivalent mud-weight pore pressure
losses during and after pumping, the density of the lead of the Albian formation (9.6 lbm/gal [1,150 kg/m3]). The
slurry will be 11.5 lbm/gal [1,380 kg/m3]. The density rheological properties of the drilling fluid are reportedly
of the tail slurry will be 15.9 lbm/gal [1,905 kg/m3]. good, with a plastic viscosity of 29 cp [29 MPa-s] and a
Because salt formations are present all along the open Bingham yield of 11 lbf/100 ft2 [5.3 Pa], although the
hole, both the lead and tail slurries are mixed with 18% Herschel-Bulkley rheological model represents a better
salt water. fit and is used for the computer simulations.
The static temperature at shoe is 176°F [80°C], cor-
responding to a geothermal gradient of 1.5°F/100 ft
[2.7°C/100 m]. The calculated BHCT (from API tables)
is 123°F [50°C]. Circulating the well at 6.3 bbl/min

450 Well Cementing

 85

80

Geothermal profile
75

70

65
Mud temperature
history Annulus
(°C) 60
API BHCT
55

50

45

40
0 20 40 60 80 100 120 140 160 180 200
Time (min)

Fig. 12-12. Temperature versus time at shoe. Depth is 1,973 m.

Annulus

500

Depth 1,000
(m)
Geothermal profile
Tubular fluids

1,500

2,000
25 30 35 40 45 50 55 60 65 70 75 80 85
Temperature (°C)

Fig. 12-13. Temperature profile after 3 hr of mud injection.

Chapter 12 Primary Cement Job Design 451

12-7.2 Slurry formulation 40
Both the lead and tail slurries require the incorporation
of fluid-loss additives to maintain the designed slurry 30
properties and avoid bridging across the permeable
zones of this long interval. Dispersants will also be Lead slurry
required for optimal fluid-loss control and because the Shear 20
stress
friction pressures generated by the viscous slurries (Pa) Laminar spacer
could pose a risk to weak zones. Additionally, the disper-
10
sant will facilitate the mixing of the lead slurry, which is
designed with a very low water-to-solid ratio. Drilling fluid

The duration of the job is likely to be 4 to 5 hr. Adding 0

a 50% safety margin, the minimum thickening time of 0 20 40 60 80 100 120 140 160
the lead slurry should be 6 to 8 hr, and less for the tail Shear rate (sec ) –1

slurry (5 to 6 hr). A retarder will then be added to

achieve adequate placement time. Fig. 12-14. Recommended spacer rheology.
The design criteria for each slurry will then be the fol-
lowing:
■ Lead slurry (11.5 lbm/gal [1,380 kg/m3]) A neural-network expert system translates the
required rheology for this cementing operation into a
Special lightweight blend compatible combination of products to achieve the
+ dispersant required performance under downhole conditions. This
+ fluid-loss additive considerably reduces the amount of laboratory testing
+ salt (18% by weight of water) required to fine-tune the spacer composition. Finally,
the mud removal criteria are verified for each fluid inter-
+ retarder (if necessary) for thickening time of 7–8 hr face along and around the entire annulus as shown in
■ Tail slurry (15.9 lbm/gal [1,905 kg/m3])
Fig. 12-15.
API Class G cement To further improve the cementing operation, some
+ fluid-loss additive to achieve API-estimated fluid chemical wash will be pumped ahead of the spacer to
loss rate of 80–100 mL /30 min thin the drilling fluid and facilitate its removal. Owing to
pore-pressure requirements and the need for a saline
+ dispersant fluid, the density of the chemical wash is 9.6 lbm/gal
+ antisettling viscosifier [1,150 kg/m3]. It is important to perform tests to verify
+ salt (18% by weight of water) that the spacer and wash are compatible with the
+ retarder (if necessary) for thickening time of 5–6 hr drilling fluid.

12-7.4 Job design

12-7.3 Mud removal strategy and spacer
formulation As discussed previously, adequate casing centralization
is required for efficient drilling-fluid displacement and
The cement slurries will not be pumped in turbulent proper cement-slurry placement throughout the inter-
flow, because of hole size and the presence of weak val, from shoe to surface. Laboratory testing is per-
zones. This, and the risk of mud contamination, suggests formed to design slurry and spacer formulations that
the use of spacers or a combination of spacers and chem- meet the required performance specifications. Such
ical washes to achieve good mud removal. The spacer pilot testing also provides data concerning the rheologi-
density would be between that of the mud and the lead cal properties of the slurries, spacers, and mud at both
cement. surface and downhole conditions. These data (Table 12-4
Simulators calculate the optimal rheology for the and Table 12-5) are then used in the final job design.
spacer to efficiently displace the drilling fluid in the hole
at the given pipe standoff and with the planned range of
pump rates (and their corresponding shear rates in the
eccentered annulus), as shown in Fig. 12-14.

452 Well Cementing

Fig. 12-15. Mud removal efficiency advisor.

Table 12-4. Lead Slurry Properties The fluid volumes, density, rheology, and annular
API/ISO fluid loss rate 24 mL/30 min cement fill (along with information about the position of
other fluids in the annulus at the end of the job) planned
Free water 0 mL for this cementing operation are shown in Table 12-6. A
Rheology at 100°F [37.8°C] Herschel-Bulkley model
graphical representation of the pressure margins
(hydrostatic pressure only at this stage) with such a
Consistency index 0.00396 lbf-sn/ft2 [0.189 Pa-s] fluid column is shown in Fig. 12-16.
Power-law index 0.914

Yield point 16.3 lbf/100 ft2 [7.8 Pa]

10-min gel strength 42 lbf/100 ft2 [21 Pa]

Thickening time 8 hr, 5 min

24-hr compressive strength 3,320 psi [22.9 MPa]

Table 12-5. General Fluid Data

Fluid Sequence
Name Volume Annular Top Density Downhole Rheology
(m3) Length (m) (kg/m3)
(m) Consistency Index Power-Law Index Yield Point Value
(Pa-sn) (Pa)
Chemical wash 3.2 0.0 – 1,150 0.005 1 0

Spacer 14.3 0.0 – 1,320 2.490 0.300 1.69

Lead slurry 62.3 1,789 0 1,378 0.189 0.914 7.79

Tail slurry 6.4 185 1,789 1,900 0.396 0.694 7.08

Mud 73.3 – 0 1,174 0.151 0.747 3.48

Chapter 12 Primary Cement Job Design 453

Table 12-6. Job Schedule of Example Well
Name Flow Rate Volume Stage Time Elapsed Time Cumulative Volume Comments
(L/min) (m3) (min) (min) (m3)
Start job Prejob safety meeting.
Check data monitoring
and recording equipment.
Pressure-test lines.

Chemical wash 477 3.2 6.6 6.6 3.2 Pump chemical wash.

Spacer 954 14.3 15.0 21.6 14.3 Pump spacer.

Pause 0 0 5.0 26.6 0 Drop bottom plug.

Lead slurry 635 62.3 98.2 124.8 62.3 Mix and pump lead slurry.

Tail slurry 795 6.4 8.0 132.8 6.4 Mix and pump tail slurry.

Pause 0 0 5.0 137.8 0 Drop top plug.

Mud 954 47.7 50.0 187.8 47.7 Displace with mud.

Mud 636 11.1 17.5 205.3 58.8 Slow down rate in stages.

Mud 477 7.9 16.6 221.9 66.7 Slow down toward end displacement.

Mud 318 6.6 20.7 242.6 73.3 Slow down further—bump plug.

End job Pressure-test casing.

Bleed off pressure.
Check returns.

Fluid Sequence Static Well Security

0
Hydrostatic
pressure

500
Mud
(displacement)
Fracturing
pressure
Depth 1,000
(m)

1,500 Lead
slurry

Pore
Tail pressure
slurry
2,000
0 10 20 30 40
Pressure (MPa)

Fig. 12-16. Downhole pressure-density plot.

454 Well Cementing

 A simulation of the actual operation, including shut- Figure 12-19 provides data similar to those given by
downs, rate changes, U-tubing, etc., is shown in Fig. 12-16. In this case, dynamic frictional components
Fig. 12-17 and Fig. 12-18. Figure 12-17 illustrates the fact are considered, as are all worst-case scenarios through-
that flow rates in and out of the well are not equivalent out the duration of the entire pumping operation. This
during a large part of the job. Sudden increases or graph reveals everything the engineer needs to know
decreases in rate, resulting from fluids of varying density about well control. The minimum pressure safety mar-
moving from the casing into the annulus, can be pre- gins are displayed in Table 12-7 for each of the parame-
dicted ahead of time. Knowledge of the magnitude of ters.
these fluctuations, and the times at which they are Once it has been verified that the planned conditions
expected, can help allay fears that well control is threat- for the cement job will not create well-control problems,
ened or that serious losses have occurred. the final mud-removal parameters are verified with the

Fluids at 1,973 m
1,000
Inward flow

800 Outward flow

600
Flow
rate
(L/min) 400

200

0
0 30 60 90 120 150 180 210 240 270
Time (min)

Fig. 12-17. Flow-rate comparison at shoe.

18 32

17 30
Fracturing Fracturing
16 pressure 28 pressure
Dynamic Hydrostatic
pressure
Annular 15 pressure Annular 26 Dynamic
pressure pressure pressure Hydrostatic
(MPa) 14 (MPa) 24 pressure

13 22
Pore pressure Pore pressure
12 20

11 18
0 30 60 90 120 150 180 210 240 270 0 30 60 90 120 150 180 210 240 270
Time (min) Time (min)

Fig. 12-18. Placement pressures at 1,200 m (left) and at shoe (1,973 m) (right).

Chapter 12 Primary Cement Job Design 455

 Dynamic Well Security help of a 2D numerical simulator to visualize all the
0
Hydrostatic pressure aspects of the mud-displacement process. For example,
Minimum hydrostatic pressure it is important to verify that one fluid does not bypass or
Maximum dynamic pressure
Minimum dynamic pressure
fall through another, and that in the end the cement
500 Fracturing pressure slurry is in direct contact with the drilling fluid.
Pore pressure Output from the simulator can be an animation of the
entire process or a picture at any given time.
Figure 12-20 displays the situation in the annulus 2 hr
Depth 1,000
(m)
and 34 min after the beginning of the operation (refer to
the pump schedule). It is essential to verify that,
throughout the entire simulation, mud conditioning,
pipe centralization, preflush, and slurry densities, rheol-
1,500
ogy, and pump rates have been correctly chosen. This
will ensure that the interfaces between fluids remain as
flat as possible and the cement is never in contact with
2,000 the mud.
0 4 8 12 16 20 24 28 32
Pressure (MPa)

Fig. 12-19. Well control.

Table 12-7. Minimum Safety Pressure Checks in Dynamic Conditions

Parameter Min. Absolute Pressure Found at Depth Min. Relative Pressure Found at Depth
(MPa [psi]) (m) (%) (m)
Fracture gradient 0.25 [36] 750 2.1 1,078

Pore pressure 0.15 [22] 750 1.7 759

Collapse rating 18.3 [2,650] 1,945

Burst rating 40.5 [5,874] 0

Velocity (m/s) Fluid Concentrations Risk Mud on Wall Flow Regime

0
200
400
600
800
1,000
1,200
1,400
1,600
1,800
2,000
Wide Narrow Wide Wide Narrow Wide Wide Narrow Wide Wide Narrow Wide

1.37–1.55 Mud High Turbulent flow

1.03–1.20 Tail Medium No flow
0.69–0.86 Lead Low Laminar flow
0.34–0.52 Spacer None
0.00–0.17 Chemical wash

Fig. 12-20. Snapshot of the annulus at 2 hr 34 min.

456 Well Cementing

 At the end of the simulated operation, the situation in Cement may also become contaminated in the course
the annulus in terms of cement coverage and expected of shipment, and should be sampled at the wellsite.
isolation should be compatible with the objectives of the Typical contaminants found in cement include bentonite
operation (Fig. 12-21). If the results are incompatible, (gel), barite, and silica. All can have a negative impact
the job parameters must be further optimized. on slurry performance. To ensure that the materials at
the wellsite perform properly, both the service company
and the operating company should implement strict
Fluids Concentrations quality-control procedures. Additive drums, sacked
0 0 materials, and silos bearing cement blends should be
200 clearly marked to avoid mistakes.
Cement
coverage 400 Special instructions or procedures that are important
500
Standoff 600
to the success of the job must be communicated to the
between wellsite. It is imperative that personnel from the operat-
centralizers 800
Depth 1,000 Depth ing company and service company understand and agree
1,000
(m) (m) on the exact details of the job. Recommendations to rec-
1,200 iprocate or rotate pipe, install a given number of cen-
1,400 tralizers per joint, or pump at certain rates for specific
1,500
1,600 periods of time are made with good reason, and every
1,800 effort should be made to follow these instructions at the
2,000 2,000 wellsite.
50 60 70 80 90 100 The use of computer-based data acquisition systems
Wide

Narrow

Wide

% at the wellsite allows accurate monitoring of the opera-

tion. On-the-job recordings of pump rate, surface pres-
sure, and fluid density can be plotted and compared with
Fig. 12-21. Annulus fill-up and cement coverage at end of job. the original design simulations. This helps to ensure that
jobs are executed as designed and also to identify anom-
alous well conditions in the course of the operation.
Ultimately, such information can be used to help “close
12-8 Preparing for the job the loop” between design and execution, permitting the
A satisfactory job design is one that not only satisfies the engineer to modify future designs to achieve optimal
requirements of slurry performance and displacement results. Before evaluating the cementing operation with
mechanics, but is also practical to execute in the field. a sonic or ultrasonic log (Chapter 15), the playback of
Therefore, an essential part of the design process is to data recorded during the cement job helps the engineer
review equipment requirements and availability, bulk identify unforeseen problems that might have occurred
storage capacity, rig facilities, and space availability. The (Smith et al., 1987; Beirute, 1987; Piot and Loizzo, 1998).
practicalities of using different dry cement blends or of
mixing various additive formulations for jobs featuring
multiple slurries, spacers, and chemical washes must be 12-9 Conclusions
considered. In many situations, using liquid additives
Designing a cement job is an iterative and collaborative
and a single cement or blend may prove beneficial, par-
process that starts long before the cementing operation.
ticularly in locations in which storage capacity and
Ideally, the collaboration between the operating com-
space are limited (e.g., offshore).
pany and the service company begins at the well plan-
Cement-performance tests are performed with the
ning stage and continues throughout the drilling stage.
field-cement blends and field mix water before the job
As drilling progresses, the well data are continuously
to verify that the performance of the materials at the
updated, and the cementing options continuously evolve.
wellsite is consistent with the job-design criteria devel-
Then a final decision is made, specifying a cost-effective
oped during the pilot testing phase. Many location
solution that is compatible with the job objectives.
waters contain organic residues or dissolved salts detri-
This process encompasses the following steps (each
mental to the performance of some slurries (Kieffer and
may be repeated several times):
Rae, 1987). The quality of the water may be highly vari-
■ determining and refining objectives
able throughout the seasons because of currents, mete-
orological factors, or runoff. Hence, it is important to ■ planning for health, safety, and environmental issues
sample the water that will be used to mix the additives
and cement slurries.

Chapter 12 Primary Cement Job Design 457

■ determining 12-10 Acronym list
– well geometries 2D Two-dimensional
– deviation 3D Three-dimensional
– static temperature gradient ANSI American National Standards Institute
– temperatures API American Petroleum Institute
■ studying problems in offset wells BHCT Bottomhole circulating temperature
■ selecting appropriate equipment and materials BHST Bottomhole static temperature
■ finalizing job procedure options ECD Equivalent circulating density
■ finalizing slurry design options HPHT High-pressure, high-temperature
■ selecting mud displacement strategy ISO International Organization for Standardization
■ selecting LWD Logging while drilling
– spacers MD Measured depth
– cement systems MWD Measurement while drilling
– procedures OBM Oil-base mud
■ finalizing the program. TOC Top of cement
Common sense is the engineer’s greatest asset in TVD True vertical depth
cement-job design. Considerable time can be saved by
quickly distinguishing the relevant well characteristics
from those that are not.
The engineering team must recognize any factors that
may affect the final quality of the cement job. Otherwise,
the final objectives of the cementing operation are less
likely to be achieved.
Delivering a smooth wellbore with a smooth trajec-
tory and a stable nongelling drilling fluid creates ideal
cementing conditions. The farther the well conditions
deviate from the ideal situation, the more difficult it will
be to achieve the cement-job objectives.
Tools such as 2D numerical fluid displacement simu-
lators are instrumental in analyzing and understanding
the effects of each well parameter and can be a power-
ful aid to optimize the cementing operation in ever more
challenging wells.
The objective of well cementing no longer stops at the
end of cement placement but encompasses the mainte-
nance of zonal isolation throughout the life of the well.

458 Well Cementing

13-1 Introduction
Primary Cementing Techniques
Bernard Piot and Gérard Cuvillier—Schlumberger

Primary cementing is a technique for placing cement

slurries in the annular space between the casing and the
13
accommodate the expected completion for the antici-
pated production rate, and the design works backwards
to the surface, adding one more casing size as particular
borehole. After placement, the cement hardens to form zones must be protected or isolated. The complete pro-
a hydraulic seal in the wellbore, preventing the migra- gram, which depends on the availability of drilling and
tion of formation fluids in the annulus. Therefore, pri- completion tools and equipment for each size, is
mary cementing is one of the most critical stages during designed to ensure that the well is drilled and operated
the drilling and completion of a well. This procedure safely throughout its lifetime.
must be planned and executed carefully, because there Casings are primarily classified by their size (outside
is only one chance to complete the job successfully. diameter [OD], or outside diameter of the body of the
In addition to providing zonal isolation, the set- pipe), material (carbon steel or corrosion-resistant alloy
cement sheath should anchor and support the casing material), weight (lbm/ft or kg/m), yield strength (psi or
string (preventing formation sloughing or caving into the N/mm2), type of connection (thread or coupling), and
wellbore) and protect the casing string against corrosion range (length of each casing joint). These characteris-
by formation fluids. Uncemented steel casing can cor- tics give the casing string certain mechanical properties,
rode rapidly when exposed to hot formation brines and such as tensile-, burst-, and collapse-pressure ratings.
hydrogen sulfide. It can also be subjected to erosion by The most common casing connections are threaded con-
the high velocity of produced fluids, particularly when nections, although quick-makeup mechanical connec-
solid particles such as formation sand are being trans- tors are available in the large sizes.
ported. Lateral loads on poorly cemented casing strings Figure 13-1 also shows casing strings known as liners.
can result in buckling or collapse because of overloading A liner is a string of standard casing that does not extend
at certain points. On the other hand, properly cemented all the way to the surface; instead, it is hung from inside
casing is subjected to a nearly uniform loading approxi- the previous casing string. Liners are covered in detail
mately equal to the overburden pressure. later in this chapter.
In principle, primary cementing techniques are the The string type and its function govern the casing-
same regardless of casing-string purpose and size. The string design. The selection of a particular casing pri-
cement slurry is pumped down inside the string to be marily depends on the
■ depth
cemented, exits the bottom, and displaces drilling mud
as it moves up the annulus. Details vary from casing to ■ hole sizes in which the casing string is to be set
casing, and the differences in placement technique are ■ required tightness and torque resistance of connec-
discussed in this chapter. It is assumed that the reader tions
is familiar with the previously presented supporting ■ mud-column and formation pressures
material: Chapters 5 (Mud Removal), 9 (Annular
■ condition of the formation
Formation Fluid Migration), 11 (Cementing Equipment
and Casing Hardware), and 12 (Primary Cement Job ■ drilling objectives.
Design) in particular.
The casing is designed to withstand the mechanical
and chemical stresses in the well (Lubinski, 1951;
Oil-Well Cementing Practices in the United States,
13-2 Classification of casing strings American Petroleum Institute [API], 1959; Smith, 1987).
A series of casing strings is necessary to complete most A methodology known as the Service Life Model
oil and gas wells and produce the desired fluids success- (Klementich and Jellison, 1986) analyzes the influence
fully (Fig. 13-1). The design of the casing program is con- of any drilling, completion, or production event on the
tingent upon several factors. It starts with the produc- burst, collapse, and tension and/or compression resis-
tion casing, which must be large enough to

Well Cementing ■ Chapter 13 Primary Cementing Techniques 459

 28-in. hole 30-in. conductor
Conductor casing

26-in. hole 22-in. surface casing

Surface casing
20-in. hole 16-in. intermediate casing

171⁄ 2-in. hole 133⁄ 8-in. intermediate casing

Intermediate casing
14-in. hole 111⁄ 4-in. contingency liner

12 3⁄ 4-in. hole 9 7⁄ 8-in. production casing

81⁄ 2-in. hole 75⁄ 8-in. contingency liner

Production liner

81⁄ 2-in. hole 7 or 51⁄ 2-in. production liner

Simple Program Complex and Heavy Program in Deep Well

Fig. 13-1. Example of simple and complex casing programs.

tance of the casing. A model by Johnson et al. (1987) Expandable casing tubulars are an alternative to
considers triaxial loading, and one by Rodriguez et al. conventional casings. They are covered in detail in
(2003) simulates the influence of cement mechanical Section 13-6.2.
properties on the casing-collapse load.
The casing must be designed to withstand all loads
imposed during installation and throughout the lifetime 13-2.1 Conductor pipe
of the well. In particular, the innermost string of casing The conductor is usually the first and shortest casing
(hereafter defined as production casing) is an integral string. Its purpose is to protect shallow formations from
part of the well architecture, designed to maintain well being contaminated by drilling fluids and help prevent
integrity under severe conditions. Such conditions washouts that can easily occur near the surface in
include: a well full of gas (subsequent to a tubing fail- unconsolidated topsoils and sediments. The conductor
ure), an empty well, hot flowing production fluids, and pipe also serves as a conduit to raise the circulating fluid
injection of cold treatment fluids. The cements support- high enough to return to the mud system and enables a
ing and isolating these respective casings must also diverter to be installed should gas sands, for example, be
withstand the same conditions throughout the life of the encountered at a shallow level. The conductor pipe is
well. In this section, the functions of the casing strings, designed to provide structural support for all subsequent
the depths to which they are normally set, and special casing and tubing strings and blowout preventers
considerations for each are discussed. (BOPs) as well as the wellhead when the ground support

460 Well Cementing

is inadequate. Therefore, the conductor must be set casing. Most often, a guide shoe is welded on to help
deeply enough into firm ground that it will not subside lower the conductor into the well. Cementing of the con-
when subjected to the additional loading. Offshore, the ductor is performed through a swage, which is screwed
conductor should be designed to withstand loads to the top of the conductor. The cement slurry is pumped
imposed by waves and currents. through the swage and into the pipe. If the length of the
At certain offshore locations or during swamp barge conductor is short, the annular and pipe volumes are rel-
operations, driving or jetting the conductor into the atively small, and cement slurry is pumped until returns
ground is a common practice. A pile-driving hammer is are observed at the surface. Cement slurry is then dis-
used, and sections of conductor casing are welded placed from the casing without the use of plugs.
together as they are driven into the ground. Increasingly, Alternatively, and particularly for longer conductors, the
mechanical connectors are used because they offer sig- stab-in technique or inner string cementing can be used.
nificant rig time savings over welded connections. These techniques allow pumping of a minimum volume
However, the mechanical connectors must be driveable. of cement slurry and limit slurry contamination during
The setting depth is usually less than 300 ft [91 m]. It is displacement (see Section 13-3.1).
often determined by the limitations of the pile driver as In deepwater wells, conductor casings were initially
the conductor begins to encounter firmer ground. Once drilled and cemented, because techniques for driving
driven to the maximum depth (depth of refusal), the piles at such great depth were not available.
conductor is then cut to the appropriate height below Unfortunately, this often resulted in large, washed-out
the drill floor (and above the water line in offshore holes owing to the presence of soft and fragile surface
applications), and a casing head is welded into sediments. It was not uncommon to pump 300% excess
place (Fig. 13-2). cement slurry yet not observe cement returns on the
The hole for the conductor is sometimes drilled (e.g., seafloor. Today, the conductor casing, such as a 30-in.
a 36-in. hole for a 30-in. conductor), and the pipe is made casing, is more frequently attached to a 26-in. or 24-in.
up and lowered in a manner similar to conventional bit and drilling assembly and lowered at the same time
the hole section is drilled. This jetting casing-drilling
operation minimizes the disturbance of the soft forma-
Drill floor tions. At casing depth, the conductor is simply discon-
nected from the drilling assembly and left uncemented
as a guide for the fluids. Drilling of the 20-in. or 18 5⁄8-in.
Rotary table surface-casing section can then proceed without delay.
In shallow-casing jobs, washouts and lost circulation
often prevent the cement from reaching the surface.
Normally, the necessary volume of cement slurry is esti-
Flowline mated before the job, then mixed and pumped down-
Fill line hole. If the washouts prevent the cement slurry from
reaching the desired height, a top-up job must be per-
Annular-type formed (Section 13-3.2). If lost circulation occurs after
blowout mixing is completed, the casing volume must be dis-
preventer placed, requiring the pumping of large quantities of
cement into the loss zone (Chapter 6).
Casing head
13-2.2 Surface casing
The second string of casing, which serves to case off rel-
atively shallow unconsolidated formations and aquifers,
Ground level is known as surface casing (Fig. 13-3). In addition to
maintaining hole integrity, the surface casing prevents
Cellar (optional)
the contamination of fresh groundwater by drilling
fluids, subterranean brines, oil, or gas. Depending on the
country, there are usually government regulations stipu-
Conductor lating minimum casing requirements and set-cement
properties (Chapter 12), particularly regarding the pro-
tection of aquifers.
Fig. 13-2. Conductor ready for drillout with the casing head welded
on and the BOP connected.

Chapter 13 Primary Cementing Techniques 461

 Circulating Mud Pumping Wash, Spacer, and Cement Slurry
Plug releasing pins

Bottom cementing plug Top cementing plug Tail slurry

Lead slurry

Spacer
Centralizers
Wash

Float collar

Shoe

Displacement Displacement

Displacement fluid

End of Job

Fig. 13-3. Typical one-stage primary cement job on a surface casing string.

Surface casing must be set within competent forma- sizes and setting depths vary considerably; generally
tions to allow the installation of BOPs (Fig. 13-4) before speaking, diameters range from 7 to 26 in. and depths
drilling into potentially hydrocarbon-bearing and pres- can reach 12,550 ft [3,825 m] (Fontenot, 1986).
sured formations. It is the first casing string for pres- A major problem associated with cementing surface
sure-control purposes. Therefore, the selected casing casing is placing the required annular height of cement
must be strong enough to support a BOP and withstand slurry (often to surface) when the hydrostatic pressures
the fluid pressures that may be encountered. Surface of the slurries exceed the formation fracture pressure.
casing also provides a solid anchor for the casing head The use of low-density slurries and foamed cement slur-
when the well is put on production. The surface-casing ries is becoming more common in such circumstances

462 Well Cementing

 and the large annular cross-sectional area (even in a
Rig floor Rotary table gauge hole) often prevents achieving the flow properties
required to ensure efficient mud removal. High-solids,
high-performance lightweight cements, which combine
Fill line
fast setting characteristics and high viscosity at low den-
Flowline sity (Chapter 7), are particularly useful in deepwater
Annular preventer wells. Effective laminar flow is the recommended dis-
placement regime in such large annuli (Chapter 5).
Choke line Large-diameter casings, with ODs equal to or greater
Kill line
than 18 5⁄8 in., are subject to large upward forces during
Blind rams cementing. Such casings have a large cross-sectional
area upon which the pumping pressure can act. Unless
preventive measures are taken, the upward forces may
exceed the buoyed weight of the casing, and the casing
Pipe rams will rise out of the hole (Appendix C). To prevent such
problems, the pump pressure can be controlled, the den-
sity of the mud used to displace the slurry can be
adjusted, or the casing can be physically restrained by
various methods.
Surface casing head
Conductor pipe
Surface casing
13-2.3 Intermediate casing
The intermediate casing isolates hydrocarbon-bearing,
Cement abnormally pressured, fractured, and lost-circulation
zones, as well as problem shales and similar geological
horizons. Intermediate casings are set within more com-
Figure 13-4. Surface casing support of blowout prevention equipment. petent formations than the surface casing, providing
greater blowout protection when drilling deeper than
the previous casing would allow. Any casing string run
between the surface casing and the production casing is
an intermediate casing string.
(Chapters 3 and 7). Washouts are another frequent prob-
One or several intermediate casings are often neces-
lem. When the borehole is enlarged by washouts, its size
sary to maintain borehole integrity at greater drilling
often exceeds the measuring capability of caliper tools.
depths. Typical casing sizes range from 6 5⁄8 in. to 16 in.,
Unless adequate information is available from similar
and the depth can vary from 1,000 to more than 17,000 ft
offset wells, accurate hole volumes may not be obtain-
[305 to more than 5,180 m]. Intermediate casing is usu-
able.
ally run to surface and once again anchors and connects
A through-drillpipe stab-in cementing system can be
BOPs with a higher pressure rating for subsequent
used in some surface casing cementing operations, but
drilling.
often this is not possible when using smaller-size surface
Depending on well depths and prevailing pressure
casing or when larger sizes are run beyond 4,500 ft
regimes, intermediate casings can fulfill several func-
[1,500 m]. Drilling rig design constraints become the
tions.
limiting factor in these applications.
■ One or several intermediate casings are generally
Cementing of surface casings traditionally has been
performed in stages when severe lost-circulation zones employed to seal off weak zones that might fracture
or other troublesome intervals were encountered. Today, and cause lost circulation in the presence of high-
they are frequently cemented in a single-stage operation density mud. Such muds are usually required as well
using high-performance, low-density cement systems depth increases.
(Chapter 7; Al-Suwaidi et al., 2001). ■ Occasionally, salt or anhydrite formations might dis-
Surface casing strings often encounter sloughing solve and leach out, causing drilling-fluid contamina-
shales and shallow gas pockets (Chapter 9). These are tion or pipe sticking. Salt formations have a natural
some of the most difficult casing strings to cement suc- tendency to creep or swell, a phenomenon likely to
cessfully. Low formation temperatures prolong the thick- cause stuck pipe while drilling and casing collapse
ening times of conventionally extended cement slurries, unless precautions are taken (Chapters 6 and 7).

Chapter 13 Primary Cementing Techniques 463

■ Sometimes an intermediate string seals off older pro- should be sufficient to drill the next hole section without
ducing zones, allowing the operator to drill deeper. the hydrostatic pressure of mud exceeding the formation
■ Intermediate strings may seal shallow permeable for- fracture pressure.
mations containing uneconomical hydrocarbons or With the advent of high-performance lightweight
permeable zones that have been charged by gas leak- cements, the casing string can usually be cemented in a
ing through poorly cemented annuli in nearby wells. single-stage primary job. However, a multistage job may
■
be necessary under certain circumstances:
Intermediate casings protect the hole in deviated sec-
■ if the column of cement slurry would exert a hydro-
tions, particularly through unstable formations.
■
static pressure greater than the formation-fracture
High-pressure, noncommercial fluid zones, which
pressure
may be encountered well above the targeted pay
■ for the isolation of a lost circulation zone
zone, can be hydraulically sealed.
■ for imperatively isolating an hydrocarbon layer.
An intermediate casing also provides better protec-
tion against well pressure than the surface string, owing Figure 13-3 is an illustration of a single-stage primary
to its smaller diameter and the availability of higher job. Figure 13-5 depicts a typical two-stage cement job
steel grades. The setting depth of an intermediate casing on an intermediate casing string.

First-Stage Displacement Opening Bomb

Seated in Stage Collar

Displacement fluid

First-stage Opening
sealing plug bomb

First-stage
slurry with
excess cement

Wash-ahead
spacer

Circulating First-Stage Cementing Start Second Stage Stage Collar Closed

Upper Section Sequence Completed 1,500 psi

First-stage
excess slurry

Ports opened

Cement basket
or packer

Fig. 13-5. Typical two-stage cementing sequence of an intermediate casing string.

464 Well Cementing

13-2.4 Production casing lost to the zone. Good fluid-loss control minimizes
The production casing or production liner is the last damage to the critical wellbore matrix and prevents pre-
tubular element in the well. It isolates the different mature slurry dehydration in the annulus, which could
zones above and within the production zone and with- result in annular bridging and a failed cement job. Fluid-
stands all of the anticipated loads during testing opera- loss rates should be less than 100 mL/30 min, preferably
tions and production throughout the well’s life. Any in the 50 mL/30 min range (Chapter 6). In the case of
casing or liner that creates an annular space with the high-pressure, high-temperature wells, the fluid-loss rate
production tubing is treated as a production casing or must be less than 50 mL/30 min.
liner. Achieving hydraulic isolation across, above, and
Setting the production casing string is one of the prin- below the pay zone is essential. With conventional
cipal objectives when drilling a well. In many respects, cement slurries, an important property of the set cement
the production string is the oil well. It is the protective is compressive strength. The set cement must also have
housing for the tubing and other equipment used in a low permeability to prevent fluid invasion. Set-cement
well. Tubing may be pulled out of the hole for change or permeability is directly related to water content and
inspection, or if there is a completion equipment mal- inversely related to cement content and strength. The
function, but the production string is cemented in place. general guideline for adequate zonal isolation is
Indeed, special attention is usually devoted to ensure a 1,000-psi [7.0-MPa] compressive strength and less than
pressure-tight bond between the formation and the pro- 0.1 mD water permeability. Strength retrogression must
duction string. Common sizes range from 4 1⁄2 to 9 5⁄8 in. be prevented when the bottomhole static temperature
Depths can vary from 1,500 to more than 25,000 ft [460 exceeds 230°F [110°C] (Chapter 10). This condition,
to more than 7,620 m]. which also applies to any cement used in surface and
The production casing is normally run and cemented intermediate strings that will be exposed to producing
through the zone to be produced and then perforated to temperatures greater than 230°F [110°C], is generally
allow communication with the formation of interest fulfilled by adding 30% to 40% by weight of cement silica
(for producing reservoir fluids or injecting fluids). flour to the Portland cement.
Sometimes it is set just above the zone of interest, and When pressure and temperature variations are antic-
an openhole completion is performed. The casing should ipated during production, injection, or well mainte-
be the best-quality pipe that is appropriate for the con- nance, special cements with tailored mechanical prop-
ditions involved, including the potential for long-term erties (Young’s modulus, Poisson’s ratio, and tensile
corrosion. A small leak can develop into a blowout; strength) are available to maintain isolation throughout
therefore, the threaded connections should be appropri- the life of the well (Chapter 7).
ate for the anticipated pressures. To guard against leaks, In shallow wells that are not prolific producers, a
the casing joints should be carefully connected as the small cemented casing can be substituted for the pro-
casing is run into the well. Torque should be applied cor- duction tubing. Such tubingless completions are less
responding to the thread compound used. If casing rota- expensive. Sometimes this technique is used in minia-
tion is an element of the mud-removal strategy, premium turized dual completions, in which two separate casing
connections that can withstand the torque should be strings are run and cemented in the same wellbore. Such
used. Different designs with proprietary pipe-thread completions are cemented in a single operation using
geometries and metal-to-metal seals are available for gas completion brine as the displacement fluid.
tightness and should be used whenever the operating
conditions demand reliable pressure-tight sealing and
100% connection efficiency. 13-2.5 Combination casing strings
Excellent primary cementing of production casings is In casing-string design, it is common to vary the casing
essential. The cement slurry must be designed to keep weights (internal diameters) or grades within a nominal
the producing zone under control by providing adequate size range because of load considerations, cost savings,
hydrostatic pressure. Preflushes and spacers run ahead and the available casing inventory. However, to minimize
of the slurry must also be checked to maintain hydrosta- the risk of errors in the pipe running sequence, one
tic overbalance and to maintain well control at all times should not use more than two types of casing joints in a
(Chapter 12). given casing string. These factors must be known when
Zonal isolation is imperative both to protect the pay designing the cement job, because burst and collapse
reservoir from fluid migration and crossflow and to iso- ratings are affected, internal diameters vary, and thread
late the pay hydraulically for any future stimulation treat- connections may change within a particular string.
ments. In addition, the cement slurry must have ade- Another technique is to actually vary the nominal
quate fluid-loss control to minimize the amount of filtrate casing size. Commonly used combinations include

Chapter 13 Primary Cementing Techniques 465

10 3⁄4 in. and 9 5⁄8 in. and 7 5⁄8 in. and 7 in. Installing such annulus (also called reverse circulation cementing) is
combinations requires the casing to remain stationary in sometimes performed in the extreme case in which
the rotary table for a substantial amount of time—the important lost circulation zones or fragile formations
time required to change the casing handling equipment occur near the shoe but some cement is required to seal
(slips and elevators). The end result is similar to a liner off an upper interval. This is typically a last-resort
completion. The larger-internal-diameter (ID) casing option, because fluid placement is largely uncontrolled
may be desirable in dual completions or in gas wells in and the shoe is never cemented. Modifying the float
which additional tubular completion equipment is equipment is necessary to allow careful control and pre-
required, such as side-pocket injection mandrels or in cise monitoring of the returns through the casing.
offshore wells requiring subsurface safety valves. For large-diameter casings, the traditional cementing
Figure 13-6 depicts a typical tapered string comple- technique is frequently inadequate; consequently,
tion. This particular example was completed to accom- cementing through the drillpipe or a grouting technique
modate an additional tubing string that could be run to in which the cement is circulated into place by pumping
the bottom of the 7 5⁄8-in. casing. the slurry down one or more small-diameter pipes
placed in the annular gap is performed. When cementing
intermediate or production casings, well conditions and
the length of interval to be cemented influence the
placement technique to be used. Usually, the maximum
permissible downhole pressures determine whether a
job should be performed in a single stage or in multiple
stages. In this section, the most common procedures are
7 5⁄8-in. casing described.

9 7⁄8-in. open hole

13-3.1 Cementing through drillpipe (stab-in or
Crossover swage
inner-string cementing)
As discussed earlier, performing the job through
8 3⁄4-in. open hole drillpipe can prevent many problems related to cement-
ing large casings. With the stab-in technique, which can
7-in. casing only be used on stationary rigs (land rigs and jackup or
platform rigs), the casing is run in place with a stab-in
float shoe. The casing is set in the casing slips, thus sus-
pending the string off bottom. Drillpipe made up with a
stab-in stinger (Fig. 13-7) is then run in the casing until
Fig. 13-6. Typical tapered string completion. it is approximately 3 ft [1 m] above the float shoe.
Circulation with the drilling fluid is then established,
and returns come from the annulus between the
drillpipe and the casing. Circulation is stopped and the
The only special cementing considerations for combi-
drillpipe is lowered, enabling the stinger to stab or screw
nation strings are for the displacement plug. The most
into and seal in the float shoe. While a constant watch is
common practice involves modifying the displacement
maintained on the fluid level in the casing-drillpipe
plug of the larger-diameter casing. The core of the plug
annulus, which must remain stationary, circulation is
is machined to a size less than the ID of the smaller
broken again, and one typically observes returns flowing
casing. As a result, the wiper fins aid displacement in the
between the conductor pipe and the casing. Cement is
larger-diameter casing. Because of their flexibility, the
mixed and pumped through the drillpipe and up the
plug will fold and pass through the smaller casing. First-
annulus until the slurry reaches the surface. As soon as
stage wiper plugs of the type used in stage cementing
mud contamination is no longer evident in the cement
could also be considered.
returns, mixing can be stopped and the drillpipe volume
displaced.
If lost circulation is noticed before the cement reaches
13-3 Cement placement procedures the surface, mixing should be stopped and the cement
Most primary cement jobs are performed by pumping the displaced, avoiding the pumping of large quantities of
cement slurry down through the casing and up the annu- cement into the fractured zone. Also, care must be taken
lus. Other techniques exist for solving various well-com- to avoid casing collapse because of excessive differential
pletion problems. For example, cementing through the pressure between the outer annulus and the drillpipe-

466 Well Cementing

 Float
collar

Float
shoe

Circulating Mud Pumping Cement Dropping Dart and Dart Landed; Pulling Out Stinger
Displacement End of Cementing

Fig. 13-7. Through-drillpipe stab-in cementing.

casing annular space. Special packoff cement head to the right for several turns, and the coarse threads
assemblies can be used to seal the drillpipe-casing annu- release the stab-in tool. Simpler stab-in tools are also
lus and allow pressure to be applied. This pressure serves commonly used that omit the latch-in design and simply
to offset the pump pressure that creates collapse loading rely on the drillpipe weight to hold the stinger in place
whenever inner-string cementing operations are con- while cementing. Special drillpipe centralizers central-
ducted. Alternatively, mud of an adequate weight can be ize the stinger and the last few joints of drillpipe, partic-
pumped in the drillpipe-casing annulus before stabbing. ularly in deviated wells.
Through-drillpipe cementing has several advantages. Collapsing the casing is the greatest risk in stab-in
Accurate hole volumes (most often unknown in conduc- cementing operations. This may occur if the annulus
tor or surface holes) are not required, because the becomes blocked for any reason. A preferred adaptation
cement slurry is mixed and pumped until returns are of through-drillpipe stab-in cementing is therefore
observed at the surface. This procedure optimizes the offered by using a cementing mandrel (Fig. 13-8) with
total volume of cement mixed and pumped and virtually the drillpipe (or tubing) hanging freely to within 15 to
eliminates the possibility of overdisplacement because 30 ft [4.6 to 9.2 m] of the shoe or collar. This type of
the subsequent volume displaced from the drillpipe is arrangement, often called inner string cementing, offers
negligible. This method also eliminates the need for the additional possibility of casing reciprocation. In
large-diameter swages or cement heads, as well as large- addition, unlike the stab-in technique, it can be used on
casing wiper plugs. Also, minimal cement contamination a floating rig, in which the drillpipe hangs underneath
occurs during through-drillpipe cementing. the conductor (or surface casing) wellhead-housing run-
Various options are possible with the through- ning tool. Above all, it eliminates the possibility of casing
drillpipe stab-in technique. A backup check valve (float collapse, because the pressures in the annulus and
collar and float shoe) can be run as depicted in Fig. 13-7. within the casing are equal. Pressure inside the casing
Alternatively, a stab-in float shoe alone could be used. (the drillpipe-casing annulus) can be monitored at the
The types of available stab-in tools offer the possibility to packoff head (on a stationary rig). However, during
latch into the float collar or shoe, thus preventing U-tubing, the column of fluid in the drillpipe-casing
pumpout of the stinger while cementing. Upon comple- annulus is not controlled, resulting in possible cement-
tion of the cementing operation, the drillpipe is rotated slurry contamination.

Chapter 13 Primary Cementing Techniques 467

 tional manner and pumped—often with centrifugal
pumps—only until cement slurry is circulated to the
surface. The lines and tubing are flushed with water, and
the tubing (if still hanging freely) is withdrawn from the
annulus (Fig. 13-9).
18-in. combination lift

Macaroni (small-ID) tubing

Mandrel

11-in. diameter
casing bore (279.4 mm)
O-ring
Tong area
Macaroni tubing
moved during job
Circulating port
Drillpipe to casing
adapter Fig. 13-9. Top-up cementing.

Collar
The cement slurry can also be mixed and pumped
directly into the annulus with the tubing string in place.
Free-hanging drillpipe In extreme cases, such operations may have to be
above shoe or collar repeated several times until the cement slurry returns to
Casing string to the surface and sufficient gel strength builds to support
be cemented the slurry until it sets. However, when attempting to fill
Fig. 13-8. Cementing mandrel.
the casing annulus from the surface, there is no method
to determine how deeply the cement has fallen, and the
casing annulus may not be uniformly cemented.
Small-diameter tubings are not as easy to use off-
13-3.2 Grouting (top-up cementing) shore. Therefore, a special tool, called a Titus Assembly,†
When lost circulation occurs during large-casing slurry is run as a contingency measure on the landing joint at
displacement, the immediate solution is to recement the same time as the casing. The primary cementation is
down the annulus. On land, a small-diameter tubing performed through the swivel in the open position. After
string is run down the annulus between the casing and the primary job is complete, a ball is dropped, which
the open hole (17⁄8-in. [5-cm] tubing is a common size). diverts the flow through the swivel down a hose to the
Several joints can be made up together and pushed down Titus Ring.† Slow circulation for approximately 3 hr,
the annulus as far as possible. The tubing string is then while waiting for the cement at the shoe to set, removes
connected to the cementing unit through a high-pres- any contaminated cement from the top of the annulus.
sure treating line, and circulation with drilling mud or Several 50-bbl [8-m3] top-up jobs are then performed
water is established. Caution must be exercised, because though the ring to ensure the placement of strong
friction pressures will be high because of the small cement to the surface.
tubing ID. Cement slurry is then mixed in the conven-
† Mark of Titus Tools

468 Well Cementing

 Exposed weak formations make cementing to the sur- role of the bottom wiper plug is to prevent heavy fluids
face problematic. When a top-up job is anticipated, (spacer or cement slurry) from falling through lighter
cement baskets are positioned within the previous ones (chemical washes and drilling fluid).
casing. They allow a first cementing operation to form a When a spacer is used, sinking of the cement slurry
ring of cement, used later after its initial set to support through the spacer will occur to a lesser extent, because
further top jobs until the cement reaches the surface. the density differential between the two is usually not
very large. If a plug is run between the spacer and the
slurry but not between the spacer and the mud, the
13-3.3 Single-stage cementing spacer will become contaminated with mud during the
With the development of new ultralow-density cement trip down the casing. In addition, the plug will sweep the
systems (Chapters 3 and 7), the need for multistage casing wall clean, pushing accumulated mud film ahead
cementing has been significantly reduced, if not elimi- and contaminating the last portion of the spacer. Once
nated. A long column of low-density, high-solids, or the bottom-plug diaphragm breaks, a mud-contaminated
foamed cement can often be placed in the annulus in spacer will be in contact with the cement slurry—a sit-
one stage without the risk of breaking down weak for- uation the spacer was supposed to prevent. Depending
mations. on the type of mud and its compatibility with spacer or
with cement, this situation can be circumvented by
13-3.3.1 Mud conditioning pumping a sacrificial volume of spacer or cement.
Clearly, the ideal situation would be to use two bottom
After the casing is in place, the mud is circulated as long plugs to avoid intermixing of all fluids during the trip
as necessary to remove high-gel-strength mud pockets down the casing. Three-plug cement heads are used at
formed during the semistatic period of removing the times (Fraser et al., 1996); however, these are not
drillpipe, logging, or running the casing. Mud circulation common. With present plug containers (cement heads),
is usually performed through the cement head to avoid more than one shutdown would be necessary to load the
stopping for an excessive period of time after the mud plugs.
has been conditioned. Under static conditions, mud gel The following sequences are normally recommended
strength can develop quickly and may greatly reduce the when only one bottom plug is used.
mud removal efficiency (Chapter 5). In case of gas cut- ■ Bottom plug—spacer—cement slurry
ting, mud circulation must continue until a steady in-out
■ Wash—bottom plug—spacer—cement slurry
mud weight is obtained. In addition, cementing should
not take place before the well is completely static. ■ Wash—bottom plug—cement slurry
If a single-plug cement head is used, circulation must
As a general practice, the bottom wiper plug should
be stopped before cementing to load both cement plugs.
be placed between the fluids of highest density differ-
The bottom plug must be placed below the lower 2-in.
ence, provided a sacrificial volume is tolerated for the
[5-cm] inlet to allow room for the upper plug between
contamination between the other fluids, usually spacer
the two inlets. If a double-plug cement head is used,
and cement. These latter fluids are designed to be rheo-
both cement plugs can be loaded before starting mud
logically compatible, but contamination of the cement
circulation.
by the spacer may result in retardation. Therefore, the
bottom plug is often dropped in the last portion of the
13-3.3.2 Bottom wiper plug spacer, meaning that the actual sequence is
The bottom cement wiper plug is a key element in a ■ Spacer—bottom plug—spacer
cementing operation. It serves two functions: it prevents (last portion or sacrificial volume)—cement slurry.
the intermixing of fluids and it sweeps clean the inner
wall of the casing. The most obvious function of the
bottom plug is to prevent mud from contaminating the 13-3.3.3 Displacement procedures
cement slurry. Fraser et al. (1996) and Griffin and Valkó Dropping the top wiper plug is an easy operation and
(1997) demonstrated that failing to use a bottom wiper should not take longer than the time needed to open and
plug was the cause of many primary cementing failures. close the valves at the cement head. Cement heads are
However, if the bottom wiper plug is properly located, it very reliable tools under normal working conditions and
can also contribute to the conservation of spacer or are designed to minimize time delays. Stopping circula-
slurry properties. As discussed in Chapter 5, rheology tion for long periods of time (5 to 10 min) allows down-
has a crucial effect on the ability of a spacer fluid to hole fluids to develop high gel strength. Remotely oper-
remove mud. A small amount of mud contamination can ated cement heads (Lavaure and Galiana, 1991)
alter characteristics and reduce effectiveness. Another facilitate the wiper plug dropping exercise and increase

Chapter 13 Primary Cementing Techniques 469

safety by not requiring someone to climb the mast to Surface pressure is then released, and backflow is
operate the valves and launch the plugs. There are two observed to test the functioning of the float equipment.
principal consequences when plug launching is delayed. If no returns are observed, the line is left open during
■ Additional applied pressure may be required to the waiting-on-cement (WOC) period. If the float-collar
restart fluid movement. In extreme cases, this pres- valve does not hold the backpressure, the fluid returned
sure may exceed the fracture pressure resulting in during the test must be pumped back into the well, leav-
lost circulation. ing the casing pressurized until the cement gels and
■ Poor removal of gelled mud may occur, leading to poor
loses mobility. However, it is very important to bleed off
the excess casing pressure (caused by thermal expan-
zonal isolation.
sion) before the cement begins to develop compressive
The spacer and slurries are then displaced through strength; otherwise, a microannulus may form because
the casing, isolated between the two wiper plugs. In real- of expansion and contraction of the casing.
ity, because of U-tubing (Chapter 12), the top of the Once the cement slurry has started to set, the normal
cement column may be at a considerable depth below practices and operations of putting the string in tension
the surface at the time the top plug is released. and landing it in the wellhead can be performed accord-
Depending upon the mud density and the cement ing to the casing landing procedure (see Section 13-7.8).
volume and density, the first part of the cement column
may have already rounded the shoe. Computer programs
for cement-job design (Chapter 12) can predict such 13-3.4 Multistage cementing
phenomena. The rate at which the cement is displaced As mentioned previously, high-performance lightweight
into the annulus is not the same as the pumping rate; cements have made multiple-stage cementing a rela-
instead, it varies depending upon the different fluid den- tively rare event. However, multiple-stage cementing
sities and volumes. This phenomenon continues until may still be necessary if
the fluid level inside the casing reaches the surface. ■ downhole formations are unable to support the hydro-
Continuous flow can then take place. static pressure exerted by a long column of cement
In general, there is a tendency to disregard the impor- ■ the upper portion of the annulus must be cemented
tance of displacement rates during the period of with a higher-density cement system
U-tubing. If the job is designed for turbulent flow, the
■ cement is not required between widely separated
maximum pump rate possible is recommended during
this period, because downhole fluid velocities are proba- intervals.
bly one-third to one-half of the surface pumping rate. For Most multiple-stage cement jobs are performed to
jobs pumped in laminar flow, the surface rate must be alleviate high hydrostatic pressure. It is not uncommon
controlled to maintain the desired flow regime. Again, to cement a long string to the surface to protect the
computer programs can calculate the optimal pump casing from corrosion. Another scenario might be the
rates according to the well geometry and fluid proper- presence of lost circulation zones below the last casing
ties. A tool to control U-tubing was introduced by Head shoe that prevents cement slurries from reaching the
et al., 1995. It did not gain popularity because the main surface. Two-stage cementing, with the top of the first-
use was in large casings. To be efficient, such tools must stage covering the weak zone, will permit safer, more
restrict the flow rate; as a result, the duration of the complete filling of the total annular space.
cement job increases and, owing to the low flow rates, Three standard multistage techniques are commonly
mud removal may be questionable. employed.
Once continuous flow is restored, the annular flow ■ regular two-stage cementing, in which the cementing
rate is equal to the pump rate, and the surface pressure of each stage is a distinct and separate operation
begins to increase as the rest of the cement is placed
■ continuous two-stage cementing, with both stages
behind the casing. Displacement then continues at the
programmed rate until the top wiper plug bumps in the cemented in one continuous operation
float collar. However, the pump rate is usually reduced at ■ three-stage cementing, in which each stage is

the end of the displacement to avoid a sharp pressure cemented as a separate operation
increase when the plug reaches the collar. On bumping The longer execution time of stage cementing
the plug, one should watch for leaks. If pressure holds increases the rig time. In addition, most cement heads
after bumping, the casing can be immediately pressure cannot accommodate the preloading of all the plugs and
tested, provided that the plugs and collar have been bombs required in the operation sequence. As a result,
selected to withstand such extra differential pressure the cement head must be opened to release the opening
without collapsing or breaking.

470 Well Cementing

bomb, assuming the first-stage plug was preloaded. The the opening bomb to reach the collar by gravity and seal
shutoff plug could be loaded after the bomb is released, properly to operate the opening sleeve. However, even
but caution should be exercised to ensure that all of the when using a hydraulically operated tool, it is prudent to
plugs and bombs are compatible with the cement head. have an opening bomb available as a contingency.
In addition, the plugs should be carefully checked for Figure 13-10 shows the operation of mechanical and
correct fitting in the head. hydraulic stage tools.
Hydraulically operated stage collars are commonly
used in deviated wells in which it would be difficult for

Closing
plug
Brass shear balls

Closing sleeve

Opening sleeve Fluid ports Double

Brass shear seals
balls
Antirotation Broken shear Lock
mechanism balls ring
Free-fall
Body connection
opening device
with metal-to-
metal seal and
elastomer
backup

Running Position Opened Position Closed Position

Closing
plug
Brass shear balls

Closing sleeve

Differential Fluid ports Double

opening sleeve seals

Brass shear
balls Broken shear Lock
balls ring

Body connection
with metal-to-
metal seal and Shouldered
elastomer antirotation
backup mechanism
(not shown)

Running Position Opened Position Closed Position

Fig. 13-10. Operation of mechanical (top) and hydraulic (bottom) stage collars.

Chapter 13 Primary Cementing Techniques 471

13-3.4.1 Conventional two-stage cementing cementing jobs. The first-stage slurry should cover the
In addition to conventional casing equipment (e.g., stage collar. Any excess can be circulated out when the
guide shoes and float collars), a stage-cementing collar ports are opened. Some types of stage collars allow the
(Chapter 11) is run to the desired depth. There are sev- use of first-stage wiper plugs. First-stage displacement
eral types of two-stage collars. It is important to be com- plugs are mandatory and must be compatible with the
pletely familiar with the operation of the selected type original stage collar and the float collar. Plugs and stage
and to follow the manufacturer’s operating recommen- collars from different manufacturers should never be
dations. All stage collars must be handled with care, as mixed.
they are manufactured to close tolerances. Smooth slid- It is a common mistake to leave a portion of the first-
ing and sealing of the concentric sleeves is necessary for stage openhole section uncemented. Unless an annular
proper operation. Rough handling before or during packer is coupled to the stage tool, the second-stage
installation can “egg” or misalign the moving parts, caus- cement slurry will fall downhole, resulting in a poor
ing a job-execution failure. One must also be absolutely second-stage cement job.
sure that the float collar and the stage collar are com- In case of high incompatibility between the cement
patible. The first-stage wiper plug (if used) and the first- and mud, it may be desirable to run a bottom wiper plug
stage displacement plug must fit and seal against the ahead of the slurry in the first stage. To do so, the fol-
float collar. lowing additional equipment must be used.
To explain the sequence of stage cementing opera- ■ Flexible plug: This special wiper plug is pumped

tions, a brief explanation of the equipment is necessary ahead of the first-stage slurry.
(Fig. 13-5). Conventional stage equipment consists of ■ Bypass insert: This part, located above the float collar
the following. or float shoe, provides a seat for the flexible plug but
■ A stage cementing collar is basically a casing joint allows continued circulation of slurry through its
with ports, which are opened and closed or sealed off ports.
by pressure-operated sleeves. ■ Special insert collar: This collar, located one casing
■ A rubber sealoff plate is a part installed in the top joint above the bypass insert, provides a seat for the
float collar to assure a positive shutoff. special first-stage plug that follows the cement.
■ A first-stage plug is a rubber plug used to separate the ■ Special first-stage plug: This plug, provided with a

slurry from the displacement fluid; it gives a positive head that is shaped to seal off in the insert collar,
indication of the end of displacement. replaces the first-stage plug in conventional stage
■ An opening bomb is a device that is dropped after the equipment.
first stage and allowed to gravitate to the opening The sequence of operations is similar to that of the
seat in the stage collar. Subsequent application of conventional two-stage cementing procedure, except
pressure will move the sleeve downward, opening the that the additional wiper plug is launched ahead of the
collar’s ports. first-stage slurry or spacer.
■ A closing plug is a rubber plug that is pumped to a
shutoff on the closing seat. 13-3.4.1.2 Cementing the second stage
After the first stage is completed, the opening bomb of a
13-3.4.1.1 Cementing the first stage mechanically operated stage collar is dropped and
The mixing and pumping of spacers and slurries during allowed to fall by gravity to the lower seat in the stage
the first stage are similar to those of a single-stage job, collar. Once the bomb is seated, pressure is applied until
except that in most cases there is no bottom wiper plug. the retaining pins are sheared, forcing the lower sleeve
After the mixing of the slurry, the first-stage plug is to move downward and uncover the ports. Usually 1,200
dropped and displaced until it lands in the float collar. to 1,500 psi [8.4 to 10.5 MPa] will shear the retaining
When cementing production strings, some operators dis- pins. A sudden drop in surface pressure indicates the
place the first stage using two fluids, leaving the casing opening of the ports. This operation is performed as soon
below the stage collar filled with completion fluid and as possible after the completion of the first stage. Under
the upper casing filled with drilling mud. This mud is normal circumstances, the excess cement from the first
subsequently used to circulate the hole through the stage will sit above the stage collar and must be circu-
stage-collar ports. lated out of the hole before it develops excessive gel
Accurate hole volumes are necessary to determine strength.
the correct slurry height in the annulus; therefore, a
caliper log should be mandatory on all multistage

472 Well Cementing

 When using a hydraulically operated stage tool, the
displacement of the first stage must be complete; i.e.,
the first-stage plug must have landed on the float collar.
Then pressure is gradually increased until the opening
pressure is attained.
Once the stage-collar ports have been opened, the
well must be circulated until the mud is conditioned for
the second stage and the first-stage cement is set.
Otherwise, the weak zones along the first-stage section
may not withstand the hydrostatic pressure from the
second-stage cement column. For cementing the second
stage, spacers and slurries are mixed as in any single-
stage job. The closing plug is dropped after the slurry
mixing and displaced to its seat in the stage collar. After
the plug has seated, a minimum of 1,500 psi [10.5 MPa]
above the second-stage displacing pressure is required
to close the stage-collar ports. Pressure is usually
released from the casing after the ports are closed.
Most second stages of two-stage jobs are performed
using low-density filler slurries to allow circulation to
the surface. Normal-density slurries are used when a
high-pressure zone or an aquifer must be thoroughly iso-
lated. Protection of the weakest point in the casing
string, the stage collar, can be improved by simply
increasing the density of the last portion of the cement
slurry.

13-3.4.2 Continuous two-stage cementing

Sometimes the situation demands that the cement be
mixed and displaced without stopping to wait for an
opening bomb to gravitate to the seat in the stage collar. Fig. 13-11. Continuous two-stage cementing.
This is known as the continuous-stage cementing
method (Fig. 13-11). This method is presented as a his-
torical reference, because it is rarely used. This tech- off when the plug lands and permits some tolerance in
nique requires tight control of fluid volumes and does the displacement fluid volume. After this displacement
not address the objective of a multistage operation—to fluid has been pumped, the stage-collar opening plug is
isolate the top section when lower weak zones are pre- released.
sent. The second stage of cement may be pumped immedi-
The first stage of cement is mixed and pumped into ately behind the opening plug. The closing plug follows
the well. A wiper plug follows the cement to separate it this slurry. Displacement of this slurry will cause the
from the displacement fluid. Following the plug, suffi- opening plug to seat on the opening sleeve. When pres-
cient water or mud is pumped to displace the cement out sure is applied, the sleeve will open the collar ports.
of the casing below the stage collar. Allowance must be Further pumping will displace the slurry through the
made for compression and pipe stretch so that the ports and eventually land the closing plug on the closing
cement is not overdisplaced around the casing shoe. seat. Application of 1,500 psi [10.5 MPa] above the cir-
Overdisplacement is also possible owing to a bypass culation pressure will close the tool.
insert that is installed above the float collar on which
the cement wiper plug lands. This insert prevents a shut-

Chapter 13 Primary Cementing Techniques 473

13-3.4.3 Three-stage cementing to shut off at the float collar. The second stage could be
The three-stage cementing technique is also presented performed at any time after the first, depending upon
as a historical reference, because it is rarely applied. It the cement program.
multiplies the operating time, complicates the opera- A regular opening bomb is used to open the ports of
tion, increases the number of weak points in the string, the lower-stage collar. The well is circulated, and the
and significantly increases the failure risk. A single spacers and slurries are pumped through the ports. The
collar malfunction is sufficient to ruin the entire job. ports are closed using a special closing plug, which
Weak zones in deep wells, combined with gas chan- replaces the regular closing plug. This flexible type of
neling or potential casing corrosion problems, may plug passes through the top-stage collar and seats on the
require a three-stage cement job. The basic procedure is lower collar, allowing the application of pressure to close
similar to two-stage cementing (Fig. 13-12). The first the ports.
stage is performed through the shoe, the second through The final stage can also be performed at any time
a regular two-stage collar, and the final stage through a after the second one. An opening bomb (larger than the
top-stage collar. The first stage is performed through the one used for the second stage) is dropped and allowed to
shoe in the conventional manner, using a first-stage plug gravitate to the lower seat of the top-stage collar. Ports
are opened, and the final stage is performed as usual. A
special closing plug is then used to close the collar ports.
Figure 13-13 summarizes the different types of plug
used in multistage cementing operations.

Overlap to isolate
Three-Stage Continuous Two-Stage with Two-Stage
Two-Stage Plug Bypass
Third-stage collar
Third stage

Closing plug Closing plug Closing plug Closing plug

Weak Zone 2

Free-fall Pumpdown
Zone to isolate opening device opening plug Free-fall Free-fall
opening device opening device

Second-stage collar
Second stage

Closing plug

Weak Zone 1 Bypass plug First-stage First-stage

shutoff plug sealing plug

Shutoff baffle
Free-fall
opening device
First stage

First stage

First-stage
sealing plug Bypass plug

Fig. 13-13. Plugs used in multistage cementing operations.

Fig. 13-12. Three-stage cementing.

474 Well Cementing

13-4 Liners 13-4.1 Running a liner
A liner is a string of standard casing that does not extend Like a standard casing string, a liner is assembled joint-
all the way to the surface; instead, it is hung from inside by-joint at the rotary table and lowered into the well.
the previous casing string. The overlap depends on the When liner rotation is planned during the cement job,
purpose of the liner and could vary from 50 ft [15 m] for premium connections that can withstand high torque
drilling liners to as long as 500 ft [152 m] for production are required. Float equipment is also included, and
liners. Liners can be classified as follows (Fig. 13-14). sometimes a landing collar (typically one joint above the
■ Production liners are run from the last casing to total float collar) is used to provide a seat for the liner wiper
depth. They replace production casing. Cementing is plug. The dart and plug system must be compatible with
usually critical, because zonal isolation is essential the float collar.
during production and any subsequent stimulation Centralizers are essential in liner cementing.
treatments. Because annular clearances are usually small, the liner
must be kept clear of the borehole wall. The mud dis-
■ Drilling or intermediate liners are set primarily to
placement efficiency improves significantly with better
case off and isolate lost-circulation zones, overpres- centralization (Chapter 5). Centralizers also help pre-
sured zones, sloughing shales, or plastic formations. vent the liner from becoming stuck while running in the
This is necessary to resume drilling. Cementing these hole and make liner reciprocation and rotation during
liners is often difficult. cement displacement easier and more effective.
■ Tieback stub liners extend from the top of an existing A liner hanger is placed at the top of the liner
liner to a point uphole inside another casing. They (Fig. 13-15). The hanger is often combined with a packer,
are generally used to repair damaged, worn, corroded, or sometimes provision is made for a tieback receptacle
or deliberately perforated casing above the existing to allow running an isolation packer at a later date, if
liner and to provide additional protection against required. When set at the desired depth, this supports
corrosion or pressure. In addition, they must be
designed to perform within the casing-design enve-
lope for the well.

Seal
Tieback casing Wiper plug

Slips

Intermediate casing

Drilling liner

Tieback stub liner

Landing collar

Production liner
Float shoe

Fig. 13-14. Types of liners. Fig. 13-15. Liner setting tool and hanger assembly.

Chapter 13 Primary Cementing Techniques 475

the weight of the liner string. Therefore, the liner is kept setting tool are then raised slightly to verify that the set-
under tension, which prevents it from buckling under its ting tool is released from the liner. The seal assembly
own weight. The liner hangers have slips that, when set, holding the liner wiper plug is usually 10- to 15-ft [3.0- to
close tightly into the upper casing and provide 4.6-m] long to enable this operation to be performed
the anchor to suspend the liner. There are two types of without breaking the seal between the liner and the
liner hangers, depending on the hanger setting mecha- drillpipe.
nism: hydraulic (pressure-activated) or mechanical. The If reciprocation is planned during cementing, the
choice between the two types of hanger primarily hanger can only be set after cementing.
depends on the rig type (floater or stationary rig) and
the well type (deviated or not). In addition, the hanger
design may provide the ability to rotate and/or recipro- 13-4.2 Liner cementing procedure
cate the liner. The liner hanger remains permanently in 13-4.2.1 Mud removal
place once the liner is cemented. In addition, the liner The success of any cement job depends upon the mud
hanger may be coupled to or complemented with a removal efficiency. Liner cementing can be one of the
packer. most difficult cases. The annular space is small, and the
Liners are usually run into the well using drillpipe pipe may not be well centralized. This subject is covered
and a special setting tool. The setting tool is retrievable; thoroughly in Chapter 5; nevertheless, there are certain
i.e., it is pulled out of the well with the drillpipe after points that deserve reinforcement.
the liner is run and cemented. It performs the following A 5-in. OD liner, hung from a 7-in. casing inside a
functions. 6 1⁄8-in. drilled open hole, will have a maximum clearance
■ It provides a pressure-tight seal between the drillpipe of 9⁄16 in. [1.4 cm] if the liner is perfectly centered. The
and the liner. Thus, fluids pumped into the drillpipe annular clearance will be less in some parts of the hole
must circulate down inside the liner and out of the because of a thin, nonremovable mudcake on the wall
shoe before returning up the annulus. of permeable formations. Crooked hole and small
■ It holds the weight of the liner as it is run into the clearances between the casing and formation often
well. inhibit the use of normal centralizers, resulting in liner
■ It provides attachments for the liner wiper plug eccentering. Under severe conditions, actual borehole
system. The liner wiper plug, attached by shear pins, contact occurs. Under these circumstances, it becomes
has a hole through its center to allow the passage of much more difficult for cement slurries to remove mud.
fluids and cement slurry until the drillpipe pump- Small clearance centralizers, including those that allow
down dart closes it. Applied pressure will then shear liner rotation, are available and should be used to
the pins, and the wiper plug can be pumped down the increase the chance of success of this critical cementing
liner behind the cement slurry. Because this system operation.
does not accommodate a bottom plug, drilling fluid It is for these reasons that pipe movement during dis-
contamination inside the liner often occurs, resulting placement becomes critical. Bowman and Sherer (1988)
in poor isolation. This is particularly troublesome in reported that less than 20% of liner jobs include plans
long liners. Systems with bottom and top plugs, called to move the liner during cementing. There are many
two-plug-dart systems, are available but are not industry concerns about liner reciprocation and rotation:
widely used. ■ not becoming unlatched from the liner after
cementing
During the last trip out of the hole before a liner job,
■ that a large, stronger drillstring may be required for
it is a common practice to gauge or drift the landing
string to ensure the passage of the balls and pumpdown fear of drillstring parting during pipe movement
dart. ■ excessive drag caused by centralizers

With the liner at the desired depth, but before the ■ swabbing or surging the pay zone
hanger is set, connections are made and the liner and ■ hole deterioration caused by moving pipe, which
hole are completely circulated. This conditions the mud could lead to annulus bridging
and ensures that circulation is possible before the liner
■ that the liner may become stuck and have to be
is hung. In some deep liner-setting assemblies, a circu-
lation valve is included, which allows circulation to be cemented without the designed tension.
established above the liner before closing the valve.
If the liner is not or cannot be reciprocated during
cementing, the liner hanger is set, and the drillpipe and

476 Well Cementing

 In fact, the advantages of liner movement during the chances of a good cement bond. In conventional
cementing far outweigh the concerns listed above. With casing strings, contact time can be improved by simply
the hole in good condition and correctly selected cen- increasing the volume of the scavenger slurry. However,
tralizers on the liner, fewer problems should be experi- in a liner situation, slurry volumes can be critical
enced, and certainly better cementing results should be because of the formation’s hydrostatic-pressure limit.
achieved. Bowman and Sherer (1988) stated that, in Turbulent flow displacement is efficient for mud
their study of more than 300 liner jobs, the inability to removal, but care must be taken not to exceed allowable
release the liner setting tool had only occurred twice. downhole pressures. Fortunately, small annular clear-
Premature cement setting caused one failure, and the ances make it easier to accomplish turbulent flow at low
other involved a very early tool design that was subse- pump rates. If a job must be performed in laminar flow,
quently modified. spacer volumes, rheological properties, and rates may be
Rotation frequently has advantages over reciproca- adjusted to allow for the adequate mud displacement.
tion. If the liner were in contact with the hole at any
point, up-and-down motion would not remove the
drilling fluid effectively (Chapter 5). However, pipe rota- 13-4.2.2 Regular liner cementing
tion would allow slurry to be dragged behind the pipe, The liner cement head and manifold are installed on the
ensuring a cement sheath all around the liner. As stated drillpipe with the pumpdown slurry displacement dart
by Bowman and Sherer (1988), the inability to rotate placed between the two inlets. The dart-releasing stem
liners is often caused by insufficient starting torque. holds the dart in the cement head (Fig. 13-16). After the
Once this has been overcome, the torque required for cementing lines are rigged up and pressure tested, the
rotation will probably be much less (assuming good cen- chemical wash or spacer is pumped down the drillpipe.
tralizer design). Usually, a bottom wiper plug is not used ahead of the
Because of the above problems, using adequate vol- spacer or slurry. However, two-plug liner cementing sys-
umes of washes and spacers is even more critical in liner tems exist and are preferred. Figure 13-17 shows the
cementing than in casing cementing. Maximizing the single and tandem liner plugs.
contact time of the spacer or wash generally increases

Integral Liner Cement Head Topdrive Liner Cement Head System

Topdrive manifold

Manifold and Drillpipe pumpdown

dart-dropping dart (enlarged)
component

Ball-dropping
component
Dart-dropping container

Ball-dropping sub
(not pictured)

Fig. 13-16. Liner cementing heads.

Chapter 13 Primary Cementing Techniques 477

 Upper liner-wiper plug

Drillpipe pumpdown dart

Upper pumpdown plug

Lower liner-wiper plug

Lower pumpdown plug

Liner wiper plug
(with ball seat)

Single Liner Plug and Tandem Liner Plugs and

Pumpdown Dart Their Pumpdown Darts

Fig. 13-17. Liner cementing plugs and dart systems.

Figure 13-18 shows the typical steps of liner cement- liner wiper plug. The surface pressure will rise when the
ing with a single plug system, and Fig. 13-19 depicts the dart lands. Further applied pressure, approximately
operation with a two-plug liner system. If possible, the 1,200 psi [8.4 MPa], will shear the pins that hold the
cement slurry should be batch-mixed to obtain a homo- liner wiper plug in place. Such a pressure peak is often
geneous slurry at the proper density. Once the slurry is easily detectable from surface. Therefore, this is a refer-
mixed and pumped into the drillpipe, the pumpdown ence point in the displacement volume because, until
dart is dropped and displaced to the liner hanger. At this this moment, only the drillpipe has been displaced. After
point, the pumpdown dart passes through the liner set- this moment, only the liner is displaced.
ting tool and then latches into and seals the hole in the

478 Well Cementing

 Mud Circulation Pumping Wash and Cement Slurry

Plug releasing pins

Pumpdown plug Cement slurry

Liner hanger

Wash

Landing collar

Float shoe

Displacement Displacement

Displacement Setting Tool Released from Packer

Fig. 13-18. Typical liner cementing sequence.

Chapter 13 Primary Cementing Techniques 479

 Drillpipe Drillpipe

Setting collar

Liner hanger

Upper liner- Upper liner-

wiper plug wiper plug

Lower Upper pump-

pumpdown down plug
plug
Upper liner-
Lower wiper plug
liner-wiper
plug
Landing collar Lower
liner-wiper
plug
Lower
pumpdown Lower
plug pumpdown
plug
Setshoe

Launch lower pumpdown Continue pumping Displace cement in Displace cement in liner.
dart and pump cement. cement. Lower plug drillpipe. Upper dart Plug lands on float collar.
Lower dart launches lands on float collar launches in upper plug. Remove liner setting tool
in lower plug. and reverse out.

Fig. 13-19. Liner cementing with two liner plugs.

Once released, the dart-plug combination moves as into the hanger, the reversing out depends on the quan-
one plug inside the liner while displacement continues. tity of excess cement expected and whether lost circula-
When the internal volume of the liner has been com- tion is observed. This is an important decision in liner
pletely displaced, the plugs seat on the float or landing cementing design, because proper isolation of the liner
collar and another pressure rise occurs, indicating job overlap is critical. Software helps to make sure the well
completion. Monitoring the returns after bleeding off is kept under control when reversing out. If gas releases
the pressure allows one to test the float equipment. (kicks) or losses occur, the quality of isolation may be
It can be noted that the displacement of the cement compromised.
slurry for a liner generally takes place using the cement- The amount of cement excess must be carefully cal-
ing unit rather than the rig pumps, because of the rela- culated by taking into account the well conditions and
tively small volume to displace. Hence, control of the dis- operator requirements. The following factors must be
placement volume is made much easier. balanced.
If a packer-type liner hanger has been used, the ■ Sufficient excess cement slurry must be available to
packer between the liner and the upper casing may be ensure the placement of uncontaminated cement in
set at this time, the setting tool is pulled free from the the liner overlap. A caliper with at least four arms
liner hanger, and any excess cement is reversed out. If should be run before the liner operation, and the
lost circulation is observed while displacing cement, the slurry volume should be determined from the caliper
packer is not set, thus allowing eventual squeezing of logs. Graves (1985) pointed out that hole volumes
cement in the liner-casing annular space (also called could vary by as much as 31%. When the liner is not
liner overlap) (Chapter 14). If no packer is incorporated too long, it is common practice to use a slurry volume

480 Well Cementing

 corresponding to the annular volume with its normal fore, the excess slurry volume may have to be limited.
excess, plus the overlap volume between the liner and This could definitely affect the quality of cement
the casing, plus any volume above the hanger. This around the overlapped interval.
procedure typically doubles the slurry volume but sig-
Once the job is completed, the setting tool and
nificantly increases the success by reducing cement
drillpipe are pulled out, leaving the cement to cure
contamination.
throughout the recommended WOC time. A checklist for
■ Displacement efficiency also becomes a key variable running liners is published in API Bulletin D17 and is
in determining cement-slurry volumes. Although reproduced in the sidebar.
100% efficiency is the ideal, it is not uncommon to To reduce slurry contamination inside the liner, par-
have 60% to 80% displacement efficiency in liner ticularly when the liner is fairly long, a two-plug system
cementing (Smith, 1990). The efficiency suffers as called a dual liner wiper plug system or tandem liner-
the interval length increases. wiper plug system is preferred (Fig. 13-17). This system
■ If excess slurry is to be reversed out while not having features two darts to launch their corresponding plugs.
a liner top packer, weak formations could pose a prob- An inner string liner cementing process was proposed by
lem. Also the thickening time of such slurries must be Fuller et al. (1998) which, in addition to reducing conta-
extended to allow for the reversing operation. mination inside the liner during the cementing opera-
■ If reversing out is not scheduled, operators usually do tion, also allows cleaning the liner after cementing and
not want to drill out long columns of cement; there- displacing the well before tripping out of the hole with
the hanger setting assembly.

Lining Running Procedures Checklist†

The following procedure is taken from API Bulletin D17. The 6. Circulate bottoms-up with __ bbl/min to achieve __ ft/min
reader will note that it should be modified if the intent is to annular velocity (approximately equal to previous annular
reciprocate or rotate the liner. The procedure is as follows. velocities during drilling operations).
1. Run drillpipe and circulate to condition hole for running liner. 7. Cement liner as follows: _________________.
Temperature subs should be run on this trip if bottomhole circu- 8. If unable to continue circulation while cementing because of
lating temperatures are not known. Drop hollow rabbit (drift) to plugging or bridging in liner and hole wall annular area, pump
check drillpipe ID for proper pumpdown plug clearance. On trip on annulus between drillpipe and liner to maximum __ psi and
out of hole, accurately measure and isolate drillpipe to be used attempt to remove bridge. Do not overpressure and fracture
to run the liner. Tie off remaining drillpipe on the other side of the formation. If unable to regain circulation, pull out of liner
the racking board. and reverse out any cement remaining in drillpipe.
2. Run __ ft of __ liner with float shoe and float collar spaced __ 9. Slow down pump rate just before pumpdown plug receives
joints above float collar. Volume between float shoe and plug the wiper plug. Drillpipe capacity is __ bbl. Watch for plug
landing collar is __ bbl. Sandblast joints comprising the lower shear indication, recalculate or correct cement displacement,
1,000 ft and upper 1,000 ft of the liner. Run thread-locking com- and continue plug displacement plus __ bbl maximum
pound on float equipment and bottom eight joints of liner. Pump overdisplacement.
through the bottom eight joints to be certain that float equip-
ment is working. 10. If no indication of plug shear is apparent, plug calculated
displacement volume plus __ bbl (100% + 1% to 3%).
3. Fill each 1,000 ft of the liner while running, if automatic fill-up
type equipment is not used. 11. Pull out 8 to 10 drillpipe stands or above top of cement,
whichever is greatest. Hold pressure on top of cement to
4. Install liner hanger and setting tool assembly. Fill dead space (if prevent gas migration until cement sets.
packoff bushing is used in lieu of liner setting cups) between
liner setting tool and liner hanger assembly with inert gel to 12. Trip out of hole.
prevent solids from settling around the setting tool. 13. Wait on cement __ hr.
5. Run liner on __ (size, type connection, weight, and grade) 14. Run __-in.-OD bit and fill cement to top of liner. Test liner over-
drillpipe with __ pounds minimum overpull rating. Run in hole at lap with differential test, if possible. Trip out of hole.
1 to 2 min per stand in casing and 2 to 3 min per stand in open
hole. Circulate last joint to bottom with cement manifold 15. Run __-in.-OD bit or mill and drill out cement inside liner as
installed. Shut down pump. Hang liner 5 ft off bottom. Release necessary. Displace hole for further drilling. Spot perforating
liner setting tool and leave 10,000 lbm of drillpipe weight on set- fluid (if in production liner) or other conditioning procedures
ting tool and liner top. as desired.

†From Bowman and Sherer (1988). Reprinted with permission from World Oil.

Chapter 13 Primary Cementing Techniques 481

13-4.2.3 Planned squeeze The advantages of the method are
Long liners can be cemented in two stages when they ■ it avoids damaging weak productive formations
traverse weak formations that would not withstand the ■ uncontaminated cement is placed at the liner hanger
hydrostatic pressure of the entire cement column. As ■ no excess cement is necessary.
illustrated in Fig. 13-20, the first stage is performed
through the shoe using a limited, calculated amount of Disadvantages include
cement to cover the weak zone and place the cement top ■ the complete annular space around the liner in the
as close to the last casing shoe as possible, but not in the overlap may not be cemented
overlap. After the first stage is completed, the setting ■ it may not be possible to inject anything (at safe pres-
tool and drillpipe are pulled out of the hole and the sures)
cement is allowed to cure. Drillpipe with a standard
■ the technique is more expensive.
squeeze packer is then run into the hole, and the packer
is set two or three joints above the liner hanger. The With the advent of high-performance ultralow-density
second stage consists of squeezing a premixed amount of and foamed cement systems, this procedure is rarely
cement around the liner hanger. To be able to squeeze performed.
cement into the annular space, the formation fracture
pressure in the openhole section must be overcome.
13-4.2.4 WOC for liners
When long liners are set, there may be a considerable
First Stage Squeeze Stage temperature difference between the bottom and the top
of the liner. A slurry designed to have sufficient thicken-
Squeeze ing time at the total depth may take longer to set at the
packer liner top. A range of cement retarders (Nelson and
Casabonne, 1992) allows the cement slurry to set and
Liner develop strength quickly, even when a large temperature
hanger difference exists between the circulating temperature
and the static temperature at the liner top (Chapter 3).
Drilling of cement must be performed after the cement
develops the minimum compressive strength at the top
of the liner to withstand the shock caused by drilling
tools.

13-4.2.5 Tieback liners

There are situations when it may be necessary to extend
an existing liner farther uphole, with a tieback “stub”
liner, or to surface with a tieback casing string. For
example, from a well-architecture and casing-design
point of view, if a well must be cased from bottom to sur-
face, it will be much easier and safer to run and cement
the string as a liner first. Later, one can connect and
eventually cement a tieback string. Other reasons for
running tieback stub liners or tieback casings include
Landing
■ covering up damaged casing above the top of an exist-
collar
ing liner
■ installing larger-diameter casing above the existing
liner to allow for multiple production strings
■ selective testing of multiple zones to design future
Float production assemblies and the production casing size
shoe

Fig. 13-20. Planned squeeze after liner cementing.

482 Well Cementing

■ cementing troublesome intervals (e.g., high pressure, Tieback sealing nipple
sloughing shales) before running the casing string to Run at the bottom of the tieback stub liner or casing, the
surface. tieback sealing nipple (Fig. 13-22) has multiple packing
To accomplish this, special tools must be used to con- and sealing elements, which provide a seal against the
nect the two liner strings. polished surface of the tieback sleeve. The arrange-
ments differ depending on pressure and temperature
Tieback sleeve conditions and whether an additional metal-to-metal
Also called a tieback receptacle, polished-bore recepta- seal is required.
cle, or packer-bore receptacle, the tieback sleeve Tieback casings are usually cemented by convention-
(Fig. 13-21) is either installed on top of the liner hanger ally circulating the slurries. The job is performed before
or it comes as an integral part of the liner packer itself. landing the seal nipple into the tieback sleeve. However,
The tieback sleeve provides a receptacle for the sealing the cementing may also be conducted with the tieback
nipple. Its internal surface is usually polished and casing in place, using a stage collar above the sealing
beveled on the top to guide the entry of the different nipple.
tools used during the operation.

Standalone Sleeves

Receptacle
(polished
bore area)

Sleeve Integral to Liner Packer

Receptacle
(polished
bore area)

Fig. 13-22. Different seal arrangements for the tieback sealing

nipple.

Mechanically set
liner packer
Hydraulically set
liner packer

Fig. 13-21. Examples of tieback sleeves.

Chapter 13 Primary Cementing Techniques 483

 Tieback liners must be cemented after their liner formed between casings. The same care must be exer-
hangers have been set with the seal nipple landed into cised with slurry designs and cementing procedures as
the sleeve (Fig. 13-23). A stage collar can be run on top for a regular primary cementing operation. A tieback
of the seal nipple in the open position. The liner wiper casing is often a key design element for well integrity.
plug must be able to land on the upper seal and close the The temperature difference between the bottom and top
collar ports. of the tieback string may be significant. Slurries must be
Apart from the special procedures given above, the stable at all downhole temperatures. Although light-
considerations applicable to all cement jobs also apply weight slurries are more economical, normal-density
to tieback liner cementing. In most cases, hydrostatic slurries are preferred because they are much more effi-
pressures are not significant because cementing is per- cient in terms of displacement mechanics.
The use of washes and spacers ahead of cement slur-
ries will prevent mud and/or cement contamination and
Cementing Stub Liner Cementing Tieback Casing help to remove the mud from the annular space. This is
especially important in tieback liner cementing, in which
Completion
no bottom plug separates the mud from cement inside
fluid the liner. If there is completion fluid in the hole, com-
Completion patibility with the cement must be checked or large vol-
fluid umes of fresh water must be pumped ahead of the slurry.
Setting tool Salts used in completion brines may drastically affect the
cement slurry’s thickening time, causing a premature set
Tieback or delaying compressive strength development.
casing

Liner hanger
Cement 13-5 Special offshore techniques
slurry
As discussed in Chapter 12, the logistics of offshore
cementing operations are often very different from those
Cement
slurry
for land-based operations, but the cementing procedure
Tieback
seating employed on offshore drilling rigs (such as jackups) or
Stub liner nipple platforms fixed to the seabed is similar to primary
cementing operations on land. However, considerable
differences exist in the plug release technique used on
floating rigs.
Special Figure 13-24 illustrates the general arrangement of
stage the subsea cementing system with respect to the subsea
collar
wellhead system.
Liner
Tieback hanger
sealing nipple
13-5.1 Conventional subsea plug system
Tieback sleeve The subsea plug system is similar in operation to the
Cement Cement liner plug system, apart from the larger size of the plugs.
Except for the largest size, it consists of a top and
Liner
bottom plug. The bottom plug is normally launched with
Liner a ball gravitating through the drillpipe to the seabed.
Some recent models use a two-dart plug launching
Liner hanger system instead of a ball and a dart. The technology devel-
oped for these subsea release plugs has been adapted
to liner cementing as a single-plug system in 8 5⁄8-in.
[219-mm] liners and larger. Most of the models available
today are top-drive compatible (Fig. 13-25).
Fig. 13-23. Tieback liner cementing.

484 Well Cementing

 Top plug launching dart (a) (a)
Dart release
Bottom plug launching ball Cement/
mud inlet
Ball release (b) (b)

Drillpipe

Rig floor

Installation tool
(h) (c)
Casing hanger
Running mandrel
Swivel (g)
Ocean
Top cementing plug (six shear pins) floor
(d)
Dart seat (c)
Plug connector (d)
Bottom cementing plug (three shear pins) (e)
(e)
Ball catcher (f)
Casing to be cemented
Outer casing (cemented)
(f)

Fig. 13-24. Single-stage subsea cementing system.

Figure 13-24 illustrates a conventional subsea plug

release system. It consists of a special subsea assembly
in the casing below the casing hanger. The cementing
head on the floating drilling vessel is made up to the
drillpipe serving as the casing landing string and con-
trols the cementing plug release. The head contains a
launching ball and dart, while the subsea assembly con-
tains the top and bottom casing plugs. Modern heads are
top-drive compatible. Referring to Fig. 13-24, and by
chronological order of usage, (b) is the bottom plug-
launching ball, which, when released before pumping
the cement slurry, seats in the bottom plug (e). A 100-
to 275-psi [0.7- to 1.9-MPa] pressure increase allows
the connector pins to be sheared (d), and permits the
bottom plug (e) to travel down the casing until it bumps
on the float collar and casing shoe.

Ball and Dart Two Darts

Fig. 13-25. Top-drive subsea cement heads.

Chapter 13 Primary Cementing Techniques 485

 Extra pump or hydrostatic pressure extrudes the ball Surface Dart Launcher Subsea Tool
(b) through its orifice seat, and cement displacement
continues. A ball catcher attached to the lower end of
the bottomhole plug retains the ball.
Once the cement slurry has been pumped, the top-
plug launching dart (a) is released. It will seat into the
body of the top cement plug (c). Increased circulation
pressure will then shear the retaining pins and release
the top plug (c) from the launching mandrel. The
cementing operation thus continues. At the end of the
slurry displacement, the top plug (c) bumps on the float.
In the subsea assembly, before cementing, the top
plug is pinned at the lower end of a running mandrel
that has a swivel (g). This avoids any rotation of the
cementing plugs inside the casing that could damage the
shear pins. The upper part of the mandrel is made up to
the lower part of the casing hanger running tool (h).
This casing hanger running tool is a tool connecting the
casing hanger to the drillpipe of the casing landing
string. One of the major limitations, and often a source
of cementing failure, is the reduced flow area through
the plug-retaining mandrel. This restricted flow area is
susceptible to fluid erosion and failure owing to high
pump rates (often required for turbulent flow) and large
fluid volumes of either mud or cement. Fig. 13-26. Two-plug top-drive remote-controlled subsea cement
head system.

13-5.2 Innovative system for launching plugs with

subsea wellheads The darts release casing wiper plugs when they reach
Brandt et al. (1998) introduced a two-plug launching the subsea tool. Because the dart launcher is modular,
system (Fig. 13-26) adapted to subsea operations from adding segments is easy. This increases the number of
its original fixed installation concept (Lavaure and darts that can be launched.
Galiana, 1991). This cementing head offers enhanced
reliability because of a simpler cement-plug design in
which the plugs are isolated in their container from the 13-5.2.2 Subsea tool
main fluid flow. Plugs are released from the subsea tool After dart launch, mud flows down the drillpipe, through
without physical contact between the darts and the the sliding sleeve of the subsea tool, and out of the ori-
plugs, avoiding dart-to-plug or ball-to-plug sealing prob- fices. The dart lands on a rod, and continued pumping
lems. forces the dart and rod down, pushing the plug out of a
basket. A spring retracts the sliding sleeve, allowing
complete, unobstructed flow through orifices.
13-5.2.1 Dart launcher A pressure differential resists rapid rod motion and
A surface dart launcher contains two identical darts stops rod movement after the plug releases. Combined
launched during cementing. Darts are remotely with plug friction, this pressure differential increases
released, reducing rig time and increasing the safety of pumping pressure and provides a positive indication at
the operation. Using darts instead of a free-fall ball helps the surface of plug launch. Physical spacers prevent the
prevent contamination, provides positive fluid displace- plugs from sticking and are retrieved with the tool.
ment, wipes the drillpipe clean, and saves time. The downhole subsea tool encases the plugs inside a
Uninterrupted high pump rates improve mud removal, container, eliminating difficulties associated with pump-
because the fluid is neither resting nor allowed to gel at ing fluids through the inside of plugs. Simplified plug
any time. design allows use of high-performance, but easily drill-
able, plugs.

486 Well Cementing

 The dart launcher has a high-pressure rating, and the Water Flow Zones in Deep Water Wells (RP 65, 2003),
plugs have a high collapse resistance; therefore, this which gives a number of recommendations for managing
system allows the operator to conduct a casing pressure the well and minimizing hazards before, during, and
test immediately upon bumping the plug. This saves rig after the cement job. Apart from needing ideal mud dis-
time and reduces the risk of creating a microannulus or placement to remove the viscous kill fluid in the often
damaging the cement sheath—a danger when this oper- enlarged holes, shallow water formations require a
ation is carried out at a later stage (Carré et al., 2002). cement slurry that sets very quickly to prevent water
migration into the annulus. When a low fracture gradi-
ent imposes the need for low-density slurry, foamed
13-5.3 Deepwater cementing considerations cement or high-performance lightweight cement pro-
Other significant considerations in subsea cementing vides the required level of performance (Cuvillier et al.,
are hydrostatic pressure and temperature. The hydrosta- 2000). A foam cement job on an offshore rig is more dif-
tic pressure of the seawater above the subsea assembly ficult, especially in remote areas or with difficult
can be a significant factor. In addition, the temperature weather conditions. In these situations, a high-perfor-
at the sea bottom and the first several hundred meters mance lightweight cement system is often preferred
of hole must be considered (Ward et al., 2003). when adequate bulk storage is available. However, the
Frequently, the temperature is close to freezing at the two options should be weighed and compared on techni-
ocean floor, and the underlying formation is substan- cal and economical merits for each local situation and
tially cooler than what is observed on land. This has for each well requirement, because both can be success-
implications for numerical simulations to predict the ful (Carré et al., 2002).
temperature profile during the cementing operation and Deepwater development often requires extended-
laboratory testing to design cement systems appropriate reach wells to minimize the number of subsea templates
to the conditions. Therefore, the International Organ- or to allow a dry tree development using a surface well-
ization for Standardization (ISO) (2003) and API (2004) head platform (tension-leg platform, deep-draft caisson
issued special recommendations for testing cement sys- vessel). In West Africa, for example, the departure of the
tems for deepwater applications (Appendix B). longest extended-reach subsea well is about 8,200 ft
One of the major difficulties in drilling deepwater [2,500 m] in the Girassol field in Angola (Anres, 2003).
wells is the small pore- and fracture-pressure window. Extended-reach well technologies under development
Enabling technology such as dual gradient drilling is will allow subsea wells to reach between 26,000 and
emerging for deepwater and ultradeepwater wells. This 30,000 ft [7,925 and 9,144 m] of departure (Anres et al.,
method eliminates the excessive hydrostatic pressures 2003). Such developments may require changes in the
caused by heavy mud and cuttings in the marine riser. way casings are run and cemented.
The riser is filled with seawater; therefore, the subsea
wellbore is exposed to the hydrostatic pressure of sea-
water. This technology is not only important from a
drilling and well architecture point of view (Cuvillier
13-6 Cementing techniques and new
et al., 2000), but the reduced hydrostatic pressure also enablers
reduces the risk of losses during the cementing opera- A number of emerging techniques enable the industry to
tion. The Subsea Mudlift Drilling Joint Industry Pro- drill more complex wells and produce hydrocarbons
ject (Eggemeyer et al., 2001) developed the first dual- from the reservoir in a more efficient manner. This short
gradient drilling system in the industry. From a cementing section does not attempt reviewing and detailing them,
perspective, this technology requires the use of a special but rather describes a few particulars regarding cement-
float collar and, above all, a flow restrictor to limit any ing operations in these special wells.
free fall of the heavy cement slurry (Schumacher et al.,
2002). This technology has been successfully used in a
well in 910 ft [277 m] of water, but more work is required 13-6.1 Extended-reach wells
to develop a system that can tolerate the severe free fall Extended-reach wells are highly deviated wells with a
in wells drilled in 10,000 ft [3,048 m] or more of water. high measured-depth (MD)/total-vertical-depth (TVD)
Shallow water-producing formations compound the ratio. Horizontal wells are a subset of extended-reach
already difficult problem of cementing deepwater wells. wells in which part of the wellbore is inclined 90° from
For such situations, API issued Cementing Shallow vertical. With existing technology, the length of the

Chapter 13 Primary Cementing Techniques 487

extended-reach section decreases with increasing TVD There are numerous technical challenges associated
(Fig. 13-27). Well departures can be more than 33,000 ft with extended-reach wells (Fig. 13-28). Hole cleaning
[10 km] at vertical depths ranging from about 5,000 ft (Cunha et al., 2002), hole stability, and running casing
[1,500 m] in the United Kingdom, to about 5,250 ft (Mason et al., 2003) are issues that strongly influence
[1,600 m] in Argentina, to about 7,200 ft [2,200 m] in cementing. Mud removal is also a key issue during the
Germany. Companies are working to push this limit cementing operation. Special centralizers that signifi-
beyond the 50,000 ft [15 km] mark. cantly reduce drag are used to help run the casing
Many applications exist in which extended-reach and (Chapter 11). When the last cemented casing is not the
horizontal wells can achieve production more economi- production pipe, only a short section near the shoe is
cally than vertical wells: cemented to minimize the equivalent circulating density
■ gas and water coning (ECD) and avoid losses. When long sections of the pro-
■ tight reservoirs and heavy oil
duction casing must be cemented, more precautions are
taken, and high-performance lightweight cements are
■ fractured reservoirs
used to achieve successful cementing of the entire sec-
■ edge-water or gas-drive reservoirs tion without losses (Frank et al., 1998).
■ inaccessible reservoirs Running the casing to bottom is the next important
■ enhanced oil recovery
challenge in extended-reach wells (Mason, 2003). When
the casing cannot be run to bottom by ordinary means
■ reducing the number of platforms and wells needed
like centralizers and mud lubricants, flotation is one of
to develop a field.

5,000

10,000
Industry extended
True vertical depth reach drilling
below rotary table 15,000 envelope
(ft)
20,000
Technology challenge is
outside the black line
25,000

30,000
0 5,000 10,000 15,000 20,000 25,000 30,000 35,000 40,000
Departure (ft)

Fig. 13-27. Technical envelope for extended-reach wells (from Rae et al., 2004). Reprinted with permission of SPE.

No vertical exaggeration
Feet
(MD) 30-in. casing shoe at 651 ft MD
0
Seabed
–500
16-in. casing shoe at 1,713 ft MD
–1,000 1,000
95⁄8-in. casing shoe at 7,262 ft MD 8.5-in. hole TD at 15,360 ft MD
True vertical –1,500 Top chalk
00
depth subsea 2,0
(ft) –2,000
00

11,000
10,000

13,000

14,000
12,000

15,000

–2,500
8,000
5,000

9,000
4,000

6,000

7,000
3,0

–3,000
–3,500 Upper Captain sand

Fig. 13-28. Example of an extended-reach well (from Rae et al., 2003). MD stands for measured depth.
Reprinted with permission of SPE.

488 Well Cementing

the simplest mechanical means to reduce effective drag 13-6.2 Solid expandable tubulars
and assist in running the casing (Mason et al., 1999). Solid expandable tubulars were first tried as a contin-
This imposes some rig modification as well as the use of gency measure to allow drilling to proceed through prob-
a specific flotation sleeve and flotation collar, a sacrifi- lematic zones (heavy losses or unstable formations)
cial plug, and a three-plug cement head to minimize without having to set an intermediate drilling liner. This
manual handling (Rae et al., 2004). Figure 13-29 shows technology prevents the loss of one casing diameter.
the string configuration and position of fluids when the Today, the technology has matured sufficiently to allow
casing is landed and after bursting the flotation collar. well architects to plan and engineer cost-effective wells.
After the sacrificial plug picks up the flotation sleeve Demong and Rivenbark (2003) and Demong et al. (2004)
and opens the flotation collar, mud circulation adjusts showed that this technology could be successfully
mud properties, because the well could not be circulated applied in extended-reach wells, allowing the operator
until this moment. Then the cementing operation is per- to significantly reduce the torque and drag without
formed normally, often with high-performance light- losing casing size.
weight cements because the pressure-operating window
is narrow in such long horizontal wells.

Landing the Casing Bursting the Flotation Collar

Vent line to choke-and-kill manifold
Surface cement head 9 5⁄8-in., 47-lbm/ft premium connection
113⁄4-in., 40-lbm/ft premium connection 13 3⁄8-in. surface cement
113⁄4-in. head (loaded with sacrificial plug)
Mud inlet
landing
string 113⁄4-in. landing string
Fullbore running tool
with hanger
Fluid level in casing drops Fullbore running tool
with hanger
113⁄4-in.,
40-lbm/ft
premium Air
connection

9 5⁄8-in. × 113⁄4-in. crossover

at 2,226 ft MDRT
Mud-filled 9 5⁄8-in. × 113⁄4-in. crossover
(active drilling at 2,226 ft MDRT
fluid) Selective flotation
collar joint Theoretical top of mud at 3,244 ft MDRT
Mud-filled
4,000-ft air-filled casing (active drilling fluid)
Air at Selective flotation collar
atmospheric 9 5⁄8-in., 47-lbm/ft at 4,000-ft MDRT
pressure premium connector

Float collar
200-ft filled active mud
Float collar
Connection
Active drilling mud

Active drilling fluid

9 5⁄8-in. eccentric shoe 9 5⁄8-in. shoe joint with eccentric shoe
Air
at 7,894 ft MDRT

Fig. 13-29. Landing the casing and bursting the flotation collar in extended-reach wells (from Rae et al., 2004).
MDRT means measured depth below rotary table. Reprinted with permission of SPE.

Chapter 13 Primary Cementing Techniques 489

 Campo et al. (2003) realized the ultimate goal for cement after inflation is below the overlap; otherwise,
which solid expandable tubulars were invented: they the expansion and sealing of the hanger may be affected.
successfully completed the first monobore well. This After mixing the slurry, the latch-down plug is dropped.
resulted in significant savings compared to conventional When it reaches and latches onto the launcher, increas-
wells or even slim wells for the same production-casing ing pressure initiates the expansion process. The liner is
size. Figure 13-30 compares the monobore architecture then expanded from the bottom up in stages of one stand
with that of conventional and slim wells. at a time. The cement slurry must remain stable and
Figure 13-31 details the installation process for a fully liquid until the expansion process is completed;
solid expandable tube and shows the cementing opera- otherwise, the expansion may be impaired by gelled
tion. With the tube on bottom before inflation, the slurry or by solids deposits from settled or dehydrated
cement slurry is pumped down the drillpipe and up the slurry. The entire expansion may require several hours,
annulus like a normal inner-string cementing operation. as much as 10 hr for long liners, after which the slurry
The slurry volume corresponds to the annular volume of should set and develop strength quickly.
the inflated tube. One must ensure that the top of

Conventional Plan Slim-Well Plan Monodiameter Plan

36 in. 30 in. 30 in.

30 in. 22 in. 22 in.

26 in. 16 in. 16 in./141⁄2 in.

20 in. SET 133⁄8 in. × 16 in. SET 113⁄4 in. × 141⁄2 in.

16 in. 133⁄8 in. 113⁄4 in.

133⁄8 in. SET 113⁄4 in. × 133⁄8 in.

9 5⁄8 in. MDDL

113⁄4 in. Nested 9 5⁄8 in. × 113⁄4 in. 9 5⁄8 in. MDDL

95⁄8 in. 9 5⁄8 in. 9 5⁄8 in.

SET 75⁄8 in. × 9 5⁄8 in. SET 75⁄8 in. × 9 5⁄8 in.

75⁄8 in. 75⁄8 in. 75⁄8 in.

7 in. FJ 7 in. FJ

Fig. 13-30. Monobore well with conventional and slim-well plans (from Campo et al., 2003). SET means solid expandable
tublar; FJ means flush joint; and MDDL means monodiameter drill liner. Reprinted with permission of SPE.

490 Well Cementing

 Drill Hole Run Condition Pump Plug Latch Plug; Expand Drill Out
Expandable Mud; Start Expansion Hanger Shoe
Liner Cement Liner Joint

Fig. 13-31. Installation sequence of a solid expandable tubular (from Demong and Rivenbark, 2003).
Reprinted with permission of SPE.

13-6.3 Multilaterals The junctions in Levels 2 to 5 are in fact laterals side-

Multilaterals are multiple wellbores (or multiple hori- tracked from the cased and cemented mother wellbore.
zontal wells) that emanate from a single wellbore Then a completion with varying levels of complexity
(Economides et al., 1998). They were invented in the assures mechanical or pressure integrity at the junction.
1920s and installed in Russia in the 1950s. However, it With respect to the mother wellbore, one must ensure
was not until the 1990s that interest in this technique that the cement sheath is resilient enough to resist the
became widespread and the technology was accepted by shocks incurred while the junction window is milled out.
the industry for effective reservoir management. A joint If the cement around the junction is brittle, it may frac-
industry project of North Sea operators, called ture in chunks or blocks. Such blocks may fail at the
Technology Advancement for Multilaterals (TAML) wrong time, increasing the risk of an expensive bottom-
(Diggins, 1997), classified multilateral wells in six dif- hole assembly (BHA) or completion string becoming
ferent categories (Chambers, 1998; MacKenzie and stuck. Figure 13-33 shows the result of milling a window
Hogg, 1999) depending on the mechanical integrity and with conventional cement and a special cement system
pressure integrity required at the junction. The junction toughened by adding fibers (Chapter 7).
is defined as the point at which two separate wellbores Fiber cements have very good impact resistance and
merge. As seen in Fig. 13-32, the TAML classification good postfailure behavior. However, such systems are
ranges from Level 1, with an openhole sidetrack or expensive and, if a small volume is used, placing it
unsupported junction, to Level 6, which includes both exactly at the right place may prove difficult. On the
reformable and nonreformable splitters to provide pres- other hand, flexible systems (Chapter 7) are more
sure integrity. Level 5 provides pressure integrity with affordable. They are highly elastic, deform easily without
the downhole completion equipment. Levels 3 to 6 pro- cracking, and have relatively good shock resistance.
vide mechanical strength at the junction. Latex-modified cements (Chapter 7) also provide

Chapter 13 Primary Cementing Techniques 491

 TAML Classification System
Level Description Level Description
1. Open unsupported junction 4. Motherbore and lateral
Barefoot motherbore and lateral or cased and cemented
slotted liner hung off in either bore Both bores cemented at the junction

2. Motherbore cased and cemented, 5. Pressure integrity at the junction

lateral open Pressure integrity provided by using
Lateral either barefoot or with slotted straddle packers
liner hung off in open hole

3. Motherbore cased and cemented, 6 Pressure integrity at the junction

lateral cased but not cemented Integral mechanical casing seal
Lateral liner anchored to motherbore (cement alone is NOT sufficient)
with a liner hanger but not cemented Includes reformable junctions as well as
nonreformable, full-diameter splitters
that require larger-diameter wellbores

Fig. 13-32. Classification of multilaterals (from Moritis, 2003). Reprinted with permission from Oil & Gas Journal.

Conventional Cement Tough Cement

Fig. 13-33. Cement sheaths after milling junction window in a multilateral well.

increased flexural and tensile strength and have found Throughout the industry, the emphasis is now shifting
application in cementing some multilateral wells. from providing an open, unsupported Level 1 junction in
However, in a multilateral Level 4 system, when the competent, consolidated formations to high-end Level 5
liner hanger of the lateral bore is drilled out after and 6 systems that offer full hydraulic and mechanical
cementing to leave the main bore entirely free, only junction integrity (MacKenzie and Hogg, 1999). Even for
tough and resilient cement can provide the adequate the Level 6 system, in which pressure integrity does not by
mechanical support without failing. design depend on the cement job quality at the junction

492 Well Cementing

(Hogg, 2002), it still seems preferable to use a small Retrieving head
volume of tougher cement around a short junction at the
bottom of the motherbore. This would ensure that the
formed-metal structure of the splitter remains fully sup-
Seal elements
ported at all times when drilling and completing the two
legs. Then pressure integrity at the junction can be
assured. Therefore, all the different types of multilateral Bypass ports
Centralizer
wells benefit from having an advanced tough cement
around the junction. Axial lock
Stop dogs
Axial and
13-6.4 Casing drilling torque lock
Torque lock
The concept of screwing a bit on a joint of casing, then
drilling into the ground and cementing through the bit
has been around for more than 100 years. It started in
Russia at the beginning of the twentieth century. In the Spacer collar
1950s and 1960s, this technique was also used in the
United States. However, the technique did not become Stabilizer
widespread, because it was more suitable for one-trip
applications. Therefore, it was limited to softer forma- Casing shoe
tions, relatively shallow wells (as in the mining indus-
Underreamer
try), and wells in which there were no openhole evalua-
tion requirements. Bit
The industry tried for many years to develop effective
retrievable systems but it is only in the last few years Fig. 13-34. Wireline-retrievable casing drilling BHA (from Warren et
that mechanically reliable, retrievable drilling systems al., 2000). Reprinted with permission of SPE.
appeared on the market. Casing drilling could therefore
be rejuvenated (Tessari and Madell, 1999) as a tech-
nique to eliminate the use of drillpipe, significantly ultimate size of the production casing or tubing may be
reduce tripping times, and reduce lost time caused by smaller than planned. In some cases, that limitation is
unscheduled events such as reaming, fishing, and taking not acceptable. During drilling with a shoe, the stuck
kicks while tripping. At about the same time, significant casing is cemented in place as usual though the drilling
progress was made with a drillshoe system attached to shoe. However, during drilling with a retrievable system,
the end of the casing. Although casing drilling technol- the BHA must first be recovered on wireline if it is not
ogy is still in its infancy, experience so far has demon- stuck. The use of casing drilling technology is expanding
strated that standard oilfield casing can be used to to directional wells (Warren et al., 2003) and deepwater
simultaneously drill and case the well. wells (Galloway, 2003).
Casing drilling is performed by attaching the bit to Cementing during a casing drilling operation differs
the bottom of the casing or through a BHA that extends slightly from conventional cementing. First, from a
beyond the casing. Bit rotation originates from rotating cement-slurry placement point of view, the casing is
the casing string using the topdrive or by using downhole poorly centralized because only hard-facing wear bands
motors. The latter method employs wireline-retrievable and sometimes solid hydroformed centralizers are used.
BHAs so that bits or BHAs can be changed without trip- Both are crimped onto the casing. For a production
ping the casing (Fig. 13-34). casing string, when the string is drilled to total depth
Surprisingly, casing drilling can reduce some lost-cir- with a conventional bit attached to the bottom of the
culation problems. It appears that the casing drilling casing, stabilizers are sometimes used near the bottom
process consolidates and strengthens the wellbore by to provide a stiffening effect and maintain verticality.
plastering the cuttings into the borehole walls. The stabilizers are not designed for cementing central-
The occurrence of stuck pipe, long a major drilling ization, so the rest of the casing string is not centralized.
problem, is also relatively rare during casing drilling Wear bands and stabilizers are solid and do not have
(Fontenot et al., 2003). When it does occur, the casing any restoring force. They are sufficient to prevent the
is cemented in place and drilling can continue with the casing from touching the borehole walls and being worn
next size; however, prematurely giving up a casing size out during the drilling operation. For cementing pur-
may have a negative impact on production, because the poses, this is likely to be sufficient in a gauge hole, but

Chapter 13 Primary Cementing Techniques 493

may become problematic and largely inadequate for tools, it is normally first pulled out by wireline
casing centralization in a washed-out hole. This imposes (Warren et al., 2000). However, in this situation, the
constraints on the cement slurry, which must efficiently casing does not have a float collar to land the cement
remove the drilling fluid at pressures below the fracture plug. To overcome this problem, special bottom and
gradient. Diaz et al. (2004) showed that friction losses top wiper plugs (Tessari and Madell, 1999) were
and the equivalent circulating density can be fairly well designed; however, this equipment was not suffi-
estimated; therefore, flow rates, rheologies, and fluid ciently robust. Today, a pump-down float (Vert and
densities can be adjusted to minimize the annular fric- Angman, 2001) is pumped in place before the cement
tion pressures and avoid losses during cementing. injection (Warren et al., 2000). A pressure increase
Although a few directional wells have been drilled confirms that the float has correctly landed in its pro-
with casing drilling technology, most of the experience file nipple. An alternative solution (Skinazi et al.,
today is with vertical or near-vertical wellbore sections, 2000) provides a single cement top wiper plug that
in which the casing is not centralized as discussed locks into its landing collar (Fig. 13-36). The cement-
above. Little information is available regarding cement- ing operation can now proceed conventionally, and
ing operations in directional wells drilled with casing. the casing can be rotated and reciprocated. In opera-
The cementing procedure differs with the type of bit tions in which a poor bond was obtained across gas
installed on the casing. intervals, casing rotation greatly improved zonal iso-
■ If the bit is attached to the casing, a special casing- lation in subsequent operations.
drilling bit (Fig. 13-35), derived from an expandable
drill bit (Brown and Gledhill, 2003), drills like a con-
ventional polycrystalline diamond compact (PDC) Water
bit. At casing depth, a pressure cycle displaces the
PDC cutters into the borehole wall or extrudes them
to a greater diameter than the next hole section,
opening the bottom of the casing. The casing is then
cemented normally, but this technique does not allow
rotation or reciprocation. Following the cement job,
the rest of the bit assembly can be drilled out with a
conventional PDC bit.
■ After casing has been drilled to its setting depth,
using a retrievable bit assembly that can incorporate Locking
cement
measurement while drilling and logging while drilling plug

Cement plug
Landing lands and
collar locks into
collar
Cement

Fig. 13-36. Example of cement wiper plug and lock collar used in
casing drilling operation (Skinazi et al., 2000). Reprinted with per-
mission of SPE.

Fig. 13-35. Expandable bit for casing drilling (from Brown and
Gledhill, 2003); reprinted with permission of SPE.

494 Well Cementing

■ If a pumpdown cement float or a locking cement plug ends when the top plug lands on the float collar, an event
is not available for the casing size used, then the commonly known as bumping the plug. During some
casing is cemented through a composite retainer run cementing operations, the plug does not bump after
on drillpipe (Warren et al., 2001a and 2001b). Floats pumping the theoretical casing displacement volume.
and retainers can also be run on wireline. As a con- Under these circumstances, to avoid leaving a wet shoe
tingency measure, the cement slurry can also be (shoe surrounded by unset cement), an additional
placed volumetrically, a common approach on surface volume corresponding to half the shoe track volume is
casing strings. However, this should be the last resort, often pumped before stopping the operation. However, in
because fluid intermixing and bypassing inside the most cases, a significant volume of cement is found
larger casing will occur. above the float collar.
The displacement volume is normally calculated
During the drilling of the last string in the well with
according to the nominal casing capacity for its size and
casing, a knock-off sub is commonly used to drop the bit
weight. However, casing manufacturers do not make
before performing any other cementing or completion
casing for a fixed internal diameter. Casing specifica-
operation.
tions include mechanical properties, outside diameters,
wall thicknesses, and drift diameters, but not the inter-
nal diameters. Not accounting for the true casing diam-
13-7 Operational considerations eter is the principal cause of not bumping the plug. For
Planning is basic to successful primary cementing. It example, the internal displacement volume of 9,843 ft
begins with accurate knowledge of the well conditions. [3,000 m] of a 9 5⁄8-in., 47-lbm/ft casing is 720.5 bbl
The cement job is designed for these conditions, and job [114.6 m3]. This volume would be 735 bbl [116.9 m3] if
parameters must be monitored and recorded during job the internal diameter is 1% larger than the nominal
execution. Later, the actual job can be compared to the value. This difference exceeds 12.6 bbl [2 m3], and a
design. two-joint shoe track has a capacity of 6.3 bbl [1 m3].
Therefore, to avoid such costly errors, some operators
have adopted a policy to systematically gauge either all
13-7.1 Calculations the casing joints or a statistically significant number of
13-7.1.1 Slurry volume joints in the string.
Because of the difficulty of gauging large open holes, sur- A second cause of displacement-volume errors is fail-
face casing hole volumes are rarely known. The volume ing to account for the compressibility of the displace-
of cement slurry must be based on common field prac- ment fluid, especially when using oil-base muds.
tice in the area. If this is not known, excess slurry vol- Although pressure and temperature work in opposite
umes of 50% to 100% should be used. Excess slurry vol- directions, the displacement-volume error can be signif-
umes of up to 200% are common in some areas; the icant (several barrels). Drilling fluid companies can
volumes may exceed 300% for the top sections of some provide the compressibilities of their muds, allowing
deepwater operations. the operator to calculate the correct volumes to be
Even when a caliper is run and the theoretical volume safely pumped to achieve bumping the plug.
is calculated, an excess volume is often required to
ensure proper fill-up. As much as 50% in excess of the 13-7.2 Hole condition
calipered hole volume may be used. In many countries, In addition to the physical hole parameters (i.e., depth,
the volumes of slurries are governed by regulations that diameter and direction), the drilling reports should be
can be very stringent (Chapter 12). reviewed to identify potential problems that could affect
the cement job. Hole washouts, lost circulation, and
13-7.1.2 Displacement tight spots should be noted and addressed in the design.
The displacement fluid pushes the cement slurries out of A caliper log should be mandatory on most jobs. When a
the casing to fill up the annulus and isolate the respec- caliper cannot be run, pumping a marker (such as a car-
tive horizons. It is important to achieve complete dis- bide pill) while drilling is always a good practice to
placement not only to ensure that correct fill-up and iso- assess the degree of washout before cementing.
lation are achieved in the annulus, but also to avoid Drilling mud types and properties have a significant
wasting time drilling out excess cement left in the effect on the state of the borehole wall and the quantity
casing. In addition, incomplete displacement eliminates of cuttings left downhole. This necessitates conditioning
the ability to pressure test the casing. Displacement the hole before the cement job and influences the per-
formance of the cement slurry. Drilling muds are

Chapter 13 Primary Cementing Techniques 495

designed to drill the hole efficiently, but the mud engi- the capabilities of temperature simulators to deepwater
neer rarely considers the cement job. Therefore, after operations, which were validated with field measure-
drilling, the mud properties must be optimized to satisfy ments (Ward et al., 2003).
cementing requirements. Proper practices must also be When strength development time is a key parameter,
implemented to ensure that the modified mud effec- it is equally important to know the time required for the
tively displaces gelled mud left over from the drilling cement-slurry temperature to reach the static tempera-
process. ture. Only temperature simulators can provide such
information, which is used by laboratory personnel to
test the performance of the cement systems under real-
13-7.3 Temperature istic conditions. Deepwater wells particularly benefit
Knowledge of the bottomhole circulating temperature from this capability.
(BHCT) is vital. The cement-slurry pumping time is a When dry cement is added to water, the subsequent
direct function of the bottomhole temperature. Overly hydration generates heat that raises the temperature of
long slurry pumping times can be as disastrous to pri- the cement slurry. The cement-slurry temperature also
mary cementing as those that are too short. depends on the temperature of the dry cement and the
Temperature also affects the cement and mud charac- mix water. The nomograph shown in Fig. 13-37 has been
teristics; consequently, the flow regimes, U-tubing effect, used for many years to estimate the cement-slurry tem-
and friction pressures are all directly affected (Chapters perature during mixing. Today, this nomograph is incor-
4, 5, and 12). If the BHCT is unknown, it can be deter- porated in cement-slurry design software applications.
mined through logging, circulating temperature probes The initial cement-slurry temperature can have a sig-
(Jones, 1986), or mathematical simulation of cementing nificant effect on the bottomhole circulating tempera-
circulating temperature (Beirute, 1991; Mitchell and ture during a cement job. If a job is performed in a hot
Wedelich, 1989). Temperature simulators (Guillot et al., or cold climate and the cement and mix-water tempera-
1993) have been introduced and validated (Merlo et al., tures differ significantly from 80°F [27°C], prejob labo-
1994; Davies et al., 1994) with field measurements on ratory testing should reflect these conditions. Jestes
land wells. Romero and Touboul (1998) further extended et al. (2003) presented a case study that illustrates the

Slurry Temperature for Various Temperatures

of Water and Cement
130
Based on 46% water
by weight of cement
120 (neat slurries) This is recommended
temperature limit
for receiving cement
140
110 130
Above this temperature 120
excessive gelation and 110
100 mixing troubles may be 100
encountered 90
80
90 70 Cement
Slurry 60
temperature 50
temperature
(°F) 40 (°F)
80 30
Acceptable slurry 20
temperature range 10
70 0
–10
–20
–30
60

Below this temperature,

50 many cement additives
lose their effectiveness

40

40 50 60 70 80 90 100
Water temperature (°F)

Fig. 13-37. Nomograph to estimate cement-slurry mixing temperature.

496 Well Cementing

importance of the cement and mix-water temperatures 13-7.5 Quality control
on slurry performance. If these temperatures are not A quality control program should be employed to test all
taken into account, WOC times may be too long in the materials before cementing. Laboratory conditions
winter and too short in the summer. The performance of should simulate the job as closely as possible from
cement additives, particularly accelerators, can vary known well conditions. Actual field samples of cement,
widely with temperature. additives, and mix water should be used for testing.

13-7.4 Pressure Because ISO and API specifications for cements are nec-
essarily broad in scope, additional testing should be per-
Accurate knowledge of downhole pressure is necessary
formed whenever the cement quality is suspect. ISO and
for well control and successful primary cementing. A
API rheology tests may help to identify potential prob-
minimum slurry density is required for well control
lems. Liquid additives should also be checked and thor-
during and after placement, and slurry rheology governs
oughly blended with the mix water before cementing.
the friction losses during placement. Excessive slurry
Certain dry additives are prone to separation (particu-
density together with a high displacement rate can lead
larly weighting agents), and care should be taken to
to fractured formations and lost circulation: when lost
verify proper blending with the dry cement exists before
circulation is feared, the ECD in the annulus must be
the job (Gerke et al., 1990).
projected at the design stage. A typical intermediate
casing string cement job, with its minimum and maxi-
mum hydrostatic pressures, is shown in Fig. 13-38. This 13-7.6 Casing movement
type of plot should be generated for all primary cement Casing movement—reciprocation, rotation, or both—
jobs (Chapter 12). positively improves the quality of primary cement jobs
Well control is also of concern after displacing the (Fig. 13-39). Casing movement breaks up areas of stag-
cement slurry, especially while WOC. On stationary rigs nant mud, which can cause cement channeling.
(land, jackups) using conventional wellheads, it is Scratchers and wipers are of little benefit, unless they
extremely important to WOC before lifting the BOPs for are put to work by casing movement.
slips and packoff installation.

Fluid Sequence Static Well Security

0
Hydrostatic
pressure

600
Mud
(displacement)
Fracturing
pressure
Depth 1,200
(m)

1,500 Lead
slurry

Pore
Tail pressure
slurry
2,000
0 10 20 30 40
Pressure (MPa)

Fig. 13-38. Pressure plot for intermediate casing cement job.

Chapter 13 Primary Cementing Techniques 497

 13-7.7 Cement job monitoring
Double-plug Rotation
cementing head The recording of critical parameters during cementing,
from the commencement of mixing cement to the final
displacement, is paramount. Pressure, slurry rate, den-
sity, and integrated volume are factors that must be
known in real time. These data should also be recorded
for future playback and analysis to evaluate and optimize
the design of future jobs (Piot and Loizzo, 1998).
Reciprocation Recording devices also verify that the correct volumes
and densities of preflushes, spacers, and cement slurry
were pumped into the well. Before cementing takes
place, it is particularly important to ensure that the
pressure-monitoring equipment is functioning correctly.
A typical recording device output is shown in Fig. 13-40.
The sensor package is equally important. For pres-
sure, accurate and fast-responding electronic transduc-
ers have replaced traditional hydraulic gauges. It is
important to monitor and record the pressure until the
Swivel sub
Reciprocation top plug is bumped. Slurry densities and flow rates are
accurately recorded with Coriolis mass-flow vibrating
Bails tube devices (Benabdelkarim and Galiana, 1991;
Chapter 11). Such devices have largely replaced tradi-
Elevator tional radioactive densitometers.
Traditionally, pressure, rate, and density are recorded
during a cement job. In 2003, Vigneaux et al. introduced
Casing joint
a cement monitoring and control system based on the
Rotation slurry volumetric balance. The slurry quality is guar-
anteed by assuring that it contains the designed quan-
Fig. 13-39. Rotating and reciprocating a casing head during cementing. tity of cementitious solids, regardless of the slurry
density. Although originally engineered for high-perfor-
mance, low-density cements, this monitoring system
Casing that cannot be moved before cementing con- (which records several other process parameters) offers
firms that something is wrong. Often, not much can be improved accuracy at all slurry densities and for all types
done at this point other than to cement the casing in of cement slurries.
place; however, the chances of a successful cement job
are diminished before even mixing the slurry.
Some operators prohibit the use of reciprocation 13-7.8 Casing operational sequence
when cementing from a floating rig, for fear of the casing 13-7.8.1 Stationary rig (e.g., land rig, jackup rig)
getting stuck protruding above the subsea wellhead.
Also, it should be noted that reciprocating is particularly When the conductor pipe has been driven to the desired
difficult, not to say impossible, in deviated wells. depth (or cemented), a mud return line or flowline is
Although it can be easily accomplished on almost installed underneath the rig floor so that drilling fluid
every drilling project, casing rotation is in fact seldom can return to the pits. The hole is then drilled to the
performed because of either wellhead design or other depth required for the surface casing. After the casing
equipment or rig limitation. From the very beginning, has been run and cemented (always to surface), it is cut
the overall well design and equipment selection should off underneath the rig floor at the desired height. The
aim towards making casing movement possible. When casing head (which will enable the next size of casing to
conventional casings are being cemented, the highest be hung and connects to the BOPs) is then welded to the
torque load is generally reached when the lead end of surface casing, inside and outside. Some casing heads
the cement reaches the shoe. Therefore, the casing con- screw onto a casing thread.
nection strength requirement must be calculated to con- The BOPs, with the connections for the kill line and
sider this load, not just the drilling load. choke line, are flanged onto the casing head. Before
drilling can continue, the BOPs must be tested to the

498 Well Cementing

 Cementing Job Report
Well Well #1 Client Company, Inc.
Field Gas field SIR No. 123456789
Engineer Name Job Type 4.5-in. production casing
Country Country name Job Date 02-11-2002
Time Pressure Rate Density Messages

14:59:14
Pressure test lines

15:14:00
Start pumping wash
Start pumping spacer
15:29:00

Start cement slurry

15:44:00

15:59:00

16:14:00

16:29:00

16:44:00

16:59:00

17:14:00

17:29:00

Wash up lines
17:44:00
Start displacement

17:59:00

18:14:00
End displacement
18:29:00

18:43:42

hh:mm:ss 0.00 1,000 2,000 3,000 4,000 5,000 6,000 0.00 2.0 4.0 6.0 8.0 10.0 8.0 10.0 12.0 14.0 16.0 18.0
(psi) (bbl/min) (lbm/gal)

Fig. 13-40. Recording device output from a primary cement job.

desired pressure. A sealing plug is run into the casing on

drillpipe, and the BOPs are closed one by one. The BOP
elements are pressure tested one after the other. Intermediate casing
will be hung from here
All BOPs and wellhead connections must hold pres-
sure before drilling can continue (Fig. 13-41). Then, Surface casing head
when the next size of hole has been drilled to the desired
depth, the next size of casing is run in and cemented.
Cement may or may not be required to reach the surface. Surface casing
The cement is usually allowed to set while the casing
is hanging from the elevators. After it has set and while
the casing is still hung from the elevators, the BOPs are
lifted from the casing head and suspended from the sub-
structure. Slips and the packoff assembly are then set
between the casing and the casing head.
To avoid buckling of the casing downhole, it is very
important that the casing be set with the same weight Fig. 13-41. Connection of casing strings.
hanging from the slips as that hanging from the elevator.
The casing can then be cut off at the same level as the

Chapter 13 Primary Cementing Techniques 499

casing-head flange, or 1 or 2 ft higher. A sealing mecha- Each intermediate casing and the production casing
nism is normally placed above the slips to seal the annu- are similarly run and cemented. They are connected to
lus between the two casing strings. Then, a new casing the top of a casing hanger, which lands in the wellhead
head is flanged to the previous head, the BOPs are reat- housing, and an annular seal assembly is installed to seal
tached (or replaced by BOPs with a higher pressure the annulus. The successive casing hangers pile up on
rating), and (after pressure testing the connection) top of each other.
drilling can recommence. This entire process is known
as nippling up.
In this way, each time a new string of casing is run, it 13-8 Conclusion
is hung from a casing head that was attached to the pre-
The basic mechanics of common primary cementing
vious casing head. The production casing will have a
techniques have been presented in this chapter. When
head from which to hang the tubing—the tubing head.
decoupled from the other related issues, such as fluid
Therefore, the weight of all the strings is partly sup-
rheology, cement-slurry design, cement-slurry mixing
ported by the surface casing (Fig. 13-42).
procedures, and annular gas migration, these proce-
dures may appear to be very simple. Such an impression
13-7.8.2 Floating rig (e.g., semisubmersible, is deceptive. It is essential that the engineer be inti-
drillship) mately familiar with the procedures and devices that are
On a floating rig, the wellhead and the BOPs are at the used for primary cementing. In addition, it is critical
seabed. In some ways, the sequence is simpler than on a that the engineer verify that all equipment is in proper
stationary rig because the BOPs are not removed each working order before the job; otherwise, the long and
time a casing is run and cemented. meticulous planning process before each job may be
The surface casing is run and connected to the top of wasted.
the wellhead housing. The subsea BOPs are then run and
connected to the wellhead housing by a hydraulic con-
nector. Drilling can continue after pressure testing of
the wellhead-BOP system.

Surface Wellhead Subsea Wellhead

Tubing hanger The wellhead

Tubing head system
Tubing
15-in.
wellhead
Inner casing hanger connector

Inner casing head

Inner casing
Intermediate casing

Sealing medium
Intermediate
casing hanger
Surface casing
head

Surface casing

Fig. 13-42. Typical surface and subsea wellhead assembly on wells.

500 Well Cementing

13-9 Acronym list
API American Petroleum Institute
BHA Bottomhole assembly
BOP Blowout preventer
ID Inside diameter
ISO International Organization for Standardization
MD Measured depth
MDRT Measured depth below rotary table
OD Outside diameter
TAML Technology advancement for multilaterals
TVD True vertical depth
WOC Waiting on cement

Chapter 13 Primary Cementing Techniques 501

14-1 Introduction
Remedial cementing is a general term to describe oper-
ations that employ cementitious fluids to cure a variety
Remedial Cementing
Gérard Daccord, Dominique Guillot, and
Simon James—Schlumberger
14
the variety of problems that may occur, a wide range of
techniques has been developed combining placement
methods, specific slurry formulations, and execution
of well problems. Such problems may occur at any time procedures. Successful remedial cementing requires
during the life of the well, from well construction to well careful analysis of the downhole problem(s), meticulous
stimulation, production, and abandonment. Remedial planning to formulate the best solution, and proper job
cementing is commonly divided into two broad cate- execution (Jones and Watters, 1998).
gories: plug cementing and squeeze cementing. Plug A wide variety of cement slurries has been developed
cementing consists of placing cement slurry in a well- to address almost any type of problem. Unless special
bore and allowing it to set. Squeeze cementing consists formulations are used, the following points must be
of forcing cement slurry through holes, splits, or fissures noted.
■ Cement is an intrinsically brittle material, so it will
in the casing/wellbore annular space. This chapter
describes each category and provides fundamental pro- have less resistance to mechanical stresses than
cedures and practices. purely mechanical solutions.
During well construction, remedial operations may be ■ Cement slurries are suspensions, so they have at best
required to maintain wellbore integrity during drilling, small penetration into narrow gaps and can repair
to cure drilling problems, or to repair defective primary only near-wellbore problems.
cement jobs. Wellbore integrity may be compromised
when drilling through mechanically weak formations, These limitations can be overcome by using alterna-
leading to hole enlargement. Cement slurries can be tive solids-free remedial techniques; thus, remedial
used to seal and consolidate the borehole walls. cementing techniques are not the only solution to many
Remedial cementing is often the only way to repair of these well problems. A few of these techniques will be
defective primary cement jobs, to either allow further briefly discussed at the end of the chapter.
drilling to proceed or to provide adequate zonal isolation This chapter has the following organization. A pre-
for efficient well production. Lost circulation is another liminary section presents the types of well problems that
serious drilling problem that can be cured with cement can be cured by remedial cementing. To avoid confusion,
slurries. Such applications are described in Chapter 6. detailed discussion of remedial cementing operations
During well production, remedial cementing opera- and techniques is kept to a minimum. Thorough discus-
tions may be performed to restore production, change sions of plug- and squeeze-cementing operations and
production characteristics (to alter the gas/oil ratio techniques are then presented separately.
[GOR] or control water production), or repair corroded
tubulars. During a stimulation treatment, the treatment
fluids must enter the target zones and not leak behind 14-1.1 Overview of remedial cementing operations
the casing (Chapter 1). If poor zonal isolation behind the Plug cementing consists of placing an amount of cement
production casing is suspected, a remedial cementing slurry in a wellbore and allowing it to set. The plug may
treatment may be necessary. be temporary or permanent, and it may be placed in an
Finally, well abandonment frequently involves plac- open hole or a cased section.
ing cement plugs to ensure long-term isolation between Squeeze cementing is the process of forcing a cement
geological formations, replicating the previous natural slurry under pressure through holes, splits, or fissures in
barriers between zones. However, before a well can be the casing/wellbore annular space. When the slurry is
abandoned, annular leaks must be sealed. Squeeze forced against a permeable formation, the solid particles
cementing techniques are applied for this purpose. filter out onto the formation face as the aqueous phase
Each candidate for remedial cementing presents a (cement filtrate) enters the formation matrix. The slurry
unique challenge to the engineer; therefore, to address is pumped until the wellbore pressure reaches a prede-

Well Cementing ■ Chapter 14 Remedial Cementing 503

termined level. An impermeable filtercake, not the set- During squeeze cementing treatments, most of the
ting of cement, allows the well to withstand this cement slurry is forced into the formation or behind the
increased pressure. When highly permeable media are casing (Fig. 14-1c). Excess slurry that did not enter the
present, such as gravel packs, the entire cement slurry formation may be removed from the wellbore. This is the
may invade the porous medium. A properly designed cleaning phase, often performed by reverse circulation
squeeze job causes the resulting cement filtercake to fill (Fig. 14-1d). There are many variations on this general
the opening(s) between the formation and the casing. theme that are specifically tailored to the operational
Upon curing, the cake forms a nearly impenetrable solid objectives, available equipment, and local practice.
(Suman and Ellis, 1977). In cases in which the slurry is
placed into a fractured interval, the cement solids must
bridge the fracture and then develop a filtercake on the 14-1.2 Methodology
fracture face. The selection and implementation of a specific remedial
The execution of a remedial cementing operation operation follows a logical series of steps.
begins by preparing the well. In particular, if the cement ■ Problem identification
slurry must be placed off-bottom and there is a risk for This step may be either obvious (e.g., holes in casing
the slurry to fall under gravity down the well, a plug is detected following a corrosion log) or difficult to
placed to serve as a cement base. The plug may be a vis- detect (e.g., identification of a gas leak resulting in
cous or reactive fluid (sometimes weighted to a density sustained casing pressure). The importance of this
halfway between that of the wellbore fluid and the step is often overlooked.
cement slurry) or a mechanical device (Fig. 14-1a). ■ Selection of remedial method
The cement slurry is then prepared at the surface and Once the root cause of the problem is identified,
pumped through drillpipe or coiled tubing to fill the selection of the appropriate solution is easier. Often,
wellbore. The tubular is usually pulled back out of the several radically different techniques are potentially
cement slurry. On rare occasions, the pipe is left in the applicable. Both economic and technical factors are
cement (Fig. 14-1b). Before setting, the cement plug considered.
may be slightly squeezed into the formation, but most of
the cement remains in the hole.

a b

c d
Figure 14-1. Principal steps of a remedial cementing operation (refer to text for a detailed description).

504 Well Cementing

■ Selection of placement method and mechanical Very often, only a good knowledge of the root causes can
setup ensure that the selected remedial treatment will have an
The simplicity of the operation is balanced with acceptable chance of success. In this section, the
potential risks. common types of problems are presented along with a
■ Selection of fluid volumes short description of the appropriate remedial treat-
At this stage, more detailed information about well ments. Table 14-1 provides a concise summary.
properties is required. Sometimes different scenarios
may be prepared, and the treatment choice is defer- 14-2.1 Well construction
red until just before the job, based on last-minute well
data. During well construction (i.e., drilling, casing, cement-
ing, and completion), a drilling rig is available on loca-
■ Detailed job design tion. This allows one to consider all of the standard
In this step, the job is simulated to verify that all well- methods for curing the problem.
safety issues are considered. Alternative treatment
designs are formulated as a contingency in case the
treatment does not go as expected. 14-2.1.1 Lost circulation
■ Execution Lost circulation is considered in detail in Chapter 6.
Execution requires close monitoring of pressure and Lost-circulation problems are detected by comparing
rate data, as well as comparison with design data. mud return rates with pump rates or by monitoring the
Significant discrepancies may result in pumping- fluid level in the mud tanks. When adding chemicals,
schedule changes. lost-circulation materials (LCMs), or fibers during
■ Evaluation drilling is unsuccessful, various types of cement plugs
The evaluation of job success is based on comparing can be placed and slightly squeezed to ensure good
the actual results with the initial job objectives. This adhesion between the cement and the formation. Once
step is especially important to improve future jobs in the cement has set, drilling can resume.
nearby wells. Two key questions must be answered before the job.
■ What is the nature of the leak: permeable formation,
natural fissures, induced fractures, caverns?
■ At which depths are the loss zones located?
14-2 Problem identification—Well
problems cured by remedial cementing Natural fractures and caverns are more common in
A wide variety of problems may be faced during the life carbonate rocks. Permeable zones are encountered more
of a well. Although the type of problem is often easy to frequently in sandstones. The depth is measured by cor-
determine, identifying its root causes may be difficult. relating drilling reports with electrical imaging logs.

Table 14-1. Principal Activities that Can Be Accomplished by Plug Cementing and Squeeze Cementing
Plug Cementing Squeeze Cementing
Seal lost-circulation zones Seal lost-circulation zones

Sidetrack around a fish (lost object or other debris Repair primary cement jobs (owing to mud channeling, insufficient cement
in the hole) height in the annulus, or leaking liner top)

Initiate directional drilling Eliminate water intrusion from above, below, or within the hydrocarbon
producing zone

Seal a depleted zone Reduce the producing GOR or water/oil ratio (WOR) by isolating the gas
or water zones from adjacent oil intervals

Protect a low-pressure zone during a workover treatment Repair casing leaks caused by corroded or split pipe

Provide an anchor for openhole tests Abandon a nonproductive or depleted zone

Seal wells for abandonment Plug one or more zones in a multizone injection well to direct the injection
into the desired intervals

Chapter 14 Remedial Cementing 505

 An appropriate cement slurry is then set across the (Fig. 14-3). Under these circumstances it is common to
lost-circulation (or “thief”) zone. Although some of the set a whipstock plug (also called a kickoff plug) across
slurry may be lost to the thief zone, it will harden and the zone to achieve the desired course and target. Also,
consolidate the formation (Fig. 14-2). A cement plug can when fishing cannot be performed economically, the
also be set above a zone to protect it from being fractured only remaining solution may be to plug the hole with
by the hydrostatic pressure developed while cementing a cement and sidetrack the well above the fish.
casing string. LCMs are often added to the plug slurry to To design an appropriate cement system for a whip-
maximize the chance for success (Chapter 6). Cements stock plug, it is essential to know the properties of the
containing fibers have also been used. The fibers not only formation in which the plug will be set. To effectively
form a “mat” to inhibit loss of cement slurry to the for- divert the drillbit from the previous hole, the formation
mation, but also prevent cement fragments from falling must be easier to drill than the set cement.
on the bottomhole assembly (BHA) (van Vliet et al.,
1995). The combination of fibers and cements with opti-
mized particle-size distributions has produced accept-
able solutions in difficult situations (Elmoniem et al., Kickoff point
2000). In the extreme case, when the wellbore offers no
support (e.g., for caves or caverns), inflatable cement
bags have been proposed (Piot et al., 1994).
Many alternatives to cement plugs are available to Cement plug
control lost circulation, such as deep-penetrating fluids
that plug fissures and fractures (Chapter 6). Such mate-
rials are beyond the scope of this discussion.

New
hole

Drillpipe

Open hole
Fig. 14-3. Sidetrack (whipstock) plug.

Drill bit
One of the factors that helps kicking off is to make the
Thief zone compressive strength of the plug greater than that of the
Cement formation or to increase the toughness of the plug. The
plug usual compressive-strength requirement for a kickoff
plug is 5,000 to 7,000 psi [35 to 49 MPa]. If it is not pos-
sible to prepare a cement system with a higher com-
pressive strength than the formation, the toughness of
the plug can be increased by reinforcing the cement
matrix with materials such as polymer fibers (Loveland
Fig. 14-2. Lost-circulation plug.
and Bond, 1996) or metallic microribbons (Al-Suwaldi
et al., 2001; Chapter 3). However, Trabelsi and Al-
Samarraie (1999) raised concerns about potential degra-
14-2.1.2 Sidetrack and directional drilling dation of some cement properties (porosity, permeabil-
ity, and compressive strength) when fibers are added.
Sidetracking is a method to bypass an existing hole and Other factors that determine the chances of kicking off
continue drilling a new one. When performing this type relate to the BHA used, the well deviation and the
of directional drilling through soft formations, it may be drilling direction.
difficult to achieve the correct angle and direction

506 Well Cementing

14-2.1.3 Protective plugs If a problem has been detected, the root causes must
In addition to repairing defects, temporary cement plugs be investigated. Common causes include the following.
are used to protect a weak zone at the bottom of a well ■ Poor mud displacement

while performing high-pressure operations in upper ■ Poor knowledge of the well before cementing
zones. Cement plugs are used when the hole is not (e.g., hole geometry)
cased; otherwise, mechanical bridge plugs are a better ■ Cement-placement problems such as lost circulation
solution. The cement plug is called a test anchor
(Fig. 14-4). Poor mud displacement during primary cementing
causes the cement slurry to channel through the drilling
mud or a mud film to remain on the walls. Consequently,
pockets or channels of mud are left behind the casing
(Fig. 14-5), and the principal goal of primary cement-
ing—zonal isolation—is not achieved. Should such
zonal-isolation defects not be corrected, serious prob-
lems may occur during the life of the well.
Test string ■ Stimulation treatments may not achieve the expected
results because of poor fluids-placement control
(Chapter 1).
■ Evaluation of a well’s production potential may be
inaccurate because of the parasitic effects of nearby
flowing fluids.
Zone to be tested ■ Poor well productivity may occur as a result of a high
WOR or GOR.
■ Enhanced oil recovery techniques such as water-
flooding may fail (Goolsby, 1969).

Cement
plug Cement sheath
Weak formation
Channel

Water
Fig. 14-4. Plug set as an anchor for a test.

14-2.1.4 Well construction evaluation

After a section has been cased and cemented, the
cement quality is checked before drilling is resumed. Perforation
Oil
The most common evaluation technique is the pressure
test. The pressure test involves pumping a fluid into the
wellbore to exert pressure, closing the well, and moni- Figure 14-5. A defective primary cementing job.
toring the wellbore pressure. A pressure decline indi-
cates the presence of a leak that must be plugged. More
advanced cement evaluation methods include cement
logs that are run only when good zonal isolation and The principal tool for diagnosing the root cause is a
casing mechanical support must be demonstrated. The careful comparison of the job record with design and
occurrence of such defects has decreased as better evaluation data. Once the root cause is known, a variety
cementing practices have been implemented. In excep- of cement-based repair treatments may be performed.
tional cases, some flow may be observed at the surface, ■ Squeeze the shoe
which is a clear signature of poor zonal isolation. ■ Liner top squeeze
Evaluation techniques are presented in detail in ■ Perforate and squeeze mud channels
Chapter 15.

Chapter 14 Remedial Cementing 507

■ Circulation squeeze
■ Block squeeze Perforation
■ Perforate and fill the annulus in a manner similar to
a two-stage cementing operation
Electing to perforate a cemented casing and perform
a squeeze job is a difficult decision. Any procedure that
might compromise casing integrity or further harm zonal
isolation should not be taken lightly. Defining precisely Drillable retainer
where to perforate is also critical. A thorough analysis of Channel
the primary cementing job, based on an accurate record
of all the parameters of the operation and a careful
interpretation of the logs (Chapter 15), is a key element Perforation
in the decision process. Two situations may exist behind
the casing.
■ The mud channel sits against a permeable formation.
Cement
During the squeeze job, the cement filtercake builds
and eventually fills the void.
■ Circulation is established between two sets of perfo-
rations. A “circulation” or “channel” squeeze is per-
formed to replace the mud in the channel with
cement. Basically, this is a partial or total recement- Fig. 14-6. Circulating squeeze.
ing of the interval of interest. It is important to note
that this procedure can be precarious. The risks and
possible unwanted consequences are explained later (or tubing) may become stuck in the hole. Therefore, to
in this chapter. minimize this risk, a cement retainer should be run
instead of a packer. It is easier to remove the stinger
Both of these operations can be successful only if the
assembly than the packer, because the latter has mini-
downhole treating pressure remains below the forma-
mal casing clearance. The retainer should be set as close
tion fracturing pressure. If not, fractures created during
as possible to the upper perforations. This minimizes
the job would open a preferential route through which a
exposure of the drillpipe to cement slurry that may enter
large quantity of the cement slurry can penetrate; as a
the wellbore through the upper perforations.
result, damage to the permeable interval will occur, and
the treatment objectives will not be met.
14-2.1.6 Block squeeze
14-2.1.5 Circulating squeeze Block squeezing is a method to prevent fluid migration
from either above or below a producing zone in which
The circulating squeeze, illustrated in Fig. 14-6, is often
the bond log indicates poor zonal isolation. The perme-
performed with a cement retainer in preference to a
able section below the producing zone is perforated and
packer (Chapter 11). Circulation is achieved with water
squeezed. The procedure is then repeated for the per-
or acid as a preliminary fluid. The interval is circulated
meable zone above the producing zone. The two plugs of
with a wash fluid to ensure good cleanup, and the
residual squeeze cement are then drilled out, and the
cement slurry is then pumped and displaced. No pres-
productive zone is perforated for production. This
sure buildup occurs during the job, except for an
method has frequently been successful in preventing
increase caused by the hydrostatic pressure of the
water coning.
cement column as it flows up the annulus. Once the
placement is completed, the stinger or packer is
released. The excess cement circulating out of the upper 14-2.1.7 Top of liner squeeze
perforations can be reversed out if desired. As discussed in Chapter 13, liner cementing can be diffi-
The volume of slurry required to complete a circulat- cult because of the small annular clearance and often
ing squeeze is unknown; therefore, plenty of slurry is poor centralization. The amount of cement used during
prepared. Consequently, there is a strong possibility that primary liner cementing is small, and the wellbore
some of the cement slurry may enter the casing, geometry is often not conducive to good mud removal;
drillpipe, or tubing or the annulus above the squeeze
tool during the job. Should this cement set, the drillpipe

508 Well Cementing

consequently, mud contamination can occur. In addi- the gas enters. Noise is generated by fluid flowing
tion, annular fluid migration can be a complicating through narrow channels within the cement sheath
factor (Chapter 9). (McKinley and Bower, 1979). The depth at which noise is
If the liner leaks at the overlap, the remedial proce- observed corresponds to the producing-zone location.
dure is the same as squeezing a hole in the casing. If The carbon-isotope method is a more recent tech-
good coverage is not obtained, the casing must be perfo- nique to locate the depth of a leaking gas zone. The
rated and another squeeze job performed (Smith, 1987). method is based on the recognition that, for shallow gas
reservoirs, the origin of the gas may be thermogenic or
biogenic. The relative amount of biogenic gas decreases
14-2.2 Production/injection with increasing depth, and the isotopic composition is
Several types of problems may occur during the life of a variable. The isotopic composition of thermogenic gas is
well that require remedial cementing operations, linked not.
either to poor zonal isolation behind casing or changes Thus, the 13C/ 12C ratio of various components of a gas
in the nature of the production. (mainly ethane) can be correlated with the reservoir
depth. This calibration must be performed for each
14-2.2.1 Before stimulation treatments reservoir. The current calibration method is to perform
an isotopic analysis of gas in drilling fluid as hole depth
During a stimulation treatment, the treatment fluids increases. The carbon isotope technique was developed
must enter the target zones and not leak behind the at the University of Alberta (Rich et al., 1995; Ellis et al.,
casing. If poor zonal isolation behind the production 2003) and has been used in Canada and the Gulf of
casing is suspected, a remedial cementing treatment is Mexico. The technique is currently limited to shallow-
necessary. As detailed above, the most common method gas reservoirs and must be considered a complement to
to assess zonal isolation is the cement bond log. other information sources.

14-2.2.2 Sustained casing pressure 14-2.2.3 Casing and packer leaks

A producing well that displays casing pressure on the Producing wells may suffer from corrosion caused by
surface suffers from a lack of zonal isolation. In such brines or acid gases that migrate from the formation to
cases, gas often migrates to the surface. Eventually, the tubulars. Leaks will eventually appear in the pro-
remedial treatments will be required. duction tubing or casing (Figs. 14-7 and 14-8). Leaks may
Xu and Wojtanowicz (2001) described a method to also appear at packers. In all cases, crossflows, produc-
assess the severity of the problem, based on a pressure tion of unwanted fluids, or loss of production to other
test. Once the problem has been assessed, noise logs are zones may result.
used to find the leak and specifically the depth at which

Internal External

2 1.5 1 0.5 0 –0.5 –1 –1.5 –2 2 1.5 1 0.5 0 –0.5 –1 –1.5 –2

Radius (in.) Radius (in.)

Fig. 14-7. Three-dimensional image of a hole in casing.

Chapter 14 Remedial Cementing 509

Fig. 14-8. Ultrasonic casing log showing a hole in the steel.

Under these circumstances, cement plugs are fre- Methods available to identify the root causes of these
quently more successful than squeeze cementing production problems include various logging or injection
(Nowak et al., 1996; Loveland and Bond, 1996). Old and methods.
corroded casing may suffer further damage when sub- ■ Temperature logs (Bergren and Bradley, 1988)
jected to the high treating pressures and packer-gener-
Temperature logs follow wellbore temperature varia-
ated stresses associated with squeeze cementing. Casing
tions as a cold fluid is injected through the perfora-
leaks have also been seen on new pipes, in which case a
tions and can indicate whether channels exist in the
patching job can be performed.
cement behind the casing. This procedure is also
called a pump-in temperature log. Correlating tem-
14-2.2.4 Production and injection control perature logs with cement logs allows a full assess-
The GOR or WOR may not meet expectations for many ment of the presence of channels in the cement.
reasons, including errors when perforating the reservoir ■ Boron-pulsed neutron log (Blount et al., 1991;

and reservoir depletion. Determining the root cause of Sommer and Jenkins, 1993)
excessive water production is beyond the scope of this The procedure is similar to the pump-in temperature
chapter; however, methods are available to help identify log. A boron solution is injected through the perfora-
the problem (Bailey et al., 2000). The most common tions, and a pulsed neutron log is generated. Boron
problems include a moving oil/water contact (Fig. 14-9), absorbs neutrons very efficiently; consequently, chan-
crossflow, bypass through fractures or high-permeability nels in the cement sheath can be readily detected,
streaks, and water coning. even through two strings of casing. The method is
Like production control, water-injection or steam- more sensitive than the pump-in temperature log.
injection profiles can be durably controlled by remedial
cementing methods, provided no crossflow occurs
between the zones.

510 Well Cementing

 Gas

Oil

Water

Present Future
Fig. 14-9. Water or gas invasion into an oil zone. (from Shryock and Slagle, 1968; reprinted with permission
of SPE).

■ Oxygen activation logs (Pappas et al., 1995) ■ Production history

Oxygen activation logs are used to quantify water flow If there has been any sand production or dissolution
behind casing. The logging tool uses the high-energy of the formation, there is a chance that void spaces
neutron generator to send large quantities of neu- exist behind the casing.
trons around the wellbore and the formation. The
Only a limited number of production problems can be
neutrons cause the oxygen nuclei to produce a
reliably cured by remedial cementing methods. Some
radioactive isotope of nitrogen with a 7-sec half-life.
root causes, and the remedial-cementing techniques
As the nitrogen decays, high-energy gamma rays are
employed to cure them, are listed below.
emitted. If there is water movement past the tool,
■ Breakthrough through a well-identified layer,
each gamma ray detector will detect the decay at two
different times. The two times are converted to a fluid with no crossflow:
velocity. By using more detectors or changing their The remedial treatment involves plugging the perfo-
spacing, the range of measurable velocities can be rations or part of a gravel pack in front of the problem
adapted. A significant benefit of this technology is zone.
that fluid movement can be detected and measured ■ Communication behind casing:
without direct contact with the logging tool. For
The channel is plugged by performing a low-pressure
example, it is possible to detect water flow behind
or circulation squeeze.
tubing or nonperforated casing.
■ Natural fissures or fractures and void space behind
■ Production logs (Kuchuk and Sengul, 1999;
casing:
Lenn et al., 1996; Caretta et al., 2000)
A squeeze treatment is performed with the goal of fill-
Water/oil contacts and other formation hetero-
ing the void with cement.
geneities can be detected by production logs either
■ Induced fractures or channels:
during production or shut-in periods. Such informa-
tion can help identify water entry points along pro- Treatments other than remedial cementing are often
ducing intervals. better suited for these types of problems. Never-
■ Injectivity test theless, squeeze treatments are frequently employed.
■ Production depletion:
An injectivity test consists of pumping a fluid into the
formation and analyzing the pressure response. This When production from a zone falls below a commer-
type of test is especially useful to help characterize cially viable level and other zones exist uphole that
water-control problems or to determine whether can produce economically, a cement plug can be
there is a void space behind the casing. placed across the depleted zone to isolate it from the
wellbore (Fig. 14-10).

Chapter 14 Remedial Cementing 511

 Cement plug Cement

Depleted zone

Fig. 14-11. Well-abandonment plugs.

Fig. 14-10. Plugging a depleted zone.

■ Well abandonment: 14-3 Plug cementing—Tools and

To abandon a well, several cement plugs are set at techniques
various depths, often using coiled tubing. This proce- There are several tools and techniques for placing
dure prevents interzonal communication and fluid cement plugs:
migration that might pollute underground freshwater ■ balanced plug
sources (Fig. 14-11). In many countries, oil and gas ■ dump bailer
well operators are compelled to follow abandonment
■ two-plug method
procedures dictated by government authorities (e.g.,
Texas Railroad Commission Rule 3.14: Plugging, 2002; ■ mechanically supported method (with jet-hole

Barclay et al., 2001; Nagelhout et al., 2005). The ulti- cleaning)

mate objective of these remedial treatments is to ■ flexible bags
restore the natural integrity of the formation that was ■ inflatable through-tubing packers
disrupted by drilling. The resulting hydraulic isola-
■ umbrella-shaped membranes
tion is expected to last for several decades.
■ coiled tubing placement.

512 Well Cementing

 Cement slurries are almost never pumped through The main problem that may arise during the place-
drillbits. Doing so could damage the bit bearings. ment of a balanced plug is cement contamination. To
Consequently, placing cement plugs consumes rig time minimize downward migration of the cement plug, fluids
because it requires several drillstring trips. Cement slur- with high gel strengths can be placed as a base.
ries used in plug cementing must not display sedimen- Examples of such fluids include thixotropic bentonite
tation or free water, and the thickening time must be suspensions or crosslinked polymer pills. These pills may
sufficiently long to allow thorough cleaning of the drill- be weighted to ensure better stability of the cement
string after placement. lower interface by avoiding instabilities caused by fluid
density differences. Mechanical devices are also com-
monly employed (Section 14-4.2.1).
14-3.1 Balanced plug
The most common placement method is the balanced-
plug technique (Fig. 14-12). Tubing or drillpipe is run 14-3.2 Dump bailer
into the hole to the desired depth for the plug base. To A dump bailer is a vessel that holds a measured quantity
avoid mud contamination, appropriate volumes of spacer of cement slurry. Lowered on a cable, the dump bailer
or chemical wash are pumped ahead of and behind the opens when it touches a permanent bridge plug placed
cement slurry. The volumes are such that they corre- below the desired plug interval (Chapter 11, Section
spond to the same heights in the annulus and in the 14-4.2.1). The cement slurry is dumped on the plug by
pipe. The slurry is often batch mixed for better control of raising the bailer (Fig. 14-13). Usually employed for set-
density and rheology. Basic calculations used to design ting plugs at shallow depths, the dump-bailer method
balanced plugs are presented in Appendix C. can also be used at greater depths if the cement slurry is
properly retarded. The advantages of this method are
that the depth of the cement plug is easily controlled
and it is relatively inexpensive. The principal disadvan-
tage is that the quantity of cement slurry is limited to the
volume of the dump bailer; however, multiple runs can
Displacement be made.
fluid

Spacer
fluid
Wireline

Cement slurry

Dump bailer

Cement slurry
Viscous fluid:
plug base

Fig. 14-12. Balanced plug.

Electrical or
mechanical
opening
It is common practice to slightly underdisplace the mechanism
plug (usually by 2–3 bbl). This practice avoids mud flow-
back on the rig floor after placement and allows the plug
Bridge plug
to reach a hydrostatic balance. Once the plug is bal-
anced, the pipe is slowly pulled out of the cement to a Casing
depth above the plug, and excess cement is reversed out.
One variation of this process involves placing the cement Fig. 14-13. Dump-bailer tool and method. Reproduced courtesy of
slurry in two steps—as two balanced plugs, one on top of the American Petroleum Institute.
the other.

Chapter 14 Remedial Cementing 513

 To help meet the plug objectives, the dump-bailer aluminum tailpipe becomes stuck in cement, an
method must incorporate hole and mud conditioning. increase in the pull will break the tailpipe and free
The bailer should be raised slowly for proper placement. the drillpipe.
Because the slurry is stationary during its descent, spe- 6. Excess cement is then circulated from the hole.
cial slurry-design considerations are required, especially
under high-temperature conditions. Slurry gelation or
instability must be avoided to ensure that the slurry will 14-3.4 Inflatable packers
exit the dump bailer.
Use of inflatable packers is an adaptation of the bal-
anced-plug method (Heathman et al., 1994; Jones and
14-3.3 Two-plug method Watters, 1998) and employs an inflatable-set packer to
The two-plug method uses a special tool to set a cement provide a solid bottom (Chapter 11). The packer assem-
plug at a calculated depth, with a maximum of accuracy bly includes a diverter tool that circulates fluid and
and a minimum of cement contamination. The tool washes the wellbore as the tool is run into the hole. A
essentially consists of a bottomhole sub installed at the drillable aluminum tailpipe is installed above the packer
lower end of the drillpipe, an aluminum tailpipe, a to the required plug height.
bottom wiper plug (which carries a dart), and a top When the tool reaches the target location, the inflat-
wiper plug (Fig. 14-14). Foam balls may also be used in able packer is filled with cement slurry and expands to
the place of wiper darts. form a mechanical support. Then the ports in the
The procedure includes the following steps. tailpipe are opened and the cement slurry is placed in
the same manner as a balanced plug. The tailpipe is then
1. The bottom plug is pumped ahead of the cement
disconnected and left in the plug.
slurry to clean the drillpipe wall and isolate the
More recently, this technique has been adapted to
cement from the mud.
allow packer installation and inflation with coiled tubing
2. The shear pin connecting the dart to the plug is (Wilson et al., 2004). In addition, packers have been
broken by increased pump pressure and pumped designed to withstand extreme temperatures and pres-
down through the aluminum tailpipe. sures as well as harsh chemical environments.
3. The top plug is pumped behind the cement slurry to
prevent contamination from the displacement fluid.
Increased surface pressure is observed when the plug 14-3.5 Umbrella-shaped membranes
arrives at its seat. Harestad et al. (1996) developed an umbrella-shaped
4. The drillpipe is pulled up until the lower end of the tool that is deployed below the cement plug to aid the
tailpipe reaches the calculated depth for the top of placement of balanced plugs. After deployment, the tool
the cement plug. keeps the cement slurry separated from the drilling fluid
below (Fig. 14-15). It must be emphasized that the tool
5. The shear pin between the catcher sub body and the is not a hydraulic barrier and cannot be used to control
sleeve is then broken, allowing the sleeve to slide losses farther down the well.
down and open the reverse circulating path. If the

First wiper dart Second wiper dart Second wiper dart Locator
sub with
darts
Locator sub Locator sub with Locator sub with Locator sub
first dart (sheared) darts first dart (sheared)

Cement stinger
Mud

Spacer

Cement Diverter sub

All Fluids in Pipe and First Dart in Place— Second Dart Pumped— Pressure Bled Down— Pull Above Cement
Circulating Downhole Placement Continues Pressure Increases Cement Underbalanced and Circulate

Mud Spacer Cement

Fig. 14-14. Two-plug method.

514 Well Cementing

 Release rod

Lockscrew

Release tool

Upper body

Body O-ring

Seat

Dart O-ring
Dart (dart illustrated
in open position)
Spring
Lower body

Spring retainer
Umbrella tool Diaphragm tool

Fig. 14-15. Umbrella-shaped tool and diaphragm tool for balanced

plug cementing. (Drawing courtesy of BJ Services Company.) Fig. 14-16. Internal blowout preventer (IBOP). Drawing courtesy of
WOODCO USA.

Arebrat (2004) described another tool—a diaphragm

bow—that is pumped like a cement plug through a Harrison and Blount (1986) reported that, in some
drillpipe and expands when it exits (Fig. 14-15). This instances, up to 85% savings on workover costs can be
tool is designed to prevent both fluid mixing during achieved when using coiled tubing. Coiled tubing is also
placement and fluid swapping. used extensively for squeeze cementing. It is discussed
later in this chapter.
14-3.6 Inside blowout preventer
An internal, or inside, blowout preventer (IBOP) is 14-4 Plug cementing—Placement
sometimes inserted in the drillpipe near the rig floor
when the drillpipe is pulled above the cement plug To ensure the cement plug fulfills its objectives, several
(Fig. 14-16). The IBOP is run to avoid flowback in situa- placement issues must be considered.
■ Place the plug at the correct location.
tions when, owing to a small density differential, it is dif-
ficult to balance the plug. The IBOP is activated from the ■ Prevent cement contamination.
surface. In terms of design, it has no specific impact ■ Ensure sufficient thickening time at downhole
except preventing backflow. Today this tool is mainly temperature to complete the placement and well
used in the U.S. Gulf of Mexico. cleanup.
■ Ensure that pressure and mechanical-strength limits
14-3.7 Coiled tubing placement are not exceeded during all stages.
Use of coiled tubing for remedial cementing began in the
early 1980s. Since then, the technique has gained con-
siderable popularity (Walker et al., 1992; Loveland and 14-4.1 Plug location
Bond, 1996). Coiled tubing has proved to be a very eco- Placement of abandonment plugs is often controlled by
nomical method to accurately place the small volumes of government regulations. The key objective is to restore
cement slurry required in plug cementing operations. the natural isolation between geological layers that was

Chapter 14 Remedial Cementing 515

destroyed when the well was drilled. The natural loca- ■ Contamination caused by overdisplacement
tion to place a plug is at the same depth as the caprock. ■ Failure to adhere to mud removal guidelines (cement
Kelm and Faul (1999) reviewed the best practices in this may be contaminated by a mud channel or by a mud
domain. layer not being displaced from the walls)
Lost-circulation plugs should be placed wherever loss ■ While pulling the pipe out
zones are encountered (Lindstrom et al., 1999). Whip-
■ Reversing out the excess cement
stock or kickoff plugs should not be positioned against
excessively hard formations. Each of these scenarios is covered in the following
discussion.
14-4.2 Contamination
Contamination during plug cementing has long been 14-4.2.1 Mechanical devices to prevent
recognized as a serious problem, and sets of preventive or minimize contamination
guidelines have been proposed (Dees and Spradlin,
1982; Heathman et al., 1994). The primary consequences There are several mechanical devices that can address
of cement contamination are a dramatic setting-time some of the causes of cement contamination mentioned
change and deterioration of mechanical properties. above.
Table 14-2 shows a situation in which the setting time
increased with water-base mud contamination. The situ- Bridge plugs
ation is more complex with emulsion muds. If the emul- When a cement plug is placed off-bottom, a bridge plug
sion remains stable, serious retardation may result. If (Fig. 14-17) can be used to prevent the heavy cement
the internal aqueous phase contains calcium chloride slurry from falling through the less dense fluid under-
and the emulsion breaks, acceleration may occur. neath. To prevent the releasing mechanism from being
Contamination of cement plugs may occur at differ- cemented, most operators drop one or two sacks of sand
ent times and for different reasons. on top of the retrievable bridge plug before the job.
■ During placement, inside the drillpipe, at the front Calcium carbonate is also sometimes placed on top of
and rear interfaces of the cement slurry the plug, followed by a sand cap. The calcium carbonate
■ Unstable base or top, owing to density differences
can be removed by acid to facilitate packer retrieval.
Drillable bridge plugs are normally used to isolate the
(heavy fluid on top of a lighter fluid)
casing below the zone to be treated. Their design is sim-
■ During placement with drillpipe (downward flow at
ilar to that of cement retainers (Chapter 11). They can
the drillpipe tip can cause mixing of the cement be run with a wireline or with the workstring. Bridge
slurry with other wellbore fluids) plugs do not allow flow through the tool.
■ Washed-out zones (mud removal is difficult, and the
wellbore volume may be underestimated)

Table 14-2. Effect of Mud Contamination on Cement Performance

Neat Class H Cement (16.5 lbm/gal) Compressive Strength in 18 hr at 230°F
Mud Contamination Compressive Strength Mud Contamination Normal Slurry† Reduced Water Slurry†
(% by volume) (psi at 170°F) (vol %) 15.0 lbm/gal 17.5 lbm/gal
(psi) (psi)
8 hr 16 hr
0 4,650 5,860 0 4,080 8,600

5 3,500 5,300 10 2,950 8,240

10 2,600 4,540 40 2,430 3,850

20 2,400 2,330 60 593 2,970

50 245 471
† Contains dispersant.

516 Well Cementing

 Spacer
Slump angle
Cement Viscous
Slump pill
length
Length of
good cement

Slips
Fig. 14-18. Cement slurry slumping in a horizontal well.

Sealing 27⁄8 in. or 31⁄2 in.

element tubing joint

Slips

Mechanically Hydraulically Set Cement

Set Permanent Permanent Retainer
Bridge Plug Drillable 1 in. × 6 in.
Bridge Plug cut slots ± 6 ft
(90° phasing)
Fig. 14-17. Bridge plugs and retainer. 8 holes
phased
at 45°
Bull plug
Blanked
Tailpipe or stinger mule
The drillpipe is equipped with a stinger or tailpipe (i.e., shoe
a few joints of smaller-diameter pipe). This reduces con-
Fig. 14-19. Two types of diverter subs (from Walker et al., 1992).
tamination when the drillpipe is pulled out, because the Reprinted with permission from World Oil.
fluid volumes involved are smaller. The stinger may be
made of aluminum or fiberglass pipe and may be left in
the cement plug to provide additional reinforcement
while the cement sets. In the latter case, once the Mechanical plugs, darts, and balls
cement is set, the tailpipe is broken at its top by increas- Various types of mechanical plugs, darts, and balls have
ing tension or with explosives. This method is sometimes been used to provide a barrier between fluids as they
used in horizontal wells, reducing the tendency of the flow downhole. These devices may eventually be used to
cement slurry to slump (Fig. 14-18). Van Vliet et al. accurately indicate the positions of interfaces and pre-
(1995) presented cases in which a glass-fiber-reinforced vent overdisplacement. They are an integral part of the
epoxy pipe was left in the middle of a temporary cement two-plug placement method described earlier in Section
plug to ease the plug removal process. 14-3.3.
Diverter tool
As cement slurry exits the pipe, a downward jetting
14-4.2.2 Interface stability
action may occur that may break or contaminate the vis- Cement plugs placed off-bottom have a natural tendency
cous gel that serves as the plug base. To prevent down- to be unstable, because the density of the cement slurry
ward flow, the tip of the pipe can be equipped with a is greater than the fluid underneath (Fig. 14-20).
diverter tool (combination nozzle) (Fig. 14-19). The tool Beirute (1978) was the first to clearly recognize this
redirects the flow in an outward or radial direction. instability. Since then, extensive laboratory experiments
There are many tool designs available for both drillpipe have been performed to study the phenomenon (Smith
and coiled tubing (Krause, 1991; Walker et al., 1992). et al., 1983; Calvert et al., 1995; Heathman, 1996), and

Chapter 14 Remedial Cementing 517

 0.35
Mud Mud 10° experimental
0.3 45° experimental
10° model
Chemical Unstable Weighted
45° model
wash interfaces spacer 0.25
Cement
Stable Mud 0.2
Cement interfaces yield
stress
Mud 0.15
(Pa)

0.1
Fig. 14-20. Schematic illustration of unstable and stable interfaces
for cement plugs. 0.05

0
some correlations have been found, including the effect 0 0.05 0.1 0.15 0.2 0.25 0.3 0.35
of density differential, yield stress, hole size, and devia- Cement yield stress (Pa)
tion. Although the principle of the instability is easy to
understand, its physical modeling is more difficult. Fig. 14-21. Theoretical and experimental stability results for two
Beirute (1978) presented the first analytical model hole deviations (from Crawshaw and Frigaard, 1999; reprinted with
for a perfectly vertical well. This basic model was further permission of SPE).
developed and numerically solved by Griffin and Valkó
(1997). Crawshaw and Frigaard (1999) presented a
more general solution that is valid for inclined wells, The success rates of the chemical systems have been
showing that the heavy fluid slumps along the borehole variable. A solution that succeeds in one area may give
wall and leads to a flow exchange between the cement disappointing results in another. In the case of reactive
slurry and the underlying mud. From a theoretical analy- pills, there is little guarantee that downhole mixing will
sis varying well deviation, hole size, and density differ- be sufficient to allow the reaction to proceed effectively.
ential, they determined the minimum yield stresses If there is an unstable plug interface and the cement
required to prevent the flow exchange. Wamba Fosso et slumps to the low side of the hole, no mixing can take
al. (2000) used these predictions in field operations and place.
found the success rate to be close to 100%. For large hole sizes and large density differences, the
The relevant dimensionless groups that describe the yield-stress requirement is difficult to attain, and the
process are mechanical solutions are the most reliable means to
ensure a stable plug interface (Section 14-3.5 and
(τ y ) mud (τ y )cem 14-4.2.1).
τ mud = and τ cem = , (14-1)
Δρgdhole Δρgdhole
14-4.2.3 Displacement volume for a balanced plug
in which (τy)mud and (τy)cem are the yield stresses of the
mud and cement, respectively. Δρ is the density differ- When placing a balanced plug, overdisplacement of the
ence, g is the acceleration of gravity, and dhole is the hole cement slurry may lead to severe contamination. To
diameter. The stability limit is shown in Fig. 14-21. Any avoid overdisplacement, one must have an accurate
condition located on the upper right side of the plot cor- measurement of the displacement volume.
responds to a stable situation. It appears that the pre- An ideal example of balanced plug placement is given
dictions are conservative with respect to the experi- below.
ments by a factor of approximately two. ■ A drillpipe is lowered to the planned cement-plug

Different types of viscous pills have been used to pre- bottom. The drillpipe has well-defined inside and out-
pare a base for the plug, including bentonite suspen- side diameters, and the hole is gauge and filled with
sions, silicate gels, and crosslinked polymer solutions. A a known mud.
shear-sensitive plugging fluid has been used in difficult ■ A spacer fluid, a cement slurry, and a spacer fluid are
situations (Lindstrom et al., 1999). Reactive solutions displaced by the same mud that is in the hole.
have also been applied to form a strong gel upon contact ■ The volumes of the two spacer stages are such that
with the cement slurry. their respective heights in the annulus and pipe are

518 Well Cementing

 equal. The volume of displacement mud is exactly eter variation of 0.1 in. corresponds to a volumetric
equal to the internal volume of the drillpipe up to sur- variation of about 1 bbl per 1,000 ft. Varying the dis-
face. placement volume by this amount could significantly
■ Once all the fluids are pumped, the drillpipe is slowly change the location of the fluid interfaces.
pulled out. ■ The fluids are compressible, and their volumes will
■ As the drillpipe is pulled out, the fluid interfaces fall. decrease as they are pressurized. It is often assumed
When the pipe end reaches the cement-spacer inter- that thermal expansion compensates for compress-
face, no mixing between the spacer and the cement ibility; however, this assumption is far from exact,
occurs because the interfaces in both the annulus particularly when oil-base muds are involved.
and the pipe are at the same depth. ■ Balanced plug design is particularly difficult in situa-
tions in which there is lost circulation. The well con-
In actual operations, there are many factors that lead
ditions are uncertain and the hole may only be par-
to uncertainties regarding the displacement volume.
tially filled.
■ The diameter of the pipe is not uniform when a
stinger is used. Considering these uncertainties, it is usually prefer-
■ If the plug is placed in an openhole section, its diam-
able to underdisplace the cement rather than risk
overdisplacement.
eter will not be uniform.
Software programs are available to simulate these
■ Often the only way to monitor the volume of mud
effects and determine the optimal displacement volume
pumped is to count the number of pump strokes; (Fig. 14-22). Appendix C also presents the equations for
unfortunately, the pump efficiency is rarely known. manual calculation. Treatment designs performed with
When possible, it is better to directly monitor the foamed cement require special computational methods
fluid level in a calibrated tank. (Kulakofsky and Creel, 1992).
■ The internal pipe volume is frequently unknown, Another way to eliminate the risk of overdisplace-
because the pipe may be worn or covered with a layer ment is to use mechanical plugs. When they land on a
of set cement. The nominal pipe diameter also varies downhole collar, a positive pressure change is observed
within a given range. For example, an internal-diam- at the surface.

Fluid Sequence Plug Interfaces

0
No mud overflow at the rig floor

Pipe shoe depth

Spacer (tube)
Cement slurry (in tubular)
Cement slurry (in annulus)
Spacer (in annulus)
Mud (in tubular)

Top of cement in the pipe

Depth reaches tailpipe top depth
500
(m) Spacer-mud interfaces
at the end of tailpipe

Cement-spacer interfaces
at the end of tailpipe
1,000 Mud
Spacer 0 30 60 90 120 150 180 210
Cement Depth (m)

Fig. 14-22. Example of under-displacement for cement plug placed with a tailpipe. Left diagram: fluids position
at the end of pumping. Right plot: movement of the various fluid interfaces while the pipe is pulled out.

Chapter 14 Remedial Cementing 519

14-4.2.4 Inadequate mud removal If coiled tubing is being used, the excess cement is
An accepted rule is to limit the plug length to about also contaminated and circulated out (refer to Section
1,000 ft (300 m) to minimize contamination. Longer 14-15).
plugs can be set provided special attention is paid to Typical contaminating fluids include borax solutions
controlling gel-strength development and ensuring with bentonite and biopolymer solutions. The aims are
excellent mud removal. If the wellbore is deviated, the to delay the thickening time sufficiently to allow removal
risk of contamination is higher because the plug slides of the cement slurry and to prevent solids sedimentation
downhole as the drillpipe is removed. It is preferable to in the contaminated slurry.
place several shorter plugs on top of each other, even if When the drillstring is used to inject the contami-
the cost is higher. nant, the volume is limited by the length of a pipe stand.
To prevent contamination in the annulus, rotating or In such cases, reverse circulation may be preferred. In
reciprocating the pipe during placement is essential. A all cases, the cement plug itself must not be disturbed.
diverter tool also helps to displace gelled mud and mud- To eliminate this risk, some operators prefer to allow the
cake deposited on the borehole wall. Minimizing pipe excess cement to set and then drill it out.
eccentricity and combining pipe movement and efficient
jetting through the diverter provide the best possible 14-4.4 Pressure and mechanical-strength
mud removal. constraints
Three basic safety constraints must be checked at all
14-4.2.5 Pipe removal and reversing out times during the various stages of the operation, includ-
While pulling the pipe out, cement contamination can be ing cement placement and cleanout.
prevented by using a stinger and reducing the pull-out ■ The dynamic fluid pressure in front of the formation
speed. This is especially important in highly deviated must be less than the formation fracturing pressure.
sections where gravity forces are not helpful. One solu- ■ To prevent formation-fluid entry, the static fluid pres-
tion is to leave the stinger in the cement during setting sure in front of the formation must be higher than the
and to cut it afterwards. pore pressure.
When reversing out excess cement slurry, the pipe ■ The differential pressure across the various tubulars
bottom should be moved at least 100 ft above the top of
must be less than their burst or collapse pressures.
cement to prevent its destabilization. This is discussed
When working with coiled tubing, the weight of
in the following section.
tubing filled with cement must remain within its ten-
sile-strength limits. One must also account for fatigue
14-4.3 Cleanup of the coiled tubing.
Excess cement is cleaned out by either direct or reverse Software that predicts fluid pressures is readily avail-
circulation. Cleanup is performed when the actual hole able, even for foamed fluids (Bour et al., 1990).
volume is uncertain and a large excess of cement has
been used to ensure that the top of cement lies at the
correct depth. A cleanup operation is also required to 14-5 Plug cementing—Cement-slurry
ensure that no cement remains in the drillstring.
Direct circulation involves pulling the pipe above the design
desired top of cement and circulating a contaminating Cement-slurry design is covered elsewhere in this text-
fluid. Cleanup in reverse circulation consists of pumping book (Chapters 3, 7, and 12). However, there are a few
the contaminating fluid down the annulus so that the slurry-design guidelines related to plug cementing that
excess cement slurry goes up the pipe. Each method has deserve to be mentioned here (Jones and Watters,
advantages and disadvantages. 1998).
■ The time required to circulate out a given volume of In general, the thickening time should be adequate
fluid is much shorter when reverse circulating. The for placement, but strength should develop rapidly.
thickening time of the slurry can be designed to be Slurry sedimentation or free water must be scrupulously
shorter for a reverse circulation cleanup. avoided. If weak formations are present, the slurry den-
sity should be the minimum necessary to obtain the
■ When performing reverse circulation, the smaller
desired ultimate strength; otherwise, a dense slurry can
hydraulic diameter of the pipe compared to the annu- be beneficial. To preserve the integrity of the plug, the
lus can lead to higher bottomhole pressures. If lost set cement must not shrink.
circulation is a potential problem, direct circulation
should be selected.

520 Well Cementing

 When the goal of the plug is well abandonment, the ■ A well abandonment plug is tested for impervious-
set cement must have low permeability—preferably less ness by monitoring the fluid level in the well.
than 0.001 mD. If the plug is set in an open hole, the ■ The depth of the top of the plug can be determined by
slurry should have adequate fluid-loss control. tagging the cement after the WOC time has elapsed.
For directional drilling and whipstock plugs, the set
cement should be stronger than the surrounding forma- If the job objective has not been attained or the prob-
tion. If this is not possible, adding toughening agents lem has not been satisfactorily repaired, the reasons for
such as metal fibers can be useful (Chapter 3). failure should be carefully investigated. Design modifi-
For lost-circulation plugs, low-density or thixotropic cations and improvements can then be made to improve
cement systems can be effective. Adding particulate the probability of success during subsequent treatments.
LCMs or fibers is also beneficial. As discussed earlier, the most common causes of fail-
Frequently a remedial cement system must be resis- ure are linked to cement contamination or insufficient
tant to subsequent well stimulation treatments. The cement volume. For a cement plug placed off bottom,
requirement is usually expressed in terms of solubility in contamination is likely caused by an unstable bottom
HF-HCl acid systems. Latex cement slurries and epoxy- interface. A cement plug placed in a washed-out zone is
base synthetic cements are especially resistant to these more likely to fail because mud removal is difficult and
conditions. Combining low cement permeability with the wellbore volume may be underestimated. Finally, if
small-particle-size cement and latex also provides mud removal guidelines have not been followed, the
enhanced acid resistance (Heathman et al., 1993). cement may be contaminated by a mud channel, by mud
Chemical resistance to wellbore fluids—brines and left on the borehole wall, or while pulling out the pipe.
hydrocarbons—is required for long-term durability of
abandonment plugs. Bosma et al. (1998) suggested
using a silicone-based material. For high-temperature 14-7 Squeeze cementing—Introduction
applications, special systems have been developed that As mentioned earlier, squeeze cementing has many
provide good durability in the presence of brines applications.
(Chapter 10). ■ Repairing a primary cement job that has failed
Adequate bonding of the set cement to the casing is
because of mud channeling or insufficient cement
required to ensure durable isolation. The bonding may
height in the annulus
be compromised by radial casing movement following
■ Eliminating water intrusion from above, below, or
the high mechanical stresses exerted during production
operations or negative-pressure tests. The shear within the hydrocarbon-producing zone
strength of the bond between the set cement and casing ■ Reducing the producing GOR or WOR by isolating the
must also be sufficient to withstand high differential gas or water zones from adjacent oil intervals
pressures. The bond strength of cement with steel is usu- ■ Repairing casing leaks caused by corroded or split
ally assumed to be a small fraction of the cement com- pipe
pressive strength (Chapter 6). The bond strength can be ■ Abandoning a nonproductive or depleted zone
increased by using expansive cement systems (Vijn and
■ Plugging one or more zones in a multizone injection
Fraboulet, 2001) or nonshrinking materials such as sili-
cone-cement composites (Bosma et al., 1998). well to direct the injection into the desired intervals
■ Sealing lost-circulation zones

■ Protecting against fluid migration into a producing

14-6 Cement-plug evaluation zone
After the waiting-on-cement (WOC) time has elapsed, In the following sections, a review of perforation plug-
the job results are evaluated. For cement plugs this ging theory, tools, placement techniques, slurry design,
process is fairly straightforward. job design, and job monitoring is presented. Finally job
■ The quality of a whipstock or kickoff plug is tested evaluation and reasons for failures are discussed.
while attempting to kick off.
■ The efficiency of a lost-circulation plug is evaluated
by comparing the loss rates before and after the
treatment.

Chapter 14 Remedial Cementing 521

14-8 Squeeze cementing—Perforation Cake porosity:
plugging theory
During most squeeze cementing treatments, the parti-
( )
V fc = φ × V fc + Vsolids (14-4)

cles in the cement slurry are too large to enter the for- Inserting Eq. 14-2 and 14-3 into Eq. 14-4, one obtains
mation matrix. As a result, an external cement filtercake Vfc = φVfc + f (Vfilt + Vfc) , from which one may extract the
accumulates, fills the perforations, and forms nodes that relationship giving the cake volume as a function of fil-
protrude into the wellbore (Figs. 14-23 and 14-24). trate volume.
When a volume of slurry, Vslurry, containing a solid-
volume fraction, fsV, is forced against a porous medium, fsV
V fc = V = wV filt (14-5)
a liquid filtrate of volume Vfilt passes into the medium. 1 − fsV − φ filt
The solids that remain behind produce a filtercake of
porosity φ. Mathematically, this is written as follows. The factor
Solids volume fraction: fsV
1− fsV − φ
Vsolids
fsV = (14-2)
Vslurry is called the deposition factor, w. It corresponds to the
ratio of the filtercake volume to the filtrate volume and
Conservation of volumes: can be measured by a standard American Petroleum
Institute (API) or International Organization for Stan-
Vslurry = V filt + V fc (14-3) dardization (ISO) fluid-loss test. Experimentally, one
where observes that this factor is almost constant when the
Vfc = filtercake volume. differential pressure is varied, indicating that cement
filtercakes are incompressible. A neat 15.8-lbm/gal
[1,900-kg/m3] cement slurry has a solids volume fraction
of 40%. The porosity of cement filtercakes from this
Formation system is usually about 30%. Thus, a typical value for the
Cement Primary
slurry cement deposition factor is about 1.3.
To determine the time required to build a filtercake
Filtercake
Casing of given height hfc under a constant filtration pressure,
Δp, Darcy’s law is frequently used under the assumption
that the pressure drop is constant throughout the cake.
The following relationship is obtained where the cake
permeability is kfc and the filtrate viscosity is μ filt .
Filtrate

2 k fc wΔ p
h fc = × t (14-6)
μ filt
Fig. 14-23. Filtercake buildup into a perforation channel.
Expressed in terms of the API/ISO fluid loss of the
slurry, VAPI, the cake height is
Casing Primary Formation
V API Δp t
cement h fc = w × × × , (14-7)
Dehydrated cement AAPI ΔpAPI t API

Cement where the subscript API refers to API/ISO conditions:

nodes AAPI = 3.5 in.2, Δ pAPI = 1,000 psi, and tAPI = 30 min. The
time required to build a cake of height hfc is therefore
2 2
⎛ A ⎞ ΔpAPI 6.1 × 107 ⎛ h fc ⎞
t = t API ⎜ h fc API ⎟ ≈ ⎜ ⎟ .
⎝ wV API ⎠ Δp Δp ⎝ wV API ⎠
Fig. 14-24. Perforation channel properly filled with dehydrated
cement. (14-8)

522 Well Cementing

 Using this simple approach, the time required to Table 14-3. Filtercake Permeability and Slurry-Dehydration
build a 2-in. [5.1-cm] thick cake with a slurry having an Rate†, ‡
API/ISO fluid-loss rate of 80 mL/30 min under a differ- Concentration API/ISO Permeability Time to Form
ential pressure of 500 psi [3.5 MPa] is of Fluid-Loss Fluid Loss of the Filtercake a 2-in. Thick
Additive at 1,000 psi Formed at 1,000 psi Filtercake
2
6.1 × 107 ⎛ 5 cm ⎞ (gal/sk) (mL/30 min.) (mD) (min.)
t≈ ≈ 180 min.
500 psi ⎜⎝ 1.3 × 80 mL ⎟⎠ 0.00 1,200 5 0.2

0.07 600 1.6 0.8

It is important for the reader to realize that the above
equations provide an approximate model to determine 0.13 300 0.54 3.4
the relative orders of magnitude of relevant squeeze
0.17 150 0.19 14
cementing parameters. Some caution should be exer-
cised when attempting to use these equations to 0.19 100 0.09 30
describe real downhole conditions.
Hook and Ernst (1969) gave orders of magnitude of 0.22 50 0.009 100
the time required to build 2-in. [5.1-cm] thick filtercakes 0.24 25 0.006 300
as a function of the API/ISO fluid-loss value (Table 14-3). † As a function of the fluid-loss-additive concentration. From Hook and Ernst (1969);
This cake height is accepted as a good estimate of reprinted with permission of SPE.
‡ API Class A cement with liquid fluid-loss additive and 46% water.
what is needed to fill perforations and build cement
nodes (Fig. 14-25).
Table 14-3 is a compilation of permeability measure-
ments conducted on filtercakes formed with different The data presented in Table 14-4 demonstrate the
concentrations of a fluid-loss additive in the original influence of squeeze pressure on the rate of filtercake
cement slurry. The permeability of a neat-cement filter- growth. First, it shows that varying the squeeze pressure
cake was about 5 mD—a value lower than that of many from 500 to 1,000 psi [3.5 to 6.9 MPa] does not influence
producing sandstone formations. When the slurry con- the permeability of the resulting filtercake.
tained sufficient fluid-loss additive to reduce the Nevertheless, in keeping with Darcy’s law, the data show
API/ISO fluid-loss rate to 25 mL/30 min, the resulting fil- that the flow rate of fluid through the filtercake is
tercake was approximately 1,000 times less permeable. directly proportional to the squeeze pressure.
This value approaches the permeability of matrix forma-
tions that produce very slowly and have a low injectivity.
Table 14-4. Effect of Differential Pressure on the
Permeability of Filtercakes and the Filtration Rate†, ‡
Fluid loss (1,000 psi Differential Permeability API/ISO Flow Rate
differential pressure)
of Filtercake of Filtercake Fluid Loss Through
Formation (mD) (mL/30 min) Filtercake
800 mL/30 min
(psi) (mL/min)
Slurry I§ 500 5.8 1,200 50
1,000 6.0 1,200 110
150 mL/30 min
Slurry II†† 500 1.9 600 17
Nodes 1,000 1.6 600 30
50 mL/30 min
Slurry III‡‡ 500 0.53 300 4.7
1,000 0.54 300 9.7
† From Hook and Ernst (1969); reprinted with permission of SPE.
15 mL/30 min ‡ By increasing the pressure from 500 psi to 1,000 psi, the filtration rate is increased by a

factor of about 2.
§ Class A cement, 45% water.
6-in. casing †† Class A cement with 0.5% dispersant, 0.07 gal/sk liquid fluid-loss additive, and 46% water.

‡‡ Class A cement with 0.5% dispersant, 0.13 gal/sk liquid fluid-loss additive, and 45% water.

Fig. 14-25. Node buildup after a 45-min squeeze, using slurries with
different fluid-loss rates (from Rike, 1973; reprinted with permission
of SPE).

Chapter 14 Remedial Cementing 523

 Primary cement
Equation 14-7 provides only an estimate of filtercake hnode
growth, especially when the formation has a low perme- Formation
ability. In this case, an appreciable part of the differen-
tial pressure is applied through the formation, and cake Node ri
growth is much slower than Eq. 14-6 predicts. Hook and buildup ro rperf
Ernst reported experimental data to support this fact by
building cement filtercakes on various filtration media Wellbore
Casing
Radius of perforation
(Table 14-5). These data support another conclusion by
Hook and Ernst: Cement cakes are indeed not com- hcomb
pressible, and the cake permeability does not vary with
the differential pressure. Fig. 14-26. Geometry of a cement node (from Binkley et al., 1958;
reproduced courtesy of the American Petroleum Institute).

Table 14-5. Effect of Formation Permeability on the

Rate of Filtercake Growth† Step 1. The depth of the perforation is considered
large in relation to its diameter (at least four times
Time required for the formation of a 11⁄4-in. filtercake on a 1-in.
diameter filtration surface at 1,000 psi
greater). This allows one to assume that the cake builds
up radially inside a cylindrical perforation of radius rperf.
Bandera Berea API/ISO Assuming that the pressure drop in the formation is
Sandstone Sandstone 325-Mesh
30 mD 300 mD Screen
zero, Binkley et al. demonstrated that the time required
to build a filtercake filling the perforation tunnel inside
Slurry I‡ 6 min 2.5 min 2.5 min the formation can be expressed by the following equa-
Slurry II§ 9 min 6.5 min 6.5 min tion.

Slurry III†† 5 min 2.5 min 2.5 min

μ filt ( rperf )2
t= × (14-9)
† From Hook and Ernst (1969); reprinted with permission of SPE. k form × w × Δ p 4
‡ API Class A cement with 0.5% dispersant, 0.14 gal/sk liquid fluid-loss additive, and

46% water.
§ API Class A cement with 0.5% dispersant, 0.17 gal/sk liquid fluid-loss additive, and
The ratio
μ filt
46% water.
†† API Class A cement with 0.7% solid fluid-loss additive and 46% water.

k form × w × Δ p

Consequently, it is common practice to design the contains all the variables related to the deposition prop-
fluid-loss rate of a squeeze cementing slurry according to erties of the cement slurry and is called the composition
the formation properties, with high fluid-loss rates for factor. It is interesting to note that, with the assumption
low-permeability formations and low fluid-loss rates for regarding perforation depth versus its diameter, the
high-permeability formations. For vugular zones, Grant depth of the perforation has no effect on the deposition
et al. (1990) reported success with a two-slurry squeeze process.
design. A lead slurry with a short pumping time and a Step 2. For the second step, the flow is assumed to be
fairly high API/ISO fluid-loss rate (300–500 mL /30 min) linear along the axis of the perforation. The time
is followed by a tail slurry with a longer pumping time required to build a filtercake in close contact with the
and lower fluid-loss rate. The tail slurry is used for hesi- inside of the casing can be calculated by the following
tation. equation.
Binkley et al. (1958) performed a more detailed
μ filt ⎡ (h )2 ⎤
analysis of perforation squeezing. They accounted for
the successive perforation filling followed by node t= × ⎢ comb + K × rperf × hcomb
k form × w × Δ p ⎢ 2
( )⎥⎥,
buildup according to the geometry shown in Fig. 14-26. ⎣ ⎦
The squeezing process is divided into three successive (14-10)
steps:
1. filling of perforation tunnels located inside the for- where hcomb is the combined thickness of the cement
mation sheath and casing and K is a factor that was used to fit
2. filling of perforation tunnels crossing the casing and the experimental data and was determined to be 0.25.
cement sheath
3. building cement nodes.

524 Well Cementing

 Step 3. To simplify the calculation, Binkley et al. 10,000
(1958) assumed that the nodes building inside the
casing have a spherical shape at every stage of growth
10
and that the lateral growth occurs at the same rate as 9
the vertical growth. The geometry is illustrated 1,000 8
7
in Fig. 14-26. If hnode is the height of the node building 6
5
up inside the casing, Eq. 14-11 can be derived. 4
Time to fill
3
100
dhnode k form × w × Δ p perforation
= and grow a 2

dt μ filt node inside 1

0
casing hnode
⎧ 1 ⎫ (dimension-
10
rperf
⎪⎪ ⎪⎪ less)

)( ) ( )
2 2
×⎨⎡
(
r ro ⎬
⎤ o
⎪ ⎣ hcomb + β + K × rperf ⎦ + − ro ⎪
⎪⎩ ri ( )2 ( )ri
2
⎪⎭ 1
(14-11)
where
hcomb = combined thickness of the cement sheath 0.1
and casing 0 2 4 6 8 10 12
K = 0.25 Ratio of combined casing and cement
thickness to perforation radius

ri = ( ro − hnode )2 + ( rperf )2 Fig. 14-27. Relationship between slurry property, perforation geom-
etry, and filtercake-node height inside casing (from Binkley et al.;
reproduced courtesy of the American Petroleum Institute).
( )
2
( hnode )2 + rperf + hnode
ro =
2 hnode
where pi = pressure for a step of duration t. For slightly
rperf = radius of the perforation compressible cakes, Eq. 14-6 and 14-7 can still be used
provided the cake properties are measured at the effec-
( ri )2 − ( rperf ) .
2 tive filtration pressure.
β = ri −

This equation was integrated numerically. The three

calculated times are then added, and the results are 14-9 Squeeze cementing tools and
plotted in Fig. 14-27. Using dimensionless variables, techniques
these results represent the time required to fill a perfo- There are two traditional and fundamentally different
ration and build a node, versus the ratio hcomb /rperf, for squeeze job classifications.
different node heights (hnode/rperf ). ■ Low-pressure squeeze: The bottomhole treating pres-
The bottom curve, hnode /rperf = 0, represents the times sure is maintained below the formation fracturing
required to obtain a cake that is flush with the inside of pressure.
the casing. ■ High-pressure squeeze: The bottomhole treating pres-
In actual squeeze operations, the squeeze pressure is sure exceeds the formation fracturing pressure.
increased in steps. Provided the cement cake is truly
incompressible, Eqs. 14-6 and 14-7 remain valid if the fil- Within these two classes, there are two basic tech-
tration pressure is replaced by an effective filtration niques (the Bradenhead squeeze and the squeeze-tool
pressure, Δ pe, equal to the time-average of the actual fil- technique) and two pumping methods (the running
tration pressure, squeeze and the hesitation squeeze). Each of these clas-
sifications and techniques is explained in this section.
Δpe =
∑ Δpi ti , (14-12)
These techniques are also compared and contrasted in
∑ ti Tables 14-6, 14-7, and 14-8.

Chapter 14 Remedial Cementing 525

Table 14-6. Comparison of High-Pressure and Low-Pressure Squeeze Techniques
High Pressure Low Pressure
Features Induces fractures in the formation that allow Offers the best option for squeezing the pay zone
cement to replace the dirty fluid in the wellbore Uses small volume of slurry
Avoids fracturing the formation

Applications When In
■ “dirty” wellbore fluids must be displaced to allow ■ multiple zones
placing the cement into a void ■ long intervals
■ there is mud in the hole that must be displaced ■ wells with low bottomhole pressure
before cement can enter the voids
■ there are no voids to fill, e.g., shoe squeeze,
block squeeze, or when cementing liner top

Disadvantages Larger slurry volumes will be required to fill More careful design and execution is required.
the additional void space created by the fracture.
It may be difficult to develop filtercake in the fracture.
Squeeze pressure may be difficult to attain.

Table 14-7. Comparison of Running- and Hesitation-Squeeze Techniques

Running Squeeze Hesitation Squeeze
Applications Only when All types of problems
■ fluids are clean
■ the formation is free from fractures
or interconnected voids
■ the operation can be carried out below
formation fracture pressure
Advantages Easy to apply Self-diverting technique
Simple slurry design, little fluid-loss Better way of filling all the voids
control needed and creating a complete hydraulic seal
Good chance of obtaining final squeeze Achievement of “squeeze pressure”
pressure under correct conditions a relatively good indicator of success

Disadvantages Attaining a squeeze pressure is not Job procedure is more complicated.

a reliable indicator of success. Cement placement time is longer, resulting
The method cannot be relied upon in increased time WOC.
to fill all voids. Squeeze tool placement is more critical.
The method can only be carried out with The procedure requires a skilled supervisor
clean fluids in the hole. with an ability to “feel the well” to modify job
execution, including
■ cement volume
■ pumping rate
■ stage volume
■ hesitation time.

Slurry design is more complex, because

■ pumping time testing procedure must
account for hesitation periods
■ low gel strength is needed during
hesitation periods
■ requirements for fluid loss will vary depending
on nature of voids and permeability.

526 Well Cementing

Table 14-8. Comparison of Bradenhead and Squeeze-Tool Techniques
Retrievable Packer Drillable Packer or Retainer No Packer or Bradenhead Technique
Advantages Does not have to be drilled out Can be set close to a formation Offers less chance of failure with
Can be set and released many or between perforations simple procedure and no packer to set
times on a single trip Can be set on wireline to ensure Eliminates need to squeeze fluid ahead
Can be recovered, redressed, good depth control because of cement placement just in
and reused Isolates the squeeze zone front of the perforations
Allows easy cleanup after squeezing

Disadvantages Backflow cannot be prevented. The procedure can be used The whole casing is exposed
Reversing of excess slurry can only once. to the pressure.
disturb the squeeze. Drillout takes time. The whole casing must be free
If reused on one trip, the cement An additional trip is required of leaks.
buildup in the workstring changes to set the packer. During the squeeze, the casing expands
displacement. and may partially or totally plug a
Many mechanical problems are microannulus: The flow of the cement
possible when trying to get the slurry may be much more difficult
packer through tubing. through the restricted flow channel.
It is possible that the packer could The channel will reopen once pressure
prematurely open during placement. is released.
The slurry could be contaminated
during the placement, especially if
the wellbore fluid is lighter than
the cement slurry (fluid swapping
in large sections).
The wellbore fluid should be
squeezed before the cement.

14-9.1 Low-pressure squeeze 14-9.2 High-pressure squeeze

As discussed earlier, the aim of a squeeze operation is to In some cases, a low-pressure squeeze of the perfora-
fill perforation cavities and interconnected voids with tions will not accomplish the job objective. The channels
cement filtercake, with a small “node” protruding inside behind the casing may not be directly connected to the
the casing. The cement-slurry volume is usually small, perforations. Small cracks or microannuli that may allow
because no slurry is actually pumped into the formation. gas flow do not allow the passage of a cement slurry.
Precise control of the pump pressure and the hydrosta- Such channels must be enlarged to accept viscous solids
tic pressure of the cement column is essential because carrying fluid. In addition, many low-pressure operations
excessive pressure could result in formation breakdown. cannot be performed if it is impossible to remove plug-
In low-pressure squeezes, it is essential that perfora- ging fluids or debris ahead of the cement slurry or inside
tions and channels be clear of mud or other solids. If the the perforations.
well has been producing, such openings may already be Cement-slurry placement behind the casing is accom-
free of obstructions; however, for newly completed wells, plished by breaking down, or fracturing, the formation at
it may be necessary to clean the perforations before per- or close to the perforations (Fig. 14-28). Fluids ahead of
forming the squeeze job (Section 14-9.9). the slurry are displaced in the fractures, allowing the
A properly designed slurry will leave only a small node slurry to fill the desired spaces. Further application of
of cement filtercake inside the casing. Improperly pressure dehydrates the slurry against the formation
designed systems that cause excessive filtercake devel- walls, leaving all channels (from fractures to perfora-
opment can lead to complete bridging of the inside of tions) filled with cement cake.
the casing, loss of pressure transmission to the forma- However, during a high-pressure squeeze, the loca-
tion, and insufficient contact of the cement filtercake tion and orientation of the created fracture cannot be
with the formation. controlled. Sedimentary rocks usually have a low tensile
Review of the literature shows that most authors strength and are held together primarily by the the com-
believe a low-pressure squeeze should be run whenever pressive forces of overlying formations. These cohesive
possible and that this technique has the highest success forces act in all directions to hold the rock together
rate (Rike and Rike, 1981; Goodwin, 1984; Bradford and but do not have the same magnitude in all directions.
Reiners, 1985). When sufficient hydraulic pressure is applied against a

Chapter 14 Remedial Cementing 527

 Dehydrated cement Fracturing
Mud pressure, pfrac Vertical stress, σvert
Primary cement Casing
Cement
filtercake
Filtrate Vertical
Mud fracture
Horizontal
stress,
σh1
Horizontal
stress,
σh2
pfrac pfrac

Fig. 14-28. High-pressure-squeeze technique generates a vertical

fracture.

formation, the rock fractures along the plane are per-

pendicular to the direction of the least principal stress
(Fig 4-29). A horizontal fracture is created if the frac- Induced Horizontal Fracture Induced Vertical Fracture
turing pressure is greater than the overburden pressure. pfrac > σvert pfrac > σh1 or σh2
A vertical fracture occurs if overburden pressure is σvert < σh1 or σh2 σh1 or σh2 < σvert
greater (Roegiers, 1989).
Fig. 14-29. Effect of well depth and vertical/horizontal formation
The extent of the induced fracture is a function of the stresses on the orientation of a hydraulic fracture induced by
pump rate applied after the fracture is initiated. The injected fluid. σvert = vertical stress; σh1 and σh2 = stresses in differ-
amount of slurry used depends on the way the operation ent horizontal directions (from Suman and Ellis 1977; reprinted with
is performed. High pump rates generate large fractures; permission from World Oil).
thus, large volumes of cement are required to fill them.
A properly performed high-pressure squeeze should
place the cement as close to the wellbore as possible. from 10 to 100 bbl are commonly used. A modified run-
Drilling muds or other fluids with low fluid-loss rates ning squeeze technique, using pumps that can deliver
should not be pumped ahead of the slurries. A wash with several barrels of slurry at rates as low as 0.06 bbl/min
a high fluid-loss rate, such as water or a weak hydrochlo- [10 L/min] to avoid formation fracturing, has been used
ric acid solution, not only opens smaller fractures but successfully to seal narrow microannular gaps (Slater et
also cleans perforations and the cement path. As a al., 2001).
result, the fracture initiation pressure is lower.
14-9.4 Hesitation squeeze method
14-9.3 Running squeeze method During a squeeze cementing job, the rate at which
During a running squeeze procedure, the cement slurry cement filtrate leaks into the formation is lower than the
is pumped continuously until the final desired squeeze minimum pump rate of most field equipment. Therefore,
pressure—which may be above or below the fracture maintaining a constant differential pressure is nearly
pressure—is attained. The pressure is monitored after impossible, especially when one is trying not to exceed
pumping stops. If the pressure falls owing to additional the formation fracture pressure.
filtration at the cement/formation interface, more slurry A solution to this problem is the hesitation squeeze
is pumped to maintain the final squeeze pressure. This pumping method. This procedure involves the intermit-
continues until the well maintains the squeeze pressure tent application of pressure—by pumping at a rate of 1⁄4
for several minutes without additional cement-slurry to 1⁄2 bbl/min—separated by an interval of 10 to 20 min
injection. The volume of slurry injected is usually large. for pressure falloff caused by filtrate loss to the forma-
Rike and Rike (1981) reported that volumes ranging tion. The initial leakoff is normally fast because there is

528 Well Cementing

no filtercake. As the cake builds up and the applied pres-
sure increases, the filtration periods become longer and
the differences between the initial and final pressures
become smaller. At the end of the job, the pressure
falloff becomes negligible (Fig. 14-30). The slurry vol-
umes necessary for this technique are usually much less
than those required for a running squeeze.
A loose formation normally requires a long hesitation
period to begin building squeeze pressure. A first hesita-
tion period of 30 min or more is not unreasonable. A
much shorter initial hesitation period (possibly 5 min) is
normally sufficient for tight formations (Grant and
White, 1987). Instead of prescribing pump times, the
best method is to continue pumping as long as squeeze
pressure continues to build.

A = Slurry mix water leaks off.

20 B = No slurry mix water filtrates;
the squeeze is completed. D
C = Pressure is bled.
16 D = Pressure is tested.
B C
A
Surface 12
pressure
(100 psi)
8 Spot Cement Apply Squeeze Reverse Circulate
Pressure
4 Fig. 14-31. Bradenhead squeeze technique (from Suman and Ellis,
1977). Reprinted with permission from World Oil.
0
0 20 40 60 80 100 120 140
Time (min)
14-9.6 Squeeze-tool placement technique
Fig. 14-30. Hesitation squeeze pressure behavior.
This technique can be subdivided into two parts—the
retrievable squeeze packer method and the drillable
cement retainer method. The main objective of using
squeeze tools is to isolate the casing and wellhead while
14-9.5 Bradenhead squeeze (no packer)
applying high pressure downhole.
The Bradenhead squeeze technique, illustrated in
Fig. 14-31, is a low-pressure squeeze technique practiced
when there are no doubts concerning the casing’s ability 14-9.6.1 Retrievable squeeze packer method
to withstand the squeeze pressure. No special tools are Retrievable packers with different design features are
involved, although a bridge plug may be required to iso- available (Chapter 11; Savage and Fowler, 1994).
late other open perforations farther downhole. Compression- or tension-set packers are used in squeeze
Open-ended tubing is run to the bottom of the zone to cementing. As shown in Fig. 14-32, they have a bypass
be cemented. Blowout preventer (BOP) rams are closed valve to allow the circulation of fluids while running in
over the tubing, and an injection test is performed. The the hole and after the packer is set. This feature allows
cement slurry is subsequently spotted in front of the per- tool cleaning after the cement job and reversing out
forations. Once the cement is in place, the tubing is excess slurry without excessive pressure. The bypass
pulled out to a point above the cement top, the BOPs are valve also prevents a piston or swabbing effect while run-
closed, and pressure is applied through the tubing. The ning in or out of the hole.
Bradenhead squeeze is popular because of its simplicity. The principal advantage of the retrievable packer
over the drillable retainer is its ability to set and release
many times, allowing more flexibility.

Chapter 14 Remedial Cementing 529

 Gauge
ring

Hold-down
slips
Packer

Packing Fig. 14-33. Squeeze with cement retainer.

elements

Bypass A drillable retainer gives the operator more confi-

hole dence in setting the packer closer to the perforations.
Another advantage is that a smaller volume of fluid
below the packer is displaced through the perforations
Packing ahead of the cement slurry.
elements
Slips
14-9.7 Perforation washing tool
Bypass
hole
Perforation washing before a squeeze job is a useful
Friction
method for making all perforations receptive to the
Slips squeeze cement slurry. This can be done by mechanical
pads
or chemical means.
Friction Mechanical perforation washing involves the use of a
pads washing tool coupled with back-surge techniques. The
perforation-washing tool (Fig. 14-34) isolates a small
number of perforations at a time. A wash fluid is pumped
down the tubing and forced into the perforations. It is
then forced outside the casing and back through upper
Figure 14-32. Bridge plug and squeeze packer.
perforations into the annulus. The tool is slowly moved
upward to wash the entire perforated interval. Common
wash fluids are surfactant solutions, followed by weak
acids when scales or drilling muds are to be removed.
14-9.6.2 Drillable cement retainer Organic solvents are used when paraffin deposits are
present.
Cement retainers are drillable packers that have a valve The surge tool (Fig. 14-35) is basically an air chamber
that is operated by a stinger at the end of the workstring between an upper and lower valve. The tool is run in the
(Fig. 14-33). Cement retainers are used to prevent back- hole with a packer to isolate the desired interval. Once
flow when no cement dehydration is expected or when a the packer is set, the lower valve is opened so fluids can
high negative differential pressure may disturb the enter the air chamber. The rapid depressurization of the
cement cake. In certain situations, using a packer is borehole creates a high differential pressure across the
risky because of potential communication with upper perforations, inducing the removal of debris and other
perforations. When cementing multiple zones, the plugging materials. To establish circulation after surg-
cement retainer isolates the lower perforations and sub- ing, the upper valve is opened (by means of tubing move-
sequent zone squeezing can be performed without wait- ment, tubing pressure, or disk rupturing), and the debris
ing for the slurry to set. is reverse-circulated out of the hole.

530 Well Cementing

 J-2 unloader

Upper mechanical
disc valve

Drag spring
assembly
Hydraulically operated
Built-in surge valve
bypass
Casing Packer
cups
Washing
ports

Mule shoe
ball valve
Fig. 14-35. Surge tool.

Coiled tubing has proved to be a very economical

method to accurately place the small volumes of cement
slurry usually involved in remedial cementing opera-
tions. Harrison and Blount (1986) reported that, in some
instances, up to 85% savings on workover costs can be
Fig. 14-34. Left: cup-type perforation washing tool. Center: wash achieved when using coiled tubing.
tool (courtesy of Baker Oil Tools). Right: perforation wash straddle As illustrated in Fig. 14-36, the procedure can be
packer (courtesy of BJ Services). divided into several steps.
■ A supporting column of mud (viscous pill) is injected
Chemical perforation cleaning techniques involve until its level is just below the perforations to be
pumping acids and solvents as spearhead fluids to clean squeezed. Some mud contamination may occur
the perforations ahead of the cement slurry. Sometimes during placement because of mixing with wellbore
the injectivity decreases after an acid job, owing to emul- fluids; thus, the coiled tubing string nozzle is pulled
sions formed by the formation fluid and the acid. Prejob up, and contaminated mud is circulated out. The well-
compatibility tests should be performed to prevent this bore above the mud is then loaded with water or
problem and select the best acid-surfactant combina- diesel.
tion.
■ Cement is pumped with the nozzle located just above
the mud/water interface. When the perforations are
14-9.8 Squeeze with coiled tubing covered with cement slurry, squeeze pressure is
The use of coiled tubing for remedial cementing began applied. The nozzle must be kept below the water/
in the early 1980s. Since then, the technique has gained cement interface.
considerable popularity (Walker et al., 1992; Loveland ■ After the squeeze pressure is reached, a contaminant
and Bond, 1996). Squeeze cementing with coiled tubing fluid is injected to dilute the cement slurry, the cont-
is similar to that using drillpipe. Specialized tools can be aminated cement and mud are reversed out, and the
run through tubing, such as small outside diameter (OD) wellbore is flushed clean.
and inflatable packers (Fig. 14-32) (Wilson et al., 2004).

Chapter 14 Remedial Cementing 531

 Filtered Slurry pumped at Slurry pumped at
workover fluid maximum rate/ maximum rate/
at high rate Choke pressure allowed Choke pressure allowed Choke closed; if the
open open wellbore is not full,
then, pump slowly
Packer down annulus to
Workover prevent U-tubing
fluid
Spacer/ Workover
fresh water fluid Nozzle placed
Nozzle pulled Cement above thief zone
Cement slurry up 50 ft below slurry
Wellbore clean
cement interface Workover
and packed
fluid
Wellbore Preparation Laying in Cement Slurry Placing Thixotropic Slurry

Low-rate Displacement Contaminant

continuous fluid pumped pumped at
pumping or Choke back at maximum Choke back returns, maximum Returns choked
hesitation returns, monitoring rate/pressure increasing final rate/pressure to maintain
pressure and allowed squeeze pressure pressure on
volumes squeezed zone
Workover
fluid Nozzle pulled Workover Nozzle moved
Spacer/ up >50 ft above fluid continuously
fresh water cement interface or frequently Contaminated
Spacer/ Nozzle penetrates
slurry slurry at a rate which
Cement slurry fresh water
provides a 50% mix
Cement slurry Cement of contaminant
slurry
Commencing Squeeze Completing Squeeze Contaminating Excess Slurry

Fluid pumped at Open returns Fluid pumped Open returns Fluid pumped
maximum rate/ from CT at maximum from CT at maximum Returns choked
pressure for rate/pressure rate/pressure to maintain
allowable for allowable pressure on
differential differential squeezed zone
(1,500 psi) (1,500 psi)
Nozzle penetrates Nozzle reciprocated
slurry at a rate Workover through treatment
Workover Nozzle penetrates Workover
which provides fluid zone to TD
fluid contaminated slurry fluid
a 50% mix
at a rate which
provides a 50% mix of Differential pressure
Contaminated contaminated slurry Cement maintained against
slurry and workover fluid slurry squeezed zone
Reverse Circulating Excess Slurry Reverse Circulating Live Slurry Wellbore Circulated Clean

Fig. 14-36. Coiled tubing squeeze.

This technique has suffered from a few specific draw- 14-10 Squeeze cementing—Cement-slurry
backs: poor depth control, pipe-volume errors, and fluid design
contamination (Noles et al., 1996).
■ Pessin and Boyle (1997) reviewed the various factors
For squeeze cementing, the selection of a fluid or
sequence of fluids must be tailored to the type of treat-
and causes of errors in depth measurement and pro- ment, the technique to be used, and the nature of the
posed an improved measurement that provides the voids. It is not uncommon that more than one candidate
required accuracy. slurry is prepared and the final slurry selection is made
■ A good knowledge of the coiled tubing volume is
at the wellsite based on last-minute information.
mandatory for placing small slurry volumes. Indirect Frequently, the decision is based upon the results of the
measurements are unreliable. One should directly injection test.
monitor the volume of the tubing on its reel.
■ When low volumes of slurry are placed, fluid contam-
ination may be prevented by using mechanical plugs
such as foam balls (Loveland and Bond, 1996).

532 Well Cementing

14-10.1 Slurry selection a tail slurry with a lower fluid-loss rate for hesitation
The properties of a squeeze-cement slurry must be tai- squeezing (Grant et al., 1990). Lost-circulation mate-
lored to the formation characteristics and the squeeze rials may also be added to bridge fractures. If the
technique. The slurry properties are important in three formation permeability is sufficiently high, a medium-
respects: to high-fluid-loss slurry (200 to 500 mL/30 min) will
usually permit filtercake formation and subsequent
■ to allow proper placement from surface to downhole
diversion of slurry into smaller cracks.
■ to allow fluid placement behind the casing or in the ■ If the voids behind the casing are narrow channels
perforations into which deep penetration is required, such as
■ to obtain the desired properties of the set material. channels or microfractures, then very good fluid-loss
control (<50 mL/30 min) is required to prevent early
Many types of slurries have been used for remedial slurry dehydration.
cementing, from standard Portland cement systems to
sophisticated systems adapted from civil engineering. It Appropriate fluid-loss rates are required to obtain
is generally agreed that a squeeze cementing slurry optimal filtercakes. For example, a Portland cement
should be designed to have the following general attrib- system with a low fluid-loss rate will provide a hard fil-
utes: tercake that fills a perforation. Conversely, large nodes
■ low viscosity—to allow the slurry to penetrate small resulting from slurries with an excessive fluid-loss rate
cracks will be difficult to clean and may inhibit production by
reducing the internal casing size.
■ low gel strength during placement—because a
gelling system restricts slurry movements and causes
increases in surface pressure that are difficult to 14-10.3 Rheology and sedimentation
interpret
The requirements in terms of slurry rheology are often
■ appropriate cement particle size
in conflict for squeeze jobs.
■ no free water
■ The ability of the slurry to flow into narrow channels
■ appropriate fluid-loss control—to ensure optimal fill- is a direct function of its fluidity. Thick slurries,
ing of the cracks or perforations although useful when cementing large voids, will not
■ proper thickening time—to safely meet the antici- flow into small restrictions unless they are subjected
pated job time. to high differential pressures that may exceed the
formation fracture pressure. Therefore, low-viscosity
The specifications to test a squeeze cement slurry are slurries containing dispersants are commonly used.
found in API Specification 10A, Specification for
■ Thin slurries are required when cement placement is
Cements and Materials for Well Cementing (2001)
(Appendix B). performed through coiled tubing. Nowak et al. (1996)
mention a maximum slurry viscosity of 60 cp.
■ High slurry yield stress or gelation can lead to misin-
14-10.2 Fluid-loss control and filtercake terpretation of surface-pressure readings during the
development pumping period. Gelation will prevent the application
The requirements for fluid-loss control are defined so of the full pressure through the filtercake, and cake
that the desired filtercake forms, with properties buildup will be impaired.
depending upon the formation and the type of voids to be ■ If the slurry viscosity is too low, the slurry may exhibit
filled. In addition, the relationship between API/ISO free water and sedimentation, and its downhole prop-
fluid-loss volume and filtercake properties is dependent erties will be unpredictable.
upon the slurry composition. ■ When the slurry is placed off-bottom on a lighter fluid,
■ One method is to tailor the API/ISO fluid-loss rate to
and no mechanical base is available to prevent slump-
the formation permeability, with low fluid-loss vol- ing, increasing the cement yield stress is a good way
umes (less than 100 mL/30 min) for formations with to decrease the risk of slumping.
permeabilities less than 100 mD. This rule allows
competent filtercakes to form within a reasonable
time period. 14-10.4 Density
■ If large voids or fractures exist behind the casing and
Better set-cement properties are usually correlated with
there is no flow restriction, high fluid-loss-rate slur- high slurry density. However, high-density slurries tend
ries (300 to 500 mL/30 min) are initially applied to to be viscous. The slurry density may also be limited to
allow fast filtercake buildup in the voids, followed by avoid excessive hydrostatic pressures during placement.

Chapter 14 Remedial Cementing 533

These contradictory requirements can be met by using Sometimes two slurries are used in sequence when all
engineered particle size slurries (Boisnault et al., 1999; of the required properties cannot be obtained with a
Chapter 7) that allow the following: single fluid.
■ low-density slurries with good mechanical properties ■ A deep-penetrating fluid can be followed by a con-
(Elmoniem et al., 2000) ventional cement slurry that builds a filtercake to
■ high-density slurries with relatively low viscosities obtain a squeeze pressure. The first fluid can be a
(Pokhriyal et al., 2001). small-particle-size cement slurry (Heathman et al.,
1993; Bailey et al., 2000) or a solids-free silicate-base
solution (Ness and Gatti, 1995).
14-10.5 Size of particles in the slurry ■ Grant et al. (1990) reported success with a two-slurry

The size of an opening through which a slurry can flow is squeeze design for vugular zones. A lead slurry with a
a direct function of the size of the largest particles short pumping time and a fairly high API/ISO fluid-
(Chapter 6). For unconsolidated formations, there are loss rate (300 to 500 mL/30 min) is followed by a tail
some general guidelines. slurry with a longer pumping time and a lower fluid-
■ For a sand pack, the pore size is usually estimated to
loss rate. The tail slurry is pumped using the hesita-
be one-fourth the sand-grain size. The permeability of tion-squeeze technique.
the sand pack is proportional to the square of the ■ A fast-setting slurry followed by one with a longer

pore size. thickening time (Nowak et al., 1995) has been

■ Particles will flow through pores without bridging if
applied to repair a relatively large channel behind
they are smaller than about one-fifth the pore size. the casing.

With these two guidelines, one can estimate the

largest allowable particle size versus the permeability of 14-10.6 Thickening time and temperature
the porous medium. This leads to fluid-selection guide- As with primary cementing, the temperature and pres-
lines presented in Table 14-9. sure are important factors that influence the thicken-
For plugging gravel packs, organic compounds or syn- ing time of a cement slurry. For squeeze jobs, the
thetic cements based on epoxy chemistry have been used thickening time is designed to allow for slurry place-
(Chapter 7; Chan and Griffin, 1996; Ng and Adisa, 1997), ment, squeeze time, and well cleanout. The tempera-
as well as small-particle-size cements (Chapter 7; Ewert tures encountered during squeezing can be higher
et al., 1991; Faul et al., 1999; Keese et al., 2000; Fleming than those during primary jobs, because less fluid cir-
et al., 2000; Johannessen et al., 2000). culation and well cooling occur before the job. For this
Foamed cements have been used to plug fractures reason, special API testing guidelines (Appendix B;
(Creel and Crook, 1998) and squeeze high-permeability Cowan and Sabins, 1996) exist for squeeze cement-
zones. Chmilowski and Kondratoff (1990; 1992) dis- slurry design and should be followed to prevent pre-
cussed the selection of slurry type based on the results mature thickening.
of an injection test. For highly permeable reservoirs,
foamed cement provided the highest probability of suc-
cess. 14-10.7 Chemical resistance
Small casing leaks or low-rate vent flows have been Like plug cementing systems, squeeze cementing sys-
successfully plugged in one attempt using engineered tems often must be resistant to subsequent well stimu-
microcement (Chapter 7; Farkas et al., 1999; Slater et lation treatments. The requirement is usually expressed
al., 2001). in terms of solubility in HF-HCl acid systems. As dis-

Table 14-9. Fluid-Selection Guidelines According to Particle Size

Gravel Size Gravel Properties Possible Plugging Material
Mesh Number µm Permeability Pore Size Maximum Particle Size Fluid Type
(D) (µm) (µm)
10 2,000 234 500 100 Class G cement

60 250 4 62 13 Microcement

120 125 1 31 7 Solids-free treatment

534 Well Cementing

cussed in Section 14-5, latex cement slurries and epoxy- cases, less than a barrel would be sufficient. However,
base synthetic cements are especially resistant to these for job convenience, a 5- to 15-bbl batch is normally
conditions. Achieving low cement permeabilty by com- prepared. Another approach is to use two sacks of
bining small-particle-size cement and latex also provides cement per foot of perforated interval, with a mini-
enhanced acid resistance (Heathman et al., 1993). mum amount of 50 sacks per job (Smith, 1987).
■ A high-pressure squeeze, in which the formation is
14-10.8 Economics fractured, requires a higher volume of slurry. The
volume required is a function of the width and depth
Because they are workover operations, squeeze opera- of the fractures created. Rike and Rike (1981)
tions can be costly. As shown schematically in Fig. 14-37, reported that, during some running squeezes in
the various costs to consider include the loss of produc- which fracturing was excessive, volumes exceeding
tion, problem diagnosis, implementing the repair solu- 100 bbl of slurry were injected.
tion, and evaluating the success of the job. Therefore,
■ The slurry volume can be minimized by fracturing at
one must carefully justify such treatments economically.
The cement slurry usually represents less than 10% of a low pump rate and below the fracture-propagation
the total job cost. In addition, one should choose the pressure. If the squeeze is performed at high pres-
fluid system that offers the best chance of success, even sures and pump rates, the fractures will develop
if it is more costly. accordingly, resulting in large quantities of cement
being pumped in the formation.
Other constraints that limit the slurry volume include
Perforating Squeezing the following.
Tools ■ The hydrostatic and surface pressures must be con-
trolled during the job. A high cement column during
Diagnostics Evaluation displacement could cause the breakdown of low-pres-
sure or depleted formations. When large quantities of
Interrupted cement are necessary (natural fractures), low-density
revenue or other slurries should be used.
operation Rig
■ The volume should not exceed the capacity of the run-
in string.
■ The volume should not form a column that cannot be
reversed out.
■ When placing the slurry with coiled tubing, the
weight of the tubing filled with cement must be
within the tubing tensile-strength limit.
Fig. 14-37. Example of cost breakdown of cement squeeze operation
for a hypothetical case. Equations to calculate the fluid volumes for squeeze
cementing, as well as the pressures at various points in
wellbore during the treatment, are presented in
Appendix C.
14-11 Job design
14-11.1 Volume selection 14-11.2 Spacers and washes
The optimal slurry volume for squeeze cementing tends Spacers and washes have three main roles in squeeze
to be determined by best practices and common rules cementing.
rather than rigorous methodology. This is mainly caused
■ They prevent contamination of the cement slurry.
by the large number of unknowns, especially the hole
size. ■ They help clean voids and perforations that must be

■ For filling voids behind casing, the proper use and filled with cement.
interpretation of injection tests provides a good ini- ■ Often, the tubulars have been in place for a long time

tial estimate of the void volume and hence the volume and are covered with rust, debris, inorganic scales, or
of cement slurry to use. organic deposits. Spacers and washes help remove
■ Plugging perforations and small leaks with a low-pres- such materials and push them ahead of the cement
sure squeeze requires only enough slurry to build a slurry.
cement filtercake in each perforation tunnel. In many

Chapter 14 Remedial Cementing 535

 In squeeze cementing, cement-slurry contamination ■ The duration of a squeeze operation can be long, and
may be especially troublesome because the volumes are the cement-slurry temperature will have more time to
often small, and slurry properties such as thickening equilibrate with the formation temperature.
time, fluid-loss rate, and viscosity can be seriously
For a Bradenhead squeeze, the casing may also
affected. The final two objectives can be met by per-
expand if it is unsupported on the low-pressure side. The
forming a pretreatment step during which cleaning
following example provides estimates of the expected
fluids are pumped ahead of the main treatment
volume variations owing to fluid-pressure and tempera-
sequence. Devices such as perforation washing tools are
ture changes.
used with acids to clean perforations (Section 14-9.9).
Consider a 5-in. drillpipe, 5,000 ft [1,524 m] long,
Acids also weaken the formation structure and reduce
filled with a 11.7-lbm/gal [1,400-kg/m3] fluid. The com-
the breakdown pressure for high-pressure squeeze treat-
pressibility of the fluid is 4 × 10–10 Pa–1, and its coeffi-
ments. Acids, surfactants, and viscous fluids may be used
cient of thermal expansion is 4 × 10–4 K–1. Under its own
in combination with jetting devices to clean the inside of
weight, the fluid will compress by 0.43% (16 gal
the tubulars.
[0.061 m3]). This is the extra volume required to fill
Prevention of fluid mixing in the pipe is best achieved
the pipe in a vertical well. If the fluid is heated to the
by isolating the fluids with mechanical devices (balls,
formation temperature (assuming a geothermal gradi-
darts, etc.) rather than with spacers. However, spacers
ent of 2.4°C/100 m), the fluid will expand by 0.33%
are required to help displace the fluid in the hole by
(12 gal [0.046 m3]). To apply a surface pressure of
cement. The process of mud removal is similar to pri-
3,000 psi [21 MPa], 0.50% of the pipe volume must
mary cementing (Chapters 5 and 13), and the same gen-
be pumped (18 gal [0.070 m3]). Overall, the uncertainty
eral rules concerning mud conditioning, density, and
in fluid volume is around 1%, which translates to 49 ft
rheology hierarchy still apply. However, the conditions
[15 m] in depth.
are different in a few respects.
■ The pipe tip is seldom centralized. It lies on the hole
wall, and a small amount of mud is likely to remain 14-11.4 Pressure and mechanical strength
trapped along this contact line. checks
■ To limit the amount of trapped mud, the pipe should Three basic safety limits must be monitored at all times
be rotated during placement (not reciprocated). during the various stages of the operation, including
■ When placing the slurry with coiled tubing, the gen- cement placement, squeeze, and cleanout.
eral rule is to pull the tubing at the same speed as the ■ For low-pressure squeezes, the dynamic fluid pressure

cement velocity in the annulus, keeping the coiled in front of the formation must be less than the for-
tubing tip a few tens of feet below the upper cement mation fracturing pressure. This constraint may be
interface. exceeded at the end of a successful squeeze operation.
■ The static fluid pressure in front of the formation
must be higher than the pore pressure to prevent any
14-11.3 Displacement volume for a squeeze formation fluid entry.
operation ■ The differential pressure across the various tubulars
The maximum displacement volume for a cement must be less than their burst or collapse pressures.
squeeze operation corresponds to the volume from the When working with coiled tubing, the weight of
surface down to the top perforations to be squeezed. A tubing filled with cement must remain within the
safety margin is added, usually a few barrels. Like bal- tubing tensile-strength limits. One must also account
anced plugs, the maximum displacement volume for for coiled tubing fatigue.
squeeze operations may be uncertain owing to factors
such as pump efficiency, variability of the internal pipe The differential pressure across tubulars must allow
volume, and fluid compressibility. Fluid compressibility for unexpected occurrences. For example, when using a
is particularly important for the following reasons. cement retainer for a squeeze operation, the fluid may
■ The pressure attained during a squeeze operation is
flow upward into the annulus and pressurize the outside
of the casing above the retainer. To prevent casing col-
much greater than that attained during primary
lapse in these situations, it is a common practice to pres-
cementing, so fluid compressibility plays a larger role.
surize the pipe-casing annulus (Fig. 14-38).
When the squeeze is performed in an openhole sec-
After the squeeze treatment, the pressures exerted
tion, the formation has a tendency to expand, leading
during reverse circulation must be considered. If the for-
to a hole-volume increase.
mation is weak, reverse circulation is usually not recom-
mended.

536 Well Cementing

 1,000 psi squeeze pressure 14-12.1 Determination of injection pressure
Atmospheric pressure and rate
on the annulus
The goal of the injectivity test is to collect data about the
expected pressure to be applied during a low-pressure
squeeze treatment. Depending on the measured pres-
sure, a perforation wash treatment may be applied
before the squeeze job to increase injectivity or the
squeeze-fluid composition may be altered.
The test consists of injecting a fluid such as water
1,000 psi differential through the perforations for several minutes at a rate
pressure across casing
close to that used in a squeeze treatment, until a pseu-
dosteady-state regime is reached. Care is taken to avoid
reaching the formation fracturing pressure (Chmilowski
Channel behind casing and Kondratoff, 1992). The most important factors to
consider are the following.
■ The friction pressure developed as the fluid flows
through the pipe should be calculated and subtracted
from the pump pressure. If possible, monitor the
Fig. 14-38. Illustration of the risk of casing collapse above the annular pressure.
retainer if the annulus is not pressurized. ■ To prevent perforation plugging, the injected fluids
must be clean, and the wellbore should be circulated
before injection.
■ The injection should proceed until the pressure sta-
14-12 Injection test
bilizes, typically 10–15 min.
Before mixing and pumping the cement slurry, an injec-
■ The injection rate should be maintained below the
tion test is performed. This procedure consists of pump-
ing a fluid, typically water, into the well. The injection fracturing pressure.
test is performed for several reasons: ■ Operational procedures should be consistent

■ to ensure that the perforations are open and ready to between different wells in the same field to obtain
accept fluids (for small leaks, an injection test helps meaningful comparisons.
determine whether it will be possible to inject a fluid, In general, as the injectivity decreases, the impor-
and the expected treatment pressure and rate) tance of the fluid design increases.
■ to obtain an estimate of the proper cement-slurry ■ Grant et al. (1990) found that a single slurry with
injection rate good fluid-loss control is appropriate for tight forma-
■ to estimate the pressure at which the squeeze job will tions. For loose formations, a lead slurry with a high
be performed fluid-loss rate (300–500 mL/30 min) is recommended,
■ to estimate the volume of slurry to be used (for a flow followed by a tail slurry with a low fluid-loss rate.
behind casing, an injection test helps determine the ■ Chmilowski and Kondratoff (1992) recommend a

volume of voids behind casing and the necessary Class G slurry with a low fluid-loss rate for formations
treatment-fluid volume). with low injectivity, Class G and microspheres for
intermediate injectivities, and foamed cement for
Should the fluid fail to achieve injection, acid is often high injectivities.
injected under matrix conditions. Hydrochloric and
hydrofluoric acids are commonly used.
For a water production problem, an injection test may 14-12.2 Determination of void volume behind
be an additional aid to determine whether water is flow- casing
ing through fractures or the rock pores. This test will
also aid the selection of the appropriate chemical system For flow behind casing, an injection test helps deter-
for the treatment. mine the volume of voids behind the casing and the ulti-
Grant et al. (1990) and Chmilowski and Kondratoff mate treatment-fluid volume. The concept is to pump a
(1992) presented some basic guidelines for the design, viscous fluid that creates a measurable injection-pres-
execution, and interpretation of injectivity tests. They sure increase when it reaches the formation face. The
are detailed below. void volume is approximated by the total volume
pumped when the pressure increase occurs.

Chapter 14 Remedial Cementing 537

 To perform a useful injection test, the following fac- Alternative methods are available in specific cases.
tors must be considered. When a viscous fluid is used to displace a less viscous
■ A fluid with a viscosity between 50 and 200 cp should fluid in the formation, the equations can be simplified
be used. The tighter the formation, the lower the vis- into the following form.
cosity should be. The viscosity should be measured at 4 π × k form × h form
q 1
the anticipated treatment temperature. = ×
■ The fluid should be clean to facilitate flow through pwb − pres μ ⎛V ⎞
log ⎜ cum + 1⎟
the formation. Fully hydrated biozan, xanthan, or ⎝ V ⎠
polyacrylamide fluids are suitable.
(14-14)
■ If a preliminary estimate of the void volume is
unavailable, the injection-fluid volume should be 2 to The injectivity index,
3 times the annular volume between the casing and q
the formation, using the height estimated from ,
cement logs or lithology logs. pwb − pres
■ The fluid must be pumped at a constant flow rate, and
is plotted versus an expression of the volume pumped.
the pressure must be closely monitored. Ideally, a pres is the reservoir pressure and V the wellbore volume
downhole pressure sensor should be used, especially times the porosity. Similar derivations can be made for a
if coiled tubing is involved. All acquisition systems variety of flow geometries (e.g., fractures), which allows
should be synchronized. one to select the conditions that best fit the measured
■ After the test is complete, the injection fluid should be data.
cleaned out of the well, either by producing it or break-
ing it. This prevents problems caused by incompatibil-
ity between the polymer and the treatment fluid. 14-13 Basic squeeze-job procedures
The general sequence of events during a squeeze job is
14-12.3 Determining the nature of flow paths detailed below.
behind casing 1. Zones below the interval to be squeezed are isolated
For a water production problem, flow tests have been with a retrievable or drillable bridge plug.
used to determine the type of flow in the formation. The 2. The perforations are washed with a perforation wash-
results help determine the appropriate chemical treat- ing tool or are reopened with the back-surging tech-
ment. A traditional means of interpreting pressure- nique (Sections 14-9.9 and 14-9.10).
volume data from water injection wells is the Hall plot 3. The perforation washing tool is retrieved and, if the
(Buell et al., 1987). This analysis applies to situations in packer method is chosen, it is run in the hole with the
which a fluid is injected radially in a porous medium, workstring, set at the desired depth and tested. A test
and Darcy’s law applies. The integral of the pressure is pressure of 1,000 psi [6.9 MPa] is usually sufficient. If
plotted versus the cumulative volume pumped, Vcum. The the cement is to be spotted in front of the perfora-
slope of the curve, α, provides information about the for- tions, a tailpipe covering the length of the zone plus
mation according to the following equation. 10 or 15 ft [3 or 5 m] is run below the packer.
4. An injection test is performed using clean, solids-free
∫ Δpdt = αVcum (14-13) water or brine. If a low-fluid-loss completion fluid is
with in the hole, it must be displaced from the perforations
before starting the injection test. This test gives infor-
⎛ r ⎞ mation concerning the permeability of the formation
μ ⎜ ln ext + s⎟
⎝ rwb ⎠ to the filtrate.
α= .
2 πk form h form 5. The spearhead fluid followed by the cement slurry is
circulated downhole with the packer bypass open.
In this equation, rext is the external drainage radius, This circulation is performed to avoid squeezing dam-
rwb the wellbore radius, s the skin factor, kform the for- aging fluids ahead of the slurry into the formation. A
mation permeability, and hform the formation thickness. small amount of backpressure is applied on the annu-
lus to prevent slurry free-fall as a result of the U-tube
effect.

538 Well Cementing

 If no tailpipe has been run, the packer bypass must 7. Pumping continues until no pressure leakoff occurs.
be closed 2 to 3 bbl before the slurry reaches the A further pressure test of about 500 psi [3.5 MPa]
packer. If the cement is to be spotted in front of the more than the final injection pressure indicates the
perforations with the packer unset, circulation is end of the injection process. Usually, a well-cemented
stopped as soon as the cement covers the selected perforation accepts a pressure above the formation
zone. The tailpipe is pulled out of the cement slurry, fracturing pressure, but fracturing may occur if one
and the packer is set at the desired depth. attempts to verify such a condition.
The depth at which the packer is set must be care-
fully chosen. If a tailpipe is run, the minimum dis-
tance between the perforations and the packer is lim-
Casing
ited to the length of the tailpipe. The packer must not
be set too close to the perforations, because pressure
communication through the annulus above the Packer
packer may cause casing collapse. A safe setting
depth must be selected after evaluation of the quality Cement slurry
of the cement bond with the logs (Fig. 14-39).
Possible contamination of the cement slurry by the
fluid in the hole limits the maximum spacing between
the packer and the treated zone. In Fig. 14-40, the
packer is set too high, allowing cement slurry to be
contaminated as it channels through the mud to Cement
reach the perforations. Shryock and Slagle (1968) channeled
recommended that the retrievable packer be set no through mud
more than 25 ft [8 m] above the perforations. Mud
6. Squeeze pressure is applied. If the high-pressure
squeeze technique is used, the formation is broken
Fig. 14-40. Cement-slurry contamination (from Shryock and Slagle,
down, and the cement slurry is pumped into the frac- 1968; reprinted with permission of SPE).
ture. For low-pressure squeeze jobs, hesitation pump-
ing begins as soon as the packer is set.

Running tail pipe Spotting cement Squeezing all perforations

to below the perforations after setting tail pipe and
packer above the cement

Fig. 14-39. Squeeze job with a retrievable packer and tailpipe.

Chapter 14 Remedial Cementing 539

8. Pressure is bled off and returns are checked. The 14-15 Cleanup after a squeeze job
packer bypass is opened and excess cement is When perforations and casing leaks are squeezed, the
reversed out. Washing off the cement slurry in front of excess cement remaining inside the casing may be
the perforations can be performed by releasing the cleaned up before it sets or drilled out after it sets.
packer and slowly lowering the workstring during the Removing the excess cement before setting provides a
reversing; however, there is a danger of disturbing the clear benefit in terms of cost and time; however, this
unset cement filtercake. procedure presents the risk of destroying the cement
9. Tools are retrieved, and the well is left undisturbed to nodes. In all cases, a positive pressure should be main-
allow the slurry to cure for the recommended WOC tained until the cement is fully set.
time. Excess cement in Bradenhead squeezes is easily
When preparing the slurry, the use of a recirculating cleaned up, because no hardware prevents pipe move-
or batch mixer is strongly recommended, because it ment. To accelerate the setting of the cement nodes,
ensures that the properties of the slurry pumped in the Krause and Reem (1992) described a method of spotting
well are close to those designed in the laboratory. On an accelerator (triethanolamine) in front of the
most squeeze jobs, the cement-slurry volume is small, squeezed section after the cement is cleaned out. When
but the quality requirements are high; therefore, special a retrievable cement retainer is used in a circulation
care during slurry preparation is justified. squeeze, its upper surface should be cleaned from any
cement that might have fallen on top of it.
Removing excess cement when the treatment is per-
formed with coiled tubing is more common. Different
14-14 Monitoring squeeze jobs techniques are possible.
The typical parameters that should be monitored during ■ Reverse circulate cement slurry
the job include the following.
■ Contaminate excess cement and reverse it out
■ Density and volume of fluids in various tanks
■ Contaminate excess cement and circulate it out
■ Pressure, density, flow rate, and volume at pump
directly
outlet (Fig. 14-41)
■ Contaminate excess cement and leave it in the hole
■ Volume and density of all fluids returned and pumped
through the annulus The relative advantages and drawbacks of direct and
■ Coiled tubing pressure, string weight, depth, tubing reverse circulation (Section 14-4.3) are exacerbated
OD, and tubing cycles because of the increased pressure drop and lower rates
with coiled tubing.
During the job, these data are routinely analyzed in During the contamination step, the coiled tubing is
real time to determine the subsequent steps. The opera- moved into the excess cement at a velocity adjusted to
tor is limited to a few options: continue squeezing, ensure a proper volume ratio between the cement and
increase squeeze pressure, or stop squeezing and start the contaminating fluid, typically 50% (see Fig. 14-36).
the cleanup operation. Using a diverter or jetting tool at the tip of the tubing

5 Squeezing slurry

4 Injectivity test
Pumping slurry
Pressure 3
(1,000 psi)
2

0
Time
Fig. 14-41. Typical pressure plot during a squeeze treatment.

540 Well Cementing

enhances mixing (Walker et al., 1992). Cleanup is often
performed in more than one stage. The bulk of the Test valve
cement is cleaned out first. Then, another run is per-
formed to provide a final cleaning. When the jet passes
downhole hardware, the pump rate is increased to Packer set
ensure the removal of cement from all surfaces.

Cement-filled
perforations
14-16 Squeeze job evaluation
The extent to which one must evaluate the results of a
squeeze job depends on the requirements of the subse-
quent operations to be performed on the well. As a pre-
liminary step before an evaluation, the state of the well- Fig. 14-42. Dry test.
bore is checked to detect the presence of cement nodes
that may restrict passage of downhole tools. An under-
reaming operation is eventually performed. Also, the (a)
H D A
rathole is checked for the presence of cement.
G

14-16.1 Temperature log

When a high-pressure squeeze job is performed to Pressure
F
ensure that a subsequent fracturing treatment will stim- C
E B
ulate a target zone, a temperature log can locate the
cement and indicate if any slurry was injected outside of Baseline
the perforated interval (Krause and Reem, 1992).
Goolsby (1969) evaluated squeeze results on water Typical DST chart of a perforated interval before cement squeezing
injection wells by comparing pre- and postsqueeze tem- A Initial hydrostatic pressure with packer set
perature profiles. By logging the well temperature after B Initial flowing pressure when (water cushion used)
a postsqueeze injection test, he was able to demonstrate when test valve tool opened
that the well accepted the injection water at the B-C Flow period
planned location. C Test valve closed
C-D Pressure buildup curve
D Initial shut-in pressure
E Flowing pressure after second opening of test valve tool
14-16.2 Positive and negative pressure tests E Flow period
Plugged perforations are evaluated by performing a pos- F Test valve closed
itive or negative pressure test. A positive pressure test G Second (final) shut-in pressure
subjects the well to a given pressure and determines H Final hydrostatic pressure after unsetting packer
whether fluid can be injected into the formation. A neg-
(b) E A
ative pressure test, also known as an inflow test, subjects
the well to a pressure reduction and is used to determine
how effectively the plugged perforations prevent the
ingress of fluids from the formation. The test pressure is
usually that which the well is expected to experience Pressure
during injection or production; it is determined during D C B
the job-design phase (Walker et al., 1992).
The negative pressure test is performed as follows: Baseline
■ placing a light brine across the perforations
DST chart of the same interval after a successful squeeze job
■ swabbing the well

■ running a dry test (Fig. 14-42) (Chapter 15).

A Initial hydrostatic pressure with packer set
B Flowing pressure when test valve tool opened
If the perforations are completely sealed, no inflow C Pressure at the end of flow period, test valve tool closed
D Pressure at the end of shut-in period
should be recorded on the test pressure chart E Final hydrostatic pressure after unsetting packer
(Fig. 14-43).
Fig. 14-43. Pressure chart from negative pressure drillstem test
(DST).

Chapter 14 Remedial Cementing 541

 If the pressure reduction is excessive, the resulting 14-17.1 Improper slurry selection
mechanical stresses on both the cement and casing may The cement slurry does not penetrate the pores of the
cause the nodes to fail. rock. Only the mix-water and dissolved substances pene-
trate the pores, while the solids accumulate at the for-
14-16.3 Acoustic logs mation face and form the filtercake. It would require a
permeability higher than 100 D for solids from a conven-
When the objective of the squeeze is to repair a primary
tional Class G cement to penetrate a sandstone matrix.
cementing job, cement logs should be run to evaluate
Even microfine cements have limited penetration, if any,
the effectiveness of the repair by comparing presqueeze
through porous media. The only way for a slurry to pene-
and postsqueeze logs (Chapter 15).
trate a formation is through fissures, fractures, and large
holes (vugs).
14-16.4 Production changes
One of the most common ways to evaluate a remedial 14-17.2 Excessive final squeeze pressure
treatment is to compare the well production rates before
A high final squeeze pressure does not increase the
and after the treatment (Keese et al., 2000).
chance of success; on the contrary, it increases the
chance of fracturing the formation and losing control of
14-16.5 Cement hardness the cement-slurry placement. Once created, a fracture
If well cleanup was not performed after the squeeze, the may extend across various zones, and open unwanted
cement in the wellbore is drilled out. For such jobs, channels of communication between previously isolated
Suman and Ellis (1977) reported that a good indication zones. It is important that a “think downhole” attitude
of success is the nature of the cuttings. If the cement is be developed among all personnel involved in this oper-
hard throughout, the results are usually good. However, ation.
soft spots or voids are not necessarily indicative of a fail-
ure. 14-17.3 Plugged perforations
Another common misconception concerning squeeze
14-16.6 Radioactive tracers cementing is that all perforation holes are open and
Radioactive materials may be added to the cement receptive to fluids (Rike and Rike, 1981). Such an
slurry, and subsequent tracer surveys can indicate assumption can lead to failure. The mud filtercake,
whether the cement is placed in the desired interval. which is capable of withstanding a large differential
The isotopes 131I, 192Ir, and 46Sc are appropriate because pressure when applied from the wellbore toward the for-
of their short half-lives—8 days, 75 days, and 85 days, mation, cleans up easily when submitted to a differential
respectively. The iridium and scandium radioisotopes pressure in the other direction. In addition to mudcake,
are preferable, because iodine (present as iodide) is sol- debris, scale, paraffin, formation sand, pipe dope, rust,
uble and may be squeezed out of the cement with the fil- and paint can accumulate and plug the perforations.
trate. Goodwin (1984) reported that, in a producing well, the
upper perforations are usually open, while the plugged
perforations are generally found in the lower zones.
Squeezing under such conditions will not fill all of the
14-17 Reasons for squeeze cementing perforations with cement. Following the treatment, the
failures perforations not plugged with cement will allow entry of
Whenever a squeeze treatment fails to meet its objec- formation fluids into the well.
tives, a thorough investigation must be conducted to Perforation washing before the squeeze job helps to
analyze the job, understand why a failure occurred, and render all perforations receptive to the squeeze cement
improve the design of subsequent treatments. Squeeze slurry.
failures may originate from misconceptions of what a
squeeze treatment is and what actually happens down-
hole.

542 Well Cementing

14-17.4 Improper packer location Other types of treatments such as relative perme-
If the packer is set too high above the perforations, the ability modifiers are beyond the scope of this discussion.
cement slurry will become contaminated as it channels
through the mud or completion fluid (Section 14-13).
Slurry properties such as fluid loss, thickening time, and 14-19 Conclusions
viscosity are adversely affected by contamination, and Remedial cementing operations are often performed on
slurry placement results are altered. short notice, leaving little time for planning and design.
Consequently, the success rate is frequently low. To mit-
igate this problem, general best-practice guidelines
14-18 Noncementing solutions based on long experience have been offered by various
As discussed in the introduction, there are solutions to authors (Rike, 1973; Smith et al., 1983; Toor, 1983;
many of the well problems described in this chapter Goodwin, 1984; Heathman et al., 1994; Lopez and
that do not involve cementitious systems. Although they Renshaw, 1998; Bimaputra et al., 2000). The guidelines
are somewhat beyond the scope of this chapter, they and recommendations presented in the chapter are of
deserve to be mentioned briefly. Noncementing solu- little use unless all involved agree that a detailed design
tions include mechanical tools and noncementitious is required, and there is good communication to ensure
fluids. Such solutions can be classified according to that the design is executed properly.
their penetration into the wellbore: none, limited, and For plug cementing, implementation of the simple
deep. techniques and guidelines discussed in this chapter
Solutions at the wellbore (no penetration) include improves the probability of success.
bridge plugs or mechanical plugback tools (Chapter 11). Successful squeeze cementing starts at the job-design
These devices have long been used as an alternative to stage. The questions in Table 14-10 must be answered
cement plugs for isolating bottom perforations in cased before executing the operation.
hole producers. One cannot be confident of job success without satis-
Limited penetration techniques include the follow- factory answers to these questions.
ing. A set of perforations can be plugged without plug- There is a general consensus among the authors cited
ging the entire hole by using a scab liner—a piece of in this chapter that the most successful method is the
liner that is cemented off-bottom (Loveland and Bond, low-pressure hesitation squeeze, with a low-fluid-loss
1996; Chapter 11). Daigle et al. (2000) presented case slurry and a packer or retainer as the isolation tool. The
studies of expandable liners used to repair corroded high-pressure techniques should be applied with
casing. Saltel et al. (1999) described the use of an inflat- extreme caution.
able composite sleeve to plug perforations or casing A short summary of remedial-cementing guidelines is
leaks. The same method can be used to plug small sec- presented in Table 14-11, together with additional issues
tions of open hole. Injecting high-viscosity polymers that are application dependent. These guidelines may
(Irani, 1999), polymerizable resins, or other types of also be presented in the form of a decision tree
sealants (Rusch et al., 1999) can plug gas-migration (Fig. 14-44).
channels behind the casing. Carpenter et al. (2001)
described test results of placing a metal alloy in the
annulus and fusing it to the casing. The excellent metal- 14-20 Acronym list
to-metal bond stopped gas flow. BOP Blowout preventer
Deep-penetrating polymer fluids are used to control
water or gas production by plugging fissures and frac- GOR Gas/oil ratio
tures (Eaton et al., 2000). For such applications, low- IBOP Internal, or inside, blowout preventer
viscosity solids-free systems are often used (Chan, 1989; LCM Lost-circulation materials
Shelley and Scuillo, 1995; Creel and Crook, 1997). A WOR Water/oil ratio
common example is a polyacrylamide solution contain-
ing crosslinkers such as Cr 3+ (Lai et al., 1999) or acids
(Chatterji and Borchardt, 1980). The crosslinking reac-
tion is timed to occur after fluid placement. For such
treatments, the fluid volume is based on the expected
penetration, assuming a radial displacement and know-
ing the formation porosity and thickness from cores or
logs.

Chapter 14 Remedial Cementing 543

Table 14-10. Questions Before a Squeeze-Cementing Job
Category Questions
Problem identification What is the problem?
What are the objectives of the job?

Selection of remedial method Which squeeze technique will be used?

Selection of placement method and mechanical setup Which types of tools are to be used?
At which depth should the packer be set?
To which depth should the tailpipe be lowered?
Which well preparation technique(s) are needed?

Selection of fluid type and volume Which type of fluid is in the hole?

Detailed job design What will be the maximum job pressure?

Which job procedure will be followed?
What is the estimated job time?
What are the composition and properties of the slurry?
What quantity of slurry is necessary?
How will the perforations be opened or cleaned?
What is the formation breakdown pressure?
Have other formation data (lithology, permeability, pressure, water/oil and
gas/oil contact, bottomhole temperature) been considered?
What is the WOC time?

Job evaluation How will the job be evaluated?

544 Well Cementing

Table 14-11. Summary of Best-Practice Guidelines for Remedial Cementing
General guideline for all remedial jobs Check slurry properties versus acceptance criteria: density, fluid
loss, filtercake thickness, rheology, and thickening time.
Define job objectives
Circulate the hole sufficiently. Use a low-yield-point mud with low
Determine root causes of the problem. plastic viscosity. Mud density must be sufficient to control the
Define job type and type of slurry. well.
Batch mix the cement slurry.
Collect well data
Caliper for cement placement depth and slurry volume. Prefer Pump the fluid with mechanical wiper plugs between each fluid,
placing plugs in gauged and competent locations. If no caliper especially for large depths.
is available, volume will be increased. Monitor volumes from tank level.
Determine fracture pressure, especially for injector wells for Pull out of hole very slowly, without rotating the pipe, to prevent
swabbing.
which it may be low.
If a thixotropic slurry is pumped, clean the excess before it gels.
Clean the well For a coiled-tubing job, overdisplace the slurry.
Condition mud for optimal mud removal during cement Contaminate excess cement with biopolymer or borax solution.
placement. Reverse circulate the contaminated slurry only if the formation
Remove all debris, in particular from sump hole/rathole. will withstand the pressure. Do not disturb the cement plug
during this phase.
Design job Clean the drillpipe with a wiper ball after the cement.
Set plug length between 300 and 800 ft, no less, no more.
Allow a sufficient WOC time.
Determine the base for plug (if placed off-bottom)
Prefer solid base. Evaluation
Compare results with job objectives.
Consider filling entire hole with cement, several staged plugs.
Use viscous pill or reactive viscous pill with density halfway Archive this information because it will be used for subsequent
between mud and cement, yield point of 70 lbf/100 ft2 minimum, jobs of same type.
and length of 300 ft. Additional guidelines specific to kickoff and abandonment
If no mechanical base is possible, decrease slurry density cement plugs
while maintaining good mechanical strength. Use high-yield-strength or thixotropic slurries.
If slumping is anticipated, increase plug length. Place the top of a kickoff plug in a softer formation.
Use mud-removal rules for mud displacement. If the kickoff plug is unsuccessful, do not drill through it. First,
wait a few more hours and make another attempt to sidetrack.
Low-density turbulent wash should be preferred. If this is not
possible, use a laminar spacer with density and viscosity If the second attempt is unsuccessful, set another plug using
halfway between the mud and the cement. what remains of the first plug as a solid base.
A densified cement will displace the mud more efficiently. Additional and specific guidelines for squeeze operations
Warning: A good spacer has a tendency to remove the mud-
cake, increasing the risk of differential sticking. Maintaining Problem identification
pipe rotation minimizes this risk. Estimate the most probable situations, evaluate the response
Slurry thickening time: Define temperature based on API squeeze of each situation to various treatments, and select the treat-
cementing tables. Determine thickening time and compressive ment that minimizes risks of failure.
strength at bottomhole circulating temperature. Thickening time Pretreatment
is based on placement time + 30 min to 1 hr (only for cement
Clean the perforations.
plugs).
Use ample cement volume. Perform an injectivity test. If results are less than 1 bbl/min at
2,000 psi, inject acid.
Schedule the disposal of returned fluids and contaminated
cement. If acids or solvent are used to increase injectivity, filter these
Prepare expected pressures vs. volume pumped diagram for reactive fluids.
real-time evaluation of job, especially for squeeze treatments.
Execution
Execution Kill the well before squeezing.
Use a tailpipe with length 1.5 times the plug length (not for lost Check potential for casing collapse when there is a packer
circulation plugs) to minimize the amount of mud remaining.
above the squeezed perforations, a squeeze packer, cement
Use scratchers and centralizers on the tailpipe when the hole is retainer, or production packer.
not excessively washed out.
Monitor and interpret real-time data.
Use a diverter and rotate pipe during cement placement to help
displace the mud. During a hesitation squeeze, if the pressure drops during a
While cleaning the hole, use a sacrificial wiper dart to clean the stage, another hesitation sequence should be performed.
drillpipe before the job. This also allows determination of the pipe (continued on next page)
volume for accurate displacement volume.

Chapter 14 Remedial Cementing 545

Table 14-11. Summary of Best-Practice Guidelines for Remedial Cementing (continued)
Additional specific guidelines for jobs done with coiled tubing Wait for 15 minutes and pump 1⁄6 volume—again while
Monitor internal tubing volume. Use a graduated tank to monitoring for pressure increase.
monitor the volume of fluid pumped.
Wait 15 min and apply pressure.
Perform a depth correlation to be certain of the coiled-tubing-
If no buildup in pressure is discovered, run in two stands (or as
tip location.
needed) below previous calculated cement top and reverse
Monitor the pressure drop through the coiled tubing by out excess cement.
pumping down the coiled tubing and up the annulus.
Place another balanced plug and repeat Bradenhead
Calculate the maximum pressure for preventing tubing burst procedure.
and collapse, especially when using fatigued tubing. For
Use slurry design with API/ISO fluid-loss rate <100 mL /30 min.
collapse, the maximum risk is during reverse circulation, close
to the surface. If planning to tack and squeeze an 113⁄4-in. liner, make sure
there is 500 ft of liner overlap. Weld some centralizing ribs.
Be aware of cement-slurry shear sensitivity. Check the
thickening time accordingly. Non-Bradenhead cement squeeze
Move the coiled tubing during all phases to prevent sticking. Pump adequate wash ahead to remove the mudcake.

Bradenhead cement squeeze Mix sufficient cement volume to place the entire volume into
the formation on the first squeeze. Use excess.
Pump 300 ft (up to 30 bbl) of spacer/wash ahead of cement.
Pump spacer/wash behind at volume calculated to balance. Stop and wait 15 min—pump the remaining volume in the
tubing and/or drillpipe while monitoring for pressure increase.
When possible, use a cement stinger to place the balanced
plug. If no pressure increase is discovered, pump the remaining
cement out of the tubing and reverse out to make sure there is
Place at least 300 ft of cement (500 ft is preferred).
no cement behind the workstring.
Pull two to three stands above plug and reverse out excess.
Repeat the squeeze.
Pump away one-third of the cement volume while monitoring
Use slurry design with API/ISO fluid-loss rate <100 mL/30 min.
for pressure increase as cement feeds into the formation.
If planning to tack and squeeze an 113⁄4-in. liner, make sure
Wait for 15 min and then pump 1⁄6 of volume—while monitoring
there is 500 ft of liner overlap. Weld some centralizing ribs on
for pressure increase.
the upper two joints of 113⁄4-in. casing to improve the standoff.

546 Well Cementing

 Leak identified Injector fails
in producer test

Engineer reviews well history and

status, age of casing, previous
squeezes, allocated oil production,
remaining reserves, strategic area
plan, and injection withdrawal ratio
for area. Consult team leader.

Strategic
Run casing scraper Yes reason to keep wellbore: No Plug and
and locate leak economic, technical, other abandon
with packer horizons, etc.?

No

Engineer analyzes Repair

repair options and justified?
runs economics

Yes
Go to wireline
diagnostics

Go to low Yes Squeeze?

pressures

Rerun cement log

No
No

Casing Yes Test Complete

repaired successful? repair

Fig. 14-44. Remedial cementing flowchart.

Chapter 14 Remedial Cementing 547

15-1 Introduction
Cement job evaluation is the process to determine
whether the objectives have been reached after the
Cement Job Evaluation
Mickael Allouche, Dominique Guillot, Andrew J. Hayman,
Robert J. Butsch, and Charles W. Morris—Schlumberger

15-2 Hydraulic testing

15
Hydraulic testing is a method to evaluate the isolation
provided by the cement. This can be performed after pri-
cement job has been performed. No cement-job evalua- mary cement jobs (e.g., when water zones are located
tion will be efficient if the objectives are not clear. While near the oil or gas zone). It can also be performed after
the fundamental objective is to support the casing, other remedial cementing to test if the perforations have been
objectives vary with the nature of a particular cement effectively sealed.
job. The most common techniques are pressure testing
The purpose of a primary cement sheath varies and dry testing. In some areas, the cement quality is ver-
according to the casing string it supports. For a conduc- ified through a production test or communication test-
tor casing, the main purpose of the cement job is to pre- ing through perforations.
vent formation erosion by stopping the circulation of
drilling fluids outside the casing. Surface casings must
be cemented to seal and protect water formations and 15-2.1 Pressure testing
help support deeper casing strings. Intermediate strings Pressure testing is the most common hydraulic testing
are cemented to seal abnormally pressured formations, method. It is generally performed after every surface- or
isolate incompetent formations, and shut off lost-circu- intermediate-casing cement job. A casing pressure test
lation zones. Production strings are cemented to prevent is performed first to verify the mechanical integrity of
the migration of fluids in the annulus and to ensure the casing. To avoid damaging the set cement or the
zonal isolation. Cement also provides some corrosion cement/casing bond, this should be performed soon
protection to all of the casing strings. after the top plug bumps.
For remedial cementing, the principal objectives are After the cement sets, the casing shoe is drilled out,
to improve the quality of a primary cement job, seal per- and the internal casing pressure is increased until the
forations, repair casing leaks, and isolate productive pressure at the casing shoe exceeds the pressure that
layers. will be applied at this point during the next drilling
Before the development of acoustic cement logs, phase. A casing shoe that does not hold pressure indi-
cement-job evaluation consisted of testing the hydraulic cates a poor cement job, and remedial cementing is
isolation or locating the top of the cement. A hydraulic- required. When the pressure is raised until the forma-
isolation test often requires casing perforation or addi- tion breaks down, the test is called a pressure-integrity
tional drilling. In cases in which pipe support is the only test (PIT) or leakoff test (Fig. 15-1). The objective of the
requirement, knowing the location of the top of the PIT is to determine the maximum mud weight that can
cement may be sufficient. However, to verify interzonal be used for drilling the subsequent section. Leakoff tests
isolation, more sophisticated evaluation methods must are most frequently performed on exploration wells.
be used. Procedures for conducting and interpreting a PIT are
This chapter presents the techniques presently avail- described by Postler (1997). The guidelines are summa-
able to evaluate cement jobs: hydraulic testing rized here.
(Section 15-2); nondestructive methods such as temper- 1. Prepare the test: Check valves for leaks, prepare
ature, nuclear, and noise logging (Section 15-3); and clean mud, and use a good-quality pressure gauge.
acoustic, sonic, and ultrasonic cement logging 2. Perform a casing test.
(Section 15-4). The importance of cement-job monitor-
ing is also discussed in Section 15-4.4.2.7. 3. Prepare a PIT graph.
4. Pump mud at steady rate, between 0.25 and 1 bbl/
min, and plot measured pressure data on the graph.

Well Cementing ■ Chapter 15 Cement Job Evaluation 549

 a: Pressure versus volume pumped b: Job preparation following the casing test
and versus time during falloff 2,000
Max. allowable pressure
Max. test pressure
Min. volume line
Predicted leakoff (casing test) Facture
1,500 closure
Leakoff pressure pressure
Max. volume line
Predicted 1⁄ 2 lbm/gal
Pressure Min. formation stress
Pressure (psi) 1,000

Shut-in time
(min.)

500

Min. volume 2 × min. volume 0

0 1 2 3 4 5
Volume/time
Volume (bbl)
c: Interpretation of a defective PIT as a channel behind casing
Initial Test
Operator Well name and no. Pit no. Date Casing size Shoe depth Total ver. depth
F 1 1980 133⁄8 in. 5,686 ft 5,686 ft
Mud weight Test rate TD of test Mud pumped Mud flowback Pred. frac. grade
13.0 1⁄2 5,734 ft 71⁄2 0 17.5
Repeat Test After Squeeze
Operator Well name and no. Pit no. Date Casing size Shoe depth Total ver. depth Operator Well name and no. Pit no. Date Casing size Shoe depth Total ver. depth
F 2 1980 133⁄8 in. 5,686 ft 5,686 ft F 3 1980 133⁄8 in. 5,686 ft 5,686 ft
Mud weight Test rate TD of Test Mud pumped Mud flowback Pred. frac. grade Mud weight Test rate TD of test Mud pumped Mud flowback Pred. frac. grade
3⁄4 13.0 1⁄2 5,744 ft 41⁄2 17.5
13.0 5,734 ft 93⁄4 0 17.5
17.5 lbm/gal predicted
13 13
Leakoff
12 at shoe 12 17.5 lbm/gal predicted
Repeat test Stop pump
11 11
10 10
Leakoff Leakoff
9 behind
9
pipe Stop pump
8 8
Pressure Pressure
7 7
(100 psi) (100 psi)
6 6
Initial test
5 5
4 Repeat test 4 0 2 4 6 8 10
0 2 4 6 8 10 Shut-in time (min.)
3 Initial test 3
0 2 4 6 8 10
2 2
Shut-in time (min.)
1 1
0 0
0 2 4 6 8 10 12 14 0 2 4 6 8 10 12 14
Volume (bbl) Volume (bbl)

Fig. 15-1. PIT (from Postler, 1997; reprinted with permission of SPE).

5. Use the maximum volume line to determine whether ■ Shut-in pressure does not level off.
a higher pump rate is needed.
The channel is confirmed after a repeat PIT shows no
6. When the plot deviates from the linear trend, pump a improvement in the above indicators.
small additional amount and stop pumping. Eventually, corrections to the measured surface pres-
7. Monitor pressure decline for 10–15 min. sure can be made. For example, accounting for fluid
compressibility, thermal expansion, and friction-pres-
Various patterns can be interpreted on the pressure
sure decreases allows one to derive a more accurate
graph. In particular, a channel in the cement may be
bottomhole pressure. However, this is not usually done
indicated by any one of the following.
for shallow-marine sediments. Zhou and Wojtanowicz
■ Leakoff equivalent mud weight more than 0.5 lbm/gal
(1999) developed an interpretation of PITs based on
below the predicted value. channeling between cement and formation, considering
■ Gauge pressure at minimum horizontal stress less the variation of stresses during cement hydration.
than half of the maximum gauge pressure.

550 Well Cementing

15-2.2 Inflow testing 15-3 Temperature, nuclear, and noise-
Inflow testing (also known as dry testing) is a drillstem logging measurements
test (DST) to assess the isolation provided by the
cement. The DST is essentially the opposite of a pressure 15-3.1 Temperature logging
test. The pressure inside the casing is reduced, and the Temperature logging is often used to evaluate primary
well is monitored to detect the ingress of formation cement jobs, mainly to detect the top of the cement
fluids. A successful dry test shows no downhole-pressure column. Temperature surveys are also performed to
change during the opening of the downhole valve or detect leaks or channeling.
during the following shut-in period (Fig. 15-2). Fiber optics can also be used to measure the temper-
Dry tests are particularly useful for testing the effec- ature in the wellbore. The fiber is placed inside the
tiveness of a cement squeeze or a cement seal at the top annulus (in a control line) or inside the casing. The tem-
of the liner. The dry test can also be used to test the perature is recorded versus time and depth. Such mea-
cement seal around the casing after perforating across surements provide valuable information regarding
an impermeable layer or drilling out the casing shoe. cement placement.

15-3.1.1 Cement hydration detector

E A
Temperature surveys are often used to detect cement in
the annulus several hours after cement placement. The
exothermic cement-hydration reactions (Chapter 2)
raise the wellbore temperature, resulting in a deviation
Pressure from the normal temperature gradient. Figure 15-3
D C B shows a typical temperature survey performed after a
Baseline primary cement job. Such measurements are easy to per-
A Initial hydrostatic pressure when packer set form and can accurately detect the top of the cement.
B Flowing pressure when downhole valve opened When combined with hole size, volumetric calculations
C Pressure at the end of flow period, when downhole can be performed to determine the displacement effi-
valve closed ciency.
D Pressure at the end of shut-in period
E Final hydrostatic pressure after unsetting packer

Fig. 15-2. Typical inflow test pressure chart.

700

15-2.3 Tests through perforations 800

In some areas, especially when the production interval
has a low permeability, the isolation provided by the 900
cement is evaluated after the perforation of the intervals
to be produced. The well then produces through the per- 1,000
forations and the production is analyzed. The presence Depth
of water in the produced fluid typically indicates annular (ft)
communication and the need for remedial cementing. 1,100
Probable cement top
When cement bond logs (CBLs) show poor results or
when effective isolation is required over short intervals, 1,200
the casing is perforated in two different locations. A
packer is set between both sets of perforations, and pres- 1,300
sure is applied at the lower perforations. This is a com-
munication test, as explained by Abdel-Mota’al (1983). 1,400
If pressure transmission or annular transmission is
observed, hydraulic isolation in the annulus is inade-
quate, and remedial cementing is necessary. Because 90 100 110 120
many operators are reluctant to add extra perforations
Temperature (°F)
to a casing string, this test is rarely performed today.
Fig. 15-3. Typical temperature survey.

Chapter 15 Cement Job Evaluation 551

 Field experience shows that the maximum tempera- Because the amount of generated heat is related to
ture anomalies arising from cement hydration range the cement volume, the temperature increase may be
from 10° to 40°F [18° to 72°C] and strongly depend upon greater in large annuli (e.g., in washouts) when com-
the annular thickness and the thermal conductivity of pared to holes across homogeneous formations. In some
the formation. The heat generated during cement hydra- cases, a temperature increase in a smooth hole can be
tion is also intimately related to the amount of cement caused by cement invasion into a fractured or fragile
and additives present in the slurry. When compared with formation (Suman and Ellis, 1977). On the other hand,
standard cements, low-density cements containing in narrow annuli, the heat generated during cement
extenders generate less heat per unit of volume. In this hydration may not be sufficient to significantly alter the
case, the temperature increase may not be sufficiently wellbore-temperature profile.
obvious to be detected.
Well cooling by the circulation of fluids before and
during cementing also influences the kinetics of cement 15-3.1.2 Communication indicator
hydration—the longer the circulation, the lower the (pump-in temperature log)
temperature. This leads to longer thickening times and After a well has been completed, temperature logs can
delayed and smaller temperature increases in the well. help assess whether channeling behind the casing is
Temperature logs may not be suitable for evaluating very occurring because of contamination of produced fluids
long cement columns, because there may be a large tem- or fluid injection outside of the perforated interval
perature differential between the top and the bottom of (Bergren and Bradley, 1988; Krause and Reem, 1992).
the well; also, the cement at the top of the column may Figure 15-5 is a typical example. The first temperature
require a long time to set. survey shows the temperature profile before the injec-
The influence of well temperature is illustrated in
Fig. 15-4. For the same slurry, the heat peak is larger and
sharper at higher curing temperatures. Consequently, Depth Temperature (°F)
the resulting wellbore-temperature increase may be (ft)
200 210 220
larger in hot or deep wells. In most cases, the peak tem- 10,900

perature is obtained 4 to 12 hr after cement placement, Temperature

but the temperature will remain elevated for more than static run before
Temperature pumping 80 bbl
24 hr (Suman and Ellis, 1977). For best results, the tem- static run after diesel
perature survey should be run within the first 12 to 24 hr pumping 80 bbl
after cement placement, depending upon the thickening diesel
time and downhole conditions.

11,000
200
Perforations
175

150

125
Heat Cement
flow 100 channel
Gamma ray
(mW) 85°C 11,100

75 Oil/water contact

50
45°C
25°C
25

0
0 5 10 15 20
Curing time (hr) 11,200

Fig. 15-4. Effect of temperature on the hydration kinetics of a typical

Class G cement system (15.8 lbm/gal [1,900 kg/m3]). Fig. 15-5. Temperature composite profile log before cement squeeze.

552 Well Cementing

tion of 80 bbl of diesel oil. The second temperature ing fibers in the wellbore or in the annulus between the
survey, run a short time after injection, shows a large casing and formation (Carnahan et al., 1999). The fibers
temperature decrease above the perforations and tem- are usually protected inside a small-diameter (usually
perature variations down to the oil/water contact. This 1⁄4-in.) tube. Carnahan et al. presented an example of a

may indicate communication behind the casing. steam injection well in which the sensing fibers detected
Correlating these logs with other cement logs is some- the upward migration of heated fluids above 1,200 ft and
times required to fully assess the presence of channels out of the intended injection interval (Fig. 15-6).
in the cement. In such cases, remedial cementing must A squeeze treatment was performed after careful per-
be performed to seal the annulus and reduce the water forating at 620–622 ft (to avoid damaging the fibers). As
influx. shown in Fig. 15-7, the fluid migration was stopped suc-
Most limitations of conventional temperature surveys cessfully.
can be overcome by installing optical temperature sens-

500

450

400

350

300
Temperature
(°F) 250

200

150

100

50
0 200 400 600 800 1,000 1,200 1,400 1,600
Depth (ft)

Fig. 15-6. Temperature profile before squeeze, indicating fluid migration above injection zone owing to poor
primary cement job (from Carnahan, 1999). Reprinted with permission of SPE.

500

450

400

350

300
Temperature
(°F) 250

200

150

100

50
0 200 400 600 800 1,000 1,200 1,400 1,600
Depth (ft)

Fig. 15-7. Normal temperature profile after a block squeeze indicated fluid migration has stopped
(from Carnahan, 1999). Reprinted with permission of SPE.

Chapter 15 Cement Job Evaluation 553

15-3.2 Nuclear logging The primary criterion for selecting the radioactive
Radioactive materials are frequently added to treatment tracer is its half-life. When long, the alteration to the
fluids as tracers. This technique can be used to locate original gamma ray signature in the formation will be
drilling mud. By detecting the presence of the radioac- for all practical purposes permanent. When short, the
tive material in the well returns, one can estimate circu- tracer will cease to be radioactive within a few weeks or
lation times. In stimulation, addition of a radioactive months, and the gamma ray signature will return to its
material to the treating fluids allows one to estimate the initial state.
extent of the treatment by comparing gamma ray logs A second selection criterion is the energy of the dom-
run before and after injection (Ahmed, 1989). This tech- inant gamma ray emitted by the tracer. When using a
nique is also used in cementing, mainly to locate the top spectral gamma ray log, it is possible to selectively mea-
of the cement (Fig. 15-8). It is important to stress that sure the radioactivity of the tracer, and the amount of
special safety and health precautions must be taken radioactive matter can be significantly decreased.
when handling and using radioactive materials, espe- Table 15-1 is a list of the more common radioactive trac-
cially for materials with long half-lives. ers, with information regarding their half-lives and
Kilne et al. (1986) proved that radioactive tracers can gamma ray energies.
be used to perform a quantitative cement evaluation. A
uniform concentration of radioactive material is added
to the cement slurry. Then a spectral gamma ray log is Table 15-1. Radioactive Tracers Used in Nuclear Logging
compared with the caliper log. Element Name Half-Life Main Gamma Ray Energy
Several radioactive tracers can be used in cementing. (days) (MeV)
Soluble tracers (e.g., I131) can be added to the mix water. Cr51 27.7 0.32
Sand or glass beads coated with a radioactive material
(e.g., Ir192) can also be used. The standard concentration Fe59 45 1.10
of radioactive material is about 3 mCi/m3 of mix water.
Br82 1.5 0.77

I131 8 0.36

Ir192 74 0.32
Depth Radiation intensity increases
(ft) Au198 2.7 0.41

15-3.3 Boron-pulsed neutron logs

5,900 The procedure for a boron-pulsed neutron log is similar
to that of the pump-in temperature log. A boron solution
is injected through the perforations and a pulsed neu-
tron log is generated. Boron absorbs neutrons very effi-
ciently; consequently, channels in the cement sheath
Cement top can be readily detected, even through two strings of
6,000
casing. The method is more sensitive than the pump-
in temperature log (Blount et al., 1991; Sommer and
Jenkins, 1993).

Base run
6,100
15-3.4 Oxygen activation logs
Oxygen activation logging (McKeon et al., 1991; Pappas
et al., 1995) is a technique that can quantify water flow
behind casing. High-energy neutrons interact with
oxygen nuclei in water, resulting in the emission of
After run
gamma rays. Before the well is energized with neutrons,
a background gamma ray log is generated. Then, the
Fig. 15-8. Typical radioactive-tracer survey. neutron tool is activated for a very short period (1 to

554 Well Cementing

15 sec), followed by a longer gamma ray acquisition Meaningful acoustic-log interpretation (Fig. 15-9)
period (20 to 60 sec). The flow velocity is determined requires good quality control of the log, a good estimate
from the source/detector distance and the time during of relevant cement properties, and knowledge of the
which the activated water passes the detector. ■ well and casing

■ cement-job events

15-3.5 Noise logging ■ pre- and postjob well history.

Flowing fluids (gas, water, or oil) produce sound. Noise In this section, each of these points will be covered in
logging can be used to detect fluid flow behind the detail for sonic and ultrasonic logs.
casing or fluid entry inside the wellbore. This technique
can be particularly useful for identifying communication
between two zones with different pore pressures. The
analysis of the frequency spectrum (20 to 5,000 Hz) and
Acoustic log
the amplitude of the noise enables location and identifi-
cation of flowing fluids. This technique also provides
information concerning the magnitude of the communi-
cation problem. In 1973, McKinley et al. discussed cases No Log
quality control Yes
in which noise logs could identify interzonal communi- passed
cation more accurately than temperature logs.
A noise log is a series of stationary noise measure-
ments, because it is difficult to detect formation-related No Well
Yes
sounds if the tool moves continuously. As a result, this data available
technique is marginally used in the oil field.

No Cement Yes
15-4 Acoustic logging measurements job data

15-4.1 Introduction
Acoustic logging is the most widely used and efficient No Postjob Yes
method to evaluate cement jobs. Cement-job evaluation well history
through acoustic-log interpretation seeks the relation-
ship between the tool response and the quality of the
cement job after a given time following cement place-
ment. The response of acoustic tools is related to the
acoustic properties of the surrounding environment Poor evaluation Good evaluation
(casing, cement, and formation) and the quality of the
acoustic coupling between the casing, cement, and for-
mation. Fig. 15-9. Acoustic log flowchart.
Cement is only one of many parameters that can
affect the log response. The analysis of the log must be
performed carefully to determine the origin of the log
15-4.2 Quality control
response. Most of the time, detailed information con-
cerning the well geometry, formation characteristics, All acoustic logs must meet rigorous quality control stan-
and cement job is required. dards; otherwise, they have no credibility.
A fair interpretation of an acoustic log can only be
made when it is possible to anticipate the log response. 15-4.2.1 Measurement repeatability
A valid cement-job evaluation results from analyzing dis-
crepancies between the expected and the actual log All logs should have a repeat section. A repeat section is
response. Today, one can qualify and possibly quantify a short log pass, generally over about 200 ft [61 m] of
the results of the cement job, mainly in terms of cement hole, recorded immediately before the main pass. The
quality and cement coverage. interval logged for the repeat section must also be part

Chapter 15 Cement Job Evaluation 555

of the main pass, so that the two passes can be com- 15-4.3 Acoustic properties
pared. Both logging passes must occur under identical
conditions, including tool settings and hole conditions. 15-4.3.1 Definitions
The purpose is to verify that the logging tool produces Acoustics is the science of sound. In the context of log-
the same reading repeatedly under the same conditions. ging, it pertains to the propagation of sound waves.
If the tool response is repeatable, there is no guarantee Sound propagation is the periodic compression and rar-
of accuracy; however, if it is not repeatable, there is efaction of molecules (in the case of a gas or liquid) or
most assuredly a problem. This verification is particu- the squeezing and stretching of the grain fabric (in the
larly important in well logging because of the environ- case of a solid) (Fig. 15-10). When this motion occurs in
mental extremes to which the logging tool is typically the same direction as the traveling propagation, the phe-
subjected and the need to assure that the time spent nomenon is called a compressional wave.
logging will result in good data. In a solid, a second type of wave—the shear wave—
The standard of “good repeatability” depends upon can propagate. It does not exist in fluids. When a shear
the type of tool, its operating principle, and the design. wave passes through a solid, the grain-fabric vibration is
Unlike nuclear tools, acoustic tools experience no statis- perpendicular to the direction of wave propagation. The
tical fluctuation; thus, repeatability is primarily a func- shear wave always travels more slowly than the com-
tion of tool-design quality. The conventional bond log pressional wave. Compressional and shear wave veloci-
should exhibit a virtually identical response on the ties are intimately related to the elastic properties of the
repeat and main passes. material (Young’s modulus, shear modulus, and
To maximize acoustic coupling between the casing Poisson’s ratio) and are almost independent of the fre-
and the cement, some operators pressurize the casing quency. These elastic properties relate the stress to the
during the logging procedure. The sensitivity of the strain in the material according to Hooke’s law.
acoustic measurement to changes in pressure requires
that identical pressures be established and maintained
throughout the repeat pass and the main pass. Wave Propagation Modes
Depending upon the equipment used to exert pressure
and maintain a wireline seal, this may not be possible.
Under such circumstances, the repeat section is invalid.
The only way to ensure that the repeat and main passes
are performed under identical conditions is to run the
logs without pressurized casing.

15-4.2.2 Calibration summary

A true calibration is made against an accepted reference
standard. It may be an industry-wide standard such as
the American Petroleum Institute (API) limestone mea-
surement for nuclear tools. Or, it may be a logging com-
pany-specific standard. At present, there is no industry
standard for sonic-tool calibration. Therefore, the term
“tool check” is more appropriate than “calibration,” At rest Compressional Shear
because the reference is not an industry standard. Fig. 15-10. Illustration of propagation of compressional and shear
However, the industry uses the term “calibration” for wave modes.
this check. The objective of this check is to ensure that
sonic tools perform consistently over time.
To properly interpret a log, a calibration summary
should be printed. It enables the log analyst to deter-
mine if the measurement setup was correct. Without a
calibration summary, it is impossible to know exactly
what was measured and how it was measured; thus, the
log results are doubtful and interpretation may lead to
erroneous conclusions.

556 Well Cementing

 In well logging, sound waves are generally character- pressional acoustic impedance (Z) of this material, tra-
ized by their slowness (v–1), traditionally expressed in ditionally expressed in 106 kg/m2/s. This unit is also
μs/ft or μs/m, which is the inverse of velocity. For a called the mega-Rayleigh (MRayl).
linear elastic material, the relationship between the For a homogeneous, nondissipative medium, the
compressional- and shear-wave velocities (vP, vS), acoustic impedance is given by
Young’s modulus, E, and Poisson’s ratio, ν, is
Z = ρvP , (15-3)
1 1
⎡ E(1 − ν) ⎤2 ⎡ E ⎤2 where
vP = ⎢ ⎥ and vS = ⎢ ⎥ ,
⎣ ρ (1 + ν)(1 − 2 ν) ⎦ ⎣ 2ρ (1 + ν) ⎦ vP = velocity of compressional wave (m/s)
ρ = density of material (kg/m3).
(15-1) A sound wave loses energy as it propagates through a
where material. This loss of energy, called attenuation, is a
characteristic of the material and increases with the
ρ = density of the material.
frequency of the wave. No general relationship exists
Another type of wave that is important in cement between attenuation and frequency. For a given fre-
evaluation is the plate wave, illustrated in Fig. 15-11. quency, attenuation is normally expressed in decibels
This wave propagates along the steel casing slightly (dB) per unit of distance.
more slowly than the compression wave in the steel, and
it is the basis for the CBL (Minear and Goodwin, 1998). 20 p
The velocity of the plate wave, vpl, is given by A= log 10 x , (15-4)
L px+ L
E where
v pl = . (15-2)
ρ ( 1 − ν2 ) A = signal attenuation (dB)
L = distance between the two locations
Today, for cased hole-log interpretation, one is mainly px = signal pressure (amplitude) at a location x
interested in the propagation velocity of compressional px + L = signal pressure (amplitude) at a location x + L.
waves. Knowledge of the compressional wave velocity
through a material allows one to determine the com- 15-4.3.2 Acoustic properties of formations
Acoustic properties of the formation influence acoustic
logs. The familiar terms fast formation and slow forma-
tion refer to sound velocity. Traditionally, for cement-
evaluation purposes, a formation is called “fast” when
sound travels through it faster than it does along the
casing (i.e., 17,000 ft/s [5,334 m/s]). Table 15-2 presents
the acoustic properties of common formations.
The same table cannot be constructed for attenuation
characteristics, because they are frequency dependent.
Generally speaking, attenuation is high in slow forma-
tions. Attenuation is very high in unconsolidated materi-
als (e.g., shallow sands and shales). Attenuation is neg-
ligible in strong, consolidated rocks.

Fig. 15-11. Plate wave propagation.

Chapter 15 Cement Job Evaluation 557

Table 15-2. Acoustic Properties of Common Formations
Nonporous Solids
Material Type Δt Sound Velocity Acoustic Impedence
(µs ft–1) (ft/s–1 [m/s–1]) (MRayl)
Steel (compression wave) 51.4 19,450 [5,930] 46.00

Plate wave in casing (CBL) 57.0 17,500 [5,334] 41.60

Dolomite 43.5 23,000 [7,010] 20.19

Anhydrite 50.0 20,000 [6,096] 18.17

Limestone 47.6 21,000 [6,400] 17.34

Calcite 49.7 20,100 [6,126] 16.60

Quartz 52.9 18,900 [5,760] 15.21

Gypsum 52.6 19,000 [5,791] 13.61

Halite 66.6 15,000 [4,572] 9.33

Water-Saturated Porous Rock

Material Type Porosity Δt Sound Velocity Acoustic Impedence
(%) (µs ft–1) (ft/s–1 [m/s–1]) (MRayl)
Dolomite 5 to 20 50.0 to 66.6 20,000 to 15,000 [6,096 to 4,572] 16.95 to 11.52

Limestone 5 to 20 54.0 to 76.9 18,500 to 13,000 [5,639 to 3,962] 14.83 to 9.43

Sandstone 5 to 20 62.5 to 86.9 16,000 to 11,500 [4,877 to 3,505] 12.58 to 8.20

Sand 20 to 35 86.9 to 111.1 11,500 to 9,000 [3,505 to 2,743] 8.20 to 6.0

Shale –† 58.8 to 143.0 17,000 to 7,000 [5,181 to 2,133] 12.0 to 4.3

† Not available.

15-4.3.3 Acoustic properties of cements low-density slurries with a low solid-volume fraction
Cased hole acoustic-log response depends primarily have a low acoustic impedance that can change signifi-
upon the acoustic properties of the hard set cement. The cantly after several days. The acoustic impedance of the
acoustic properties of rocks are well-known and remain higher-density slurries changes less than 20% between 1
essentially constant during the life of a well. Unlike and 7 days.
rocks, the acoustic properties of cements change during Cements extended with hollow microspheres or nitro-
this time period. This fundamental difference influences gen can have a very low acoustic impedance. As a result,
the log analysis. it may be difficult to distinguish the cement from water.
Significant improvement of the response over days or
■ Log results can change with time, because the physi-
even weeks has sometimes been observed when running
cal properties of the cement are changing with time. time-lapse logs. This has usually been attributed to the
■ The set cement may not be in the same physical state evolution of the cement acoustic impedance with time.
all along the casing string. This can produce a strong Such fluctuations often originate from overestimation of
difference in the log response on long strings for the bottomhole circulating temperature, resulting in
which a large temperature difference exists between slurry overretardation. Fluctuations may also occur
the bottom and top of the cement. when the temperature variation along a long cement
column is ignored.
The acoustic properties of various cement formula-
tions at ambient conditions are reported in Table 15-3
(Jutten et al., 1987). From these results, it appears that

558 Well Cementing

Table 15-3. Acoustic Properties of Various Cement Formulations
Slurry Type Density Time Sound Velocity Acoustic Change in Acoustic
(lbm/gal [kg/m3]) (days) in Cement Impedence Impedence Over 1 Day
(m/s) (MRayl) (%)
Neat Class G 15.8 [1.89] 1 3,000 5.68 0
2 3,250 6.16 8
7 3,400 6.44 13

Class G + latex + 11.2 [1.34] 1 1,650 2.21 0

hollow silica 2 2,200 2.95 33
microspheres 7 2,500 3.36 52

Class G + soluble 12.0 [1.44] 1 1,600 2.30 0

silicate extender 2 1,750 2.52 9
7 2,000 2.88 25

Class G + hollow silica 12.0 [1.44] 1 2,600 3.74 0

microspheres + 2 2,800 4.03 8
4% CaCl2 (BWOC†) 7 3,000 4.32 16

Class G + soluble 13.3 [1.59] 1 1,750 2.79 0

silicate extender 2 2,200 3.51 26
7 2,500 3.99 43

Class G + latex 15.8 [1.89] 1 2,900 5.49 0

2 3,150 5.97 9
7 3,350 6.35 16

Class G + 18% NaCl (BWOW‡) 16.1 [1.93] 1 2,850 5.50 0

2 3,200 6.18 12
7 3,375 6.51 18

Class G + hematite weighting agent 19.0 [2.28] 1 3,300 7.59 0

2 3,400 7.74 2
7 3,530 8.04 6

36% quality foam 10.0 [1.20] 7 2,300 2.76 –§

Conventional low-density system 12.51 [1.50] 7 2,000 3 –

Engineered-particle-size 10.0 [1.20] 7 2,900 3.48 –

low-density system

Engineered-particle-size 8.61 [1.03] 7 2,790 2.87 –

ultralow-density system
† Byweight of cement
‡ Byweight of water
§ Not available.

15-4.3.4 Acoustic properties of fluids At ultrasonic frequencies, the acoustic properties of

An accurate knowledge of logging-fluid properties is vital weighted drilling muds containing suspended solid par-
for ultrasonic cement evaluation (Section 15-4.5.6). ticles depend on inertial and viscous effects. Laboratory
Information concerning other fluids that may be encoun- measurements shown in Table 15-5 (Hayman, 1989) have
tered in the annulus is also important. Table 15-4 lists shown that the effective density measured by the ultra-
the properties of some homogeneous fluids, brines, and sound can be as much as 20% less than the static density.
oils. The acoustic impedance of a homogeneous fluid can
be calculated using Eq. 15-3.

Chapter 15 Cement Job Evaluation 559

Table 15-4. Acoustic Properties of Various Homogeneous Fluids
Material Density Slowness Velocity Acoustic Impedance
(lbm/gal [kg/m3]) (µs/ft) ft/sec [m/sec]) (MRayl)
Water 8.33 [998] 206 4,860 [1,482] 1.48

Water + 10% NaCl 8.98 [1,075] 193 5,180 [1,580] 1.70

Water + 25% NaCl 9.90 [1,186] 175 5,710 [1,740] 2.06

Water + 36% CaCl2 11.3 [1,350] 170 5,870 [1,790] 2.42

Water + KCl 9.18 [1,100] 189 5,280 [1,610] 1.77

Water + 58% CaBr2 15.2 [1,824] 179 5,580 [1,700] 3.10

Sea water 8.56 [1,025] 199 5,020 [1,531] 1.57

Kerosene 6.74 [808] 230 4,340 [1,324] 1.07

Diesel 7.09 [850] 221 4,530 [1,380] 1.17

Air at 15 psi, 32°F [0°C] 0.01 [1.3] 920 1,090 [331] 0.0004

Air at 3,000 psi, 212°F [100°C] 1.59 [190] 780 1,280 [390] 0.1

Table 15-5. Acoustic Properties of Various Drilling Muds

Material Density Slowness Velocity Acoustic Impedance
(lbm/gal [kg/m3]) (µs/ft) (ft/sec [m/sec]) (MRayl)
Low Frequency At 0.5 MHz
WBM† 12.6 [1,510] 215 4,660 [1,420] 2.14 1.85

WBM 15.4 [1,850] 216 4,630 [1,410] 2.60 2.14

Oil-base fluid 7.79 [933] 231 4,330 [1,320] 1.23 1.25

OBM‡ 12.6 [1,510] 245 4,070 [1,240] 1.87 1.54

† Water-base mud
‡ Oil-base mud

The impedance of a weighted mud is given by The attenuation of ultrasound in a weighted mud is
sometimes a limiting factor in the use of ultrasonic tech-
Z mud = B × ρ mud × vmud , (15-5) niques. The attenuation increases with increasing fre-
where quency and with increasing solid content (or density).
B = correction factor (usually between 0.8 and 1.0) OBMs usually have greater attenuation than WBMs of
vmud = acoustic velocity through the mud the same density. The attenuation is difficult to predict
Zmud = acoustic impedance accurately; it depends on mud composition, tempera-
ρmud = static mud density. ture, and pressure.
Figure 15-12 shows a semiempirical approximation
for B that depends on the solid and liquid densities and
volume fractions. Proprietary software from the logging
company can be used to calculate fluid properties,
including the B factor.

560 Well Cementing

 1.00 when various gate settings were used and suggested the
0.95
need for some kind of waveform presentation to elimi-
WBM nate these ambiguities.
0.90 In 1964, Carter and Evans provided information from
Impedence
correction 0.85 experiments on the pipe-to-cement bond, with consider-
factor, OBM ation given to casing-surface finishes, cement placement
0.80
B techniques, and the timing of operations in the
0.75 cemented casing. In a paper directed primarily at for-
0.70
mation evaluation, Anderson and Riddle (1964) pointed
900 1,400 1,900 2,400 out the efficacy of a direct attenuation measurement
using two receivers in place of a single-receiver ampli-
Mud density (kg/m )
–3
tude measurement. In 1968, Walker presented an
Fig. 15-12. Typical acoustic impedance correction factors for WBM overview of the CBL and the economic considerations in
and OBM. the decision on whether to run one. Brown et al. (1970)
showed the advantages to be gained from a shorter spac-
ing between the transmitter and receiver for both ampli-
tude measurements and full-waveform measurements.
15-4.4 Sonic cement evaluation techniques They suggested 3 ft [0.9 m] and 5 ft [1.5 m], respectively,
15-4.4.1 History as good compromises.
A paper published in 1974 by Fertl et al. was oriented
Cement bond logging was developed in the late 1950s toward operational considerations: It included a check-
and is still widely used to evaluate cement jobs. Its use list and descriptions of tool-design differences and their
can be traced to 1960, when Tixier et al. found that the resulting effects on the log. The authors lauded the
transit-time curve on openhole sonic logs often “cycle potential of the bond-log measurement and lamented
skipped” when run inside the casing because of consid- the lack of standardization.
erable attenuation of the acoustic signal. A resurgence of CBL-related research occurred in the
In 1961, two major service companies published the 1980s. McGhee and Vacca (1980) summarized the possi-
results of laboratory and field testing of tools designed ble causes of a microannulus and emphasized the impor-
to measure the actual amplitude of the acoustic signal tance of dealing with a microannulus if it should occur.
(Grosmangin et al., 1961; Anderson and Walker, 1961). In 1982, Gollwitzer and Masson presented the justifica-
One company reported that more than 500 CBLs had tion for a borehole-compensated CBL that directly mea-
been run between April 1959 and November 1960 in sured attenuation by employing multiple receivers in the
North and South America. Both observed many of the logging sonde. In 1983, Bruckdorfer et al. extended the
interpretation complexities still discussed today, such as available empirical correlations between cement com-
those linked to cement-sheath thickness, pipe eccenter- pressive strength and the CBL attenuation rate to
ing, bond to pipe versus bond to formation, and fast include foamed cements.
formation (Section 15-4.4.2.4). In 1984, Nayfeh et al. published the results of an
In 1962, Winn et al. reported the results of a labora- experimental and numerical study of the effects of well-
tory study concerning the effect of various cement bore fluids on CBL amplitude measurements.
compositions on the attenuation of a sonic signal and Considerable differences in free-pipe amplitude were
compared the acoustic measurement to hydraulic com- shown to occur for various brines. Pressure and temper-
munication. In 1963, Bade reported a field study involv- ature corrections were given for the transducers used in
ing approximately 250 CBLs run in the Williston Basin of CBL tools. In 1985, Allen and Wood offered a solution to
central North America. Four commercially available the operational problem of assuring that the amplitude
tools were used. Bade observed the need for proper cen- measurement gate is correctly set by proposing the
tralization as well as the need for taking measurements simultaneous recording of sliding and fixed-gate CBL
within short time intervals (known as gates) to exclude amplitudes (Section 15-4.4.2.5). In 1985, Bigelow pre-
the formation signals. In 1963, Pardue et al. provided sented a comprehensive guide to CBL interpretation,
cement-to-log measurement correlations that are still with an emphasis on qualitative interpretation of the
used in interpretation charts, along with a theoretical full-waveform presentation based on understanding the
model employing a thin steel plate coupled to an infinite basic acoustic processes involved and the cement
half-space. Also in 1963, Pickett argued for some form of job/well completion history. Bigelow emphasized the
full-waveform presentation on both openhole and cased need for intelligent interpretation. Tubman et al. (1986)
hole CBL logs. Using data from Bade’s field study, Pickett provided a theoretical foundation for the response of
showed the wide variation in amplitude measurements acoustic logs in cased boreholes.

Chapter 15 Cement Job Evaluation 561

 In 1987, Leslie et al. analyzed the theoretical
response of the borehole-compensated bond log and Casing fluid
separated a “coupling attenuation,” insensitive to a
microannulus, from the “propagation attenuation,” Casing collar
which requires shear coupling between the casing and locator Casing
the cement. Jutten and Parcevaux (1987) and Jutten
and Corrigall (1988) advanced the understanding of the
effects of borehole geometrical parameters and cement- Gamma ray Cement
slurry composition and mechanical properties on bond- detector
log response.
In 1989, Lester developed a new cement-sheath eval-
uation tool featuring pad-mounted acoustic transducers. Formation
This technique provides compensated attenuation mea-
surements within 6 distinct 60° sectors. The operating Transmitter
frequency is 100 kHz. Sonic wave
In 2000, Frisch et al. proposed a new interpretation path
technique that utilizes both CBL and ultrasonic data
to evaluate cements. This technique uses statistical
analysis to compute a CBL variance that allows differ- 3-ft receiver
ences between free, partially bonded, and bonded
pipe. Statistical approaches are described later in this
chapter. 5-ft receiver
A variable-density log (VDL), or presentation of the
acoustic waveform at a receiver of a sonic or ultrasonic
measurement, is commonly used as an adjunct to the
CBL and offers better insights into its interpretation.

Fig. 15-13. CBL-VDL tool configuration.

15-4.4.2 The CBL
15-4.4.2.1Description of the conventional bond-logging tool
Figure 15-13 shows a schematic diagram of a represen-
tative bond-log tool, together with a cross section of a each sonic burst is about 20 kHz for larger-diameter
cased and cemented well. The tool has an acoustic trans- tools (larger than 3 in. [8 cm]) and about 30 kHz for
mitter that is usually made of a piezoelectric ceramic. smaller-diameter tools (less than 2 in. [5 cm]). One com-
There are two receivers, also of piezoelectric ceramics, pany offers a tool with a frequency of 12 kHz. In the time
in most tools. Some designs incorporate a single interval between transmitter bursts, the receiver picks
receiver. In the former case, the two receivers are gen- up the signal and makes the bond-log measurements.
erally located 3 ft [0.9 m] and 5 ft [1.5 m] from the trans- Most of the signal of interest arrives at the receiver
mitter. In the latter, the single receiver is 4 ft [1.2 m] between 1 and 2 ms after the transmitted burst.
from the transmitter. Some hostile-environment tools The transmitter burst creates an approximately
use magnetorestrictive transducers rather than those spherical wave front that expands away from the tool in
made of ceramics. This requires different pressure and all directions. As the wave front strikes the inside wall of
temperature corrections (Nayfeh et al., 1984). Not the casing, it is refracted according to Snell’s law. There
shown in the figure, but required adjuncts for the tool, is one particular direction of travel of the wave front that
are a sufficient number of centralizers to ensure that the will result in refraction straight down the casing. Known
transmitter/receiver section of the tool remains cen- as the critical angle, it is about 16.5° with fresh water in
tered in the pipe. the hole. The portion of the wave front that is refracted
straight down the pipe ultimately determines the ampli-
15-4.4.2.2 Acoustics of the bond-log measurement tude and transit-time measurements that appear on the
The transmitter repeatedly emits short bursts of acousti- log. Some parts of the original wave front travel directly
cal energy. The duration of each burst is about 50 μs, and through the mud, and some parts refract their way into
the repetition rate is between 10 and 60 Hz, depending the annulus and formation. Part of the latter ultimately
upon the particular tool design (or a frequency selected shows up at the receiver as a formation signal and the
by the logging engineer, in some cases). The frequency of former shows up as mud waves.

562 Well Cementing

 Figure 15-14 is a schematic representation of the var- annulus is the variable. The magnitude of the loss is rep-
ious paths the original burst can follow and still arrive at resented by the amplitude or attenuation appearing on
the receiver. The waveforms in the figure are meant to the log. If there is fluid in the annulus, there will be little
convey the relative times of arrival of the acoustic energy attenuation of the casing signal. In fact, all fluids would
that has traveled along the various paths. The wave that be expected to look alike because there is no shear cou-
refracted directly down the casing wall usually arrives pling for any fluid. This is also the reason that even a
first because of the high sound velocity in steel and the microscopic gap of a few thousandths of an inch between
relatively short distance. The mud wave has the shortest the pipe and a cement sheath, known as a microannulus,
distance to travel; however, because of the low sound has a strong effect on the signal (Section 15-4.4.2.13).
velocity in fluids, it arrives very late. The arrival time of
the shear and compressional formation waves is highly 15-4.4.2.3 Description of the full acoustic wave display: VDL
variable (Section 15-4.4.2.5). The signal arriving at the The presentation of the full waveform provides some
receiver will be a mixture, or composite, of waves from information about the cement job. The actual composite
all these paths. The interpretation of the actual bond-log signal is presented in Fig. 15-15. While its use in bond-log
measurement (as opposed to a picture of the entire com- interpretation is primarily qualitative, it contains all of
posite wave itself) depends upon the casing wave arriv- the available information in picture form.
ing before anything else. There are two ways to present or display this signal on
The casing wave is the portion of the original acoustic the log itself. One way is to display an actual waveform.
burst that propagates directly down the casing wall. It The disadvantage of this approach is that one does not
loses energy to the annulus and borehole as it propa- see a continuous display with depth; instead, one wave-
gates, owing to shear coupling with the adjacent materi- form for every 2 to 4 ft [0.6 to 1.2 m] is presented.
als. The greater the shear coupling, the greater the Consequently, this presentation is rarely seen today.
energy lost into the adjacent materials. The loss to the
formation is low and constant; therefore, the loss to the

Transmitter pulse Signal at receiver

Mud

Casing

Cement

Mud

Casing Formation

Cement
E1
Formation Composite

Fig. 15-14. Sonic-wave paths.

Chapter 15 Cement Job Evaluation 563

 Variable
density
log

Amplitude

Black

t0
Medium gray
Transmitter Time
firing White
Casing Formation Formation Mud Stoneley
waves compressional shear waves waves
waves waves

Min. Amplitude Max.

Variable Density Log

200 (µs) 1,200

Fig. 15-15. Presentations of the complete waveform signal from the CBL tool.

The other presentation is the variable intensity dis- In multiple-receiver tools, the full-wave display gen-
play, in which the amplitudes of the waveforms are con- erally comes from the 5-ft [1.5-m] spaced receiver.
verted to a gray or color scale. Figure 15-15 is an illus- Increasing the spacing between the transmitter and
tration of how the amplitude information is transformed receiver is advantageous, because the various con-
to the intensity information. An amplitude of zero is dis- stituents of the composite wave (casing, formation com-
played as medium gray. Positive amplitudes become pressional, and formation shear) are spread apart. As
blacker as they increase. Negative amplitudes become the distance increases, the velocity differences become
whiter as they decrease. Continuous and discrete (five- more pronounced. However, increasing the spacing can
level) intensity scales are used in the industry. The color be problematic because the received wave is more atten-
display uses a spectrum. The minimum amplitude (neg- uated; thus, a compromise is necessary. Because the full-
ative) is displayed as dark blue, the maximum amplitude wave display is used qualitatively in most cases, the very
(positive) is red, and zero amplitude is green. This dis- high attenuation at 5 ft [1.5 m] is not a problem, because
play is continuous with depth and easy to read. However, the qualitative characteristics are still distinct.
the resolution depends upon the dynamic range of the
received signals. A VDL example is shown in Fig. 15-16.

564 Well Cementing

 PIP Summary
Casing collars
Time mark every 60 s Cement isolation marker
Transit Time (Sliding Gate) (TTSL)
400 (µs) 200
Transit Time (TT)
400 (µs) 200
Gamma Ray (GR)
0 (gAPI) 150
Tension Min. Amplitude Max.
CCL (CCLU) (TENS) CBL Amplitude (CBL)
–1 (....) 1 (lbf) 0 (mV) 50 Variable Density Log
2,000 0 200 (µs) 1,200

600

625

Fig. 15-16. Full acoustic-wave display (CBL amplitude and VDL).

Chapter 15 Cement Job Evaluation 565

15-4.4.2.4 CBL/VDL: Qualitative interpretation Wavy patterns on the received waveforms are the perfect
The analysis of the full wave display gives only qualita- illustration of this. They represent a qualitative indica-
tive information about the cement job. If the cement-to- tion of good acoustic coupling between the cement and
casing bond is good, most of the sonic energy will leave formation (Fig. 15-17).
the casing and pass into the cement; thus, the casing Several special cases should be taken into considera-
waves will have a low amplitude. Likewise, if the cement- tion.
to-formation bond is good, the energy will go through the Unconsolidated formations: These shallow forma-
cement into the formation. The sonic waves (compres- tions strongly attenuate the sound. The VDL does not
sional and shear) will then propagate and attenuate show any formation waves because their amplitude is too
through the formation. Because formations are never low.
perfectly homogeneous, their acoustic properties vary.

3,175

Formation arrivals:
wavy patterns

3,200

Low CBL amplitude

3,225

First reading GR

Fig. 15-17. Wavy patterns on the VDL represent formation arrival. In this particular example, the absence
of a casing arrival and the low CBL amplitude are indicative of good acoustic coupling between the
cement and the casing.

566 Well Cementing

 Fast formations are those through which sound trav- The underlying premise in quantitative bond logging
els faster than along the casing. The wavy pattern on the is that the strength of the casing signal is a function of
VDL is seen earlier than the casing arrivals. Obviously, the material adjacent to the casing. From this premise,
both the casing-to-cement bond and the cement-to- and the considerations in the preceding paragraph, it
formation bond are generally good. follows that the earliest detectable wave, E1, should be
Salt formations, which are highly plastic, have little measured. However, this is not the first arrival because it
heterogeneity. Across such a zone, the VDL is very regu- is preceded by a much smaller single-cycle precursor.
lar most of the time, sometimes appearing similar to that Many schemes have been implemented to quantify the
of free pipe. strength of the casing signal, including measuring the
Concentric casings: If the inner casing is well peak amplitudes of E1, E2, and E3, the area under the
cemented, the VDL will show the quality of the previous half-cycles, and the area under multiple half-cycles
cement job but often displays parallel stripes as in free (Bade, 1963; Pickett, 1963). For this reason, the log
pipe. The chevron patterns at the casing collars are visi- curves generated by different tool designs in the same
ble. When the annular gap between both casings is well well at the same time can look completely different.
cemented, the casing collars from the previous casing Because there is no analytical description of the
can also be seen on the VDL. Often in this type of situa- received waveform in terms of the geometry and physical
tion, an apparent frequency increase is seen on the parameters, there is no definitively correct measure-
signal, which is visualized on the VDL by a larger num- ment scheme to be applied to the waveform. However,
ber of thinner black-and-white stripes (Jutten, 1988). laboratory work by Jutten and Parcevaux (1987) and
Intimate contact between the casing and formation: Jutten and Corrigall (1988) indicated that even the tail
When the casing is lying against the formation wall, even end of the E1 half-cycle may be affected by the well
when it is not well cemented, formation arrivals may be geometry. Therefore, measuring the amplitude of E1 is
seen on the VDL. The casing signals are also very strong. necessary.
The first quantitative measurement performed on the
15-4.4.2.5 Quantitative data from the acoustic wave full wave is the elapsed time between the transmitter
The first few cycles of the received waveform are shown firing and the arrival of the first part of the wave exceed-
in stylized form in Fig. 15-18. The traditional convention ing a preset amplitude threshold (Fig. 15-19). It is
in acoustic logging is to label the half-cycles as E1, E2, known as the transit time. The actual measurement
E3, etc., with odd numbers referring to positive peaks value depends on (1) the threshold level and (2) the use
and even numbers to negative peaks. Ideally, the first of positive or negative peaks.
cycles are from casing signal. In the case of free pipe
(i.e., a fluid-filled annulus) this is true, but in cemented
pipe the picture is not so clear, because it depends on
the formation properties and the thickness of the
cement sheath (Section 15-4.4.2.10).
E1 Poor bond
Amplitude Detection
(mV) level E3
Good bond
Travel
time t0 Time
Transit
Amplitude Stretch E2
E1 E3 E5 time
(mV)
Amplitude
ffr Time
t0 Fig. 15-19. Transit time stretch in well-bonded casing.
E2 E4 (µs)

Fig. 15-18. Sonic waveform nomenclature. The time measured will also depend on the size of the
inside diameter of the casing and the outside diameter
(OD) of the tool, as well as the speed of sound in the
borehole fluid. Because of the threshold amplitude (nec-
essary to avoid random noise), the measured time will
increase slightly as the amplitude of the relevant half-
cycle (E1 in Fig. 15-19) drops close to the threshold level

Chapter 15 Cement Job Evaluation 567

(also known as the detection level or bias level). When measured. However, this is not the case when the ampli-
the amplitude drops below the threshold, the arrival tude of the actual casing signal falls below the bias set-
time of the next half-cycle will be measured, e.g., E3. ting (detection level) of the sliding gate. Because this
Thus, the transit-time measurement responds to technique appears to be obsolete as a primary means of
changes in amplitude in a crude way. It is sometimes measuring amplitude and although its use was proposed
mistakenly assumed that the transit-time curve says (Allen and Wood, 1985) to verify the positioning of the
something about the cement. In fact, the most valuable fixed gate, it will not be considered further here.
(and essential) function of the transit-time curve is
quality control (Section 15-4.4.2.9). As long as the ampli- 15-4.4.2.6 CBL attenuation rate
tude is well above the threshold level (usually on the Many parameters have an influence on the amplitude
order of 5 mV), the transit-time curve is a highly sensi- measurement: calibration of the logging tool, centering
tive measure of tool centering in the pipe. of the tool, pressure, temperature, wellbore fluid, casing
The second quantitative measurement is the wave size and thickness, cement thickness, microannulus
amplitude from which a quantitative cement evaluation presence and size, and the spacing between transmitter
can be made. Because the arrival time of the peaks is and receiver. To reduce the measurement sensitivity and
related both to the geometry of the CBL tool and the quantify the results as a function of the cement, it is nec-
casing and to the wellbore fluid properties, the ampli- essary to speak in terms of the attenuation rate.
tude measurement can be made in two ways—using a A theoretical expression of the CBL attenuation rate
fixed gate or a sliding gate. The technique commonly was first given by Pardue et al. in 1963. It showed that,
employed today is to position a fixed gate (or time inter- for a thin, flat plate surrounded by a solid or a liquid, the
val) during which the wave amplitude is measured, attenuation rate can be approximated by the following
preferably an E1 half-cycle. For example, as shown in equation.
Fig. 15-20, the time interval is held constant relative to
the moment the transmitter is fired. To ensure precise ⎛ρ ⎞
placement of this time window over the desired part of 52.2 ⎜ 2 ⎟
⎝ ρ1 ⎠
the wave, the gate is set in the well to be logged using the α= ,
⎧⎡ ⎡⎛ v ⎞ 2 ⎤ ⎫⎪
least-bonded pipe to be found. Note, however, that the 0.5 0.5
⎪ ⎛ v pl ⎞
2 ⎤
h ⎨ ⎢⎜ ⎥ ⎢ pl − 1 ⎥ ⎬
gate must be reset for changes in pipe size and even fine-
⎢ ⎟ − 1⎥ + ⎢⎜ ⎟ ⎥ ⎪
tuned for changes in pipe weight. The correct setting of ⎪ ⎣⎝ vP ⎠ ⎦ ⎣⎝ vS ⎠ ⎦ ⎭
the gate is a crucial step in obtaining a valid bond-log ⎩
amplitude curve. (15-6)
where
Detection level h = thickness of plate (in.)
E3 vP = compressional wave velocity in annular material
Amplitude E1 vpl = wave velocity in plate
vS = shear wave velocity in annular material
E1 amplitude α = attenuation rate (dB/ft)
ρ1 = density of plate
t0 Time ρ2 = density of annular material.
Δt This formula shows the following.
E2 ■ The primary cement variable affecting attenuation is
the wave velocity.
Fixed
gate ■ The attenuation is inversely proportional to the
casing-wall thickness and proportional to the cement
Fig. 15-20. Fixed amplitude gate. density.
■ The attenuation is independent of the frequency.

■ Shear coupling is required at the casing-cement

When a sliding gate is used, the CBL curve shows the interface to produce full attenuation.
maximum amplitude measured within a time interval
Equation 15-6 ignores the effects of plate curvature,
positioned immediately after the detection of a signal
the fluid within the pipe, and the thickness of the
larger than a preset detection level. The measurement is
cement sheath.
quantitatively useful only when that first half-cycle is

568 Well Cementing

 The CBL attenuation rate can also be calculated from
two amplitude measurements of the same transmitted
signal made by two different receivers, according to the
following formulas.
Transmitter 1
20 (E )
α=– log 10 1 2 , (15-7)
L ( E1 )1 2.4 Receiver 1
ft
where 3.4 Mode 2
ft VDL
(E1)1 = amplitude of E1 at Receiver 1 Receiver 2
(E1)2 = amplitude of E1 at Receiver 2 5.8 Mode 1
L = distance (spacing) between Receivers 1 and 2 (ft). ft
Mode 3
Receiver 3
In the case of tools with a single receiver, an approxi- 3.4
mate attenuation rate can still be computed using the ft Mode 4
following expression. 2.4
ft
20 E1
α=– log 10 , (15-8)
L ( E1 ) fp Transmitter 2

where
E1 = amplitude of E1
(E1)fp = amplitude of E1 in free pipe
L = spacing between transmitter and receiver (ft).
This last expression, although still sensitive to tool
calibration and wellbore parameters (it requires the Fig. 15-21. Special CBL sonde with two transmitters and two
free-pipe reading in millivolts), can be used. receivers
To improve the accuracy of the attenuation-rate
determination, special CBL sondes have been developed
with two transmitters and two or three receivers L1 = 2.4 ft
(Fig. 15-21). The general principle is the same as a con- L2 = 3.4 ft
ventional CBL, but the design allows computing a bot- pi = acoustic pressure at transmitter i
tomhole-compensated attenuation rate on which many Sj = sensitivity of receiver i
environmental effects (e.g., pressure, temperature, well- αBHC = bottomhole compensated attenuation rate
bore fluid characteristics) have almost no effect with pi and Sj temperature- and pressure-dependent.
(Gollwitzer and Masson, 1982).
Four amplitude measurements are required. The 15-4.4.2.7 CBL bond index
upper and lower transmitters are separated by 5.8 ft
[1.8 m], and the three receivers are placed at 0.8 Early experiments (Grosmangin et al., 1961) found that
(short), 2.4 (near), and 3.4 ft (far) [0.2, 0.7, and 1.0 m, the attenuation rate is linearly related to the percentage
respectively] from the upper transmitter. The measure- of the casing circumference bonded by the cement
ments made at 2.4- and 3.4-ft spacing are used to com- (Fig. 15-22). These experiments were made with 5.5-in.
pute the bottomhole-compensated attenuation rate, diameter casings and a 25-kHz source. The bond index
which is quasi-independent of temperature and pres- (BI) was derived from this concept (Pardue et al., 1963).
sure. Its validity was later extended by Jutten and Parcevaux
(1987) to the percentage of bonded cemented area,
−10 A ×A regardless of the shape of the noncemented area and the
α BHC = log 10 12 21 (15-9) nature of the fluid behind the casing. The Jutten and
L2 − L1 A11 × A22
Parcevaux experiments used a 4.5-in. casing and 17-kHz
where transducers. The BI, symbolized by Ibond, requires knowl-
A = amplitude edge of the log response in the well-cemented section,
⎛ αh j ⎞
which is used as a reference for the computation of
−⎜ ⎟ the attenuation corresponding to 100% cementation
⎝ 20 ⎠
Aij = pi S j 10 (α100%cem).

Chapter 15 Cement Job Evaluation 569

 I bond =
)
α ( x − α fp
, (15-10)
1
1.5-ft receiver
α100% cem − α fp 0.8
Measured

where α is the CBL attenuation rate, and α fp is the CBL 0.6

3-ft receiver
attenuation rate with free pipe. Bond
index
0.4

100
Vertical channels 0.2
90
80 BPI
0
70 0 20 40 60 80 100
CBL 60 Helical channels
E1 (mV)
attenuation 50
rate Fig. 15-23. Ibond computation (from Gai and Lockyear, 1992).
(%) 40 Grosmangin et al., 1961 Reprinted with permission of SPE.
30
20
10 Series Model Parallel Model
0
0 10 20 30 40 50 60 70 80 90 100
Bonding (%)
Bonded
Fig. 15-22. Relationship between percent bonding and CBL attenu- cement
ation rate.

Unbonded
cement
In 1992, Gai and Lockyear claimed that the conven- or fluid
tional BI calculation does not have correct theoretical
grounds or physical intuition. Based on theory, labora-
tory experiments (in 7-in. casing) and case studies,
these authors proposed that the correlation between the
percentage bonding (BPI, symbolized by I%bond) and the Acoustic wave Acoustic wave
CBL amplitude (E1 peak) is Fig. 15-24. Serial and parallel models for computing the BI.
E fp − E meas
I% bond = (15-11)
E fp − E 100% cem
Figure 15-25 shows results of recent unpublished
where modeling work that takes wave-propagation effects into
account. The model computes the amplitude at the 3-ft
E100%cem = CBL amplitude corresponding to 100%
receiver and the 2.4- to 3.4-ft attenuation rate for a given
cementation
cement coverage. The results showed that the BPI rela-
Efp = CBL amplitude with free pipe
tionship overestimates the true cement coverage, and
Emeas = measured CBL amplitude. the BI equation (from both the CBL amplitude and
Their results are graphically presented in Fig. 15-23. attenuation) slightly underestimates the cement cover-
If the acoustic wave energy is assumed to travel in age. The same study found that the tiny precursor before
straight lines parallel to the axis of the pipe, the Pardue the conventionally used E1 arrival follows the BPI rela-
equation (Eq. 15-10) represents the unbonded and tionship more closely than the BI relationship.
bonded sections as sequential or serial paths for The discrepancies in the results of different experi-
the acoustic wave, while the Gai and Lockyear equation mental and theoretical BI values are probably due in
(Eq. 15-11) represents them as parallel paths part to the different test conditions—different trans-
(Fig. 15-24). The former represents measured attenua- ducer frequencies, transducer pulse shapes, transducer
tion as a weighted sum of bonded and unbonded attenu- polarities, receiver spacings, casing diameters, and
ations; the latter represents measured amplitude as a casing thicknesses. No systematic experimental BI stud-
sum of bonded and unbonded amplitudes. ies have been reported over a range of casing sizes using

570 Well Cementing

 1 Casing Size (mm)
194
140 273
0.8 116 178 340 Acoustic
BPI dB/m
Impedance of
0.6 Cement
Bond 1 (MRayl)
BI (CBL) 70 4
index
0.4 2
BI (special 50 8 8
CBL sonde) 3
0.2 40
12
30 4 7
0
0 0.2 0.4 0.6 0.8 1 16
20 5 Casing
Cement coverage Thickness
6 6
15 20 (mm) (in.)
Fig. 15-25. Comparison of BI calculations from modeling results in 15 0.6
10 7
5.5-in. casing (from unpublished Schlumberger results). 9 24
0.5
8
8
7 5
6 28 10 0.4
9
a commercial tool in well-controlled conditions. In view CBL
5

of the uncertainties surrounding the BI calculation, it amplitude 4

10
32 8
0.3
should be treated as a semiquantitative interpretation (mV) 3
7
4
tool valid for fluid channels in well-bonded cement. 36
11
6
In 1963, Pardue et al. studied how the attenuation 2 12
40
rate varies as a function of cement composition, casing 5 0.2
size and thickness. They constructed a famous nomo- 13
44
gram known as the “CBL Interpretation Chart.” This 1
3
14
nomogram, which shows the correlation between the 48
CBL signal and the cement compressive strength, was 15
later modified for lightweight cements (Bruckdorfer et 0.5 52
al., 1983). In 1987, Jutten and Parcevaux showed that, 16

over a large variety of cement-slurry formulations, the 17 56 2

CBL attenuation rate is related to the acoustic imped-
ance of the set cement, not the compressive strength. 0.2
18

Provided the cement is set, these experimental results, 41⁄2 7 133⁄8

103⁄4
also valid for foamed cements, could be used to modify 51⁄2 75⁄8
the CBL interpretation chart that now shows a relation- Casing size (in.)
ship between cement acoustic impedance (Z) and the
CBL signal (Fig. 15-26). Fig. 15-26. CBL interpretation chart (3-ft spacing).
Unfortunately, a CBL-VDL interpretation does not
give direct access to hydraulic isolation information.
From local experience, empirical charts have been to make the best possible evaluation from a CBL. Well
developed that give the minimum length required for parameters and cement-job events are necessary to com-
zonal isolation as a function of the BI and the casing pute the expected log response, while the cement job
size. These charts, established without any theoretical recording and pre- and postjob events are required most
or experimental background, should be used with of the time to understand if the discrepancies between
extreme caution. the expected response and the actual log, if any, are
The CBL-VDL interpretation is thus restricted to an caused by the cement, incomplete mud displacement, or
evaluation of the cement placement, in relationship to a microannulus. Figure 15-27 shows a flowchart for CBL
the quality of the cement. Well parameters, cement-job interpretation that covers the points discussed above.
events, and pre- and postjob well histories are required

Chapter 15 Cement Job Evaluation 571

 CBL
quality-control
approved

No Yes
Like caliper

No Correlates Yes
formation

No Job execution Yes

problem

No Postjob Yes
problem

Undetermined Decide whether to Density problem? Check for: Most probably:

problem squeeze Displacement – Fast formation incomplete mud
Check BI What are the problem? – Porous formation displacement
stresses? Check BI – Local contamination Check BI
Is there a – Local interactions
microannulus?

Fig. 15-27. CBL interpretation—general flowchart.

15-4.4.2.8 Bond-log presentation formats sit time is 200 to 400 μs. This has the advantage of a
Bond logs are presented on a standard three-track log single scale for almost all casing sizes. However, the
format with the depth track between Data Tracks 1 and small changes in time that correspond to major eccen-
2 (Fig. 15-28). Track 3 contains the full-wave display, tering (4 to 5 μs are recommended limits) require a
either as the waveforms or a variable-intensity display. more sensitive scale (100-μs width).
The common scale is 200 to 1,200 μs, although other Track 2 contains the amplitude curve, attenuation-
scales are available for special cases such as extra-large rate curve, or both. The attenuation-rate curve is usually
holes or extra-slow formation sound speeds. presented on a scale of 0-to-20 dB/ft. Amplitude-curve
Track 1 traditionally contains the transit-time mea- scales are not standardized, although a 0-to-100 or 0-
surement (or possibly a derivative thereof for the bore- to-50 mV scale is very common, with an amplified curve
hole-compensated tools), as well as a correlation curve presented on a 0-to-20 or 0-to-10 mV scale. The double
(gamma ray or neutron). Casing collars are usually indi- scale is very important because free-pipe readings can
cated here but may also appear in the depth track or approach 100 mV or more, while fine resolution of per-
Track 2. A common scale for the conventional 3-ft tran- haps 1 mV or less may be required at low amplitudes.

572 Well Cementing

 PIP Summary
Casing collars
Cement isolation marker
Time mark every 60 s
Transit Time (Sliding Gate)
400 (µs) 200
Transit Time
400 (µs) 200
Gamma Ray
0 (gAPI) 150
Tension Min. Amplitude Max.
CCL (lbf) CBL Amplitude
–1 (....) 1 0 (mV) 50 Variable Density Log
2,000 0 200 (µs) 1,200

325

Fig. 15-28. Standard three-track bond-log format.

Chapter 15 Cement Job Evaluation 573

15-4.4.2.9 Quality control of CBL/VDL logs recommendation is reasonable. There are several rami-
Quality control can easily be divided into a step-by-step fications of the 4-μs limit.
procedure (Fig. 15-29). If the first two steps apply for ■ The measurement resolution of the logging instru-
every log, the CBL has a special curve for quality con- ment must be 1 μs or better.
trol—the transit-time curve (Fitzgerald et al., 1983; ■ The log display must allow for ease of readability
Bigelow, 1985). If no transit-time curve is present on the down to a 1- or 2-μs visual resolution, leading to the
log, no quality control of the log is possible, and the eval- recommendation of using a 100-μs scale for the tran-
uation will be very limited. The next part of this chapter sit time across the track (Fertl et al., 1974; Fitzgerald
will assume that a transit-time curve is present on the et al., 1983). In the case of sonde eccentering, it is
log. impossible to quantify the results of the cement job
By comparing the measured transit time with the with the CBL. The influence of a fast formation, often
expected transit time (the time required for the sound also seen as a decrease in transit time, is discussed in
wave to travel from the transmitter to the receiver Section 15-4.4.2.13.
through the mud and along the casing), the following
conclusions can be drawn. Slightly longer transit time (stretching) is generally
Shorter transit time is an indication of either poor an indication of a good bond and should correspond to
sonde centralization or a fast formation. On the subject reasonably low amplitudes (Fig. 15-31). The BI concept
of eccentering, there have been recommendations that a is applicable. If the CBL amplitude is still high, check for
4-μs transit-time decrease or less is acceptable. This cor- interference caused by small cement-sheath thickness
responds to an eccentering of about 1⁄8 in. [0.32 cm] in and high acoustic-impedance contrast at the cement/
fresh water. Because this amount of eccentering reduces formation interface.
the amplitude by more than 25% (Fig. 15-30), this

CBL/VDL
quality-control

No Good Yes
repeat
section

No Calibration Yes
summary

No Transit time Yes

(TT) curve

No Transit time ≥ Yes

minimum TT

No Yes
Environment problem

No log
Log of Log of Valuable
quality control
limited value limited value data on log
Bad log

Fig. 15-29. Quality-control flowchart.

574 Well Cementing

 Eccentering
1.0

0.8

0.6

0.4

0.2

1 3⁄4 1⁄2 1⁄4 0 1⁄4 1⁄2 3⁄4 1

Inches Off Center

Transit Time
400.00 (µs) 200.00 3-ft Receiver 5-ft Receiver
CCL Amplitude
–19.00 (µs) 1.0000 0.0 (mV) 1.0000
Gamma Ray Amplitude VDL
0.00 (µs) 100.00 0.00 (mV) 100.00 200.00 (µs) 1,200

Expected Measured Measured Expected amplitude

transit time transit time amplitude

Fig. 15-30. Effects of eccentering on the CBL (recorded in 75⁄8-in. casing).

E1 Free pipe E1 E3 Free pipe

signal signal
t0 Threshold t0 Threshold

t t’ t t’
Very
Good bond good
Δt bond
signal E2
Fig. 15-31. Stretching effect. Fig. 15-32. Cycle skipping associated with good cement bond.
Transit time is much longer than expected (>15 µs).

Longer transit times (Δt > 15 μs) are called skips fairly common to have stable skips of more than 20 μs
(Fig. 15-32). In this case, E1 is normally too small to be but less than 50 μs. This is caused by energy reflections
detected; thus, a good bond exists between the cement at the cement/formation interface, enhanced by large
and casing. A cycle skip refers to a cycle of the original acoustic impedance contrasts as in concentric strings
wave (50 μs for a 20-kHz signal). In this case, in fixed- (Jutten and Corrigall, 1988). If applied in this situation,
gate mode, the CBL amplitude must be below the detec- the BI concept will lead to erroneous conclusions,
tion level, and the BI concept applies. However, it is because the amplitude measured was not E1.

Chapter 15 Cement Job Evaluation 575

15-4.4.2.10 Influence of well parameters on the CBL Casing size and thickness—As the casing size
Temperature and pressure—The deeper a well, the increases, so does the path through the wellbore fluid.
higher the temperature and pressure will be. In addition This leads to some signal attenuation and decreased
to the pipe, cement, and formation, the velocity and free-pipe amplitudes. However, for pipes cemented with
attenuation of sound through wellbore fluids will be the same cement system, the CBL amplitude typically
affected by downhole conditions (Fig. 15-33). The increases with casing size. This can easily be explained
response of the transducers also will vary. In 1984, by the increase of steel thickness, providing a smaller
Nayfeh et al. published pressure and temperature cor- attenuation rate.
rections for the transducers used in CBL tools. Cement thickness—When the cement thickness is
too small, energy reflections at the cement/formation
interface can interfere with the casing signal. These
1,560 interferences are seen mainly in concentric strings,
1,550
cylindrical holes with a small annular clearance, and
well-centralized pipe (Jutten and Parcevaux, 1987). To
1,540
determine if reflections interfere with E1, it is necessary
1,530
to accurately measure the openhole size and the
Sound 1,520
velocity
acoustic properties of the cement when the log is run.
1,510 Typical cement thicknesses that do not interfere with
(m/s)
1,500 E1 (20 kHz signal) can vary from 1 in. [2.54 cm] to 3 in.
1,490 [7.6 cm], depending on the compressional-wave velocity
1,470 through the cement. If the measuring gate is much
1,460 larger than half of the period of the original signal, inter-
0 20 40 60 80 100 120 140 160 ference may induce erroneous amplitude measurements
Temperature (°C) for even larger cement thicknesses.
In the special case of concentric strings (e.g., top of
Fig. 15-33. Effect of temperature on velocity of sound through water the liner), the resonance of the external casing induces
(3,200 psi). strong signal perturbations, leading to an apparent fre-
quency increase of the first few arches of the waveform.
Experiments confirmed with field logs (Jutten and
Wellbore-fluid properties—Wellbore-fluid properties Corrigall, 1988) proved that high CBL amplitudes
have an effect on both the transit time and the CBL obtained in concentric casings are often artifacts caused
amplitude. Nayfeh et al. (1984) performed experimental by an excellent cement job between both casings and a
and numerical work to study the effects of wellbore measuring gate with an excessive amplitude. Some tools
fluids on CBL-amplitude measurements. There are con- allow solving this problem by shortening the measuring-
siderable differences in free-pipe amplitude between gate width. In the log shown in Fig. 15-35, the change
various brines (Fig. 15-34). from concentric casing to open hole is clearly shown by
the frequency change in the VDL. In addition, the ampli-
tude curve in the double-string section seems to indicate
Clear water almost no cement, while an ultrasonic log would actually
120 CaCl2 11.5 lbm/gal show cement and no difference between the double-
110 CaCl2 10.5 lbm/gal
CaCl2 9.5 lbm/gal casing zone and the single-casing zone.
Free 100
pipe 90
amplitude
(mV) 80
70
60
50
40
4 5 6 7 8 9 10 11 12
Casing OD (in.)

Fig. 15-34. CBL free-pipe amplitude for different brines as a function

of casing OD.

576 Well Cementing

 CBL Amplitude
0 (mV) 50
Min. Amplitude Max.
CBL Amplitude Variable Density
0 (mV) 10 200 (µs) 1,200

3,025
Signal frequency
increased: larger
number of thinner
stripes

Double
string

3,050 Cased and

cemented
open hole

Cased and
cemented
open hole

3,175

Fig. 15-35. CBL/VDL in concentric casing.

Chapter 15 Cement Job Evaluation 577

 Fast formations—The well-known “fast-formation 15-4.4.2.11 Influence of cement job parameters on the CBL
effect” is a transit-time decrease. When the formations The most common cause of cement-job failure is poor
have higher velocities than the casing (slowness less mud removal, and poor mud removal will lead to a com-
than 57 μs/ft), the refracted formation compressional plicated CBL. Some causes of poor mud removal are
waves may arrive before the casing wave. In this case, obvious: casing not centralized, cement-slurry density
the transit time is shorter than expected, and the ampli- lower than that of the drilling mud, and low-viscosity
tude characterizes the formation instead of the cement slurry pumped behind a viscous drilling fluid.
casing/cement bond. Therefore, the BI concept cannot Others are more complex and require the use of
be applied in this case. Qualitative evaluation is possi- advanced cement-placement simulators such as those
ble. If there is sufficient sound energy propagating described in Chapter 5. Slurry-placement evaluation
through the formation to interfere with the early part of should be performed before a cement log analysis. This
the waveform, it indicates the presence of some solid requires actual job data (flow rate, pressure, and den-
material between the casing and the formation. sity) during the entire job and other information, such as
In some cases, the transit time does not reflect the openhole size and casing centralization.
fast formation while the VDL clearly shows a formation Because slurries of different densities normally have
arrival before the casing arrival. At the 5-ft [1.5-m] different acoustic properties, it should be easy to detect
receiver, the formation signal has clearly overtaken the the transition between the lead and tail slurries on the
casing signal. There could be a “fast-formation” phe- log. This concept is of prime importance for cement cov-
nomenon at the 5-ft receiver, but not at the 3-ft [0.9-m] erage estimation, because 100% bonding across the lead
receiver. It depends on the annular thickness, the speed section corresponds to a much lower CBL attenuation
of sound in the cement, and the difference in the speed rate than 100% bonding across the tail section. This is
of sound between the rock and the casing. It is impor- mainly caused by the difference in acoustic impedance,
tant to be aware of this subtlety, because the formation directly linked to density. A BI log should be computed
signal can be present in the fixed gate and thereby section by section, without forgetting that the minimum
increase the amplitude considerably while the transit cement thickness required to apply these rules also
time is reading a larger value than in free pipe. An erro- depends on the acoustic properties of the cement.
neous interpretation will result. Sometimes it is possible to estimate mud removal by
For example, given a bias-level threshold of 5 mV and comparing the expected top of the cement with the one
an amplitude of 1 mV in a “slow formation,” both quite computed from the hole geometry and volumes pumped.
realistic values, the earliest formation arrival in an adja- However, such an estimation must be done quite care-
cent fast formation could drive the amplitude reading on fully, because many parameters are involved, including
the log up to as much as 5 mV without causing a transit- the accuracy of the caliper and flowmeters as well as
time decrease. However, standard interpretation tech- volume changes caused by fluid loss and lost circulation.
niques (BI) would indicate a channel covering 35% of Cement slurries prepared at lower-than-designed
the circumference of the pipe. One must not make the densities often exhibit more free water and sedimenta-
common mistake of assuming that, because the transit tion, longer thickening times, and lower acoustic imped-
time has not decreased, the amplitude is uncontami- ance. This can also be seen on the log.
nated by a fast formation.
On some borehole-compensated bond-logging tools, 15-4.4.2.12 Influence of post-job events on the CBL
an additional spacing of 0.8 ft [0.2 m] exists to minimize Several post-job events can influence the CBL results.
the effects of fast formations in casing sizes smaller than Any pressure and temperature change applied inside the
7 in. [18 cm]. When the formation has no influence on casing will induce casing deformations that modify the
the measured peaks, the attenuation rate should be con- stresses in the cement and at the cement-to-formation
stant at all spacings. In the presence of a fast formation, and cement-to-casing interfaces. Such deformations can
the measured attenuation rate decreases with increas- break these bonds, leading to the creation of a microan-
ing spacing, because the increasing part of the sound nulus.
energy arrives in the early portion of the waveform. In
this case, the 0.8-ft [0.2-m] spacing attenuation rate is
larger than other attenuation rates (or the amplitude is
lower).

578 Well Cementing

 Work by Leslie et al. (1987) showed that the ampli- and Evans (1964) detailed the diameter expansion of
tude reduction of the sonic signal depends not only on unsupported pipe when pressurized. Often, 1,000-psi dif-
the attenuation along the casing, but also on the effi- ferential pressure is sufficient to create a fairly large
ciency of the acoustic coupling between the transducers microannulus, especially for large casing sizes.
and the casing wave. In the presence of a microannulus,
the shear coupling is lost and the attenuation along the Mechanical fatigue
casing is negligible; however, the coupling is not lost In deviated wells and on intermediate strings, drilling
when a fluid is in the microannulus. Using multiple- can produce a great amount of vibration and mechanical
receiver tools, it is possible in theory to separate the cou- stress, concentrated in special places (e.g., kickoff
pling and attenuation rate, detect a microannulus, and points). These can damage the quality of the bond
even quantify the cement coverage behind the pipe. between the casing and cement.
In the absence of experimental work, knowledge of For all of the cases mentioned above, if the cement is
microannuli is based on field experience and common strong enough to withstand the deformation, the bond
sense. Pilkington (1988) described the origin of will not be affected. If the cement is still plastic when
microannuli in great detail, with the effects produced on the stress is applied, the annular geometry will change.
the CBL. Guidelines were also given for running a CBL in If the cement is not set when the stress is released, the
the presence of a microannulus. However, for cement bond should not be affected; however, if the cement
job evaluation purposes, a CBL performed under pres- hardens while the casing is significantly expanded, it
sure may be contraindicated because of the potential may not follow the casing back when the stress is
detrimental effects on hydraulic isolation. One should released, leading to the formation of a microannulus.
first try to analyze the origin of the potential microan-
nulus. Several cases should be taken into consideration. 15-4.4.2.13 CBL/VDL examples
Well cemented section—A 7-in. [18-cm] casing (23 lbm /ft)
Thermal expansion or contraction
was cemented at a shallow depth. The average hole size
As cement sets, heat is generated, which increases the was between 12 and 17 in. [30 and 43 cm] for a bit size
temperature in the wellbore. As explained at the begin- of 9 7⁄8 in. [23.7 cm]. The casing was cemented using two
ning of this chapter, this heat is sometimes measured to different slurries—a lead mixed at 10.6 lbm/gal
detect the top of the cement. It will also cause expansion [1,270 kg/m3] and extended with hollow silica micros-
of the casing inside the wellbore. An approximate value pheres and a tail mixed at 15.8 lbm/gal [1,890 kg/m3].
is given by the following formula. Both formulations contained 35% silica flour BWOC.

( )
The maximum displacement rate during this job was
Δ d = 6.9 × 106 × Ccsg × Δ T (15-12) 2 bbl/min, achieving good mud removal.
The CBL was run several weeks after the job. The
where
selected section shows the transition between the tail
Ccsg = casing circumference (in.) and the lead slurry (Fig. 15-36). The CBL amplitude is
Δd = diameter change (in.) around 1 mV between 370 and 420 ft [113 and 128 m]
ΔT = temperature change (°F). across the tail and between 8 and 14 mV across the lead.
During the life of the well, the production of hot fluids At that time, the estimated compressive strengths were
or the injection of cold or hot fluids can also produce in excess of 5,000 psi for the tail and about 1,000 psi for
expansion or contraction of the tubular goods. The above the lead cement. As determined by using the CBL inter-
formula can be used to estimate the magnitude of the pretation chart, the CBL amplitudes were expected to be
induced geometrical change. less than 1 mV for the tail and around 4.0 mV for the
lead, which would give a pessimistic BI of 65% on the
Mechanical expansion or contraction section showing a 12-mV CBL amplitude. When using a
Mechanical effects may result from internal casing pres- modified CBL chart, with measured acoustic imped-
sure applied during pressure tests, remedial cementing, ances of 6.0 MRayl for the tail and 3.2 MRayl for the lead,
or stimulation jobs. Sometimes the casing is maintained attenuation rates were computed and extrapolated to be
under pressure while the cement sets, because of a float- about 1 mV for the tail and 8 mV for the lead. The CBL
equipment leak. After the cementing of a production interpretation is similar for the tail cement across the
string, it is also fairly common to replace the drilling bottom section. However, the discrepancy becomes crit-
mud by a lower-density completion fluid. The hydrostatic ical for the lead, because the relationship between the
pressure reduction can produce a significant casing con- CBL attenuation rate and cement acoustic impedance
traction and induce a microannulus if the bond between enables one to compute a more realistic BI of 85%.
the casing and cement is not sufficiently strong. Carter

Chapter 15 Cement Job Evaluation 579

 Fast formation—The log example in Fig. 15-37 shows
both a short interval of free pipe from uphole and
another portion of the well. The VDL display contains
much information. The free-pipe VDL character is dis-
tinct: very straight parallel bands and chevron-shaped
X300
diffraction patterns at the collars. In contrast, the rest
shows wavy bands. The pattern of the bands corresponds
to changes in the rock, as indicated by comparing the
VDL data in Track 3 to the gamma ray curve in Track 1.
These are formation signals that interfere with the
attenuated casing arrival, preventing the performance of
a quantitative evaluation. Nevertheless, the presence of
these formation signals indicates that some solid mater-
ial fills the annulus.
The amplitude curve is presented in Track 2. In the
short free-pipe interval at the top, the amplitude is high
and steady at about 60 mV, with a “kick” to the left at the
collar. Below, the amplitude varies widely—from 2 mV to
30 mV. The high amplitudes are caused by a fast forma-
tion, not by a lack of cement. This is confirmed by the
VDL display: the first red band in the fast-formation
interval is farther left than the first red band in the free-
pipe interval. The first formation arrival occurs earlier
than the first pipe arrival in the fast formation. The tran-
X400 sit-time curves track the fast formation very well. The
sliding-gate transit time decreases in front of the fast
formation.
The important point is that the amplitude measure-
ment in a fast formation is meaningless. The amplitude
is some part of the formation signal, not the first positive
peak, E1, of the free-pipe signal.
Change in casing weight—In the log shown in
Fig. 15-38, there is an abrupt change in casing weight at
6,953 ft. The 5 1⁄2-in. [14-cm] casing is 17 lbm/ft above the
change and 23 lbm/ft below. Note the difference in CBL
amplitude between the two weights—2 mV in the lighter
casing and 5 mV in the heavier one. If a BI is run in this
well, the reference point for 100% cemented must be
changed for each pipe weight.

Transit Time CBL

400.00 (µs) 200.00 0.0 (mV) 100.00
Gamma Ray CBL
0.0 (gAPI) 150.00 0.0 (mV) 20.00

Fig. 15-36. CBL/VDL section across slurry change (10.6-lbm/gal

slurry above 370 ft; 15.8-lbm/gal slurry below 370 ft).

580 Well Cementing

 Transit Time
400 (µs) 200
Transit Time (Sliding Gate) Fluid-Compensated CBL Amplitude
400 (µs) 200 Tension 0 (mV) 100 Min. Amplitude Max.
(lbf)
Spectroscopy Gamma Ray Fluid-Compensated CBL Amplitude Variable Density Log
0 (gAPI) 150 2,000 0 (mV) 20 200 (µs) 1,200
5,000

Casing
collar

Free pipe

Chevron-shaped
patterns

2,525

Fast formations

Fast formations

2,550

Fast formations

Fig. 15-37. CBL/VDL example showing the effect of a fast formation.

Chapter 15 Cement Job Evaluation 581

 Transit Time Azimuth of
400.00 (µs) 200.00
Eccentricity CCL
CCL CBL
–19.00 (µs) 1.0000 0.0 (mV) 20.000 0 (°) 360 –35 (----) 5
Gamma Ray CBL VDL
0.0 (gAPI) 150.00 0.0 (mV) 100.00 0.0 (µs) 100.00 Tension Transit Time
10,000 (lbf) 0 400 (µs) 200
Fluid-Compensated CBL Min. Amplitude Max.
Eccentricity Amplitude
0 (in.) 0.5 0 (mV) 50 Sonic VDL Curve
200 (µs) 1,200

6,900

15,900
7,000

Fig. 15-38. CBL/VDL examples showing abrupt change in casing

weight.

Microannulus—The two log sections shown in Fig. 15-39. CBL/VDL example, showing effect of a microannulus.
Fig. 15-39 demonstrate the effect of a microannulus on a
CBL. The first section was logged without additional
pressure at the surface. either disappeared or is significantly reduced, and the
■ The pipe signals are visible in the VDL display as amplitudes have decreased to much lower values. In this
straight parallel bands at the earliest time. size and weight of casing, the expansion caused by the
■ The formation signals appear later in the VDL as wavy
1,500-psi increase is about 0.0008 in. [20 μm] of radius.
bands.
■ The amplitude is erratic at moderate values.

The second log is a pass over the same interval with

1,500 psi applied at the surface (Fig. 15-40). The color
contrast on the VDL display for early pipe arrivals has

582 Well Cementing

 Azimuth of The alternative is to use ultrasonic tools to evaluate
Eccentricity CCL the cement job. These tools, described in the next sec-
0 (°) 360 –35 (----) 5
tion, can differentiate between channeling and a
Tension Transit Time microannulus without applying internal casing pressure.
10,000 (lbf) 0 400 (µs) 200
Fluid-Compensated CBL Min. Amplitude Max.
Eccentricity Amplitude
15-4.4.2.14 Pad-type sonic tools for cement evaluation
0 (in.) 0.5 0 (mV) 50
200
Sonic VDL Curve
(µs) 1,200
This category of sonic cement evaluation tools features
multiple transmitters and receivers mounted on pads
that are in contact with the casing wall. An example of
this tool is described by Lester (1989). Lester’s device
has six pads, each containing one receiver and one trans-
mitter (Fig. 15-41).
The attenuation measurement is very similar to the
CBL attenuation rate measurement described in Section
15-4.4.2.6 (Fig. 15-42). As mentioned earlier, compen-
sated attenuation measurements do not require calibra-
tion and are independent of temperature and pressure.

Pads 1 2 3 4 5 6

T T T
15,900
R R R
Segments 6 1 2 3 4 5

R R R

T T T

Fig. 15-41. Unfolded view of the six-pad configuration (from Lester,

1989). Reprinted courtesy of the Canadian Well Logging Society.

T1
Fig. 15-40. CBL/VDL example showing effect of pressurized casing.
A12

Limitations R2
A13
Unfortunately, with traditional cement-bond logs, high
amplitude over a cemented section may be caused by A42
R3
either channeling or to a microannulus. In both cases,
the VDL will show strong casing signals (parallel stripes) A43
and weak formation arrivals. The only way to differenti-
ate both cases is to run a CBL under internal casing T4
pressure. If there is a microannulus, there will be a sig-
nificant amplitude reduction. If the CBL does not
improve when pressure is applied, it can be either a ⎡ ( A12 × A 43) ⎤
Attenuation = 10 log ⎢ ⎥
large microannulus or a channel. Zonal isolation has ⎣ ( A13 × A 42 ) ⎦
probably not been achieved.
Fig. 15-42. Attenuation-compensated measurement in pad-type tool
(from Lester, 1989). Reprinted courtesy of the Canadian Well
Logging Society.

Chapter 15 Cement Job Evaluation 583

 Multiple-pad transducers are beneficial in that they indicates that no cement is present. Black indicates an
provide azimuthal information (60° coverage in this 80% BI. The conventional CBL display is shown as well.
case) and have very low sensitivity to logging-fluid The amplitude curve is generated in the following
effects. While tool eccentricity can have a large effect on manner. Six measured attenuations values are averaged
conventional CBLs, the pad tool is much less sensitive. around the borehole and then averaged over 3 ft of
The distance between the transmitter and receiver is depth. This average attenuation is equivalent to the
about 6 in., and the operating frequency is 100 kHz. attenuation that would have been measured by a per-
Because of the short distance between transmitters and fectly centered standard CBL tool operating in an ideal
receivers (6 in. versus 36 in. for CBL), this tool is less fluid. From this attenuation, the amplitude that would
affected by fast formations. Although the wavelength is have been measured is calculated. Both the amplitude
five times shorter than the conventional CBL, attenua- and the attenuation are presented. A VDL is also dis-
tions measured by both techniques are very close. played, thanks to a transducer-receiver spacing of 5 ft.
The log presentation format presents the six mea-
sured attenuations with a tool orientation trace overlay
(Fig. 15-43). A cement map is also presented. The white

Gamma Ray
0 (gAPI) 100
Min. Delta-T
140 (µs/ft) 40
Max. Delta-T Min. Attenuation Cement Map SBT Wave
140 (µs/ft) 40 20 (dB/ft) 0 0 360 200 1,200
Diff. Tension (ten)
2,400 100 Depth Relative Bearing Average Attenuation
(lbf) (ft) 0 (°) 360 20 (dB/ft) 0 20 80 2 25

Good
X600

Partial

Free X700
pipe

Partial

Good

Fig. 15-43. Log display from multiple-pad cement evaluation tool (image courtesy of Baker Atlas).

584 Well Cementing

15-4.5 Ultrasonic pulse-echo cement evaluation proposed a statistical interpretation method for the USI
tool that combined conventional and statistical process-
15-4.5.1 History ing to discriminate between cement with gas (gas-cut
Ultrasonic pulse-echo technology was first applied to cement) and fluid channels. An API report, Technical
well logging in the 1960s, when Zemanek and Caldwell Report on Cement Sheath Evaluation (TR 10TR1, pub-
(1969) designed a tool called the Borehole Televiewer lished in 1996), provided many examples of tool
for imaging the borehole wall. Ultrasonic tools operate at responses to materials in the annulus and stressed the
much higher sonic frequencies than acoustic tools, usu- usefulness of material “signatures” and statistical meth-
ally between 200 and 700 kHz. ods. This report made many useful recommendations for
Havira (1979; 1982) investigated ultrasonic methods running and interpreting acoustic logs. Goodwin (1989)
to evaluate cement behind pipe and developed a tool made suggestions similar to those in the API report. The
called the CET† Cement Evaluation Tool. Froelich et al. interpretation of Butsch’s statistical technique has since
(1981) and Dumont et al. (1984) described this tool and been extended from gas-cut cement to any type of irreg-
presented field test results. The CET device has 8 trans- ularly bonded cement (Morris et al. 2002; Butsch et al.
ducers arranged at 45° intervals around the circumfer- 2002). The latter two papers described the evaluation of
ence of the tool to map the acoustic impedance of the low-density and other specialized cements using con-
material around the pipe. Finlayson et al. (1984) ventional and statistical techniques. Frisch et al. (1999;
described the results of later field tests, along with 2000) described combined conventional and statistical
results from an interpretation model incorporating gas processing methods developed for the CAST-V tool and
bubbles in the annulus. In 1984, Catala et al. proposed presented log examples for foamed and conventional
merging the interpretations of the CET and CBL tools, cements.
using a combined display so that the individual limita-
tions of each log could be overcome.
Later, Sheives et al. (1986) introduced an ultrasonic 15-4.5.2 Advantages and disadvantages
cement evaluation tool, the PET‡ Pulse Echo Tool. A of ultrasonic tools
comparison between the PET log and conventional and Ultrasonic tools make a small portion of the casing
compensated bond logs, based on runs in a test well con- resonate by transmitting a broadband pulse (200 to
structed by the United States Environmental Protection 700 kHz) normal to the casing wall. Two major advan-
Agency, was presented by Albert et al. in 1987. tages of this technique compared to sonic methods) are
In the early 1990s, Hayman et al. (1991) introduced ■ less sensitivity to a liquid-filled microannulus
a second-generation ultrasonic tool called the USI† between the pipe and cement
UltraSonic Imager. It uses a single rotating transducer ■ good spatial resolution—about 1 in.2 of cross-
to provide full coverage of the pipe circumference with
sectional area—combined with full coverage by
greater accuracy. Comparisons with other logging tools
the second-generation rotating tools.
were made in a full-scale simulator (Hayman et al.,
1991). Graham et al. (1997) developed a similar tool, the However, disadvantages arise from the high frequency
CAST-V§ Circumferential Acoustic Scanning Tool— compared to sonic tools:
Visualization. The early CET and PET tools are now ■ pipe rugosity or roughness caused by corrosion or
becoming obsolete. deposits can affect the cement measurement
Basic interpretation of the acoustic-impedance maps ■ performance is limited in heavy solid-weighted muds,
or images is performed by setting impedance thresholds
especially OBMs, because of acoustic-wave attenua-
to discriminate between liquids, solids, and gases.
tion.
Goodwin (1989) suggested using the character of the
acoustic measurements, in addition to their absolute
values, to help discriminate between solids and fluids in 15-4.5.3 Tools
the annulus. He observed that the measured imped-
ances of solids are usually more variable than those of The first-generation tools (CET and PET) consist of
liquids. Statistical processing of ultrasonic images has eight ultrasonic transducers arranged in a helical spiral
become increasingly common since the early 1990s. around the sonde, each facing outward (Fig. 15-44).
Harness et al. (1992) and Graham et al. (1994) made They are evenly spaced around the circumference of the
statistical analyses of PET data to improve the evalua- tool body—one transducer every 45°. A ninth transducer,
tion of low-density and complex cements. Butsch (1995) facing a reflector in the tool body, measures the speed of
sound in the borehole fluid.
† Mark of Schlumberger
‡ Mark of Gearhart Industries
§ Mark of Halliburton

Chapter 15 Cement Job Evaluation 585

 Sonde

Compensating device

Motor assembly
Gearbox assembly
Rotating electrical
connection

Centralizer

Rotating shaft with

built-in electronics

Rotating seal

Transducer
Interchangeable
rotating sub

∼7.5 rps

Fig. 15-45. USI tool.

Fig. 15-44. CET drawing.

15-4.5.4 Principle of ultrasonic cement evaluation
The basic idea behind the ultrasonic technique is to
The second-generation tools (USI and CAST-V make a small area of the casing resonate through its
devices) use a single rotating transducer to achieve full thickness. The transducer sends out a short pulse of
coverage of the pipe wall. Measurements are made at a ultrasound and listens for the echo containing the reso-
large number of points around the circumference: 36 or nance. If there is fluid behind the casing, it will tend to
72 for the USI tool and 100 for the CAST-V device. resonate or “ring,” but if there is solid cement behind
Depending on the pipe diameter, several rotating sub- the casing, the resonance will be damped. The reso-
assemblies are available to optimize the distance nance is analyzed to determine the cement’s acoustic
between the transducer and the pipe. While running into impedance.
the well, the USI tool (Fig. 15-45) measures the acoustic
properties of the well fluid (velocity and impedance) by
pointing the transducer towards a built-in reflector. The 15-4.5.5 Measurements and processing
CAST-V tool has a secondary fixed transducer facing a As illustrated in Fig. 15-46, four measurements are made
flat reflector to measure the acoustic velocity of the well on the ultrasonic echoes.
fluid while logging. ■ Echo amplitude—an indicator of casing condition
Because the second-generation tools have largely
■ Internal radius of the casing—calculated from the
replaced the earlier tools, the rest of this section will
show examples for the newer tools. transit time of the main echo

586 Well Cementing

 Mud Casing Cement
Transducer

Amplitude

Time

Echo Transit Thickness Cement

amplitude time impedance

(Internal casing Internal

condition) radius

Fig. 15-46. Ultrasonic principle and measurements.

■ Casing thickness—calculated from the resonant Z2 − Z1

frequency K ref = , (15-13)
Z2 + Z1
■ Acoustic impedance of the material behind the
casing—calculated from the form of the resonance. and
2 Z2
K trans = 1 + K ref = , (15-14)
The physics can be explained by assuming that the
ultrasonic wave is planar and traveling perpendicular to
( Z2 − Z1 )
a flat plate representing the casing. Figure 15-46 shows
where Kref is the reflection coefficient at the boundary
the wave paths and the echo train obtained from an
between two materials of acoustic impedance Z1 and Z2,
ideal, infinitely short transmitted impulse. At the bound- and Ktrans is the transmission coefficient at the same
ary between the pipe and the borehole fluid, most of the boundary. The pressure transmission coefficient can be
incident energy is reflected, and the balance is trans-
mitted into the pipe wall. The fractions of incident
acoustic pressure that are transmitted and reflected are
given by the following formulas.

Chapter 15 Cement Job Evaluation 587

greater than 1, but energy is always conserved because Frequency (kHz)
the intensity, i, of a wave in a medium of impedance Z is 0 0.1 0.2 0.3 0.4 0.5 0.6 0.70 0.8
given by
p2
i= , (15-15)
Z
where p is the acoustic pressure.
The first, large reflection at the pipe wall returns to
the transducer and provides a measurement of the pipe
radius. The small fraction of the energy that is transmit-
ted into the pipe bounces back and forth inside the pipe
wall, losing energy into the annulus and hole at every
bounce. The pressure-reflection coefficient inside the
pipe is negative. Thus, the echo train consists of a large
reflection from the internal surface of the pipe followed
by an exponentially decaying series of inverted pulses.
The time separation of the train of negative pulses is 10 15 20 25 30 35 40 45 50
equal to the go-and-return time, Δt, in the pipe. Time (µs)
2 hcsg Fig. 15-47. Typical emitted ultrasonic pulse and its frequency spectrum.
Δ t= , (15-16)
vsteel

where hcsg is the casing thickness and vsteel is the

acoustic velocity in steel (19,450 ft/s [5,930 m/s]).
The resonant frequency of the pipe, f0, is given by
1 vsteel
f0 = = . (15-17)
Δ t 2 hcsg

Figure 15-47 shows the signal transmitted by a USI

transducer along with its frequency spectrum (i.e., the
magnitude of the Fourier transform). The frequency
spectrum shows that the pulse contains sound energy in
the range of 200 to 650 kHz. This is the range corre-
sponding to the inherent thickness mode resonant fre- Free
quencies of oilfield casings from 0.18 to 0.6 in. [4.57 to pipe
15.24 mm].
Figure 15-48 shows typical echoes from a casing
backed by water and cement. The resonance has been
magnified to show more detail. This diagram illustrates
two points.
■ Cement behind the pipe damps the resonance.

■ The resonant frequency (shown by the spacing

between the peaks in the resonance) is the same with
water or cement behind the pipe.
Good
While the basic principle of all ultrasonic tools is the cement
same, there are differences in the way the signal pro-
cessing is done, particularly for the cement impedance.
The USI processing is shown in Fig. 15-49. The time from 40 50 60 70 80 90
transducer firing to the main echo is used to calculate a Time (μs)
radius. The distance that the tool is eccentered in the
Fig. 15-48. Ultrasonic response to different materials in the annulus.
pipe is calculated from the radius measurements, and The resonance (red) is magnified for clarity.
then the measurements are corrected for eccentering.

588 Well Cementing

 Rugosity
Radius
Amplitude
Time Group
delay Δf Cement
impedence

fo Thickness

Time Frequency

Fig. 15-49. USI signal processing schematic.

The magnitude of the main echo is measured. It is a 15-4.5.6 Response to materials in the annulus
qualitative indicator of pipe rugosity and useful for qual- Gas
ity control. In tools other than the USI device, the ampli-
tude forms part of the cement processing, in which it is Gas has an acoustic impedance below 0.1 MRayl. It acts
used as a first-order correction for mud attenuation, tool as a barrier that totally reflects the ultrasound.
eccentering, and pipe rugosity.
Liquid
The resonance signal is analyzed to calculate the
thickness of the pipe and the impedance of the material Liquids have impedances in the range 1 to 3 MRayl
behind the pipe. The thickness is determined from the (Tables 15-5 and 15-6).
resonance frequency (Eq. 15-17). The cement-imped-
ance processing is different in each tool. The first-gen- Table 15-6. Sound-Velocity Values for Cement Systems
eration tools (CET and PET) measured the resonance with Different Densities
energy in predetermined time windows. The USI pro-
cessing is performed in the frequency domain, as out- Neat Class G Cement 1 Cement 2 Cement 3
lined below. Density 1,890 1,115 1,115 1,000
The early part of the signal, covering the main echo (kg/m3)
plus roughly the first seven cycles of resonance, is Compressional 3,600 3,000 2,700 2,400
selected by the dotted yellow window function shown in velocity (m/s)
Fig. 15-49. The group delay spectrum of this signal is cal-
culated as follows: Shear velocity 2,000 1,550 1,429 1,250
(m/s)
dφ
τ=− , (15-18)
dω
where φ(ω) is the phase spectrum and ω = 2πf is the High-quality cement
angular frequency. The group delay spectrum shows the Neat cement has an impedance of about 6 MRayl. Light
pipe resonance as a dip. The frequency of the resonance cements can have impedance as low as 2.5 MRayl, over-
f0 gives the pipe thickness (Eq. 15-17). The width of the lapping the liquid range (Table 15-3). As with acoustic
resonance Δ f/ f0 depends on the cement impedance. logging, foamed cements have particularly low imped-
The thickness and impedance are determined by ance and pose special interpretation problems.
matching the planar model (Fig. 15-46) to the measured If the shear bond between the cement and casing is
resonance using an iterative technique. The pipe prop- good, the USI tool impedance reading will be higher than
erties are known, and, using a built-in casing sample, the the compression-wave impedance. This occurs because
logging fluid velocity and impedance are determined the shear coupling introduces additional resonance
while running into the well. Therefore, the only damping, as illustrated in Fig. 15-50.
unknowns are the pipe thickness and cement imped-
ance. The final step of the processing is to apply tabu-
lated corrections for casing diameter, transducer size,
and fluid velocity, which are calculated using a three-
dimensional model (Randall and Stanke, 1988).

Chapter 15 Cement Job Evaluation 589

 12
13°

10

8
Cement Solid Cement
acoustic 6 20 mm
impedance
(MRayl) 4 120 mm 40 mm
Channel
Fluid
2 78°
26°
0
0 2 4 6 8
0
Acoustic impedance (MRayl) 8
7 Z = 6.6
Fig. 15-50. Shear coupling effect. Impedance reading is higher for
Cement
well-bonded solids than liquids. 6 Z = 5.0
Cement 5
acoustic 4
Cement with channels impedance 3 Slurry
(MRayl)
Figure 15-51 shows a USI laboratory measurement in 2
which a casing was initially immersed in water and then Channel
1 Water Z=0
surrounded by neat cement with artificial gas channels. 0
The casing was logged as the cement set. The gas 0 60 120 180 240 300 360
(0 MRayl), water (1.5 MRayl), and slurry (3 MRayl) read-
Angle (°)
ings were close to the expected values. The solid cement
reading was higher than its compression-wave imped- Fig. 15-51. USI laboratory measurements within a casing with artifi-
ance (measured simultaneously on a separate sample) cial gas channels.
because of the shear coupling effect. The narrowest
20-mm channel was indicated clearly, though the read-
ing was 1 MRayl higher, and the 40-mm channel was 8
Cement
measured correctly. 7
6 Model 12 mm
Microannulus
For the purposes of discussing ultrasonic logs, a Cement 5
microannulus will be defined as a small fluid-filled gap, acoustic 4 Model 6 mm
less than a few hundred micrometers thick, between the impedance
casing and cement. Such gaps are most often created by (MRayl) 3
Experimental 6 mm
pressure and temperature changes, or by a mud or oil 2
film left on the casing. Thicker fluid layers will be Water
1
referred to as a “mud layer” or as a channel.
The effect of a microannulus on the ultrasonic pulse- 0
echo signal depends on whether it is filled with gas or 0 50 100 150 200
liquid. For a water-filled microannulus, the measure- Microannulus (µm)
ment is weakly affected. Hayman et al. (1991) reported
Fig. 15-52. Effect of a liquid-filled microannulus on the USI signal for
experiments and theory for a microannulus between a 7-in. diameter, 6-mm thick casing (experiment and theory) and a
a 6-mm thick casing and a neat Class G cement 12-mm thick casing (theory only).
(Fig. 15-52). The experiment was performed by pulling a
slightly conical casing vertically out of the cement.
Initially, the impedance reading is 20% high because of ble to distinguish water from cement, although the mea-
shear coupling. Once the bond between the casing and sured impedance is 50% lower. Modeling shows that
cement is broken, the microannulus fills with liquid, and a 12-mm casing resonates at lower frequency than
the correct impedance is measured. Further increases in the 6-mm casing and is proportionately less affected by
the size of the microannulus decrease the measured a microannulus.
value. For a 0.1-mm thick microannulus, it is still possi-

590 Well Cementing

 Table 15-6 displays results for the modeling of liquid- The impedance of neat cement is reduced by 50%
filled microannuli behind 8- and 12-mm thick casings for (staying above 3 MRayl) when the size of the
four types of cement (Table 15-6). The results are shown microannulus is greater than 0.15 mm, depending on
in Figs. 15-53 and 15-54. casing thickness.
The conclusions from both liquid-filled microannulus 3. When lower-density cements are present, the drop in
studies are as follows. measured impedance is less than 0.2 MRayl when the
1. The USI measurement is insensitive to small amounts size of the microannulus is greater than 0.2 mm,
of liquid-filled microannulus. depending on casing thickness.
2. The measured impedance of neat cement is more sen- 4. When the cement is perfectly bonded to the casing,
sitive to a microannulus than lower-density cements. shear coupling boosts the impedance.

10
Class G
9 Cement 1
Cement 2
8 Cement 3
7
6
Acoustic
impedance 5
(MRayl) 4
3
2
1
0
0 100 200 300 400 500 600 700 800 900 1,000
Microannulus thickness (µm)
Fig. 15-53. Modeled effect of a liquid-filled microannulus on the USI signal for a 7-in. diameter, 8-mm thick
casing and four types of cement.

10
Class G
9 Cement 1
Cement 2
8 Cement 3

Acoustic 6
impedance
(MRayl) 5

1
0 100 200 300 400 500 600 700 800 900 1,000
Microannulus thickness (µm)
Fig. 15-54. Modeled effect of a liquid-filled microannulus on the USI signal for a 7-in. diameter, 12-mm thick
casing and four types of cement.

Chapter 15 Cement Job Evaluation 591

 A log example of a liquid-filled microannulus is shown ■ large acoustic impedance contrast between the
in Section 15-4.6.1. cement and formation.
A gas-filled microannulus affects the ultrasonic mea-
A well-cemented liner inside an outer casing is par-
surement much more strongly than a liquid-filled
ticularly likely to be affected by reflections.
microannulus, because the wavelength in gas is shorter
The formation reflections may be in or out of phase
and the gas impedance is much lower. Figure 15-55
with the desired casing resonance, depending on the
shows microannulus effects for air and water, measured
cement thickness and velocity, and the impedance mea-
by Jutten and Hayman (1993). The ultrasonic measure-
surement will be reduced or increased accordingly.
ment fails to indicate cement as soon as a tiny gap of less
Because the casing is usually eccentered in the annulus,
than 1 μm is created.
the thickness of the annulus is not constant, and the for-
Gas microannulus examples are shown later in
mation reflections create a pattern of interference
Section 15-4.6.1.
fringes centered on the narrow part of the annulus
(Fig. 15-56). Successive fringes indicate a change of one-
half wavelength in the annulus thickness, where the
8
wavelength, λ, is given by
7
vcem 2 × vcem × hcsg
6 λ= = . (15-19)
5 Air f0 vsteel
Acoustic Water
impedence 4 For example, if there is neat cement behind a 9-mm
(MRayl) 3 thick casing, λ/2 = 5 mm. These patterns can be easily
recognized on the cement images produced by the
2
second-generation USI and CAST-V tools.
1
0
0 1.0 10.0 100.0 1,000
Microannulus size (µm)

Fig. 15-55. Experimental liquid and gas microannulus effects on the

USI measurement, using a 4.7-in. diameter, 9-mm thick casing. (n + 2)λ/2
Reprinted with permission of SPE.

Mud layer (n + 1)λ/2

A mud layer, defined as a liquid-filled gap between the

casing and cement more than a few hundred microme- Destructive
ters thick, gives low impedance readings in the liquid interference
range. Such layers can be created by cement fingering Annulus
through the mud, leaving mud on both the casing and thickness
the formation. A figure in Section 15-4.6.1 shows a log nλ/2
with a suspected mud layer. Fig. 15-56. Interference caused by formation reflections when the
pipe is eccentered in the hole, where n is an integer denoting the
Thin cement number of wavelengths, denoted by λ . Destructive interference
The processing of ultrasonic measurements is performed occurs when the annulus thickness is n λ /2, reducing the reso-
nance amplitude and increasing the measured impedance. The
with a model in which only mud, pipe, and annulus mate- thinner the annulus, the more severe the effect is. A figure in
rial are present. However, the annulus material has a Section 15-4.6.1 shows an example of reflections from an outer
finite thickness, and the reflection of energy occurs at casing.
the annulus/formation or annulus/outer-casing bound-
ary. This reflection can affect the measured echo, espe-
cially if certain conditions exist:
■ thin cement sheath

■ smooth surface on the wall of the hole

592 Well Cementing

15-4.5.7 Other factors affecting tool response 15-4.5.8 Basic interpretation
Casing shape At the time the first logging tools were developed, the
Most oilfield casings are manufactured by forging. The industry traditionally described cement in terms of com-
external surface is usually extremely smooth and cylin- pressive strength. Laboratory testing was performed on
drical, but the internal surface has characteristic wave different formulations to determine empirical correla-
patterns with peak-trough heights on the order of 10% of tions between the acoustic impedance and compressive
the thickness. These manufacturing patterns often strength. Empirical correlations have extremely limited
affect the cement image but do not usually affect the validity, and only the measured acoustic-impedance
interpretation. They should be examined during quality value should be used for interpretation.
control operations. A figure in Section 15-4.6.1 shows a
15-4.5.8.1 Impedance thresholds
log affected by casing shape.
The cement impedance can be presented as a raw
Tool eccentering impedance map. Basic interpretation of the impedance
Excessive tool eccentering can affect the measurements, maps or images is performed by setting thresholds to dis-
especially in places in which the ultrasonic beam is no criminate between gas, liquids, and solids. Currently
longer perpendicular to the casing wall. The recom- recommended impedance thresholds for the USI maps
mended eccentering limit for the USI tool is are shown in Table 15-7 and illustrated in Fig. 15-57.

Lecc < 0.1 dh (15-20)

Table 15-7. Impedance Image Thresholds for the USI Tool
where
Threshold Value (MRayl)
d = diameter (in.) and h = thickness (in.).
Excessive eccentering will produce vertical stripes on Gas (or dry microannulus)/liquid 0.3
the cement log. Solid/liquid Impedance of mud in annulus
+ 0.5 (default 2.6)
Rugose casings
Maximum impedance 4 to 8 depending on expected
Rugosity caused by corrosion or deposits can degrade cement impedance
results significantly. The amplitude image is the best
indicator of this problem. A log affected by rugosity is
shown in a series of log examples presented in Section
15-4.6.1. 8
Raw Interpreted
image image
Heavy muds
Heavy muds, especially OBMs, attenuate ultrasound
strongly and degrade the signal-to-noise ratio. The USI 6
log is relatively immune to mud attenuation compared to Cement
the CET log, and it will work up to about 16 lbm/gal
Standard

[1,920 kg/m3] in WBM and 13 lbm/gal [1,560 kg/m3] in

Light

Acoustic
OBM. The exact limits depend on mud composition, impedance 4 Maximum
pressure, and temperature and are difficult to predict. impedance
(MRayl)

Pipe coatings Solid/liquid

A complication closely related to the microannulus is threshold
2
thin coatings on the outside of the pipe (e.g., mill varnish Liquid
and epoxy/sand coatings). An epoxy thickness of up to
0.4 mm (0.02 in.) can be tolerated with little effect on Gas/liquid
the log. 0 threshold
Gas or dry
microannulus
Casing thickness
The CET log is affected by casing-weight changes and Fig. 15-57. Impedance thresholds to distinguish between gas (or dry
needs calibration for each weight. The USI log process- microannulus), liquid, and solid.
ing avoids this difficulty.

Chapter 15 Cement Job Evaluation 593

 The liquid/solid threshold should be set 0.5 MRayl called the “gas-cut algorithm,” is now called the
above the impedance of the mud in the annulus, whose “microdebonding algorithm” because partial, uneven,
value can either be taken from the fluid-properties mea- and localized debonding is thought to be the most
surement if the same mud is used for logging or can be common cause of variability in measured cement
estimated using software developed by the various log- impedance. Gas influx can cause debonding because
ging companies. The upper limit of the display scale of dehydration and fractures, but it is just one of sev-
should be set according to the expected cement imped- eral possible mechanisms for variable bond. Another
ance, which can be measured using the ultrasonic processing method, described by Frisch (1999, 2000), is
cement analyzer (Appendix B) or estimated from the called advanced cement evaluation or statistical varia-
various proprietary software packages. tion processing.
Published thresholds for the CAST-V tool (Frisch et The USI microdebonding algorithm is illustrated in
al., 2000) are reproduced in Table 15-8. Additional Fig. 15-58. It calculates the standard deviation of the
thresholds are used to distinguish liquid containing gas cement impedance in four directions around each image
and the solid/liquid transition. pixel. If all four standard deviations are higher than set
thresholds, and the cement impedance is low (in the gas
or liquid classes), the current pixel is considered to be
Table 15-8. Impedance Image Ranges for an Ultrasonic locally debonded and is colored green. Usually, two
Imaging Tool images are presented: the raw acoustic impedance and
the interpreted image with microdebonding.
Material Color Impedance Image
Range (MRayl)
Gas Red 0–0.38

Gas-cut liquid Light blue 0.38–1.15

Liquid Blue 1.15–2.30 Vertical

deviation
Solid-liquid transition Yellow 2.30–2.70

Low-impedance cement Light brown 2.70–3.85

Diagonal 1
Medium-impedance cement Dark brown 3.85–5.00 Transducer
“spot” size
High-impedance cement Black >5.00

Horizontal Diagonal 2
deviation
15-4.5.8.2 Statistical approaches
The simple-threshold interpretation may be ambiguous,
because a number of conditions may cause solid materi-
als to have a low impedance as measured ultrasonically.
These conditions include genuine low-impedance
cement (e.g., foam), poor bonding (especially dry
microannulus), and mud contamination. Goodwin
(1989) pioneered the use of acoustic-impedance varia-
Fig. 15-58. Illustration of the microdebonding processing for a USI
tions to help distinguish solids from fluids in the annu- measurement pattern with 36 radial samples and a 1.5-in. vertical
lus. The guiding principle is that a liquid or gas in the sampling rate inside a 5.5-in. casing.
annulus has a constant, uniformly low impedance, while
a “low-impedance” solid material usually has a more
variable measured impedance.
Goodwin initially used the eight raw impedance For one ultrasonic imaging tool, statistical processing
curves provided by the first-generation tools. With the calculates the vertical variance (or derivative) of the
much larger number of measurements provided by the impedance. Three images are presented: a standard
second-generation tools, it is more convenient to interpreted cement map, an impedance variance map,
process the images than to look at curves. Both the USI and a combined map. The most recent method of con-
and CAST-V logs have statistical image processing algo- structing the combined map (Frisch, 2000) is a logical
rithms. The USI algorithm (Butsch, 1995) originally combination of the other two maps—showing cement if

594 Well Cementing

there is either high impedance or high variance and 9,000
showing fluid otherwise.
Statistical processing has certain limitations. 9,800
■ The variance thresholds are empirical. Sometimes
too much “liquid” (blue) is indicated when the bond-
10,600
ing is obviously patchy. On the other hand, casing- Tool
shape effects, casing rugosity, or irregular tool move- depth
ment sometimes provoke variations in measured (ft) 11,400
impedance that are interpreted as debonding. Such
artifacts must be detected by quality control. The 12,200
variance thresholds should ideally be tuned in a sim-
ilar type of casing surrounded by known material,
such as free pipe. 13,000
180 204 228 252 276 300
■ Solid cement can be associated with a low-variance,
low-impedance image. For example, an extended dry Fluid velocity (µs/ft)
microannulus has flat, near-zero impedance and 9,000
remains red on the cement map. Homogeneous, well-
bonded, light cement can also have a low variance. 9,800
Statistical processing log examples are shown in
Section 15-4.6.1. 10,600
Tool
depth
15-4.5.9 Log presentation (ft) 11,400
Log presentations vary from region to region and
between service companies. Schemes used for labeling 12,200
curves also differ. The data can be classed into six types.
Figures 15-59 and 15-60 show presentation examples,
and Fig. 15-61 shows an example of a full log with API 13,000
0 0.6 1.2 1.8 2.4 3
headers.
Acoustic impedance of fluid (MRayl)
Borehole fluid data
Fig. 15-59. USI fluid-properties-measurement (FPM) plots: fluid
The measured speed of sound in the borehole fluid is slowness (labeled velocity) and fluid impedance.
required for the radius calculations. It may also be dis-
played on the log. It is presented as slowness, the inverse
of velocity. This curve indicates density changes with Other information includes the inclinometer data,
high sensitivity. Occasionally, it proves useful in locating the conventional magnetic collar locator, and a gamma
fluid changes in the well (e.g., oil on top of water and salt ray log for depth correlation. The inclinometer, which
influx). may be included in the tool or in a separate tool (in the
The USI tool measures not only the acoustic velocity, case of the USI device), indicates the hole deviation and
but also the acoustic impedance of the fluid, because the the rotation of the tool compared to vertical in deviated
latter is needed for the cement impedance processing. wells. The rotation curve is labeled “relative bearing,”
Both measurements are made while running the tool and it allows the images to be oriented with the low side
into the well, and the measurements are shown in cross- of the well in the center of the image. The high-low ori-
plots labeled fluid properties measurement (FPM), as in entation is extremely useful in interpretation of primary
Fig. 15-59. cement jobs. The casing data described in the next sec-
tion also form an essential part of quality control.
Quality control and auxiliary data
Quality control data include the tool eccentering curve
and processing flags. The recommended eccentering
limit for the USI tool is 0.1dh. The USI flags indicate
problems with the acquisition and processing. They are
displayed as a map.

Chapter 15 Cement Job Evaluation 595

 QC Casing Cement
1,500

Standard API header

1,550
–500.0000 –500.0000 –1,000.0000
–0.0780 –0.0780 –500.0000
–500.0000 0.3000
–0.0680 –0.0680 0.5000
–0.0600 –0.0600 2.6000
1.0000 3.0909
–0.0520 –0.0520 1.5000
–0.0440 –0.0440 3.5818
2.0000 4.0727
–0.0380 –0.0380 2.5000
–0.0280 –0.0280 4.5636
3.0000 5.0546
0.5000 –0.0200 –0.0200 3.5000
–0.0120 –0.0120 5.5455
1.5000 4.0000 6.0364
–0.0040 –0.0040 4.5000
2.5000 6.5273
0.0040 0.0040 5.0000
3.5000 External External 0.0120 0.0120 7.0182
6.5000 Min. of 5.5000 7.5091
Eccentricity Radius Radius 0.0200 Min. of 0.0200 6.0000 Bonded
Amplitude 0.0280 0.0280 8.0000
0 (in.) 0.5 Average Average Thickness 6.5000
Process 0 (dB) 75 0.0360 0.0360 7.0000
flags 4.5 (in.) 3 3 (in.) 4. 0.0440 0.1 (in.) 0.6 0.0440 7.5000 Cement
0.0520 0.0520
(----) 0.0600
0.0680
0.0760
0.0600
0.0680
0.0760
8.0000
Map with
Impedance
Cement header
Raw Classification
Acoustic
(----)
Internal Thickness Impedance
Radii Minus Minus (----)
Average Average
(----) (----)

Revolving Internal Internal Gas or Dry

Average of
Speed Radius Radius Average of Microannulus
Amplitude
6 (rps) 8 Average Average Thickness
0 (dB) 75
4.5 (in.) 3 3 (in.) 4.5 0.1 (in.) 0.6

Internal Internal
Revolving
Max. of Radius Radius Liquid
Speed Max. of
Amplitude Max. Max.
–8 (rps) –6
0 (dB) 75 4.5 (in.) 3 3 (in.) 4.5
Thickness
0.1 (in.) 0.6
Cement log
–500.0000
–6.0000
–5.6000
–5.2000
–4.8000
–4.4000
–4.0000
–3.8000
–3.2000
–2.8000
–2.4000 Min. of Min. of
Cable –2.0000
Internal Internal Micro-
–1.6000
Speed –1.2000 Radius Radius debonding
0 (ft/hr) 2,000 –0.8000
–0.4000 4.5 (in.) 3 3 (in.) 4.5
0.5000

Amplitude of
Echo Minus
Max. Composite/LQC log
(----)

Fig. 15-60. USI log presentation with both casing and cement data.

Repeat section (cement log)

Fluid properties plots

Standard API tail

Fig. 15-61. Standard USI presentation.

596 Well Cementing

Casing data CBL/VDL data
The echo amplitude is presented as an image in which When the ultrasonic tool is run in combination with a
dark colors indicate low amplitude. Curves of minimum, sonic log, the CBL/VDL data are usually presented
mean, and maximum amplitude are also shown. alongside for easy comparison.
The measured casing radii are corrected for eccen-
tering and presented as an image. Commonly, the image Openhole and cementing data
shows the radius compared to the mean value at each For effective cement evaluation, well data and cement-
depth, with blue indicating smaller radii and red larger ing data are just as important as the acoustic logs.
radii. Thus, casing corrosion and wear show up as red Specialized presentations add openhole data, such as
areas. The casing thickness data are usually presented in calipers and lithology, and cementing data, such as pre-
a similar fashion, with red indicating thickness below dicted casing standoff and predicted cement impedance.
average and blue thickness above average. The internal Special log headers allow inclusion of detailed cement
radius and thickness are summed to calculate the exter- reports.
nal radius. Internal- and external-radius curves are
shown together, with shading to indicate the casing
cross section. Minimum-, mean-, and maximum-thick- 15-4.6 Quality control
ness curves are also plotted separately. A step-by-step procedure should be followed to control
the quality of the cement-evaluation log. The following is
Cement/annular data a general procedure.
The primary purpose of the cement imaging tools is to ■ The repeat section and main log must be consistent.
determine what is in the annulus. There are four ■ Acquisition and processing parameters must be
common types of cement images. included on the log. Check the solid/liquid impedance
1. The raw impedance image on a white-yellow-brown threshold and the maximum impedance of the color
color scale. The white color corresponds to 0 MRayl, scale. These values should also be indicated on the
while the black upper limit is usually adjusted to a color scale.
value in the range of 4 to 8 MRayl according to ■ Check that eccentering is inside the recommended
cement impedance. tolerance.
2. The conventional interpreted cement image using ■ Verify that the casing radius and thickness are close
impedance thresholds to distinguish gas (red), liquid to nominal values in uncorroded areas.
(blue), and cement (yellow-brown). Typical thresh-
■ The casing must be in good condition, and radius and
olds have been listed in Tables 15-5 and 15-6.
thickness must be accurate for a good cement log.
3. The standard deviation or variance image, in which The echo-amplitude image is a good indicator of
higher variability regions are darker. casing condition. Corrosion and deposits show up as
4. The combined conventional/variance image, provid- low-amplitude (dark) areas.
ing the most complete interpretation. This presenta- ■ If a free-pipe section is present, the mean acoustic
tion is labeled “cement with microdebonding,” and impedance should equal the expected impedance of
microdebonded areas are colored green. the fluid in the annulus, plus or minus the tool accu-
Usually, two or three of these images are shown on the racy.
right-hand side of the log, as in Fig. 15-60. Figure 15-59 shows a typical fluid properties mea-
If the hole deviates from vertical orientation, the dis- surement with a fluid-velocity crossplot and a fluid
play can be adjusted to present the image from the low impedance crossplot. The fluid velocity curve should be
side of the hole in the middle of the track and the high smooth and consistent with the fluid type (Table 15-9).
side at the edges.
An ultrasonic BI track can be included, showing the
percentage of cement, liquid, and gas at each depth.
This allows a quick comparison with the CBL BI. In addi-
tion, curves of minimum, mean, and maximum acoustic
impedance are typically shown.

Chapter 15 Cement Job Evaluation 597

Table 15-9. Typical Borehole-Fluid Slowness If the measured and theoretical fluid-impedance
Fluid Slowness (µs/ft)
values differ, then the theoretical value should generally
be used in brine or oil. The fluid-velocity measurement
Brine 160–200 used to calculate the theoretical value is more accurate
Water 200 than the fluid-impedance measurement. In weighted
mud, the theoretical value should be used; however, two
WBM 180–210 points must be considered: (1) the B factor is accurate
Oil and OBM 210–250
to only 5–10%, and (2) there may be sedimentation in
the well. The log must be checked in a known free pipe
or a gas zone, if possible, to verify that the Zmud value is
correct. An error of 0.1 MRayl in Zmud changes the cal-
The measured fluid impedance should have low dis- culated cement impedance by about 0.5 MRayl.
persion (below about 0.1 MRayl) and be within 10% of
the theoretical value, calculated as follows. 15-4.6.1 Combined interpretation examples
Clear fluids:
The ultrasonic image shows the acoustic impedance and
⎡ 304.8 ⎤ the distribution of materials around the casing; there-
Z mth = ρ ⎢ ⎥ (15-21) fore, it is usually much easier to interpret than a
⎢ ( v fluid )–1 ⎥ CBL/VDL log. Most of the following examples are com-
⎣ ⎦
bined ultrasonic-sonic logs. A combined interpretation
Weighted muds: provides a coherent picture and compensates for limita-
tions of the individual logs. As already discussed, the
⎡ 304.8 ⎤ CBL-VDL has an ambiguous response to channeling,
Z mth = Bρ ⎢ ⎥, (15-22) liquid-filled microannuli, and contamination. A pressure
⎢ ( v fluid )–1 ⎥ pass is needed to diagnose a microannulus. The ultra-
⎣ ⎦ sonic log is very sensitive to gas and dry debonding. It is
where affected by poor casing conditions and attenuation in
B = correction factor in the range 0.85 to 1.0 very heavy mud. Table 15-10 compares the responses of
(vfluid)–1 = fluid slowness (μs/ft) CBL-VDL and ultrasonic measurements.
Zmth = theoretical mud impedance (MRayl)
ρ = fluid density (g/cm3).
An empirical formula for the B factor is available in
proprietary logging company software. Figure 15-12
shows examples of B factor for typical WBMs and OBMs.
The value of fluid impedance, Zmud, used for process-
ing the log must be verified. It should be on the parame-
ter list at the bottom of the log. If the parameters have
been zoned versus depth, then the zoned values of
(vfluid)–1 and Zmud will be at the top of the log.

598 Well Cementing

Table 15-10. USI and CBL-VDL Interpretation Guide
USI Tool CBL-VDL
Resolution 1.2 in. 360° × 3 ft

Gas Gas (0 MRayl) Higher amplitude than free pipe

Liquid Liquid (1–3 MRayl) Free pipe

Well-bonded cement Cement (4–8 MRayl) Good bond

Very light cement Low contrast from liquid Low contrast from liquid

Dry microannulus Dry microannulus or gas Good-to-fair bond

(debonded cement) (0 MRayl)

Wet microannulus Medium impedance cement Poor bond (ambiguous)

Pressure pass needed

Mud layer on sector Liquid channel Fair bond (ambiguous)

Contaminated cement Low-impedance cement Fair bond (ambiguous)

Mud channel Channel Fair bond (ambiguous)

Narrow gas channel Gas channel Cement

Cement/formation bond Not seen VDL qualitative indication

Reflections from outer Slightly affected Strongly affected

casing or hard formation (interference patterns) (poor bond shown)

Casing condition Very sensitive Slightly sensitive

Mud-density limit < 16 lbm/gal (WBM) No limit

< 13 lbm/gal (OBM)

Chapter 15 Cement Job Evaluation 599

Good cement
Figure 15-62 is a combined USI and CBL-VDL log in high-
quality, well-bonded neat cement. The ultrasonic image
shows cement all around the casing with a mean imped-
ance of 8 MRayl. The CBL amplitude is low and flat,
and the VDL indicates a weak casing arrival and a strong
formation arrival.

–500.0000
0.3000
0.3000
2.6000
2.6000
CCL Gamma Ray 3.0000

–20 (----) 20 0
3.5000 Liquid Index
(gAPI) 100 4.0000
4.5000 0 (----) 1
5.0000
Cable Fluid-Compenstated 5.5000
Speed CBL Amplitude 6.0000 Contaminated
6.5000
0 (m/hr)1,000 0 (in.) 10 7.0000
Max. of AI Cement Index
7.5000
8.0000
0 (MRayl) 10 0 (----) 1
Revolving Fluid-Compenstated 8.1000
Speed CBL Amplitude 8.2000 Average of AI Bond Index Min. Amplitude Max.
8.3000
0 (rps) 10 0 (in.) 50 9999.9902 0 (MRayl) 10 1 (----) 0
Acoustic Impedance Variable Density Log
Eccentricity Contaminated with Thresholds Minimum of AI Gas Index 200 (µs) 1,200
0 (in.) 0.5 –19 (in.) 1 (----) 0 (MRayl) 10 0 (----) 1

QC CBL USI Tool VDL

Fig. 15-62. A USI and CBL-VDL log of good cement. AI means acoustic impedance.

600 Well Cementing

Mud channel whose colors are allocated according to the measured
The USI image in Fig. 15-63 shows a wide, 180° mud impedance. The CBL-VDL log is coherent with the USI
channel (blue). The cement color (medium brown) indi- log, although on its own it would be ambiguous. The CBL
cates an impedance of about 5 MRayl, which is lower amplitude is high and variable, and the VDL indicates a
than expected for neat cement and suggests mud conta- strong casing arrival and a weak formation arrival.
mination. The image is a conventional interpreted map

Joint

X000

Joint

QC CBL USI Tool VDL

Fig. 15-63. USI and CBL-VDL log of a mud channel.

Chapter 15 Cement Job Evaluation 601

Cement top Low-density cement
At the top of the job shown in the previous example, the Figure 15-65 shows the USI log of a mixed aggregate 11-
USI image indicates mud and traces of low-impedance lbm/gal cement job in 7-in. casing at 73° deviation. The
(yellow) contaminated cement, as shown in Fig. 15-64. wellbore fluid while drilling and cementing was an 8.8-
The CBL amplitude is high and flat, as it would be in free lbm/gal OBM. The log was run 4 days after cementing.
pipe. Despite the strong casing arrival on the VDL, there The gamma ray curve indicates the production interval
is a weak formation arrival consistent with traces of where perforations were planned, a shale section that
cement on the USI image. As is often the case, there is must provide the isolation, and the water sand at the
no well-defined cement top. bottom. The USI log showed full cement coverage in the
shale interval with impedances in the range of 3 to 5
MRayl. Production tests showed no water production,
confirming zonal isolation.

X000

QC CBL USI Tool VDL

Fig. 15-64. USI and CBL-VDL log of cement top.

602 Well Cementing

 Microdebonding

Gas or Dry
Microanulus

–500.0000 Liquid
0.5000 –500.0000
External Radius Min. of Internal 0.3000
1.0000
Average Radius 1.5000 2.6000
3.7 (in.) 2.7 2.7 (in.) 3.7 2.0000 3.0000
2.5000 3.5000
3.0000
Bonded 4.0000
Min. of Internal Internal Radius 3.5000 4.5000
Max. of AI 5.0000
Radius Max. 4.0000
4.5000 0 (MRayl) 10 5.5000
3.7 (in.) 2.7 2.7 (in.) 3.7 5.0000 6.0000
5.5000 6.5000
Gas or Dry 7.0000
Internal Radius Internal Radius 6.0000 Microannulus
6.5000 7.5000
Gamma Ray Max. Average 7.0000
Average of AI Ratio 8.0000
0 (gAPI) 100 3.7 (in.) 2.7 2.7 (in.) 3.7 7.5000 0 (MRayl) 10 1 (----) 0
8.0000
Cement Map with
Internal Radius External Radius Microdebonding Impedance Classification
CCL Average Average Raw Acoustic Impedance Min. of AI Ratio
(----)
–20 (----) 20 3.7 (in.) 2.7 2.7 (in.) 3.7 (----) 0 (MRayl) 10 1 (----) 0

Joint

Productive
Centralizers interval

Shale

Water sand

Fig. 15-65. USI log showing low-density cement.

Chapter 15 Cement Job Evaluation 603

Mud layer channel. Further simulations suggested that the most
The USI and VDL data in Fig. 15-66 indicate a wide chan- likely explanation was cement fingering through the
nel. However, in this case the cement simulations using mud, leaving a few millimeters of mud on the casing and
cement-placement software gave no reason to suspect a formation.

Microdebonding

–500.0000 –1,000.0000
0.5000 –500.0000
1.0000
0.3000
1.5000
2.0000 3.6000
2.5000 4.0000 Liquid
3.0000 4.4000
3.5000 5.2000
4.0000 5.6000
4.5000
6.0000
5.0000
5.5000 6.4000
Gas or Dry
6.0000 7.2000
6.5000
Max. of AI Microanulus
7.6000
7.0000 0 (MRayl) 10 8.0000
7.5000
8.0000 Average of AI
Min. Amplitude Max.
0 (MRayl) 10 Cement Map
Raw Acoustic with Impedance Bonded
Impedance Min. of AI Classification Sonic VDL Curve
(----) 0 (MRayl) 10 (----) 200 (µs) 1,200

Centralizer

Mud layer

Fig. 15-66. USI and VDL log showing probable mud layer.

604 Well Cementing

Wet microannulus Gas channel and microannulus
In Fig. 15-67, the USI tool shows medium-impedance This example concerns an old gas-storage well from
cement, but the CBL is high and flat and the VDL shows which the gas was leaking to the surface inside a 9.625-
a strong casing arrival, just like free pipe. This is inter- in. casing. The USI log in the 7-in. liner indicated
preted as a liquid-filled microannulus about 0.1 mm patches of gas microannulus and good cement with gas
thick. The immunity of the ultrasonic measurement to a communication throughout. The section in Fig. 15-68
liquid-filled microannulus is one of its great strengths. shows narrow gas channels plus large areas of gas
With the CBL alone, a repeat pressure pass would be microannulus. The well was originally cemented under
needed to diagnose a microannulus. In the presence of a gas pressure, and the CBL log generated at the time of
liquid-filled microannulus, the CBL amplitude is usually cementation showed almost 100% bond. The interpreta-
slightly lower than in this example, and some weak for- tion is that the cement debonded from the 7-in. liner,
mation arrivals should be visible. creating a gas microannulus through which the gas
escaped to the top of the liner.

Bonded
Bonded
Raw Liquid
Acoustic Raw Liquid Cement
Gas or Dry Acoustic Map with
Impedance Microannulus VDL Gas or Dry
Impedance Microannulus Thresholds CCL

CBL

Gas
microannulus

Gas
channel

Fig. 15-68. USI log showing a gas channel and a gas microannulus.

Fig. 15-67. USI and CBL-VDL log showing a wet microannulus all
around the pipe.

Chapter 15 Cement Job Evaluation 605

Various materials with microdebonding processing on the images. Track 6 is the conventional interpreted
Before looking at examples in which the statistical image based on thresholds. Track 5 is the map including
microdebonding processing is useful, the method and its microdebonding logic that indicates microdebonding in
response will be presented. Figure 15-69 shows how dif- green. The last track is the BI track, which shows the
ferent materials and bonding conditions are seen by percentage of bonded cement, microdebonded cement,
both conventional threshold-based interpretation and by liquid, and gas.
the microdebonding processing. The first four tracks are Free pipe has impedance in the liquid range and with
images of the impedance standard deviations computed low standard deviations; therefore, it is unaltered by the
in four directions. Curves of the average, minimum, and microdebonding processing. Bonded cement has high
maximum standard deviation at each depth are overlaid impedance with moderately high standard deviations,

Bonded
Micro-
debonding

STDV STDV STDV STDV Micro- Map with Liquid

Vertical Horizontal Diagonal 1 Diagonal 2 debonding Thresholds Gas or Dry
Microannulus

4,950

Free pipe

5,300
Bonded
cement

5,900

Extended gas
microannulus

6,500
Microdebonded
cement

Fig. 15-69. USI microdebonding processing, showing different materials in the annulus. STDV means standard deviation.

606 Well Cementing

and it is also left unchanged by the special processing. microannulus can occur in places where gas entry is
An extended gas microannulus has an impedance near impossible, such as double-casing strings, in which the
zero and low standard deviations. It is also not affected gas in the microannulus is probably water vapor.
by the processing (except that the centralizers can trig-
ger artifacts). Microdebonded cement has high standard Patchy bond (microdebonding)
deviation, and parts of the image with low impedance Figure 15-70 shows part of a USI and CBL-VDL log in a
(classified as liquid or gas on the conventional image) well cemented with 12.0-lbm/gal high-performance low-
are reclassified as microdebonded cement and shown in density cement. A very good bond was achieved above
green. X000 ft. Below this depth, the cement BI indicates 20%
Dry microannulus, gas microannulus, and dry to 80% bond, while the VDL is variable. The convention-
debonding are alternative labels for a common phenom- ally interpreted USI map, produced by setting imped-
enon that can be extensive or patchy, as shown in ance thresholds (Track 3), shows patches of a gas
Figs. 15-68 to 15-71. Contrary to common belief, a dry microannulus, liquid, and cement in a distribution that

Cement Map
with Map with
Microdebonding Thresholds
–1,000.0000
Bonded
–500.0000 –500.0000
0.3000 0.3000
3.6000 3.6000 Microdebonding
VDL 4.0000
4.4000
4.0000
4.4000

Min. Amplitude Max. 5.2000 5.2000 Liquid

5.6000 5.6000
6.0000 6.0000
6.4000 6.4000
Variable Density Log 7.2000 7.2000 Gas or Dry
200 (µs) 1,200
7.6000
8.0000
7.6000
8.0000
Microannulus

X000

X100

Bond
index

Fig. 15-70. A USI and CBL-VDL log of microdebonded low-density cement.

Chapter 15 Cement Job Evaluation 607

is difficult to explain physically. This appearance is typi- Acoustic
cal of variable cement bonding. The microdebonding Impedance
processing classifies such areas of variable impedance –999.0
as weakly bonded solids (green in Track 2). The 0.5
microdebonding image and the BI (Track 4) indi- 1.0
cate 80–100% cement around the casing over much of 1.5
2.0
the interval. Remaining red patches are larger areas of 2.5 Cement Map
dry debonding and should also be counted as cement. It 3.0 with Map with
Microdebonding Thresholds
can be concluded that there is good zonal isolation 3.5
4.0 –1,000.0000 Bonded
across the interval, and a squeeze operation would not 4.5 –500.0000 –500.0000
be recommended. 5.0
0.3000
3.6000
0.3000
3.6000
Microdebonding
5.5 4.0000 4.0000
4.4000 4.4000 Liquid
Gas-affected low-density cement 6.0 5.2000 5.2000
6.5 5.6000 5.6000 Gas
In this well, a 7-in. casing was cemented using 12.2lbm/ 7.0 6.0000 6.0000
6.4000 6.4000
gal mixed-aggregate cement. Figure 15-71 shows the USI 7.5 7.2000 7.2000 BI
and CBL log over a 140-ft interval with a known gas sand 8.0 7.6000 7.6000 1 0
8.0000 8.0000
across the top section. The CBL BI (black line in Track
4) is near 100% over most of the interval. In the lower
section, the USI log indicates 4 to 6 MRayl cement, but
in the upper section the raw image (Track 1) has vari-
able impedance from 0 to 6 MRayl. The conventional
interpreted image (Track 3) shows patches of “liquid”
and gas. The microdebonding processing (Track 2) indi-
cates poorly bonded cement in the upper section, and
the total-cement index (yellow plus green) in Track 4 is
near 100%, agreeing with the CBL. In this example, there
was a known gas source; therefore, the conclusion was
that the cement bond was affected by gas influx.

Casing shape
Manufacturing patterns on the inside of forged casings
often affect the cement image slightly. Figure 15-72
shows a log in which near-vertical stripes on the ampli-
tude and internal-radius images correlate with stripes
on the cement image. While it is important to recognize
these artifacts, they do not impede the interpretation
except in exceptional cases. Despite the strong effect
shown in this example, the well-bonded cement in the
upper part of the image can easily be distinguished from
poorly bonded cement in the lower part. In this example,
there is also some rugosity, indicated by black spots on
the amplitude map and blue flags.

Fig. 15-71. USI log showing gas-affected low-density cement.

608 Well Cementing

 –500.0000
Azimuth of –6.0000
Eccentricity –5.6000
–5.2000
0 (°) 300 –4.8000
–4.4000
–3.6000
CCL –3.2000
–2.6000
–20 (----) 30 –2.4000
–2.0000
–1.6000
Cable –1.2000
Speed –0.6000
(ft/hr) –0.4000
0.5000
0 2,000
Min. of Min. of –500.0000 –500.0000 –1,000.0000
Revolving Amplitude of Internal Internal –0.0780 –0.0780 –500.0000
–500.0000
Echo Minus –0.0680 –0.0680 Micro- 0.30000
Speed Radius Radius 0.5000
debonding 2.5000
Max. –0.0520 –0.0520 1.0000
–0.0440 –0.0440 3.0000
–8 (rps) –8 4.5 (in.) 3 3 (in.) 4.5 1.5000 3.5000
(----) –0.0380 –0.0380 2.0000 4.0000
–0.0280 –0.0280 2.5000 4.5000
Internal Internal –0.0200 –0.0200 3.0000 5.0000
0.5000 –0.0120 –0.0120 3.5000
Revolving 1.5000 Max. of Radius Radius Max. of 5.5000
–0.0040 –0.0040 4.0000 6.0000
Speed 2.5000 Amplitude Max. Max. 0.0040 Thickness 0.0040 4.5000 Liquid
3.5000 6.5000
0.0120 0.0120 5.0000 7.0000
–8 (rps) –8 6.5000 0 dB 75 4.5 (in.) 3 3 (in.) 4.5 0.0200 0.1 (in.) 0.8 5.5000
0.0200 7.5000
0.0280 0.0280 6.0000 8.0000
0.0360 6.5000
Process Internal Internal 0.0360 7.0000
0.0440 0.0440 Gas or
Revolving Flags Average of Radius Radius 0.0520 Average of 0.0520
7.5000 Cement
Speed Amplitude Average Average 0.0600 Thickness 0.0600
8.0000 Dry Map with
(----) 0.0680 Microannulus Impedance
0.0680
0 (rps) 8 0 dB 75 4.5 (in.) 3 3 (in.) 4.5 0.0760 0.1 (in.) 0.6 0.0760
Raw Classification
External External Internal Thickness Acoustic (----)
Min. of Radius Radius Radius minus Min. of minus Impedance
Eccentricity Amplitude Average Average Average Thickness Average Bonded
(----)
0 (in.) 8 0 dB 75 4.5 (in.) 3 3 (in.) 4.5 (----) 0.1 (in.) 0.6 (----)

QC Casing Cement

Fig. 15-72. USI cement log affected by casing shape and rugosity. Casing shape is shown in the internal
radius image (Track 6), and it also affects the amplitude (Track 3). Rugosity appears as black spots on the
amplitude image. Both effects correlate with patterns on the cement images.

Deposits smaller channel, while a second repeat found the chan-

Several conditions, such as corrosion, drillpipe wear, nel had almost disappeared. The casing data revealed a
mud deposits on the low side of the hole, and cement vertical low-amplitude streak (Track 3) coinciding with
deposits, can create the illusion of false channels or gas. the “channel.” The conclusions were twofold. First,
Ultrasonic logs must not be interpreted blindly by look- deposits on the low side of the hole affected the imped-
ing at the final cement image; quality control checks, ance measurement to create the apparent “channel.”
including casing condition, are essential. Second, the deposits were scraped away during the log-
Figure 15-73 shows a log in a high-deviation section ging runs by the large number of centralizers on the log-
that seemed to indicate a narrow channel on the low ging string. This example illustrates how standard qual-
side of the hole. The log was oriented with the low side ity control using the casing and repeat-pass data can
in the center of the image. A first repeat log showed a prevent a false interpretation.

Chapter 15 Cement Job Evaluation 609

 Bonded
Micro-
debonding Cement
Map
Liquid with
SP Echo Internal Raw Micro-
Gas or Dry
Auxiliary Flags Amplitude Cross Section Radius Thickness Thickness Impedance Microannulus debond

X000 Log 1

Log 2
X100

X200 Log 3

QC Casing Cement

Fig. 15-73. USI log showing false channel because of deposits on the low side of the hole.

Reflections from an outer casing string spiral around the casing owing to uncorrected tool rota-
The log in Fig. 15-74 was recorded in an 8.625-in. liner tion. The patterns are best observed on the raw imped-
inside a 10.75-in. casing. Strong reflections from the ance image (Track 3). In the lower part of the image, in
outer casing created marked “galaxy” interference pat- which there is good cement bond, the pattern occupies
terns on the narrow side of the annulus. These patterns about half the circumference. However, in the upper part

610 Well Cementing

 Bonded
Micro-
debonding

Raw Liquid Cement Map

Casing Acoustic AI with
SP Gas or Dry
Auxiliary Cross Section Impedance 0 10 Microannulus Microdebond

X400

X500

X600

Fig. 15-74. USI log showing interference patterns resulting from reflections from an outer casing string
through mud (top) and cement (bottom).

in which there is free pipe, the pattern is narrower. ference pattern does cause the microdebonding algo-
Interference patterns are usually indicative of good rithm (green color in Tracks 5 and 6) to falsely indicate
cement, and they occur in free pipe only where the annu- debonding in the cement map, just as it causes the false-
lus is very thin. In this example, the liner was not cen- gas (red) coloring. No reflections can occur through an
tralized. The mean acoustic impedance (Track 4, black) extended gas microannulus, because it acts as a barrier
is hardly affected by reflections because the interfer- to ultrasound.
ence effects balance out on average. However, the inter-

Chapter 15 Cement Job Evaluation 611

15-5 Conclusions 15-6 Acronym list
The acoustic impedance of the material in the annulus is AI Acoustic impedance
one of many parameters that influences acoustic mea- API American Petroleum Institute
surements. In particular, both sonic and ultrasonic tools
BI Bond index
are sensitive to the bond between the material and the
pipe, but in different ways. Ultrasonic logs are generally BPI Percentage bonding
easier to interpret and less ambiguous than sonic logs, BWOC By weight of cement
but the combination of sonic and ultrasonic cement logs BWOW By weight of water
provides more information than either log alone. For
CBL Cement bond logs
optimal evaluation, knowledge of well data, cement job
events, and pre- and postjob well histories is vital. DST Drillstem test
FPM Fluid properties measurement
OBM Oil-base mud
OD Outside diameter
PIT Pressure-integrity test
TT Transit time
VDL Variable-density log
WBM Water-base mud

612 Well Cementing

For symbol definitions, please see the nomenclature,
p. 679.
Digest of Rheological Equations
Dominique Guillot—Schlumberger

A
Table A-1. Definitions of the Main Parameters
Pipe Flow Annular Flow Slot Approximation
2 τw 2 τw
Fanning friction factor ffr = b ffr = b
ρv( ) 2
( )2
ρv

where b † = 154.84 b = 154.84

dw ⎛ dp ⎞ d o − dw ⎛ dp ⎞
shear stress at the wall τw = b ⎜⎝ dz ⎟⎠ τw = b ⎜⎝ dz ⎟⎠
4 f
4 f

b = 1.2 b = 1.2

4q 4q
and v =b v =b
fluid mean velocity ( )
π dw
2
( ) − (dw )2 ⎤⎥⎦
π ⎡⎢ do
2

⎣
b = 808.50
b = 808.50

Reynolds number in
laminar flow
(NRe )MR = 16
f
(NRe )AN = 24
f
fr fr
†b is the oilfield unit conversion factor

Well Cementing ■ Appendix A Digest of Rheological Equations 613

Table A-2. Main Equations for Laminar Flow
Pipe Flow Annular Flow Slot Approximation
8v 12 v
γ NW = b γ NW = b
Newtonian shear rate dw d o − dw
at the wall
b = 0.2 b = 0.2
3n ′ + 1  2n ′ + 1 
Shear rate at the wall γ w = γ γ w = γ
4n ′ NW 3n ′ NW

where the local n′ =

dlog τw ( ) n′ =
dlog τw ( )
power-law index (
dlog γ NW ) (
dlog γ NW )
τw τw
d d −d
Velocity profile† if τ(x) ≤ τy v (x ) = constant = b w
2 τw
τy
∫ γ (τ)dτ v (x ) = constant = b o w
4τw ∫ γ (τ)dτ
τy

τw τw
d d −d
Velocity profile if τ(x) ≥ τy v (x ) = b w
2 τw ∫ γ (τ)dτ
τ( x )
v (x ) = b o w
4 τw ∫ γ (τ)dτ
τ( x )

b =5 b =5

Note: x is the normalized pipe axis, i.e. plane of symmetry of the slot:
distance to the:
2r 2r
x= x=
dw d o − dw

and shear stress at x is equal to τ(x ) = x τw τ(x ) = xτw

τw τw
4 3
( ) ∫ ∫
Newtonian shear-rate/ γ NW (τw ) = γ (τ)τ 2dτ γ NW (τw ) = γ (τ)τdτ
shear-stress relationship τw
3
τy ( )
τw
2
τy

† For annular flow, the integral is calculated from the plane of symmetry of the slot to one of the walls.

614 Well Cementing

Table A-3. Definition of the Local Power-Law Parameters and Friction Factor/Reynolds Number Relationships
Pipe Flow Annular Flow Slot Approximation
All flow regimes

( )2−n ′ (dw )n ′ ( )2−n ′ (do − dw )

n′
ρv ρv
Metzner and Reed (NRe )MR = b1 (b2 ) n′
×
8n ′−1 × k pipe
′
(NRe )AN = b1 (b2 ) n′
×
12n ′−1 × ka′nn
Reynolds number†

b1 = 0.000064584 b1 = 0.000064584

b2 = 5 b2 = 5

where n′ =
( )
dlog τw
n′ =
( )
dlog τw
⎛ 8v ⎞ ⎛ 12v Lam ⎞
dlog⎜ lam ⎟ dlog⎜ ⎟
⎝ dw ⎠ ⎝ d o − dw ⎠

−n ′ −n ′
⎛ 8v lam ⎞ ⎛ 12v lam ⎞
and ′ = b1 b2
k pipe ( )−n ′ × τw ⎜⎝ dw ⎟⎠
′ = b1 b2
kann ( )−n ′ × τw ⎜⎝ d ⎟
o − dw ⎠

b1 = 0.01 b1 = 0.001

b2 = 0.2 b2 = 0.2

Laminar flow (NRe )MR ≤ NRe1 = 3250 − (1150 × n ′ ) (NRe )AN ≤ NRe1 = 3250 − (1150 × n ′ )
16 24
Fanning friction ffr = ffr =
factor (NRe )MR (NRe )AN
Transitional flow ( )MR ≤ NRe2
NRe1 ≤ NRe NRe1 ≤ NRe ( )AN ≤ NRe2
⎛f ⎞ ⎛f ⎞
log⎜ fr 2 ⎟ log⎜ fr 2 ⎟
Fanning friction
⎛f ⎞
log⎜ fr ⎟ =
⎝ ffr 1 ⎠ ⎡ N ( )
× log ⎢ Re MR
⎤
⎥
⎛f ⎞
log⎜ fr ⎟ =
⎝ ffr 1 ⎠ ⎡ N ( )
× log ⎢ Re AN
⎤
⎥
factor ⎝ ffr 1 ⎠ ⎛ NRe 2 ⎞ ⎢⎣ NRe1 ⎥⎦ ⎝ ffr 1 ⎠ ⎛ NRe 2 ⎞ ⎢⎣ NRe1 ⎥⎦
log⎜ ⎟ log⎜ ⎟
⎝ Re1 ⎠
N ⎝ Re1 ⎠
N

16 24
where ffr 1 = ffr 1 =
NRe1 NRe1

⎧ ⎡ ⎛ n '⎞ ⎤⎫ ⎧
⎪ 4.0 ⎡2 ⎛ n′ ⎞ ⎤⎫
1 ⎪ 4.0
=⎨ × log ⎢NRe 2 × ffr 2 ( )1−⎜⎝ 2 ⎟⎠ ⎥ ⎪⎬ − 0.4 1
=⎨ × log ⎢ NRe 2 × ffr 2 ( )1−⎜⎝ 2 ⎟⎠ ⎥ ⎪⎬ − 0.4
ffr 2 ( )
⎪ n'
⎩
0.75
⎢⎣ ⎥⎦ ⎪
⎭ (n ') 1.2
ffr 2 ( )
⎪ n′
⎩
0.75
⎢⎣ 3 ⎥⎦ ⎪
⎭ (n ′ )1.2
Turbulent flow‡ (NRe )MR ≥ NRe2 = 4150 − (1150 × n ′ ) (NRe )AN ≥ NRe2 = 4150 − (1150 × n ′ )
⎧
⎪ 4.0 ⎡ ⎛ n′ ⎞ ⎤⎫ ⎧ ⎡2 ⎛ n′ ⎞ ⎤⎫
Fanning friction
1
=⎨ × log ⎢ NRe ( )MR × (ffr )1−⎜⎝ 2 ⎠⎟ ⎥ ⎪⎬ − 0.4 1 ⎪ 4.0
=⎨ × log ⎢ NRe ( )AN × (ffr )1−⎜⎝ 2 ⎟⎠ ⎥ ⎪⎬ − 0.4
( ) ( ) ( ) (n ′ )1.2
0.75 1.2 0.75
factor ffr ⎪ n ′ ⎢⎣ ⎥ ⎦ ⎪⎭ n′ ffr ⎪ n ′ ⎢⎣ 3 ⎥ ⎦ ⎪⎭
⎩ ⎩
–
†v is the mean velocity at which the fluid would flow (in the laminar flow regime) for a given shear stress at the wall, τw . If the flow regime is not laminar,
lam
this velocity is not the actual mean velocity of the fluid.
‡ For low values of the local power-law index, this definition of NRe2 can lead to inconsistent friction-factor values just above NRe1 (see Tables A-5, A-6a, and A-7).

Appendix A Digest of Rheological Equations 615

Table A-4. Newtonian Fluid† Relationships
Pipe Flow Annular Flow Slot Approximation

NRe = b
ρvdw
(NRe )AN = b
ρv do − dw ( )
Reynolds number μ μ

b = 15.461 b = 15.461

Laminar flow NRe ≤ NRe1 = 2100 (NRe )AN ≤ NRe1 = 2100

μNRe1 μNRe1
Max. velocity for laminar v1 = b v1 = b
flow ρdw (
ρ d o − dw )
b = 0.064677 b = 0.064677
16 24
ffr = ffr =
Fanning friction factor NRe (NRe )AN
⎛ dp ⎞ 32 μv ⎛ dp ⎞ 48μv
⎜⎝ dz ⎟⎠ = b ⎜⎝ dz ⎟⎠ = b
( ) ( )
2 2
Frictional pressure drop f dw f d o − dw

b = 0.00034809 b = 0.00034809

Normalized velocity
profile
v (x )
v
(
= 2 1− x 2 ) v (x ) 3
v
=
2
1− x 2 ( )
Turbulent flow NRe ≥ NRe 2 = 3000 (NRe )AN ≥ NRe2 = 3000
μNRe 2 μNRe 2
Min. velocity for v2 =b v2 = b
turbulent flow ρdw (
ρ d o − dw )
b = 0.064677 b = 0.064677

Fanning friction factor

1
ffr
(
= 4.0log NRe × ffr − 0.4 ) 1
ffr
⎡2
= 4.0log ⎢ NRe
⎣3
( )AN × ⎤
ffr ⎥ − 0.4
⎦

−0.26
⎡2
( )
ffr = 0.0792 NRe
−0.25
ffr = 0.0893 ⎢ NRe ( )AN ⎤⎥
Power-law approximation ⎣3 ⎦
for N Re2 ≤ N Re ≤ 105
−0.26
or ffr = 0.0992 ⎡ NRe
⎣ ( )AN ⎤⎦
ρ0.75 × μ 0.25 × (v ) ρ0.74 × μ 0.26 × (v )
1.74
⎛ dp ⎞ 1.75
⎛ dp ⎞
= b × 0.1584 = b × 0.1984
⎝⎜ dz ⎠⎟ f
Friction pressure drop
⎝⎜ dz ⎠⎟ f dw ( )
1.25
d o − dw (
1.26
)
using the power-law
approximation
b = 0.0027141 b = 0.0026408

Transitional flow NRe1 ≤ NRe ≤ NRe 2 NRe1 ≤ NRe( )AN ≤ NRe2

⎛f ⎞ ⎛f ⎞
log⎜ fr 2 ⎟ log⎜ fr 2 ⎟
Fanning friction factor
⎛f ⎞
log⎜ fr ⎟ =
⎝ ffr 1 ⎠ ⎛N ⎞
× log⎜ Re ⎟
⎛f ⎞
log⎜ fr ⎟ =
⎝ ffr 1 ⎠ ⎡ N AN
× log ⎢
Re ( ) ⎤
⎥
⎝ ffr 1 ⎠ ⎛ NRe 2 ⎞ ⎝ NRe1 ⎠ ⎝ ffr 1 ⎠ ⎛ NRe 2 ⎞ ⎢ NRe1 ⎥
log⎜ ⎟ log⎜ ⎟ ⎣ ⎦
⎝ NRe1 ⎠ ⎝ NRe1 ⎠

16 24
where ffr 1 = ffr 1 =
NRe1 NRe1

1 1 ⎡2 ⎤
= 4.0 × log ⎡NRe 2 × ffr 2 ⎤ − 0.4 = 4.0 × log ⎢ NRe 2 × ffr 2 ⎥ − 0.4
⎣ ⎦
ffr 2 ffr 2 ⎣3 ⎦

( ) ( )
−0.25 −0.26
or if the power-law ffr 2 = 0.0792 NRe 2 ffr 2 = 0.0992 NRe 2
approximation can
( ) ( )AN ⎤⎦
be used in turbulent 0.9525 1.3593
ffr = 5.2178 × 10−6 NRe ffr = 2.3226 × 10−7 ⎡ NRe
flow ⎣

⎡ 1.9525 2.9525 ⎤ ⎡ ⎤
Friction pressure drop ⎛ dp ⎞ −5 ⎢ ρ × (v ) ⎥ ⎛ dp ⎞ −7 ⎢ ρ2.3593 × (v )
3.3593
⎥
⎜⎝ dz ⎟⎠ = b × 1.0436 × 10 ⎢ 0.9525 0.0475 ⎥ ⎜⎝ dz ⎟⎠ = b × 4.6453 × 10 ⎢ 1.3593
using the power-law
approximation in
f ⎢⎣ μ × dw ⎥⎦ ( ) f ⎢⎣ μ d o − dw
−0.3593 ⎥
( ⎥⎦ )
turbulent flow
b = 0.073064 b = 0.22259
.
† Defined as fluids for which τ = μ × γ .

616 Well Cementing

Table A-5. Power-Law Fluid† Relationships
Pipe Flow Annular Flow Slot Approximation

(dw )n (do − dw )
n
ρ(v )
2 −n
ρ(v )
2 −n
Reynolds number (NRe )MR = b1 (b2 ) n
×
8n −1 × k pipe
(NRe )AN = b1 (b2 ) n
×
12n −1 × kann

b1 = 0.000064584 b1 = 0.000064584

b2 = 5 b2 = 5

n n
⎛ 3n + 1⎞ ⎛ 2n + 1⎞
where k pipe = ⎜ k kann = ⎜ k
⎝ 4n ⎠⎟ ⎝ 3n ⎠⎟

Laminar flow (NRe )MR ≤ NRe1 = 3250 − (1150 × n ) (NRe )AN ≤ NRe1 = 3250 − (1150 × n )
1 1

Max. velocity for v 1 = b1 b2 ( )

1
2 −n
⎡ 8n −1 × k
⎢ (
pipe × NRe1 ) ⎥
⎤ 2 −n
v 1 = b1 b2 ( )
1
2 −n
⎡ 12n −1 × k × N
⎢ ann Re1 ( ) ⎤ 2 −n
⎥
⎢ ⎥ ⎢ ⎥
laminar flow ⎢⎣ ρ × dw ( )
n
⎥⎦ ⎢⎣ ρ × d o(− d w
n
) ⎥⎦

b1 = 5 b1 = 5

b2 = 619.35 b2 = 619.35

16 24
Fanning friction ffr = ffr =
factor (NRe )MR (NRe )AN
2 3n +2 × k pipe × (v ) 22n +2 × 3n × kann × (v )
n n
⎛ dp ⎞ ⎛ dp ⎞
⎜⎝ dz ⎟⎠ = b1 b2 ( )n ⎜⎝ dz ⎟⎠ = b1 b2 ( )n
(dw ) (do − dw )
n +1 n +1
f f
Frictional pressure
drop b1 = 83.333 b1 = 83.333

b2 = 0.2 b2 = 0.2

Normalized velocity
profile
v (x ) 3n + 1
v
= 1+
n + 1 1− x n
1
( ) v (x ) 2n + 1
v
= 1+
n + 1 1− x n
1
( )
Turbulent flow‡ (NRe )MR ≥ NRe2 = 4150 − (1150 × n ) (NRe )AN ≥ NRe2 = 4150 − (1150 × n )
1 1
⎡ 8n −1 × k ⎤ 2 −n ⎡ n −1 ⎤ 2 −n
pipe × NRe 2 ⎥ ⎢ 12 × kann × NRe 2 ⎥
1 1
Min. velocity for v 2 = b1 b2 ( ) 2 −n ⎢
⎢ ⎥
v 2 = b1 b2 ( ) 2 −n
⎢ ⎥
turbulent flow ⎢⎣ ρ × dw ( )
n
⎥⎦ (
⎢⎣ ρ × do − dw
n
⎥⎦ )
b1 = 5 b1 = 5

b2 = 619.35 b2 = 619.35

⎡ ⎛n⎞ ⎤ ⎡2 ⎛n⎞ ⎤
Fanning friction factor
1
=
4.0
0.75 ( )MR × (ffr )1−⎜⎝ 2 ⎟⎠ ⎥ − n01.4.2
× log ⎢ NRe
1
=
4.0
0.75
× log ⎢ NRe ( )AN × (ffr )1−⎜⎝ 2 ⎟⎠ ⎥ − n01.4.2
ffr n ⎢⎣ ⎥ ⎦ ffr n ⎢⎣ 3 ⎥ ⎦
b (n )
⎡2
( )MR ⎤⎦ ( )AN ⎤⎥
b (n )
ffr = a(n) × ⎡ NRe ffr = a(n) × ⎢ NRe
Power-law ⎣ ⎣3 ⎦
approximation
for NRe2 ≤ NRe ≤ 105 a(n) = 0.0792 + 0.0207 × log(n) a(n) = 0.0792 + 0.02207 × log(n)
b (n) = −0.251+ 0.141 × log(n) b (n) = −0.251+ 0.141 × log(n)
continued on next page

Appendix A Digest of Rheological Equations 617

Table A-5. Power-Law Fluid† Relationships (continued)
Pipe Flow Annular Flow Slot Approximation

⎛ dp ⎞ ⎡ ⎛ dp ⎞ ⎡
( )b (n) × (b3 ) ( )b (n) × (b3 )
n ×b (n ) ⎤ n ×b (n ) ⎤
⎜⎝ dz ⎟⎠ = ⎢⎣b1 × b2 ⎥⎦ × ⎜⎝ dz ⎟⎠ = ⎢⎣b1 × b2 ⎥⎦ ×
f f

2a(n) × ρ1+b (n ) × (v )
2+(2−n )b (n ) b (n ) ⎡ 2a(n) × ρ1+b (n ) × (v )2+(2−n )b (n ) ⎤
⎛ 2⎞
×⎢ ⎥
(8 ) ( ) ⎜⎝ 3 ⎟⎠
( )
b (n )
)
1−nb (n )
(
b (n ) 1−nb (n ) ⎥
Friction pressure drop n −1
× k pipe × dw ⎢ 12n −1 × k × d − d
using the power-law ⎢⎣ ann o w ⎥⎦
approximation
b1 = 0.0053820 b1 = 0.0053820

b2 = 6.4584 × 10−5 b2 = 6.4584 × 10−5

b3 = 0.2 b3 = 0.2

Transitional flow NRe1 ≤ NRe1 ( )MR ≤ NRe2 (

NRe1 ≤ NRe1 )AN ≤ NRe2
⎛f ⎞ ⎛f ⎞
log⎜ fr 2 ⎟ log⎜ fr 2 ⎟
Fanning friction factor
⎛f ⎞
log⎜ fr ⎟ =
⎝ ffr 1 ⎠ ⎛ N
× log⎜
Re MR
⎞
⎟
( ) ⎛f ⎞
log⎜ fr ⎟ =
⎝ ffr 1 ⎠ ⎛ N
× log⎜
Re AN
⎞
⎟
( )
⎝ ffr 1 ⎠ ⎛N ⎞ ⎜⎝ NRe1 ⎟⎠ ⎝ ffr 1 ⎠ ⎛N ⎞ ⎜⎝ NRe1 ⎟⎠
log⎜ Re 2 ⎟ log⎜ Re 2 ⎟
⎝ NRe1 ⎠ ⎝ NRe1 ⎠

16 24
where ffr 1 = ffr 1 =
NRe1 NRe1

⎡ ⎛n⎞ ⎤ ⎡2 ⎛n⎞ ⎤
1
=
4.0
0.75
× log ⎢NRe 2 × ffr 2 ( )1−⎝⎜ 2 ⎠⎟ ⎥ − n01.4.2 1
=
4.0
0.75
× log ⎢ NRe 2 × ffr 2 ( )1−⎝⎜ 2 ⎠⎟ ⎥ − n01.4.2
ffr 2 n ⎢⎣ ⎥⎦ ffr 2 n ⎢⎣ 3 ⎥⎦

b (n )
⎛2 ⎞
( )
b (n )
or if the power-law ffr 2 = a(n) × NRe 2 ffr 2 = a(n) × ⎜ NRe 2 ⎟
approximation can ⎝3 ⎠

( )MR ⎤⎦
be used in turbulent d (n )
ffr = c(n) × ⎡ NRe
( )AN ⎤⎦
d (n )
flow ⎣ ffr = c(n) × ⎡ NRe
⎣
⎡
( ) ⎤
b (n )
log ⎢ a(n) × NRe1 × NRe 2 ⎡ b (n ) ⎤
⎥ ⎛ ⎞
⎢⎣ ⎥⎦ ⎢ a(n) × N × 2 N ⎥
d (n) =
16 log ⎢ Re1
⎝⎜ 3 Re 2
⎠⎟ ⎥
⎛ NRe 2 ⎞ ⎢⎣ ⎥⎦
log⎜ d (n) = 24
4
⎟
⎝ NRe1 ⎠ ⎛N ⎞
log⎜ Re 2 ⎟
16 ⎝ NRe1 ⎠
c(n) =
(NRe1 )
1+d (n )
24
c(n) =
(NRe1 )
1+d (n )

.
† Defined as fluids for which τ = k γ n.
‡ Forlow values of the power-law index, this definition of N Re can lead to friction-factor values just above N Re1 that are lower than those given by the laminar flow equation.
This is physically impossible. This problem can be solved by imposing NRe2 values that are always larger than those defined by the point of intersection of the friction-factor/
Reynolds-number curves in laminar and turbulent flow.

618 Well Cementing

Table A-6a. Bingham Plastic Fluid† Relationships
Pipe Flow Annular Flow Slot Approximation

(NRe )BG = b ×
ρ × v × dw
(NRe )BG = b ×
ρ × v × d o − dw ( )
Bingham Reynolds μp μp
number
b = 15.461 b = 15.461

( ) ( )
2 2
ρ × dw × τy ρ × d o − dw × τy
NHe = b × NHe = b ×
(μ ) (μ )
2 2
Hedström number p p

b = 37, 015 b = 37, 015

τy τy
Dimensionless shear ψ= ψ=
stress τw τw

All flow regimes

⎛ ⎛
(1− ψ )⎜⎝1+ 23 ψ + 31 ψ 2 ⎞⎟⎠ (1− ψ )⎜⎝1+ 21 ψ ⎞⎟⎠
n′ = n′ =
(1+ ψ )(1+ ψ 2 ) (1+ ψ + ψ )
Local power law
2

n′ n′
1−n ′ ⎛ μp ⎞ 1−n ′ ⎛ μp ⎞
n ′ ⎛ τy
⎞ n ′ ⎛ τy
⎞
and consistency index ′ = b1 b2
k pipe ( ) ⎜ ⎟
ψ
⎜
4 1
⎟
⎜ 1− ψ + ψ 4 ⎟
′ = b1 b2
kann ( ) ⎜ ⎟
ψ
⎜
3 1
⎟
⎜ 1− ψ + ψ 3 ⎟
⎝ ⎠ ⎝ 3 ⎝ ⎠
3 ⎠ ⎝ 2 2 ⎠

b1 = 0.01 b1 = 0.01

b2 = 0.0020885 b2 = 0.0020885

Laminar flow (NRe )MR ≤ NRe1 = 3250 − (1150 × n ′ ) (NRe )AN ≤ NRe1 = 3250 − (1150 × n ′ )
16 24
Fanning friction factor ffr = ffr =
(NRe )MR (NRe )AN
τy 1 ⎛ 4 1 ⎞ τy 1 ⎛ 3 1 ⎞
Frictional pressure drop γ NW = b × 1− ψ + ψ 4 ⎟ γ NW = b × 1− ψ + ψ 3 ⎟
μ p ψ ⎜⎝ 3 3 ⎠ μ p ψ ⎜⎝ 2 2 ⎠

b = 478.80 b = 478.80

( ) ( )
4 3
16 8NHe 16 NHe 24 3NHe 4 NHe
or ffr = + − ffr = + −
(NRe )BG ( )BG ⎤⎦
3 ⎡ NRe
2
3 (NRee )BG ⎤ × (ffr )
⎡
8 3 (NRe )BG ( )BG ⎤⎦
3 ⎡ NRe
2
⎡(N ) ⎤ × (f )2
6

⎣ ⎣ ⎦ ⎣ ⎣ Re BG ⎦ fr

⎡ ⎤ ⎡ ⎤
Normalized velocity
( ) = 2 ⎢⎢ (1− ψ )2 − ( x − ψ )2 ⎥⎥
v x v x ( ) = 3 ⎢⎢ (1− ψ )2 − ( x − ψ )2 ⎥⎥
profile when τx ≥ τy v ⎢ 1− 4 ψ + 1 ψ 4 ⎥ v 2⎢ 3 1
1− ψ + ψ 3 ⎥
⎢⎣ 3 3 ⎥⎦ ⎢⎣ 2 2 ⎥⎦

and when τx ≤ τy
v x( )= 2 ( )=
v x 3
v 2 1 v 2+ψ
1+ ψ + ψ 2
3 3

Other flow regimes‡, § (NRe )MR ≥ NRe1 = 3250 − (1150 × n ′ ) (NRe )AN ≥ NRe1 = 3250 − (1150 × n ′ )
.
† Defined as fluids for which τ = τ y + μ p × γ .
‡ Flow equations are solved numerically using the local power-law approach.
§ For high values of the Hedström number, the definition of N
Re2 can lead to Bingham plastic Reynolds-number values that do not increase continuously
when the dimensionless shear stress, ψ, decreases from 1 to 0. This is not physically sound; however, in practice this is not a problem because it is
nearly impossible to reach turbulence when the Hedström number is relatively high. If necessary, NRe2 values that are much higher than those given
by the proposed linear equation should be used for low values of the local power index.

Appendix A Digest of Rheological Equations 619

Table A-6b. Approximations for Bingham Plastic Fluids when ψ << 1
Pipe Flow Annular Flow Slot Approximation

(NRe )BG (NRe )BG

Reynolds number (NRe )MR = ⎡ ⎤
(NRe )AN = ⎡ ⎤
NHe NHe
⎢1+ ⎥ ⎢1+ ⎥
⎣ ( )
⎢ 6 NRe BG ⎥
⎦
⎢ 8 NRe BG
⎣ ( ) ⎥
⎦

Laminar flow (NRe )MR ≤ NRe1 = 3250 − (1150 × n ′ ) (NRe )AN ≤ NRe1 = 3250 − (1150 × n ′ )
1 1
Local power-law index n ′ = 1− n ′ = 1−
1+
( )BG
6 NRe
1+
( )BG
8 NRe
NHe NHe

μ p × NRe1 ⎛ 2N ⎞ μ p × NRe1 ⎛ N ⎞
v1 = b × ⎜ 1+ 1+ He ⎟ v1 = b × ⎜ 1+ 1+ He ⎟
Max. velocity for
laminar flow
2 × ρ × dw ⎜⎝ 3NRe1 ⎟⎠ (
2 × ρ × d o − dw ) ⎜⎝ 2NRe1 ⎟⎠

b = 0.0064677 b = 0.064677

⎡ 32 × μ p × v ⎤ ⎛ 16τy ⎞ ⎡ 48 × μ p × v ⎤ ⎡ ⎤
⎛ dp ⎞ ⎢ ⎥+ b × ⎛ dp ⎞ ⎢ ⎥ + ⎢b × 6τy ⎥
⎜⎝ dz ⎟⎠ ⎢b1 ×
=
⎥ ⎜⎝ 2 3dw ⎟⎠ ⎜⎝ dz ⎟⎠ ⎢b1 ×
=
f ⎢⎣ dw ( )
2
⎥⎦ f ⎢⎣ (
d o − dw
2
) ⎥ ⎢ 2 d −d
⎥⎦ ⎣ o( w ) ⎥⎦
Frictional pressure
drop
b1 = 0.00034809 b1 = 0.00034809

b2 = 0.83333 b2 = 0.83333

Other flow regimes† (NRe )MR ≥ NRe1 = 3250 − (1150 × n ′ ) (NRe )AN ≥ NRe1 = 3250 − (1150 × n ′ )
† Flow equations are solved numerically using the local power-law approach.

620 Well Cementing

Table A-7. Herschel-Bulkley Fluid† Relationships
Pipe Flow Annular Flow Slot Approximation
τy τy
Dimensionless shear ψ= ψ=
stress τw τw

All flow regimes

Local power-law n ′ = n 1− ψ ×( ) n′ =
( )(
n 1 − ψ 1 + n + nψ )
index
⎡
⎢
( )( )
n + 1 2n + 1 + 2n n + 1 ψ + 2n 2 ψ 2 ( ⎤
⎥
) (1+ n + 2nψ + 2n ψ ) 2 2

⎣ ( )( ) ( )
⎢ n + 1 2n + 1 + 3n n + 1 ψ + 6n 2 ψ 2 + 6n 3 ψ 3 ⎥
⎦

Laminar flow (NRe )MR ≤ NRe1 = 3250 − (1150 × n ′ ) (NRe )AN ≤ NRe1 = 3250 − (1150 × n ′ )
⎡ 1 1 ⎤ 1

Frictional pressure γ NW
1
=bn
⎢⎛ τy ⎞ n 4n 1− ψ
× ⎢⎜ ⎟
( )1+ n ⎥⎥ γ NW
1
=bn
⎛ τy ⎞ n
×⎜ ⎟
1
drop ⎢⎝ k ⎠ ⎥ ⎝ k ⎠
⎢⎣ ψ n
⎦⎥ ⎡ 1 ⎤
( )
⎡ 1− ψ 2 2 ψ 1− ψ
ψ2 ⎥
⎤
( ) × ⎢
( 1+
) (
⎢ 3n 1− ψ n 1+ n + nψ ) ⎥⎥
×⎢ + + ⎢
1
⎥
⎢ 3n + 1
⎢⎣
2n + 1 n + 1⎥
⎥⎦ ⎢⎣ ( )( )
n + 1 2n + 1 ψ n ⎥⎦

b = 0.01
b = 0.01
1
1+

1− ⎢
⎡ x −ψ ⎤ n
⎥
( ) v (x )
(
= 2n + 1 )
Normalized velocity v (x )
=
⎢⎣ 1− ψ ⎥⎦ ( ) ⎡
v
1 1 ⎤
profile when τx ≥ τy v ⎡ n +1 2 ⎤ ⎡ 2 n +1 ( ) ( ⎤ ( ) − ( x − ψ )1+ n ⎥
⎢ 1- ψ
1+
n

( )
⎢ 3n + 1 1− ψ ⎥ + ⎢ 2n + 1 ψ 1− ψ ) ⎥ + ψ2 × ⎢ 1 ⎥
⎣ ⎦ ⎢⎣ ⎥⎦
⎣ ( )1+ n (1+ n + nψ ) ⎥⎦
⎢ 1− ψ

and when τx ≤ τy
v (x )
=
(
2n + 1 3n + 1 )( ) v (x )
=
(2n + 1)
v ⎡⎣(n + 1)(2n + 1) ⎤⎦ + ⎡⎣2n (n + 1) ψ ⎤⎦ + 2n 2 ψ 2 v (1+ n + nψ )
Other flow regimes‡, § (NRe )MR ≥ NRe1 = 3250 − (1150 × n ′ ) (NRe )AN ≥ NRe1 = 3250 − (1150 × n ′ )
.
† Defined as fluids for which τ = τy + k × γ n.
‡ Flow equations are solved numerically using the local power-law approach.
§ For low values of the power-law index, this definition of N
Re2 can lead to predictions of friction factors that are not physically sound.
The problem can be solved by using much higher N Re2 values than those given by the proposed linear equation.

Appendix A Digest of Rheological Equations 621

Table A-8. Dimensionless Breakthrough Time and Circulation Efficiency in Laminar Flow
Pipe Flow Annular Flow Slot Approximation
v ×t v ×t
Dimensionless time t* = t* =
L L
t *break t *break
Additional parameter J = 1− J = 1−
t* t*

Circulation efficiency
for t * ≤ t *break ( )
ηcirc t * = t * ( )
ηcirc t * = t *

Newtonian fluids
1 2
Dimensionless t *break = t *break =
breakthrough time 2 3

⎛ 3⎞
Circulation efficiency
for t * ≥ t *break
( ) ( )
ηcirc t * = 1− J 2 × t * = 1−
4t
1
* ( )
ηcirc t * = ⎜ 1− J 2 ⎟ × t *
⎜⎝ ⎟⎠

Power-law fluids
n +1 n +1
Dimensionless t *break = t *break =
breakthrough time 3n + 1 2n + 1

( 3n +1) ⎤ * (2n +1) ⎤ *

Circulation efficiency
for t * ≥ t *break
( ) ⎡
ηcirc t * = ⎢
⎣1 − J (n +1) ⎥⎦ × t ( ) ⎡
ηcirc t * = ⎢
⎣1 − J (n +1) ⎥⎦ × t

Bingham plastic fluids

1 1 1 2+ψ
Dimensionless t *break = + ψ + ψ 2 t *break =
breakthrough time 2 3 6 3

⎧ J ⎫ ⎧ J ⎫
⎪1− t * × ⎪ ⎪1− t * × ⎪
( ) ⎪ ⎪ *
( ) ⎪ ⎪ *
break break
Circulation efficiency ηcirc t * =⎨ ×t ηcirc t * =⎨
for t * ≥ t *break ⎡ 1 ⎤⎬ ⎡ 1 ⎤⎬ × t
4
( 1
)
⎪ ⎢ ψ + ψ 1− ψ J 2 + 1− ψ J ⎥ ⎪
⎪⎢
2
3 2
2

⎥⎦ ⎪
( ) ⎪⎢
2
3
(
⎪ ⎢ ψ + 1− ψ J 2 ⎥ ⎪
⎥⎦ ⎪
)
⎩⎣ ⎭ ⎩⎣ ⎭
Herschel-Bulkley fluids
(n + 1)(2n + 1) + 2n (n + 1) ψ + 2n 2 ψ 2 1 + n + nψ
Dimensionless t *break = t *break =
breakthrough time (
2n + 1 3n + 1 )( ) 2n + 1

⎧ J ⎫
⎪1− t * × ⎪
⎪ break ⎪ ⎧ J ⎫
⎪1− t * × ⎪
⎪⎡
( + ) ( n ⎤⎪
( ) ( )
2 n 1 ⎪ ⎪ *
Circulation efficiency ηcirc t *
⎪ 2
= ⎨⎢ ψ +
( ) )
ψ 1− ψ J n +1 + ⎥ ⎪⎬ × t * ηcirc t * = ⎨⎡
break
n ⎤⎬ × t
for t * ≥ t *break
⎪⎢⎢ 2n + 1 ⎥⎪
⎥ ⎪⎢ ψ + n(+ 1 )(
1− ψ J n +1 ⎥ ⎪ )
( )
⎪⎢ n + 1 2n
⎥⎪ ⎪⎢ (
2n + 1 ) ⎥⎦ ⎪
⎪⎢
( )
(
⎪ ⎢ 3n + 1
2
1− ψ J n +1 ) ⎥⎪
⎪
⎩ ⎣ ⎭
⎩⎣ ⎦⎥ ⎭

622 Well Cementing

A-1 Application of the local-power b) To determine the beginning of the turbulent flow
approach to Bingham plastic fluids regime, it is necessary to solve for ψ2 using the fol-
lowing set of equations (assuming the Fanning frac-
As an example of the application of the local-power tion factor/Reynolds number relationship in turbulent
model to the flow of non-Newtonian fluids in transitional flow can be described by a power-law relationship).
and turbulent flow, the equations governing the flow of
1
Bingham plastic fluids will be described hereafter. It will
⎛ ⎡ 16 ⎤2− n′ ⎞ 2+(22− n′ ) b( n′ )
be assumed that the friction factor/Reynolds number ⎜⎢ ⎟
⎥
relationship in turbulent flow can be described by a ⎜ ⎣ a( n ′ ) ⎦ ⎟
power-law approximation. N Re 2 ( n ′ ) = ⎜ n′ ⎟
⎜ ⎧⎪ N ⎡ 4 ⎫
4⎤ ⎪ ⎟
2
For flow in a pipe, the main equations are the follow-
ing. ⎝ ⎪⎩ 8 ψ 2 ⎣ 3
1
3
( )
⎜ × ⎨ He × ⎢1 − ψ 2 + ψ 2 ⎥ ⎬ ⎟
⎦ ⎪⎭ ⎠
a) To determine the end of the laminar flow regime, it is
necessary to solve for ψ1 using the following set of
equations.
)
N Re2 ( n ′ = 4150 − (1150 × n ′ )
(1 − ψ 2 ) ⎡⎢1 + 32 ψ 2 + 31 ( ψ 2 )2 ⎤⎥
2
N He ⎡ 4 4⎤
N Re 1 ( n ′ ) =
1
1− ψ + ψ
8 ψ 1 ⎢⎣ 3 1 3 1 ⎥⎦
( ) ⎣ ⎦
n′ =
( ) ( )
1 + ψ 2 ⎡1 + ψ ⎤
⎣
2
2 ⎦
)
N Re1 ( n ′ = 3250 − (1150 × n ′ )
a( n ′ ) = 0.0792 + 0.0207 × log( n ′ )

(1 − ψ 1 ) ⎡⎢⎣1 + 32 ψ 1 + 31 ( ψ 1 )2 ⎤⎥⎦ b( n ′ ) = −0.251 + 0.1441 × log( n ′ )

n′ = Once the value of ψ2 is determined, the corre-
(1 + ψ 1 ) ⎡⎣1 + ( ψ 1 )2 ⎤⎦ sponding values for n′, a(n′), and b(n′) can be calcu-
lated and the maximum velocity for the flow regime to
These three equations can be combined into one: be laminar can be derived from
2 1
N He ⎡ 4 4⎤ ⎧⎛ N ⎞ 1+(1− n′ ) b( n′ ) ⎫ 2+(2− n′ ) b( n′ )
⎢
8ψ1 ⎣ 3
1
( )
1 − ψ 1 + ψ 1 ⎥ = 3250 − 1150 ⎪ He ⎪
3 ⎦ ⎪⎜⎝ 8 ψ 2 ⎟⎠ ⎪
⎪ ⎪
⎡ 2 2⎤
( ⎣
)
3
1
1 − ψ 1 ⎢1 + ψ 1 + ψ 1 ⎥
3 ⎦.
( ) ( NRe ) =
BG 2 ⎨
⎪ ⎡ 16 ⎤
⎢ a( n ′ ) ⎥
⎪
⎬ .
× ⎪ ⎣ ⎦ ⎪
⎪×
( ⎡
1 + ψ1 1 + ψ
⎣ )2⎤
1 ⎦ ( ) ⎪ ⎡1 − 4 ψ + 1 ψ 4 ⎤
( )
n ′× b( n ′ ) ⎪
⎪
⎢ 3 2 ⎥⎦ ⎪⎭
⎩⎪ ⎣ 3
2

Once the value of ψ1 is determined, the maximum

velocity for the flow regime to be laminar can be
derived from

( NRe ) BG1 = 8 ψHe ⎡⎢1 − 34 ψ 1 + 31 ( ψ 1 )4 ⎤⎥.

N
1 ⎣ ⎦

Appendix A Digest of Rheological Equations 623

c) The friction pressure drop in turbulent flow can be These three equations can be combined into one:
derived from
⎧ ⎛ 1 ⎞⎫
N He ⎡ 3 3⎤
2
⎪⎪ 1 − ψ 1 ⎜ 1 + ψ 1 ⎟ ⎪⎪ ( )
1
( )
1 ⎝ 2 ⎠ ⎬.
1− ψ + ψ = 3250 ⎨1150 ×
⎧⎛ N ⎞ 1+(1− n′ ) b( n′ ) ⎫ 2+(2− n′ ) b( n′ ) 12 ψ 1 ⎢⎣ 2 1 2 1 ⎥⎦
⎪ He ⎪
⎪
⎪⎩
⎡1 + ψ + ψ 2 ⎤ ⎪
⎣ 1 ⎦ ⎪ ⎭ ( )
⎪⎜⎝ 8 ψ ⎟⎠
1
⎪
⎪ ⎪ Once the value of ψ1 is determined, the maximum
⎪ ⎪
( NRe ) BG = ⎨
⎡ 16 ⎤
⎢ a( n ′ ) ⎥ ⎬ velocity for the flow regime to be laminar can be
⎪ ⎣ ⎦ ⎪ derived from
⎪× n ′b( n ′ ) ⎪
⎪ ⎛1− 4 ψ + 1 ψ4⎞ ⎪
( NRe ) BG1 = 12ψHe ⎡⎢1 − 32 ψ 1 + 12 ( ψ 1 )3 ⎤⎥.
N
⎪⎩ ⎜⎝ 3 3 ⎟⎠ ⎪⎭
1 ⎣ ⎦
d) The friction pressure drop in transitional flow can be f) To determine the beginning of the turbulent flow
derived from regime, one must solve for ψ2 using the following set
1
of equations (assuming the Fanning fraction factor/
⎧⎛ N ⎞ 1+(1− n′ ) d ⎫ 2+(2− n′ ) d Reynolds number relationship in turbulent flow can
⎪ He ⎪ be described by a power-law relationship).
⎪⎜⎝ 8 ψ ⎟⎠ ⎪
⎪ ⎪⎪ 1
( NRe ) BG = ⎨ (
N Re 1 )
1+ d
⎬ . ⎛⎡
⎜ ⎢ 24 ⎛ 3 ⎞
b( n ′ ) ⎤
2− n ′ ⎞ 2+(2− n′ ) b( n′ )
⎟
⎪× ⎪ ⎥
n ′d ⎜ ⎢ a( n ′ ) ⎜⎝ 2 ⎟⎠ ⎟
⎪ ⎛ 4 1 4⎞ ⎪ ⎜ ⎣
⎥
⎦ ⎟
⎪ ⎜1− ψ + ψ ⎟ ⎪ N Re 2 ( n ′ ) =
⎜
⎪⎩ ⎝ 3 3 ⎠ ⎭ n′ ⎟
2⎫
⎜ ⎧⎪ N He ⎡ 3 3⎤ ⎪ ⎟
⎜×⎨
⎝ ⎩⎪ 12 ψ 2 ⎣ 2
⎢ 1 − ψ 2 +
1
ψ( ) ⎬ ⎟
2 2 ⎥⎦ ⎪⎭ ⎠
⎛ f fr 2 ⎞
log ⎜ ⎟ N Re 1 ( n ′ ) = 4150 – (1150 × n ′ )
⎝ f fr1 ⎠
d=
⎛N ⎞
log ⎜ Re 2 ⎟
(1 − ψ 2 )⎛⎜⎝ 1 + 12 ψ 2 ⎞⎟⎠
⎝ N Re 1 ⎠ n′ =
⎡1 + ψ + ψ 2 ⎤
⎢⎣ 2 2 ⎥
⎦
( )
For flow in a narrow concentric annulus, the main a( n ′ ) = 0.0893 + 0.0246 × log( n ′ )
equations are the following.
e) To determine the end of the laminar flow regime, it is b( n ′ ) = −0.263 + 0.138 × log( n ′ )
necessary to solve for ψ1 using the following set of
equations. Once the value of ψ2 is determined, the corre-
sponding values for n′, a(n′), and b(n′) can be calcu-
N He ⎡ 3
2 lated and the maximum velocity for the flow regime to
3⎤
N Re 1 ( n ′ ) = ⎢
12ψ 1 ⎣ 2
1
1 − ψ1 + ψ1 ⎥ ( ) be laminar can be derived from
2 ⎦ 1
⎧ b( n ′ ) 1+(1− n ′ ) b( n ′ ) ⎫ 2+( 2− n ′ ) b( n ′ )
)
N Re1 ( n ′ = 3250 − (1150 × n ′ ) ⎪⎛ 3 ⎞
⎪⎜⎝ 2 ⎟⎠
⎛ N He ⎞
×⎜ ⎟
⎪
⎪
⎝ 12ψ 2 ⎠
⎪ ⎪
⎪ ⎪
(1 − ψ 1 )⎛⎜⎝ 1 + 12 ψ 1 ⎞⎟⎠ ( NRe ) BG 2 = ⎨ ⎡ 24 ⎤
⎢ ⎥
⎬ .
⎪ ⎣ a( n ′ ) ⎦ ⎪
n′ = ⎪× n ′b( n ′ ) ⎪

⎢⎣ 1 ( )
⎡1 + ψ + ψ 2 ⎤
1 ⎥
⎦
⎪ ⎡1 − 3 ψ + 1 ψ 3 ⎤
⎪⎩ ⎢⎣ 2 2 2 2 ⎥⎦ ( ) ⎪
⎪⎭

624 Well Cementing

g) Friction pressure drop in turbulent flow can be A-2 Conversion factors
derived from The b coefficients in oilfield units were determined
1 using the following primary conversion factors.
⎧ b( n ′ ) 1+(1− n ′ ) b( n ′ ) ⎫ 2+( 2− n ′ ) b( n ′ )
⎛N ⎞
⎪⎛ 3 ⎞ × ⎜ He ⎟ ⎪
⎪⎜⎝ 2 ⎟⎠ ⎝ 12 ψ ⎠ ⎪
⎪ ⎪ Unit To be Multiplied by To Get Value in SI Unit
⎪ ⎪
( NRe ) BG = ⎨
⎡ 24 ⎤
⎢ a( n ′ ) ⎥ ⎬ . bbl 0.1589873 m3
⎪ ⎣ ⎦ ⎪
⎪ × n ′b( n ′ ) ⎪ cp 0.001 Pa-s
⎪ ⎛1− 3 ψ + 1 ψ3 ⎞ ⎪
⎪⎩ ⎜⎝ 2 2 ⎟⎠ ⎪⎭ ft 0.3048 m

gal 0.003785412 m3
h) Friction pressure drop in transitional flow can be
in. 0.0254 m
derived from
lbm 0.45359237 kg
1
⎡⎛ N ⎞ 1+(1− n′ ) d ⎤ 2+(2− n′ ) d
⎢ He ⎥
⎢⎜⎝ 12 ψ ⎟⎠ ⎥
⎢ ⎥ Coefficients equal to 1 are not reported. For SI units,
( NRe ) BG = ⎢ (
N Re 1
1+ d
) ⎥ . b coefficients are always equal to 1. The acceleration of
gravity is 9.80665 m-s–2.
⎢× ⎥
⎢ ⎛ 3 n ′d
1 3⎞ ⎥ In this appendix, oilfield units are expressed as
⎢ ⎜1− ψ + ψ ⎟ ⎥ follows.
⎢⎣ ⎝ 2 2 ⎠ ⎥⎦

⎛ f fr 2 ⎞
Oilfield Unit Clarification
log ⎜ ⎟
⎝ f fr1 ⎠ Diameter in.
d=
⎛N ⎞ Length ft
log ⎜ Re 2 ⎟
⎝ N Re 1 ⎠ Velocity ft/min

Flow rate bbl/min

Friction pressure gradient psi/1,000 ft

Pressure psi

Stress lbf/100 ft2

Density lbm/gal

Viscosity cp

Consistency index lbf-sn/ft3

Appendix A Digest of Rheological Equations 625

B-1 Introduction
Laboratory Testing, Evaluation,
and Analysis of Well Cements
Bernard Dargaud and Linda Boukhelifa—Schlumberger

Laboratory testing of cements and cementing materials

is an essential part of the cementing process. Testing
B
American Petroleum Institute (API), the International
Organization for Standardization (ISO), or ASTM
International (formerly the American Society for Testing
begins at the cement and additive manufacturing sites and Materials) for such guidance.
to monitor product quality and continues through the
slurry-design stages at the pumping service company or
operating company laboratories. Samples are frequently B-2 Relationship between the API
obtained from the bulk plant as the blend is prepared, and the ISO
and samples are taken from storage silos when the blend The API Committee on Standardization of Well Cements
is placed on location. Field samples of the dry cement (Committee 10) acts as a governing body regarding the
blends and the resulting slurries can be obtained during development of standardized testing procedures for
mixing for subsequent evaluation, either in the labora- evaluation of the performance of well cement slurries.
tory or on location using portable laboratory equipment. The procedures are published by API as API Specifi-
Laboratory examination of samples obtained from the cation 10A and API Recommended Practice (RP) 10B.
field can be used for post-treatment investigations. Equivalent publications by the ISO are ISO 10426-1 and
In general, there are two types of laboratory testing of ISO 10426-2, respectively.
cements and cementing materials: performance evalua- Specification for Cements and Materials for Well
tion and chemical characterization. The typical oilfield Cementing, API Specification 10A, and ISO 10426-1 con-
laboratory primarily evaluates cement performance tain the requirements for eight classes of well cements,
through physical measurement of specific slurry and set- including chemical and physical requirements and spe-
cement properties under simulated downhole condi- cific procedures for physical testing. These standards
tions. This type of evaluation occurs mainly during the are applicable to well cement classes A, B, C, D, E, F, G,
slurry-design and execution stages of a cementing treat- and H (Chapter 2). These documents present tests that
ment. Chemical characterization typically involves are to be performed on neat cement with no additives.
quantitative or qualitative analysis of the slurry compo- The purpose of the specification tests is to evaluate spe-
nents before and after mixing to ensure their suitability cific chemical and performance requirements of the raw
for use. Analytical techniques are used for quality-con- cement. The tests are highly specific, and no variance
trol purposes at the point of manufacture to determine from the testing protocol is allowed.
that components of a dry-blended cement system are Recommended Practice for Testing Well Cements,
present in the desired quantities and are blended thor- API RP 10B and ISO 10426-2, provides general recom-
oughly at the bulk plant. Such techniques are also used mended practices for testing cement slurries and
to monitor the quality of the mix water on location. related materials under simulated well conditions.
Correct application of a wide variety of laboratory test- These procedures are designed to simulate downhole
ing methods helps achieve a successful cementing treat- conditions for performance testing and are based on a
ment. compromise between realistic wellbore conditions and
This appendix presents a broad overview of labora- the practical limitations of the laboratory environment.
tory testing procedures and equipment for cement test- At this writing there are three additional standards
ing. Most of these techniques are readily available in that cover special aspects of well cementing. They
most well-equipped cement-testing laboratories. In can be considered as supplements to API RP 10B and
addition, a few test methods are presented that are cur- ISO 10426-2.
rently in the domain of research laboratories but may ■ ISO 10426-3, “Testing of deepwater well cement for-
become more common in the future.
mulations.” In a deepwater cementing environment,
This appendix is not intended to be a manual for
a large number of factors affect the thermal history of
cement laboratory workers. The reader must instead con-
the cement slurry (Chapter 12). These factors include
sult the official publications of organizations such as the

Well Cementing ■ Appendix B Laboratory Testing, Evaluation, and Analysis of Well Cements 627
 water depth, geothermal gradient, cement mix-water B-3 Sample preparation
temperature, bulk cement temperature, cement A meaningful laboratory evaluation requires a represen-
mixing rate, cement heat of hydration, and prior cir- tative sample of that material. Statistical considerations
culating and static event history. Test procedures that regarding the choice of sample size may be found in the
reflect the deepwater environment do not exist in ISO ASTM Standard C183, Standard Practice for Sampling
10426-2. Therefore, this standard presents special- and the Amount of Testing of Hydraulic Cement.
ized sampling and testing requirements and dis- Sampling and handling procedures for sacked and bulk
cusses the unique downhole temperature profiles cement are described in API RP 10B. The use of proper
found in deepwater wells. storage procedures is particularly important to avoid
■ ISO 10426-4, “Methods for atmospheric foamed exposure of the cement to moisture and carbon dioxide
cement-slurry preparation and testing.” Foamed in the air. Several studies regarding the sampling of
cements are routinely used in the well cementing blended cements have been reported (Pace et al., 1984;
industry (Chapter 7). Unlike conventional cement Cobb and Pace, 1985; Gerke et al., 1985; Kunze, 1986;
systems, foamed cements contain a gas phase. Bell et al., 1988).
Consequently, special laboratory preparation and The preferred sampling device for blended cements is
testing methods are required. ISO 10426-4 describes a diverted flow sampler (Fig. B-1), which permits sam-
the testing procedures to measure different proper- pling from a complete cross section of a flowing stream
ties of foamed cement, including slurry density, slurry of material. Before testing in the laboratory, field blend
stability, strength development, and permeability. samples may be split using a mechanical splitter as
Some testing cannot be performed directly on the described in ASTM Specification C702, Standard
foamed slurry, and the document indicates alternate Practice for Reducing Samples of Aggregate to Testing
methods for testing thickening time, fluid loss, and Size, because segregation of blended components may
rheology. occur during shipping to the laboratory. Similarly, a
■ ISO 10426-5: “Test methods for determination of mechanical splitter may be appropriate for use in
shrinkage and expansion of well cement formulations obtaining laboratory samples from bulk or sack quanti-
at atmospheric pressure.” The dimensional stability ties of multicomponent additives in the field.
of cement slurries during hydration and after harden-
ing is a fundamental zonal-isolation parameter. As
explained elsewhere in this textbook, the lack of Couplings
dimensional stability can lead to various problems
including
– a microannulus, leading to a bad bond log
(Chapter 15)
– interzonal communication (Chapter 9)
– lack of a hydraulic seal when using cement 5-in. valves
inflated packers (Chapter 11).
ISO 10426-5 describes standard devices and tech-
niques to evaluate internal and external dimensional
changes that occur during cement-slurry hydration,
setting, and hardening.
5-in. discharge pipe
It is not the intent of the API and ISO to simulate well
conditions. The goal is to provide a standard set of test Fig. B-1. Diverted flow sampler (top view).
procedures to allow comparison of results between vari-
ous laboratories. The procedures in API RP 10B and ISO
10426-2 may be modified to match the particular well
conditions and mixing methods used in the field. While B-4 Performance of conventional cement
the tests outlined in API RP 10B and ISO 10426-2 are slurries
applicable to field work, the engineer is responsible for
determining the applicability of a particular test and B-4.1 Slurry preparation
determining if the results are relevant to a particular The equipment specification and operational proce-
well. In the future, API will adopt all cement-related new dures for the preparation of well-cement slurries in the
ISO documents. laboratory are contained in API RP 10B. The mixing

628 Well Cementing

device is a two-speed, propeller-type mixer, shown in If the slurry is going to be batch mixed during the
Fig. B-2. Specifications are given for the propeller speeds, cementing operation, it should be transferred to a con-
mixer blade wear, batch size, and mixing time. Usually, sistometer (Section B-4.2) and stirred in a manner con-
600 mL of slurry are prepared. The mixer is operated at sistent with the expected wellsite conditions of time and
4,000 rpm for 15 sec (during which all of the cement temperature.
solids should be added to the mix water), followed by The slurry mixing procedure specified by the API and
35 sec at 12,000 rpm. Cement slurries are very abrasive; ISO is not suitable for ultralow-density systems contain-
therefore, careful monitoring of the mixer blade condi- ing microspheres or nitrogen as extenders (Chapters
tion is essential. 3 and 7). Hollow microspheres are easily broken under
With this method, dry materials are uniformly high shear; consequently, the mixer is typically operated
blended with the cement before addition to the mixing at or below 4,000 rpm. A typical mixing procedure for
fluid (mix water plus any liquid additives). If liquid addi- slurries containing microspheres involves adding the
tives are present, they should be thoroughly dispersed in solids to the mix fluid within 30 sec at 4,000 rpm, fol-
the mix water before the cement is added. In certain lowed by an additional 300 sec of mixing at 4,000 rpm.
cases, the order of liquid-additive addition to the mix Foamed cements are routinely prepared in the
water may be critical. Such special mixing procedures mixing devices depicted in Fig. B-2. A special multiblade
and mixing times should be documented. Variations in assembly can also be employed (Fig. B-3). A base
mixing procedures can significantly alter the resulting cement slurry containing surfactant is placed in the
slurry properties (Roy and Asaga, 1979); therefore, indi- blending container, the container is capped and sealed,
vidual laboratories should establish and adhere to sup- and the mixing device is operated at a high propeller
plemental procedures covering items not specifically speed until the resulting foam completely fills the mixer
addressed by API RP 10B. Variations in field mixing pro- bowl. The foamed cement density and quality is varied
cedures can produce differences in slurry properties at by adjusting the volume of base slurry added to the
the wellsite, and attempts are continually being made to blending container. The disadvantage of this procedure
define and improve the correlation between field mixing is that the foam is not prepared under simulated high-
procedures and those used in the laboratory. pressure field conditions.

Fig. B-2. Propeller-type mixing device commonly used to prepare well-cement slurries (photos courtesy
Chandler Engineering, LLC, and OFI Testing Equipment, Inc.).

Appendix B Laboratory Testing, Evaluation, and Analysis of Well Cements 629

 Figure B-4. Pressurized mud balance (photograph courtesy
Chandler Engineering, LLC).

Figure B-3. Blending container and multiblade assembly for the tion to a syringe) is attached to the cap, and pressure is
preparation of foamed cements at atmospheric pressure (from applied to collapse air bubbles entrained in the slurry.
ISO 10426-4). Photograph courtesy of the International Organization
for Standardization.
Then the device is placed on a fulcrum, and a sliding
weight is adjusted until both sides are balanced. The
slide is calibrated in units of slurry density.
However, some pressurized testing methods were
developed by de Rozières and Ferriére (1990). They are
described in Chapter 7. B-4.3 Thickening time
Thickening-time tests are designed to determine the
length of time a cement slurry remains in a pumpable,
B-4.2 Slurry density fluid state under simulated wellbore conditions of tem-
An operational procedure for determining slurry density perature and pressure. The operational procedures for
is found in API RP 10B. The procedure uses a pressur- determining the thickening time are contained in API
ized fluid density balance, shown in Fig. B-4. Slurry is RP 10B.
poured into the cup and a pressure cap is screwed on. A The test slurry is evaluated in a pressurized consis-
pressurizing plunger filled with slurry (similar in opera- tometer, shown in Fig. B-5, which measures the consis-

Fig. B-5. Pressurized consistometers (photos courtesy Chandler Engineering, LLC, and Cement Test
Equipment, Inc.).

630 Well Cementing

tency of the test slurry contained in a rotating cup while (Fig. B-6). A nonpressurized, or “atmospheric,” consis-
under simulated wellbore conditions. Most apparatuses tometer (Fig. B-7) can be used to obtain a thickening
are capable of exposing cement slurries to a maximum time for low-temperature cement systems; however,
temperature and pressure of 400°F and 25,000 psi today it is most commonly used for conditioning of slur-
[204°C and 175 MPa]; however, special units capable of ries (according to various API/ISO procedures) before
700°F and 40,000 psi [371°C and 280 MPa] are available rheology, fluid-loss, or free-fluid tests.
for simulating very high-temperature, very high-pressure The consistency of the slurry is measured in Bearden
applications. A smaller, portable consistometer that uses units (Bc), a dimensionless quantity with no direct con-
a rotating paddle in a stationary cup is also available version factor to more common units of viscosity, such as

Fig. B-6. Portable pressurized consistometers (photos courtesy Chandler Engineering, LLC, and
OFI Testing Equipment, Inc.)

Fig. B-7. Atmospheric consistometers (photos courtesy Cement Test Equipment, Inc., Chandler
Engineering, LLC, and OFI Testing Equipment, Inc.).

Appendix B Laboratory Testing, Evaluation, and Analysis of Well Cements 631

Pa-s or poise. The end of a thickening-time test occurs ipated job duration, and the effects of well operations
when the cement slurry reaches a consistency of 100 Bc; performed before the cement job. API RP 10B contains
however, 70 Bc is generally considered to be the maxi- testing guidelines for various types of cement jobs: pri-
mum pumpable consistency. Figure B-8 shows the mary cementing (casing and liner specifications) and
output from a typical thickening-time test. Thickening- squeeze cementing (continuous-pumping and hesita-
time consistency profiles often begin with a flat, low-con- tion).
sistency period that lasts a few hours. Then, as the slurry
begins to set, the consistency rises with an ever-increas-
ing slope until 100 Bc is attained. The time at which the Primary cementing
consistency begins to increase is called the point of For many years, API RP 10B contained a collection of
departure. thickening-time schedules that specified the rates at
No provision for slurry fluid loss is made in the design which temperature and pressure are increased as well
of a consistometer slurry cup; therefore, the thickening as the final temperature and pressure, during the thick-
time for a slurry in the wellbore may differ from that for ening-time test. Such schedules were derived from field
the same slurry in the laboratory, particularly if the data collected from wells with different depths and tem-
design specifies little or no fluid-loss control. perature gradients. With the advent of computer simula-
Temperature and pressure can have a pronounced tors and improved instrumentation in the field to mea-
effect on the measured thickening time. The thickening sure job parameters in real time, the primary cementing
time also depends on the rate at which the final temper- schedules have been replaced with equations to calcu-
ature and pressure are attained. During a cement job, late a custom schedule for a particular cement job. The
flowing cement slurry is exposed to continuously chang- reader is referred to API RP 10B for a complete presen-
ing pressure and temperature; consequently, measure- tation of the equations. The complete set of data from
ment of the circulating-temperature and pressure pro- which the equations were derived is presented in API
file in the wellbore can be difficult (Chapter 12). The Report 10TR3: “Technical Report on Temperatures for
rates depend on the well conditions and the cement-job API Cement Operating Thickening Time Tests” (May,
design. Common variables that must be taken into 1999).
account include well depth, geothermal gradient, antic-

Temperature and pressure schedule:

100 Bc Thickening Time: 7:13:30 185°F and 10,200 psi
Casing
400 25,000 100

360 22,500 90

320 20,000 80

280 17,500 70

240 15,000 60

Temperature Pressure Consistency Temperature

200 12,500 50
(°F) (psi) (Bc)
160 10,000 40

120 7,500 30 Pressure

Point of departure
80 80 20
Consistency
40 40 10

0 0 0
00:00 02:00 04:00 06:00 08:00
Time (hh:mm)

Fig. B-8. Typical thickening-time test output.

632 Well Cementing

 An alternate method to create a primary cementing 3. Calculate the bottomhole pressure using the follow-
schedule is to consult tables in RP 10B that give esti- ing equation.
mated bottomhole circulating temperatures (BHCT) for
wells with various depths and geothermal gradients. An pBH = g × ρdf × DTV (B-2)
example is shown in Table B-1. The thickening-time where
schedule is determined as follows. DTV =true vertical depth at the top of the cement
1. Account for surface mixing of the slurry. If batch column (m)
mixing is performed during the cementing operation, g = acceleration of free fall (m/sec2)
the slurry may be stirred in the consistometer to sim- pBH = bottomhole pressure (kPa)
ulate the time and temperature. The time and slurry ρdf = density of drilling fluid (kg/m3).
temperature may be estimated depending upon the 4. Determine the starting pressure (pi) to which the
expected conditions at the wellsite. The batch mixing leading edge of cement slurry is subjected as it leaves
simulation is performed before the start of the thick- the cementing head.
ening-time test.
5. Calculate the increase in pressure (pressure-up rate)
2. Calculate the time to displace the leading edge of the as the slurry travels to the bottom.
cement slurry to bottom. The rate of slurry displace- p − pi
ment is calculated as follows. R pu = BH (B-3)
tdisp
Vpipe
tdisp = , (B-1) where
q
pBH = bottomhole pressure (kPa)
where pi = starting pressure (kPa)
q = rate at which the fluid slurry is pumped (m3/min) Rpu = pressure-up rate (kPa/min)
tdisp = time to displace the leading edge of cement tdisp = time to displace leading edge of cement to
slurry to bottom (min) bottom (min).
Vpipe = volume of the pipe (m3). 6. Determine the BHCT by consulting the tables in API
RP 10B (e.g., Table B-1).

Table B-1. BHCT Chart for Casing Well Simulation Tests†

Temperature Gradient (°F/100 ft Depth [°C/100 m Depth])
1.6 [0.9] 2.0 [1.1] 2.4 [1.3] 2.7 [1.5] 3.1 [1.7] 3.5 [1.9]
Depth‡ (ft [m]) BHCT§ (°F [°C])
1,000 [305] 80 [27] 80 [27] 80 [27] 80 [27] 80 [27] 80 [27]

2,000 [610] 89 [32] 89 [32] 90 [32] 90 [32] 91 [33] 91 [33]

4,000 [1,020] 99 [37] 100 [38] 101 [38] 102 [39] 103 [39] 104 [40]

6,000 [1,830] 112 [44] 114 [46] 116 [47] 118 [48] 120 [49] 126 [52]

8,000 [2,440] 126 [52] 129 [54] 135 [57] 140 [60] 146 [63] 160 [71]

10,000 [3,050] 141 [61] 146 [63] 158 [70] 167 [75] 180 [82] 200 [93]

12,000 [3,660] 148 [64] 165 [74] 183 [84] 201 [94] 219 [104] 236 [113]

14,000 [4,270] 164 [73] 185 [85] 207 [97] 228 [109] 250 [121] 271 [133]

16,000 [4,880] 182 [83] 207 [97] 233 [112] 258 [126] 284 [140] 309 [154]

18,000 [5,490] 201 [94] 231 [111] 261 [127] 291 [144] 321 [161] 350 [177]

20,000 [6,100] 222 [106] 256 [124] 291 [144] 326 [163] 360 [182] 395 [202]

22,000 [6,710] 244 [118] 284 [140] 324 [162] 364 [184] 404 [207] 444 [229]
† From API RP 10B-2. Reproduced courtesy of the American Petroleum Institute.
‡ True vertical depth.

§ Bottomhole circulating temperature.

Appendix B Laboratory Testing, Evaluation, and Analysis of Well Cements 633

7. Calculate the heat-up rate by subtracting the BHCT column, following steps described in the API and ISO
from ambient temperature and dividing by the time recommendations.
to bottom.
The highest temperature and pressure to which a Squeeze cementing
given volume of slurry is exposed may not occur at the Tabular well-simulation thickening-time schedules for
same point in the wellbore; consequently, they may not both continuous-pumping and hesitation squeeze treat-
occur at the same time. Slurries that come to rest at the ments are presented in API RP 10B. An example is
top of a long cement column will almost certainly have shown in Table B-2.
been exposed to higher temperatures and pressures as Predictive equations to estimate squeeze-cementing
they circulated past deeper regions. According to API temperatures are also given in API RP 10B.
and ISO recommendations, if no measured data are Although simulating the dynamic wellbore environ-
available for the top of the cement column, then the ment in the laboratory is difficult, refinements in proce-
final BHCT specified on the schedule should be main- dures and improvements in testing equipment are help-
tained until the completion of the thickening-time test. ing to make more realistic laboratory simulations
If such data are available, the temperature and pressure possible.
may be changed to those at the top of the cement

Table B-2. Thickening-Time Schedule for a Continuous-Pumping Squeeze Well-Simulation Test†

Temperature Gradient (°F/100 ft Depth [°C/100 m Depth])
1.6 [0.9] 2.0 [1.1] 2.4 [1.3] 2.7 [1.5] 3.1 [1.7] 3.5 [1.9]
Time Pressure Temperature‡
(min) (psi [kPa]) (°F [°C])
0 500 [3,400] 80 [27] 80 [27] 80 [27] 80 [27] 80 [27] 80 [27]

1 500 [3,400] 80 [27] 80 [27] 82 [28] 83 [28] 85 [29] 86 [30]

2 520 [3,600] 80 [27] 80 [27] 82 [28] 83 [28] 85 [29] 86 [30]

3 540 [3,700] 80 [27] 80 [27] 82 [28] 83 [28] 85 [29] 86 [30]

4 560 [3,900] 80 [27] 80 [27] 82 [28] 83 [28] 85 [29] 86 [30]

5 580 [4,000] 80 [27] 80 [27] 82 [28] 83 [28] 85 [29] 86 [30]

6 600 [4,100] 80 [27] 80 [27] 82 [28] 83 [28] 85 [29] 86 [30]

7 620 [4,300] 80 [27] 80 [27] 82 [28] 83 [28] 85 [29] 86 [30]

8 640 [4,400] 80 [27] 80 [27] 82 [28] 83 [28] 85 [29] 86 [30]

9 660 [4,600] 80 [27] 80 [27] 82 [28] 83 [28] 85 [29] 86 [30]

10 680 [4,700] 80 [27] 80 [27] 82 [28] 83 [28] 85 [29] 86 [30]

11 700 [4,800] 80 [27] 80 [27] 82 [28] 83 [28] 85 [29] 86 [30]

12 720 [5,000] 80 [27] 80 [27] 82 [28] 83 [28] 85 [29] 86 [30]

13 740 [5,100] 80 [27] 80 [27] 82 [28] 83 [28] 85 [29] 86 [30]

14 760 [5,200] 80 [27] 80 [27] 82 [28] 83 [28] 85 [29] 86 [30]

15 780 [5,400] 80 [27] 80 [27] 82 [28] 83 [28] 85 [29] 86 [30]

Ramp (°F [°C])

0.00 [0.00] 0.00 [0.00] 2.00 [1.11] 3.00 [1.67] 5.00 [2.76] 6.00 [3.33]
† From API RP 10B-2. Reproduced courtesy of the American Petroleum Institute.
‡ Depth: 1,000 ft [305 m] and mud density: 9.2 lbm/gal [1.10 kg/L]. Ramp time: 1 min. Temperature dwell: 14 min.

Pressure rate (per min) = 0 psi [0 kPa] for 1 min. then 20 psi (143 kPa) for 14 min.

634 Well Cementing

B-4.4 Fluid loss
Fluid-loss tests measure the slurry dehydration during
and immediately following a cement job. API RP 10B
offers operational test procedures for determining the
fluid-loss rate. After conditioning at simulated wellbore
conditions, the test slurry is placed in a heated cell and
subjected to 1,000 psi [6.9 MPa] of differential pressure.
The filtrate loss is measured across a standard filtra-
tion medium (45-μm [325-mesh] screen supported on a
250-μm [60-mesh] screen). The filtration area is 3.5 in.2
[22.6 cm2]. After 30 min, the collected filtrate volume
is noted. The reported fluid-loss value is equal to the col-
lected filtrate volume multiplied by two. If all of the fil-
trate passes through the screen in less than 30 min, the
following equation is used to calculate the API fluid loss.

⎛ ⎞
( qAPI )calc = 2Vt ⎜⎝ 5.477 ⎟⎠ (B-4)
t
where
Vt = volume of filtrate (mL) collected at time t (min).
The test is performed either in a static heated filter-
press cell, shown in Fig. B-9, or in a stirred fluid-loss cell,
shown in Fig. B-10. Whatever the equipment—static
filter press or stirred fluid loss—the actual filtration test
always takes place with a slurry in a static state. Before
the filtration test, slurry conditioning may be performed
in a pressurized consistometer, in a stirred fluid-loss cell,
or, if the temperature is less than 194°F [90°C], in an
atmospheric consistometer. Fig. B-10. Stirred fluid-loss testing apparatuses (photographs cour-
tesy Chandler Engineering, LLC, and OFI Testing Equipment, Inc.).

Fig. B-9. Filter-press cell and assembly for API/ISO fluid-loss tests (photographs courtesy
OFI Testing Equipment, Inc.).

Appendix B Laboratory Testing, Evaluation, and Analysis of Well Cements 635

 The prescribed test evaluates slurry fluid loss under B-4.5 Strength
static conditions (immediately following placement), The procedures to determine the strength of set cement
regardless of whether the slurry is conditioned in a con- are described in API RP 10B. Two types of strength mea-
sistometer or in a stirred fluid-loss cell. This procedure surements are described:
makes no provision for the measurement of fluid loss
■ compressive strength (measured by a destructive
during placement, although results of fluid-loss determi-
nations under dynamic conditions have been reported crush test)
(Bannister, 1978). As discussed in Chapter 5, the ■ sonic strength (calculated from sonic-velocity mea-

dynamic fluid-loss rate of a given slurry is usually higher surements through the set cement).
than the static fluid-loss rate unless a filtercake has
already been deposited by the drilling fluid, a spacer, or
a chemical wash. B-4.5.1 Compressive strength
The stirred fluid-loss test equipment determines the Test cement slurries are prepared according to the
fluid-loss rate without transferring the heated slurry API/ISO mixing procedure, poured into 2-in. cubic
from one vessel to another. This removes the safety molds, and cured for various time periods at specific
hazard of transferring the heated slurry from the consis- temperatures and pressures. The set-cement cubes are
tometer to the heated filter press. Commercially avail- removed from the molds and placed in a hydraulic press,
able stirred fluid-loss equipment can operate at up to where increasing uniaxial load is exerted on each until
400°F [204°C] and 2,000 psi [14 MPa]. failure. The compressive strength is then calculated by
Static high-pressure, high-temperature filter presses dividing the load at which failure occurred by the cross-
are available for use at slurry temperatures up to 350°F sectional area of the specimen.
[177°C] for the smaller units (175 mL), and up to 500°F Figure B-12 shows a typical curing mold that makes
[260°C] for the larger units (500 mL) (Fig. B-11). All two test specimens. API/ISO procedures describe curing
units require the use of a backpressure receiver when at pressures from atmospheric pressure to 3,000 psi
the slurry temperature exceeds 200°F [93°C] to prevent [21 MPa]. For atmospheric tests, curing in a water bath
evaporation of the filtrate. or a cooling bath can simulate cold weather or per-
mafrost conditions. Pressurized curing chambers, such
as the device shown in Fig. B-13, are available in various
sizes and with varying performance capabilities. There
are commercial units that hold up to 32 specimens, with
maximum operating conditions of 600°F [315°C] and
20,000 psi [140 MPa]. Section 7.7 of API RP 10B contains
the prescribed heat-up and pressurization schedules,
which, like the thickening-time schedules, are derived
from field data and the anticipated wellbore tempera-
ture gradient. For arctic cement systems, special curing
methods are given in API RP 10B. In addition, a method
for evaluating a cement system’s resistance to freeze-
and-thaw cycling is given. When a set-cement specimen
is placed in the hydraulic press for strength measure-
ment (Fig. B-14), the loading rate is regulated according
to the anticipated strength of the specimen. For speci-
mens with a compressive strength greater than or equal
to 500 psi [3.5 MPa], the loading rate is 4,000 psi
[28 MPa]/min. For specimens with a compressive
strength less than 500 psi, the loading rate is 1,000 psi
[6.9 MPa]/min.

Fig. B-11. Filter-press assembly for high-temperature fluid-loss tests

(photograph courtesy OFI Testing Equipment, Inc.).

636 Well Cementing

 Fig. B-14. Hydraulic press for measurement of compressive
strength (photograph courtesy Chandler Engineering, LLC).

Fig. B-12. 2-in. curing molds for compressive-strength testing. B-4.5.2 Sonic strength
The strength can be estimated ultrasonically (Rao et al.,
1982). The ultrasonic cement analyzer (UCA), shown in
Fig. B-15, measures the travel time of ultrasonic energy
through a cement sample as it cures under simulated
temperature and pressure conditions. The ultrasonic
measurement is nondestructive and may be made con-
tinuously as the cement sample cures at high pressure
and elevated temperature. The sonic strength is corre-
lated to the transit time (reciprocal of ultrasonic veloc-
ity) using an empirical relationship initially established
from mechanical compressive strength and transit time
data for various slurry systems. The strength estimate
can be output directly via a preprogrammed micro-
processor. API RP 10B now includes testing procedures
for nondestructive sonic testing of cements with UCAs.
Cobb et al. (2002) developed a method to predict the
sonic strength of foamed cements. One first measures
the sonic strength of the base slurry. Using a strength-to-
foam quality correlation, one then calculates the
strength estimate.
The strength values obtained using either the
API/ISO crush test or the UCA are indicative of the
integrity of the cement under uniaxial loading (i.e., no
lateral restraint). In the wellbore, the cement is subject
to complex triaxial loading, and the failure stresses
Fig. B-13. Pressurized curing chamber for compressive-strength may be substantially different from those observed in
tests (photograph courtesy Chandler Engineering, LLC).
the standard compressive strength test (Chapter 8).
Furthermore, the strength measurement provides no
guide to the shear strength of the casing-to-cement or
the casing-to-formation bond (Parcevaux and Sault,
1984). More detailed procedures to determine cement
mechanical properties are presented in Section B-4.

Appendix B Laboratory Testing, Evaluation, and Analysis of Well Cements 637

Fig. B-15. UCAs (photos courtesy Chandler Engineering, LLC, Cement Test Equipment, Inc., and Fann
Instrument Company).

B-4.6 Free fluid and slurry sedimentation The procedure permits slurry conditioning at ele-
When a slurry is allowed to stand for a period of time vated temperatures and pressures. No provision is made
before setting, water may separate from the slurry, for fluid loss. The duration of the test is 2 hr, measured
migrate upward, and accumulate in pockets or at the top from the moment the slurry is poured into the graduated
of the column. This separation can impair zonal isola- cylinder. For temperatures less than 176°F [80°C], the
tion, particularly in a highly deviated wellbore graduated tube is placed in a preheated or precooled
(Chapter 13). The free-fluid test measures this separa- test chamber. For higher temperatures, the graduated
tion tendency in the laboratory, using a graduated cylin- tube is placed in a preheated, oil-filled heating chamber
der as a simulated wellbore. that is maintained at a pressure sufficiently high to
prevent boiling. The test procedure is documented in

638 Well Cementing

API RP 10B. To simulate deviated wellbores, many oper- B-4.7 Rheological measurements
ators orient the graduated cylinder at the angle of well- A detailed discussion of cement rheological properties
bore deviation. An increase in free fluid is usually and their significance is found in Chapter 4, and a sum-
observed in such situations; however, there is no clear mary of the pertinent mathematical relationships is pre-
understanding of how free fluid depends upon column sented in Appendix A. The use of these relationships for
height. accurate prediction of friction pressure and slurry-flow
In addition to free-fluid development, the suspended properties depends upon reliable laboratory measure-
solids may also tend to separate from the slurry and ments of the rheological parameters.
settle toward the bottom of the cement column. This Coaxial cylinder rotational viscometers are the most
separation is particularly evident in slurries containing common apparatuses to make rheological measure-
weighting agents. Such sedimentation can produce vari- ments with cement slurries. As shown in Figs. B-17 and
ations in slurry density, leading to annular invasion and B-18, rotational viscometers are designed with a rotating
possible loss of well control (Chapter 9). outer cylinder (couette type, covered by API RP 10B) or
After an appropriate conditioning period, the slurry is a rotating inner cylinder (Searl type) (Bannister, 1980).
poured into a sedimentation tube (Fig. B-16). The slurry-
filled tube is placed in a water bath or autoclave pre-
heated to the anticipated wellbore temperature or 194°F
[90°C], whichever is cooler. The slurry temperature is
then adjusted to simulate temperature changes in the
wellbore. A minimum amount of pressure is applied to
prevent slurry boiling, or if desired, the pressure can
simulate bottomhole conditions. After an appropriate
curing period (typically 24 hr, or until set), the set
cement is sliced into several segments or wafers. The
density of each segment is measured, and the percent
density difference between the liquid sample and the set
sample is calculated using the following equation.
⎛ ρ cem − ρ slurry ⎞
Δρ = ⎜ ⎟ × 100, (B-5)
⎝ ρ slurry ⎠
where
ρcem = density of cement segment
ρslurry = density of cement slurry.
The test procedure is presented in API RP 10B.

Fig. B-17. Couette-type rotational viscometer (photographs cour-

tesy Chandler Engineering, LLC, and Fann Instrument Company).

Fig. B-16. Slurry sedimentation test apparatus for use in an autoclave.

Appendix B Laboratory Testing, Evaluation, and Analysis of Well Cements 639

 recorded at various rotational speeds. As discussed in
Chapter 4, the manner in which the readings are taken
is critical. Readings must be taken first in ascending
order and then in descending order. As shown in Table
B-3, the rheological measurements are reported as an
average of the ramp-up and ramp-down readings.

Table B-3. Example of Data From a Rheological Properties

Test†, ‡, §, ††
Rotational Ramp-up Ramp-Down Reading Average
Speed r/s Reading Reading Ratio Reading
3 21 24 0.87 22.5

6 40 36 1.11 38

30 65 83 0.78 74

60 84 100 0.84 92

100 100 115 0.87 107.5

200 137 147 0.93 142

300 170 170

† From ISO 10426-2 (2004). Reproduced courtesy of the International Organization
for Standardization (ISO).
Fig. B-18. Searl-type rotational viscometer. ‡ Initialslurry temperature = 150°F [66°C].
§ Final slurry temperature = 146°F [63°C].
†† Rheological properties reported at average temperature of 148°F [65°C].

API RP 10B provides detailed guidelines concerning

slurry preparation and conditioning before rheological
measurements. The shear history that a slurry experi- In addition to friction-pressure and flow-regime cal-
ences during mixing can have a significant effect upon culations, laboratory measurements of rheological para-
the ultimate rheological properties. The slurry is pre- meters can provide information about other slurry char-
pared in a blade-type mixer according to the guidelines acteristics. A decrease in shear-stress values when
presented in API Specification 10A (summarized in measured at decreasing shear rates, compared to those
Section B-4.1). measured at increasing shear rates, may indicate that
Immediately after mixing, the slurry is poured into the slurry is thixotropic. A determination of gel strength,
the slurry container of an atmospheric or pressurized using the procedure described in API RP 10B will pro-
consistometer for preconditioning. The container tem- vide an indication of a slurry gelation tendency. The
perature must be initially ambient to avoid thermally procedure prescribes a conditioning period of 20 min
shocking temperature-sensitive additives. The slurry is before measurement of the rheological parameters;
then heated to the test temperature and stirred for a however, measurements taken immediately after
period of 20 min. If preconditioning was performed in a mixing may provide some indication of slurry mixabil-
pressurized consistometer at an elevated temperature ity, particularly in a batch mixer.
and pressure, the slurry must be cooled as quickly as
possible to 190°F [88°C] before the slurry container is B-4.8 Static gel strength
opened. The static gel strength of a cement slurry is routinely
After preconditioning, the slurry is immediately measured immediately after determining the rheological
poured into a preheated couette-viscometer cup to the properties of the slurry. A couette-type rotational vis-
fill line. With the sleeve rotating at the lowest speed, the cometer is used for the measurement. Typically, the gel
cup is raised until the liquid level reaches the inscribed strength is measured after the slurry has been left undis-
line on the sleeve. This operation minimizes slurry gela- turbed for 10 sec and again after resting for 10 min (API
tion and ensures uniform distribution of the slurry. After RP 10B).
the slurry temperature is noted, dial readings are

640 Well Cementing

 Procedures specifically developed for measuring the and curing it underwater at atmospheric pressure. The
static gel strength of a cement slurry use a consistome- cement bar is removed from the mold when it is suffi-
ter-type apparatus equipped with a low-friction mag- ciently strong, the length is carefully measured, and the
netic drive and torque-measuring system coupled to the bar is then returned to the water bath for further curing.
paddle (Sabins et al., 1980). The slow movement of the Periodically during the curing period, the bar is removed
paddle apparently does not inhibit gel strength develop- for additional length measurements. The ASTM method
ment, yet it permits torque measurements. has two major drawbacks: (1) because the cement must
It is important to note that rheometers are not appro- develop some strength before a length is measured, it is
priate for determining the transition time of a slurry impossible to obtain a reliable “zero reading,” and (2)
(Chapter 9). there is no provision for studying the effect of pressure.
New measurement techniques that rely on the analy- In 1994, the API formed a work group on cement
sis of the acoustic waveform transmitted through the shrinkage and expansion. API Technical Report 10TR2
cement have been developed (Fig. B-19). They allow the (1st ed.), July 1997, “Shrinkage and Expansion in Oilwell
evaluation of static gel strength under downhole condi- Cements,” proposed equipment and procedures to inves-
tions (pressure and temperature). Proprietary algo- tigate shrinkage and expansion in oilwell cements. The
rithms are used to transform the acoustic waveform report described four pieces of equipment to measure
transmitted through the cement into an evaluation of cement-volume changes under different boundary con-
static gel strength. ditions: the annular ring expansion mold, the cylindrical
sleeve, the membrane test, and the cement hydration
analyzer (CHA).
The annular expansion ring test measures linear
expansion under conditions of free access to water. Free
access to water is defined as an open system. The annu-
lar expansion mold (Fig B-20) simulates the annulus of
a well. The cement slurry is poured into the annular
space in the mold, and the mold is placed into a water
bath or pressurized curing vessel. Because the slurry is
in contact with water during the entire test, water entry

Fig. B-19. Static gel strength analyzer (photograph courtesy

Chandler Engineering, LLC).

The static gel strength is an important variable

related to annular fluid migration. Additional informa-
tion about static gel strength testing in the context of
annular fluid migration is presented in Chapter 9.

B-4.9 Expansion and shrinkage

Set-cement expansion can be measured using the proce-
dure described in ASTM Standard C151, Standard Test
Method for Autoclave Expansion of Portland Cement.
This procedure, developed for the concrete industry,
involves placing the cement slurry into a bar-type mold
Fig. B-20. Annular expansion molds.

Appendix B Laboratory Testing, Evaluation, and Analysis of Well Cements 641

will compensate for internal shrinkage as long as the sets of pins is taken with a micrometer. If the set cement
cement matrix is permeable. The outer ring has a verti- expands during the curing period, the diameter of the
cal slit; therefore, if the cement expands during the sleeve increases and the longitudinal slit opens. After
hydration period, the diameter of the ring will increase. the mold is removed from the curing chamber, the dis-
Conversely, if the cement shrinks, the diameter of the tances between the pins are measured again, and the
ring will decrease. The change in ring diameter is deter- amount of expansion is calculated according to the fol-
mined by measuring the distance between the two pins lowing equation.
mounted on the outside of the mold with a micrometer.
Measurements are taken before and after curing, and ⎡r ⎛ L ⎞⎤
ΔV = ⎢ i+ L ⎜ 2 − 1⎟ ⎥ , (B-7)
special care is required to ensure that both measure-
⎢⎣ ri ⎝ L1 ⎠ ⎥⎦
ments are taken at the same temperature.
The percent dimensional change (shrinkage or where
expansion) is calculated as follows: L1 = average distance between the pins when the sleeve
(
ΔV = 0.358 L f − Li , ) (B-6) is empty
L2 = average distance between the pins when the sleeve
where is expanded
Li = initial distance measurement between pins (mm) ri = inside radius of sleeve
Lf = final distance measurement between pins (mm) ri + L = plus the length of the pin.
ΔV = dimensional change of the cement sample (%). The membrane test is a measurement of bulk shrink-
This method is described in ISO 10426-5, Test meth- age or bulk expansion under impermeable conditions.
ods for determination of shrinkage and expansion of Lack of free access to water is defined as a closed
well cement formulations at atmospheric pressure. system. The slurry is poured into an impermeable flexi-
The cylindrical sleeve test is also an open-system ble membrane (i.e., a latex sheath) that is sealed by
bulk-expansion measurement with free access to water. tying a knot at the top. The membrane is attached to a
The sleeve, invented by Spangle (1983), consists of a balance hook and immersed into a temperature-con-
cylindrical mold with a vertical slit. The sleeve is trolled water bath. The slurry temperature is measured
designed to fit inside a standard 2-in. cube mold with a thermocouple immersed in the slurry. The
(Section B-4.4). As shown in Fig. B-21, three sets of pins “pseudoweight,” as measured by the balance, is the
are mounted on the outside of the sleeve, each located weight of the membrane containing cement minus the
on opposite sides of the slit and secured with a spring. buoyancy force acting on it. An increase of the
The sleeve is closed when the slurry is first introduced, pseudoweight corresponds to a decrease of the volume
and a zero reading of the distance between the three (shrinkage).
The CHA test (Fig. B-22) measures inner shrinkage
under free access to gas. The slurry is placed in a closed
cell equipped with a gas inlet. During the test, a constant
gas pressure is applied to the container. Shrinkage is
determined by measuring the volume of gas required to
maintain a constant pressure inside the test cell until
the cement hardens. Assuming that the cement matrix
remains sufficiently permeable to gas throughout the
test, this volume of gas corresponds to the inner shrink-
age of the cement. This test can also be performed using
water instead of gas. If gas is used, this method is essen-
tially a closed test. If water is used, this method is an
open test.
More recently, Bosma et al. (2000) developed an
apparatus with two cells to measure expansion and
shrinkage phenomena in either a closed or open system,
as defined by API 10TR2. Concurrently, the apparatus
can measure the hydration and setting behavior of the
cement system, and information about the mechanical
Fig. B-21. Curing sleeve for measurement of cement expansion. properties of the set cement can be acquired. In addi-
tion, the apparatus is capable of operating at elevated
temperatures and pressures.

642 Well Cementing

 of these data are then used either as input into mathe-
matical simulators or for comparing the performance
Pressure transducer differences between various slurries.
A thorough discussion of the various devices for gas-
migration testing, complete with schematic diagrams
and photographs, is presented in Chapter 9.

Test cell with

B-4.11 Permeability
cement slurry The permeability of the cement sheath strongly affects
zonal isolation. An operational procedure for determin-
ing the relative permeability of set cement to liquid or
gas is contained in API RP 10B. The procedure deter-
Floating piston mines comparative permeabilities of cement samples
rather than absolute permeability values. Fluid is
injected through the sample at either a constant flow
Hydraulic
rate or constant differential pressure, and the resulting
pressure fluid-flow rate is measured.
line A variety of permeameters is available to perform this
test. Earlier apparatuses and procedures were devel-
oped in the early 1950s (Morgan and Dumbauld, 1952).
Gas injection Today, many laboratories use a Hassler sleeve-type
valve holder (Figs. B-23 and B-24), and measurement of per-
meability to air, methane, or other gases is fairly routine
(API, 1960).

Fig. B-22. Diagram of the CHA apparatus.

B-4.10 Gas migration

At this writing, neither API nor ISO have published a
standard method or procedure for testing gas migration
control in cement slurries; therefore, there is no recog-
nized standard laboratory procedure to characterize the
ability of a cement system to prevent or reduce gas
migration.
A significant number of nonstandard laboratory tests
for gas migration has been developed by the industry
throughout the years. A wide variety of experimental
prototypes that attempt to simulate the gas migration
process are described in the literature. Two main types
of experimental simulators exist: large-scale pilot
devices, which attempt to reproduce the process as it
occurs in the wellbore, and small-scale, benchtop
models, which can be used to derive the fundamental
laws of a particular physical process under investigation.
To date, none of the simulators described in the litera-
ture permits the derivation of a physical model that
quantitatively describes gas migration over a wide range
of conditions. Recently, many investigators have begun
to rely upon measuring individual slurry properties such
as gel strength development and shrinkage. The results Fig. B-23. Apparatus to measure water permeability (photograph
courtesy Chandler Engineering, LLC).

Appendix B Laboratory Testing, Evaluation, and Analysis of Well Cements 643

 may prove to be crucial, because the sample can be dam-
aged by the drying process, leading to questionable
results. Regardless of the method chosen, care must be
taken to preserve sample quality and assure the sample
is dry before testing.
The gas permeability of the set cement is calculated
using Darcy’s law according to the following equation
(B-9).
2000 × μ gas × qgas × p × L
k= , (B-9)
(
A × ⎡⎢ pinlet
⎣
) (
2
)
− poutlet ⎤⎥
2

⎦
where
A = sample cross-sectional area (cm2)
k = permeability to gas (mD)
L = sample length (cm)
p = adjusted barometric pressure (atm)
pinlet = inlet pressure (atm)
poutlet = outlet pressure (atm)
qgas = flow rate of gas (mL/s)
μgas = viscosity of gas (cp).

B-5 Spacers and chemical washes

Fig. B-24. Apparatus to measure gas permeability (photo courtesy
OFI Testing Equipment, Inc.).
As discussed in Chapter 5, spacers and chemical washes
perform two important functions during a cementing
operation: to clean and remove the drilling fluid from
the wellbore and to minimize cement contamination by
Water permeability is usually determined by forcing the drilling fluid. To accomplish these objectives, the
water through a core of set cement at a differential pres- spacer and wash must impart some degree of cleaning
sure of 20 to 200 psi [100 to 1,400 kPa]. The water effect in the wellbore and must be compatible with both
expelled from the core is collected during a 15-min time the drilling fluid being displaced and the cement slurry
period, or until 1 mL has been accumulated in the mea- being placed in the hole. Laboratory testing of these
suring tube. Darcy’s law (Eq. B-8) is used to calculate materials must include procedures to evaluate the com-
the permeability. patibility and the cleaning effect.

⎛ q × μ water × L ⎞ B-5.1 Compatibility

k = 14,700 ⎜ , (B-8)
⎝ A × Δ p ⎟⎠ Operational procedures to evaluate compatibility are
contained in API RP 10B, Section 16. The procedures
where test the effects of a spacer or chemical wash on the
A = sample cross-sectional area (cm2) behavior of both the drilling fluid and cement slurry.
k = permeability (mD) Various volumetric ratios of drilling fluid, cement slurry,
L = sample length (cm) and spacer are specified. The thickening time, fluid-loss
Δp = differential pressure (psi) rate, and rheological characteristics of each mixture are
q = flow rate (mL/s) measured. No specific criteria are provided to aid in the
μwater = water viscosity (cp). interpretation of the test results.
To measure gas permeability, the core sample must
be dried completely before testing. The API and ISO do
not specify the drying procedure. Sample drying can be B-5.2 Cleaning effect
performed by placing the sample over a dessicant or by There are no standard procedures to evaluate the mud-
drying it in a vacuum oven or in a conventional oven at removal capability of preflushes. Following the descrip-
high temperature (Mindess and Young, 1981). This stage tion presented in Chapter 5, the cleaning process is

644 Well Cementing

achieved in two steps: bulk displacement of the mud fol- As an aqueous spacer displaces a water-in-oil emul-
lowed by cleaning of the annulus wall. Bulk displace- sion mud, both fluids come into contact in the wellbore.
ment is governed by fluid mechanics and the rheology Throughout the intermixing zone, the emulsion will
and density of the spacer or chemical wash. These prop- absorb water until the water droplets become so large
erties can be measured under simulated downhole con- that the external oil layer can no longer contain them. At
ditions using the standard API and ISO procedures if the this point, the emulsion breaks and the electrical con-
temperature does not exceed 185°F [85°C]. However, it ductivity of the mixture increases (Fig B-26).
is often necessary to extrapolate laboratory data to Heathman et al. (1999 and 2000) developed an appa-
downhole temperature and pressure conditions using ratus and test procedure aimed at comparing the ability
appropriate models. of different fluids to convert the mud to an oil-in-water
No standard method has been agreed upon to evalu- emulsion. As shown in Fig. B-27, the wettability appara-
ate the annular wall cleaning efficiency of spacers and tus is comprised of (1) a double-walled, stainless-steel
washes. As discussed in Chapter 5, the relevant phe- mixing jar with electrodes mounted inside the walls, (2)
nomena include the following. a control box containing the temperature controller and
1. When the mud is an invert emulsion, the wash or the wettability electronics, and (3) a harness connection
spacer must invert it to an oil-in-water emulsion that provides power to the heater, the thermocouple,
(Fig. B-25). and the electrodes. The apparatus allows operators to
test drilling fluid-spacer mixtures at controlled shear
2. The oil film adsorbed on the wall is displaced to leave
rates and controlled temperatures up to 190°F [88°C].
water-wet surfaces.
3. The mud layer on the walls is eroded by the tangential
flow.
Phenomena 1 and 2 are specific to invert-emulsion
muds, while tangential erosion is applicable to all mud
types.

Oil Water
phase phase

Phase
Water inversion Oil
phase phase

Fig. B-27. Apparent wettability meter for evaluating the water-

Emulsifiers wetting capacity of surfactants (photograph courtesy of Fann
Oil-External Emulsion Water-External Emulsion Instrument Company).

Fig. B-25. Phase inversion of emulsion from oil-external to water-

external (from Heathman et al., 2000). Reprinted with permission of
SPE. The interpretation of the results is straightforward,
provided the operator keeps certain points in mind.
■ Once the mud is inverted, an oil film may remain on

Fluid phase inversion the electrodes, providing incorrect measurements.

■ Even if operated at low rotary speed, the blade mixer
0 Hn 200 Hn provides a high shear rate that helps disperse and sta-
100%
drilling 100% 100%
100% bilize emulsions.
spacer
fluid oil-wet water-wet ■ The apparatus does not provide a reliable measure-
Flow
ment of the wettability of the annular wall, because
the electrode is made from a material that is different
0 25 50 75 100 from the casing and the formation rock.
Fig. B-26. Measurement of apparent wettability in Hogans (Hn) as a
function of fluid-phase inversion in the drilling fluid-spacer inter-
mixing zone (from Heathman et al., 2000). Reprinted with permission
of SPE.

Appendix B Laboratory Testing, Evaluation, and Analysis of Well Cements 645

 Various means to monitor the tangential erosion of a Analytical and numerical models help predict the
mud layer from a solid surface have been proposed. stress distribution in the cement sheath caused by
■ Boyington et al. (1989) described the use of a piece of changes in downhole conditions. Whatever the calcula-
tile, representing the formation rock, first immersed tion method, the mechanical and thermal properties for
in the mud and then rotated in the cleaning fluid. A all the involved materials—steel casing, cement, and
sample of steel pipe, preferably the same metal as the formation—are required. The results can differ signifi-
casing, can be used instead of the tile to evaluate cantly depending on the testing method used to deter-
casing cleaning effectiveness. mine these properties.
■ Davison et al. (2001) used a rotational viscometer
To provide coherent, reproducible, and comparable
results, one must define and adhere to consistent test
that was covered with sand to simulate the formation
procedures. To become familiar with mechanical prop-
wall.
erties terminology and concepts in the procedures
■ A variation of the Davison et al. (2001) method
described below, a review of Chapter 8 is recommended.
involves covering the rheometer rotor with a piece of It is important to point out that the test methods
metallic mesh to simulate the casing. assume that the cement specimens are continuous
In these methods, the mass of removed mud is moni- materials exhibiting a linear elastic mechanical behav-
tored versus time. The test results strongly depend on ior. As discussed in Chapter 8, cement systems are rather
the nature of the solid surface. Therefore, for compari- porous materials. Nevertheless, experiments described
son of different fluid systems, the solid surface must be by Boukhelifa et al. (2004) showed that the continuous
very reproducible from test to test. and linear elastic assumption was sufficient for predict-
Al Khayyat et al. (1999) proposed monitoring the ing the failure mode.
water-wettability of a steel pipe covered with mud while
cleaning it by rotation in the spacer. The wettability is B-6.2 Sample preparation
visually estimated by observing water drops poured on
the cleansed metal surface. Ali et al. (1997) described a The cement samples must be prepared with care to
procedure in which a volume of mud and a volume of ensure that their properties are representative of those
brine are shaken in a glass jar. The water wettability is that will exist in the well. The initial slurry must be
determined visually by observing how brine droplets stable with, if possible, zero free water and no sedimen-
behave on the glass surface. tation.
The above methods allow the screening of various The cement system must be cured at bottomhole
cleaning solutions. The main limitations of these meth- static temperature (BHST) for a period sufficient to
ods include the need to use various complementary allow the set cement to reach a stable compressive
methods to mimic downhole phenomena. Also, because strength. Curing time of 7 to 28 days may be required.
the tests are performed at atmospheric pressure, the For thermal cements, the samples must be cured until
test temperatures are limited to 185°F [85°C]. the appropriate cement-phase conversions have
occurred (Chapter 10). For systems that will be used in
steam injection wells, it is appropriate to first cure the
cement slurry at the BHST and then expose the set
B-6 Mechanical properties testing of cement to steam-injection temperatures.
well cements To eliminate the effects of entrained air, all systems
should be cured in a pressurized curing chamber. If pres-
B-6.1 Introduction
surized curing chambers are not available, the cement
In a wellbore, the cement sheath supports the casing slurry should be thoroughly degassed by stirring it gently
and seals the annulus between the casing and the for- under a vacuum.
mation or between two casing strings. During the life- The cement systems should be cured in molds of, or
time of the well, the cement will be exposed to stresses machined into, the appropriate geometry. During all
induced by temperature and pressure changes, e.g., operations, the samples must remain saturated with
through steam injection, fracture stimulation, or deple- water to avoid drying shrinkage that may induce frac-
tion (Chapter 8). Uniaxial compressive strength and tures within the specimen.
permeability to gas or water are the principal tests
performed on set cement. These tests do not provide
sufficient information to predict and model the behavior
of the cement sheath under the expected well stress
conditions.

646 Well Cementing

B-6.3 Cement static mechanical properties test is a cylinder that is 1 in. [25.4 mm] in diameter and
Chapter 8 discusses some key measurements to deter- 2 in. [50.8 mm] long.
mine whether a cement sheath can withstand the A standard mechanical measurement apparatus
expected temperature and pressure changes downhole. imposes a load or a deformation at a fixed rate and mag-
These key cement properties—tensile strength, com- nitude (Fig. B-28). This allows the measurement of the
pressive strength, Young’s modulus, and Poisson’s unconfined compressive strength, Young’s modulus, and
ratio—are derived from stress-strain tests performed on Poisson’s ratio.
the solid material.
Stress is a force per unit area and is expressed in the
same units as pressure. When submitted to a stress field,
a material deforms. The rupture strengths of a material
in compression and in tension are the compressive
strength and the tensile strength. Strain is the relative
change in shape or size of a material caused by exter-
nally applied forces. This is a displacement per unit
length and is a dimensionless value. These properties
are crucial, because the best sealants will be those that
can withstand the most extreme deformations and high-
est stresses. Hence, the optimal sealing materials would
feature low Young’s moduli and high compressive and
tensile strengths.
Assuming perfect linear elastic mechanical behavior,
the elastic constants are intrinsic values for homoge-
neous materials. However, when stressed in flexion,
compression, or tension, the elastic constants can
vary depending upon the stress distribution and loading
rate. This is particularly true for cements because of the
heterogeneity of their grains and pores.
The elastic constants should be measured according
to the way the cement sheath will be stressed. When
cement is in tension, a flexural test is the most appro-
priate way to measure the Young’s modulus. If cement is
under compression, a compressive test is appropriate.
However, the Young’s modulus determined in a compres-
sive test is commonly higher than the value measured in
a flexural test. Therefore, to be conservative, one can use
the value obtained in a compressive test for any stress Fig. B-28. Load frame for mechanical testing. Load and displace-
condition. Because test results may vary depending upon ments are recorded with a data acquisition system.
the test procedure used, a consistent method should be
followed when evaluating a series of cement systems.
During compression, the material will shorten in the
loaded direction (negative strains) and expand in the
B-6.3.1 Uniaxial unconfined compression test normal or unconfined direction (positive strains). Load,
The methods discussed below are more sophisticated axial displacement, and radial displacement are moni-
than the uniaxial compressive strength test described in tored throughout the test. Axial deformation is typically
Section B-4.4. When an axial load is applied on an uncon- measured with a linear variable differential transformer
fined cement sample, there are no constraints in the lat- (LVDT) set on the moveable platen. Radial displacement
eral direction. The stress distribution varies depending can be measured with strain gauges or a cantilever set
on the geometry and size of the specimen, because fric- halfway up the cylindrical sample (Figs. B-29 and B-30).
tion is generated at the interfaces between the sample Cement is a porous medium saturated with a fluid.
and platens of the press. The optimal sample height-to- Therefore, the loading rate is an important test parame-
width ratio is between 2 and 3 (Vutukuri et al., 1974). ter. As discussed in Chapter 8, drained and undrained
To reduce frictional effects at the interfaces, the rec- compressive strengths can be measured on set cement.
ommended sample geometry for a uniaxial compression A drained condition occurs when there is no change in

Appendix B Laboratory Testing, Evaluation, and Analysis of Well Cements 647

 determining the elastic constants of concrete in com-
pression and recommendations for the load rate and the
cycling procedure. The procedure suggests the applica-
tion of several load-unload cycles on the sample, remain-
ing in the elastic domain (40–50% of the uniaxial com-
pressive strength). Also, to ensure testing close to
drained conditions and for practical reasons (duration of
the test), the recommended axial deformation rate is
0.05 mm/min.
The first loading-and-unloading cycle generates more
deformation than subsequent cycles that follow the
same path. This is because of experimental artifacts
when the sample surface comes in contact with the load-
ing platen of the machine (misalignment), added to
strengthening of the material after compaction (plastic
Fig. B-29. Equipment for measuring sample deformation (LVDTs and deformation). As shown in Fig. B-31, the static Young’s
cantilever). modulus measured from the initial loading cycle (E1) is
lower than the value measured at the second loading
cycle (E2). Consequently, data from a simple crush test
give an optimistic estimate of Young’s modulus.
Subsequent loading cycles provide values very similar to
E2; therefore, E2 is a more representative value and
should be used for modeling. At least two subsequent
cycles are recommended to verify the repeatability of
the result.
Before testing, it is advised to measure and report the
set-cement density and the specimen geometry (diame-
ter and height). The faces of the specimen must be par-
allel to the platens because ridges and hollows at the
specimen ends may concentrate stresses and cause fail-
ure at a relatively low load. The use of a specific grinding
machine or a precision diamond saw may be required to
ensure parallelism of the end faces. A minimum of three
specimens per cement design is recommended.
As shown in Fig. B-31, Young’s modulus is the slope in
the linear domain of the curve of axial stress versus axial
strain. The Poisson’s ratio is the slope in the linear
domain of the radial strain versus axial strain (Fig. B-32).
The linear domain is characterized by reversible defor-
mations.
The following test method can be used at ambient
Fig. B-30. Cantilever set around a cylindrical sample for the mea- temperature to simultaneously determine the Young’s
surement of radial displacements. modulus, Poisson’s ratio, and uniaxial compressive
strength. The method requires at least four cylindrical
specimens (1 in. [25.4 mm] diameter; 2 in. [50.8 mm]
pore pressure caused by external loading. Hence, the length).
pore fluid can drain out of the sample. An undrained 1. Perform a uniaxial compression test to failure on the
condition occurs when the pore fluid cannot drain out of first sample at a constant deformation rate (e.g.,
the cement matrix because of a high loading rate. In this 0.05 mm/min). This gives a rough value of the uniax-
case, the load is shared between the pore fluid and the ial compressive strength (UCS).
cement matrix. The rate of deformation should be low 2. Place axial and radial displacement sensors on the
enough to enable unimpeded drainage. ASTM C469-02, second specimen.
Static Modulus of Elasticity and Poisson’s Ratio of
Concrete in Compression, presents a procedure for

648 Well Cementing

 20

Radial strain
16

12
Axial stress E2
(MPa) E1
8
Axial strain

0
–0.0045 –0.0035 –0.0025 –0.0015 –0.0005 0.0005
Strain (m/m)
Fig. B-31. Radial and axial strains while cycling a compressive stress on a test cement up to half its
compressive strength (ASTM Method C469). E 1 and E 2 are respectively the Young's Modulus of the
first and subsequent loadings.

0.00025

0.00020

0.00015 ν
Radial strain
(m/m) Experimental data points

0.00010

0.00005

0.00000
–0.0035 –0.0030 –0.0025 –0.0020 –0.0015 –0.0010 –0.0005 –0.0000
Axial strain (m/m)

Fig. B-32. Radial strains as a function of axial strains while cycling a compressive stress on a cement
sample. ν is the Poisson’s ratio.

Appendix B Laboratory Testing, Evaluation, and Analysis of Well Cements 649

3. Load this specimen to 50% UCS, and then unload to a pressure cell measure the axial displacement. The end
very low stress (e.g., 30 psi [0.2 MPa]). The test can platens of the cell have a connection to an independent
be recorded as a quality-control check, but the data pump that allows control of the cement pore pressure
will not be used in the analysis. during the test (i.e., a drained condition). If the system
4. Repeat the load-unload cycle in Step 3 two more were completely sealed, the pore pressure would build
times. Record the stress and the axial and radial up during the measurement (i.e., an undrained condi-
strains simultaneously. Use the loading-up portions tion), leading to an incorrect determination of the
of the curve to determine the Young’s modulus and cement properties. If triaxial testing is performed at sev-
Poisson’s ratio. eral confining pressures, the angle of internal friction
can be determined (Chapter 8).
5. Repeat steps 2–4 with the other specimens. Defor-
mations are reversible in the elastic domain, so this
test should be nondestructive. B-6.3.3 Brazilian test
6. Perform a uniaxial compression test to failure with the The uniaxial tensile strength is the stress value of failure
three undamaged specimens to determine the UCS. when the cement sample is subjected to tension (i.e.,
7. Report the mean values of the UCS and the elastic when its ends are pulled). The most common way to
constants. measure tensile strength is to anchor a cylindrical core
at both ends and stretch it to failure. However, this test
is difficult with set cement because the core frequently
B-6.3.2 Confined triaxial compression test breaks at the grip points. Therefore, an indirect method
During a confined triaxial compression test, a cylindrical known as the Brazilian test (Carneiro and Barcellos,
core is placed in a cell and subjected to constant confin- 1953) is often performed to determine uniaxial tensile
ing pressure. The pressure is hydraulic and is applied strength with set cement.
radially through a thin impermeable membrane. Platens The technique involves loading disk-shaped speci-
apply an axial load vertically until the sample fails. The mens in compression across their diameter until they
Hoek cell is an appropriate triaxial cell (Fig. B-33) for fail (Fig. B-34). A tensile stress is generated at the
investigating rocks and set cements. A special cantilever center of the disk in a direction normal to the applied
that fits inside the pressure cell measures the radial load. The use of curved platens, as described by the
deformation of the sample. As in the uniaxial compres- International Society for Rock Mechanics (ISRM)
sion test, displacement transducers located outside the Commission on Standardization of Laboratory and Field
Tests, Suggested Methods for Determining Tensile
Strength of Rock Materials (1978), is preferred to
Axial force improve the stress distribution in the sample. ASTM
International (formerly the American Society for Testing
Strain gauge and Materials) also standardized this test for determin-
lead wires ing the tensile strength of concrete—as described in
Spherically ASTM C496, Standard Test Method for Splitting Tensile
Upper cap seated platen Strength of Cylindrical Concrete Specimens.
of cell
Rubber
membrane Rock core

Annulus Strain
gauges

Oil inlet

Spherically
seated platen
Fig. B-34. Cement sample at the end of the Brazilian test.
Fig. B-33. Hoek cell for triaxial compression test.

650 Well Cementing

 However, for routine tests, simpler methods that do B-6.4.2 Dynamic elastic constants
not require curved platens can be performed on 2-in. Young’s modulus and Poisson’s ratio can also be deter-
[50.8 mm]-diameter and 1-in. [25.4 mm]-thick disk- mined by ultrasonic methods such as logging tools. The
shaped cement specimens, with at least three speci- literature shows that the values obtained from static lab-
mens per design. For testing brittle rocks, Wang and oratory testing are significantly lower than the dynamic
Xing (1999) suggested flattening the two edges of a disk- values measured with logging tools (Vutukuri et al.,
shaped specimen in contact with the press platens and 1974). Dynamic Young’s modulus and Poisson’s ratio
defined the required widths of the flat ends. Another values are calculated from the relationship between the
test procedure involves covering the steel platens of the sample density, ρ, and the velocities of compressional
press with cardboard pads to lower the compressive (P) and shear (S) waves traveling along the loading axis.
stresses at the contact points between specimen and The compressional and shear velocities (vP and vS) are
platens. The pad’s width should not exceed one-tenth of linked to the Young’s modulus, E, and Poisson’s ratio, ν,
the cylinder diameter (Pittet and Lemaître, 2000). by the relationships shown in Eqs. B-11 and B-12.
Once a test method is selected, it must remain the
same throughout a given study or for comparison pur- 1
poses. For all testing methods, Equation B-10 is used to ⎡ E(1 − ν) ⎤ 2
vP = ⎢ ⎥ (B-11)
calculate the tensile strength. Either U.S. or SI units may ⎣ ρ(1 + ν)(1 − 2 ν) ⎦
be used, so long as the choice is consistent.
1
F fail ⎡ E ⎤2
Stens = , (B-10) vS = ⎢ ⎥ (B-12)
2π × ddisk × hdisk ⎣ 2ρ(1 + ν) ⎦
where For homogeneous elastic materials such as steel, the
ddisk = disk diameter elastic properties are independent of the stress and
hdisk = disk thickness wave frequency. The elastic properties are unchanged
Ffail = load at failure whether the measurements are dynamic or static.
Stens = tensile strength. Plona and Cook (1995) showed that the elastic prop-
The test is valid when the primary fracture starts erties of rocks are stress (or strain) dependent as well as
from the center of the disk and spreads along the load- rate dependent. The higher the deformation rate, the
ing diameter. If not, the sample did not fail under ten- higher the elastic modulus and strength will be. This
sion and the test must be discarded. must be taken into account to determine whether
absolute dynamic properties can be used directly in
numerical models.
B-6.4 Dynamic mechanical properties The elastic properties of set cements are strain-
B-6.4.1 Dynamic compressive strength dependent. Figure B-35 compares the dynamic versus
static Young’s moduli of four cement systems. Samples
The UCA is an instrument used to monitor the strength were prepared, cured, and tested under the same con-
development of cements (Section B-4.4) at simulated ditions. Dynamic and static measurements were per-
downhole pressure and temperature. The strength is formed simultaneously. The solid line shows the theo-
indirectly estimated by monitoring the compressional retical perfect match between static and dynamic
velocity of ultrasonic waves transmitted through the values. Note also that dynamic properties are undrained
cement specimen as it hardens. The strength is calcu- properties if the material is saturated.
lated using proprietary empirical algorithms based on Two examples of laboratory-scale acoustic tools for
the sound-wave velocity changes for a given slurry den- examining cements are discussed below.
sity.
■ The portable ultrasonic nondestructive digital indi-
It is important to note that the dynamic strength
measured by the UCA can differ significantly from the cating tester (PUNDIT), shown in Fig. B-36, measures
static UCS (Section B-4.4). The static compression test the compressional velocity of an ultrasonic pulse
is a direct measurement of the uniaxial compressive through a material. The pulse velocity can be used to
strength; therefore, it should be used when entering assess the quality and uniformity of a material. The
data into software modeling tools. method is also useful for detecting the presence of
cracks.

Appendix B Laboratory Testing, Evaluation, and Analysis of Well Cements 651

 16
4
12
Dynamic 2
Young’s 8
modulus 3
1
(GPa)
4

0
0 2 4 6 8 10 12 14 16
Static Young’s modulus (GPa)

Fig. B-35. Dynamic versus static Young’s moduli of four cement sys-
tems (Plona and Cook, 1995). Cement 1 is a 1.5-sg flexible cement,
Cement 2 is a 1.33-sg lightweight cement, Cement 3 is a 1.32-sg 30%
foamed cement, and Cement 4 is a 1.9-sg conventional cement.
Base slurry porosities range from 42% to 58% depending on the
cement system.
Fig. B-37. Dynamic mechanical properties testing equipment (pho-
tograph courtesy Chandler Engineering, LLC).

B-6.5 Thermal properties of set cement

Changes of wellbore temperature caused by production,
cool-down from acid or stimulation treatments, or gas or
steam injection can deform the cement and generate
stress. If stresses are greater than the cement strength,
fissures can develop, compromising the integrity of the
cement sheath. Ultimately, zonal isolation can be lost.
Thermoelasticity theory provides a linear relationship
between the strains, stresses, and temperature varia-
tion, assuming that the cement is homogeneous and
isotropic. If we consider a radial geometry, the tempera-
Fig. B-36. PUNDIT tester. ture distribution as a function of time is obtained from
the heat diffusion equation, which is expressed with the
conditions that the initial and boundary conditions only
■
depend on the radius r as
Equipment that measures cement dynamic mechani-
cal properties under downhole conditions of temper- ∂2 T 1 ∂T ρC ∂T
ature and pore pressure (dynamic Young’s modulus + = , (B-13)
∂r 2 r ∂r λ ∂t
and dynamic Poisson’s ratio) is presently available
(Fig. B-37). The equipment is similar to a UCA, except where
the cell is horizontal to minimize the effects of free λ = thermal conductivity
water and sedimentation and the equipment mea- ρ = density
sures both the compressional and the shear-wave C = specific heat capacity
velocities. The elastic constants are determined from T = temperature.
the velocities and the slurry density using Eqs. B-11
and B-12. However, this equipment cannot address Owing to thermal expansion, the temperature influ-
the effect of confining pressure. ences the stress state of the various materials. Linear
thermal strain, εT , is directly proportional to the tem-
perature variation ΔT (Eq. B-14).
ε T = α × Δ T, (B-14)

652 Well Cementing

where
α = coefficient of linear thermal expansion.
Test methods are proposed below for the determina-
tion of cement thermal conductivity and coefficient of
thermal expansion.

B-6.5.1 Thermal conductivity

A procedure for the measurement of cement thermal Quartz rod
conductivity can be defined following the standard test
method ASTM D2326-70, Method of Test for Thermal
Conductivity of Cellular Plastics by Means of a Probe.
A thermal shock probe system is placed between two flat
faces of a sample of set cement. The specimen size must
be large enough to cover the probe’s surface and reduce
the boundary effects. Once thermal stability is reached,
a thermal pulse is generated to disturb the equilibrium.
Induced temperature changes are recorded as a func-
tion of time. The linear thermal conductivity of the Cement sample
cement is deduced from this relationship. To establish
reproducibility, this test should be performed on a mini-
mum of three specimens.
The greater the value of thermal conductivity, the
faster the temperature will equilibrate in the cement
sheath. Also, the cement will expand or contract more
quickly. Standard lab equipment automates the testing
process and provides a direct readout of the thermal Fig. B-39. Thermal linear expansion measurement apparatus.
conductivity (Fig. B-38).

quartz rod and a digital micrometer positioned on a

stand made of a thick stainless steel plate.
The stand sits inside a water bath. The measuring tip
of the digital indicator, extended with a thin quartz rod,
is in contact with the top of the cement sample bar. The
temperature of the water bath is increased to a value
lower than the cement curing temperature in several
steps. Temperature and sample deformation are
recorded throughout the test. The linear coefficient of
thermal expansion is calculated from the relative
sample dilatation and the temperature increase.

B-7 Cement characterization and analysis

Characterization of cement and cementing materials in
Fig. B-38. Experimental setup to measure thermal conductivity. the laboratory requires appropriate analytical tech-
niques to provide a chemical and physical description of
the sample as a whole or of the chemical constituents of
the sample. This physico-chemical identification may
B-6.5.2 Coefficient of linear thermal expansion include the following:
According to Eq. B-14, the coefficient of thermal expan- ■ qualitative determination of the chemical species
sion is a strain per unit rise in temperature. A simple present
technique to determine this coefficient is mechanical
■ quantitative measurement of the amount of one or
dilatometry. A typical apparatus (Fig. B-39) consists of a
more of these species

Appendix B Laboratory Testing, Evaluation, and Analysis of Well Cements 653

■ determination of physical properties of one or more of Specific identification of surface phases can also be
these species or of the sample as a whole. performed using scanning electron microscopy and light
microscopy. As with XRD, accurate quantitative analysis
Five types of samples are typically examined in a
requires considerable operator skill. Thermal methods
cementing laboratory: neat cement powder, additives,
such as thermogravimetric analysis (TGA) and differen-
dry-blended cement with additives, set cement, and mix
tial thermal analysis (DTA) can be used to determine
water. If a failure to obtain set cement in the wellbore is
gypsum, hemihydrate, and free lime in neat cement
suspected, a sample of fluid returned from the wellbore
(Glasser, 2001).
may be examined to determine the presence or absence
of cementitious material.
A wide variety of analytical techniques is available to B-7.2 Physical characterization of neat cement
characterize cement and cementing materials in the and cementing materials
laboratory. Obtaining an accurate analysis of a sample Physical characterization of neat cement powder in the
may require the use of several of these techniques. An laboratory usually includes measurement of the particle-
excellent review of techniques for the chemical charac- size distribution, surface area, and specific gravity. The
terization of neat and dry-blended cements was pub- determination of specific surface area is frequently per-
lished by Simpson (1988). Many techniques applicable formed using a Blaine permeameter (see ASTM C204,
to the physical characterization of these materials are Standard Test Method for Fineness of Hydraulic
contained in the ASTM standards. A general description Cement by Air Permeability Apparatus), shown in
of the most common applications for these techniques is Fig. B-40. This method is based on the principle of resis-
provided here, organized by sample type and by charac- tance to air flow through a bed of cement. The time, t,
terization classification. required to pass a fixed volume of air through a bed of
cement (12.7 mm diameter and 15 mm depth) is related
B-7.1 Chemical characterization of to the specific surface of cement by a simple relation-
Portland cement ship:
Chemical analysis of Portland cement powder ascertains
A = K t, (B-15)
the relative amounts of the four principal phases (trical-
cium silicate, dicalcium silicate, tricalcium aluminate, where K is a constant for the apparatus. The pressure
and tetracalcium aluminoferrite), gypsum, and certain steadily decreases as air flows through the specimen. A
minor oxides (Chapter 2). X-ray diffraction (XRD) deter- reference cement with standardized specific area, pro-
mines the crystalline phases qualitatively, although reli- duced by the National Institute of Standards and
able quantitative analyses are possible with careful Technology in Washington, D.C., USA, is used to calibrate
attention to sample preparation and the use of consis- the apparatus. The accuracy of this method is highly
tent standards (Aldridge, 1982; Lawrence, 2001). A more dependent on operator skill.
accurate quantitative analysis of the principal cement
phases requires first performing a complete chemical
analysis, using wet chemical methods or atomic excita-
tion methods such as atomic absorption spectrometry
(AAS), inductively coupled plasma spectroscopy (ICP),
and X-ray fluorescence (XRF). The oxide composition of
the cement is calculated, and finally the “potential
phase composition” is calculated using equations devel-
oped by Bogue (1929) and later refined by Taylor (1989).
Bogue’s method is based upon various equilibrium rela-
tionships between the clinker phases (Chapter 2).
Selective chemical extraction and complexometric
techniques facilitate the separation and subsequent
determination of individual phases (Simpson, 1988).
Some of these techniques can be applied specifically to
determine the reactive phases located on the surface of
the cement particle.
Fig. B-40. Blaine permeameter.

654 Well Cementing

 Another technique for determining the specific sur- scattering principle, a particle-size distribution can be
face area is the Wagner method, described in ASTM determined over a wide range of particle sizes, typically
C115, Standard Test Method for Fineness of Portland from 0.1 to 200 μm (Wertheimer and Wilcock, 1976;
Cement by the Turbidimeter. The cement powder is dis- Allen, 1981). Such a device is presented in Fig. B-41.
persed in kerosene and a column of the suspension is These distribution profiles are much more useful in
allowed to settle in a turbidimeter. The intensity of light explaining and predicting performance variations
transmitted through the column is recorded as a func- between cement samples than simple surface area mea-
tion of time and depth below the surface of the column. surements. A typical particle-size distribution curve for
The Wagner method assumes that the cement particles Class G cement is shown in Fig. B-42.
are spherical, and all particles with a diameter of less
than 7.5 μm are assumed to have an average diameter of
3.8 μm. As a result, this method gives a lower specific
surface than the Blaine method. Values of fineness
determined by the Wagner method are included in the
specifications for some classes of API/ISO cements
(Chapter 2).
A measurement of specific surface area that is inde-
pendent of particle shape can be obtained using a gas
adsorption technique, the Brunauer-Emmett-Teller
(BET) method (1938). The instrumentation for this mea-
surement is considerably more sophisticated and expen-
sive than a Blaine permeameter or a Wagner turbidime-
ter, so BET surface area measurements are not obtained
routinely in the field laboratory.
The above surface-area measurements provide a
determination of average particle size and surface area
within a sample, but they provide no information about Fig. B-41. Laser light scattering particle size analyzer (photograph
the range of particle sizes within that sample. Using courtesy Malvern Instruments).
more sophisticated instrumentation based on a light-

10 100
d [4, 3] = 24.19 µm
d [3, 2] = 9.28 µm
d [V, 0.1] = 3.67 µm
8 d [V, 0.5] = 19.34 µm 80
d [V, 0.9] = 48.75 µm

6 60
Volume distribution Cumulative volume
within size band smaller than stated size
(%) 4 40 (%)

2 20

0 0
0 1 10 100 1,000
Particle diameter (µm)

Fig. B-42. Typical Portland cement particle-size distribution.

Appendix B Laboratory Testing, Evaluation, and Analysis of Well Cements 655

 The specific gravity of neat cement powder may vary determinations for the separated species. Lost-circula-
from 3.10 to 3.25. This variability could result in slurry- tion materials and other additives with large particle
density deviations up to 0.27 lbm/gal [33 kg/m3] for slur- size (such as CaCl2 flakes) frequently are physically sep-
ries with a constant water-to-solids ratio. arated from a cement blend by sieving, followed by iden-
The standard method for determining the specific tification and quantification using standard techniques.
gravity of these materials in the laboratory uses the Le Determination of salts, extenders, weighting agents, and
Châtelier flask (ASTM C188, Standard Test Method for silica can be performed using a combination of XRD and
Density of Hydraulic Cement). The procedure is simple XRF techniques without separation of these materials
in concept, but considerable skill and time are required from the blend (Simpson, 1988). Accurate quantitative
for accurate results. A pycnometer can also be used results require considerable care and may not be obtain-
(ASTM C128, Standard Test Method for Density, able for small quantities of the additive.
Relative Density (Specific Gravity), and Absorption of Optical microscopy has been shown to be a useful tool
Fine Aggregate). The procedure can be performed for the qualitative analysis of cement blends (Reeves
quickly and with accuracy comparable to using the Le et al., 1983). The presence of various cement additives
Châtelier flask. New devices such as the helium pyc- can be confirmed by using a microscope equipped with
nometer (Fig. B-43) can provide rapid and fully auto- polarization and fluorescence options, combined with
matic density determination on a wide variety of materi- simple laboratory techniques.
als including powders, objects, and slurries. Such
devices are now the preferred testing apparatuses in the
laboratory for density determination. B-7.4 Chemical characterization of set cement
Many of the same techniques used to characterize neat
cement powders and dry-blended cements can be used
to examine set cement, although obtaining meaningful
descriptions of material requires skillful sample prepa-
ration and interpretation of analytical results.
Frequently, accurate quantitative analysis is possible
only with considerable knowledge about the properties
of the components of the material under investigation. A
combination of XRD and XRF techniques has been used
with some success (Simpson, 1988).

B-7.5 Analysis of cement mix water

Fig. B-43. Example of automated helium pycnometer (photograph Procedures for the analysis of oilfield waters are con-
courtesy Micromeritics). tained in API RP 45, Recommended Practice for
Analysis of Oilfield Waters. This document is directed
toward the determination of both dissolved and dis-
persed components in oilfield waters. The reader should
B-7.3 Analysis of dry-blended cements
consult API RP 45 for detailed information concerning
Chemical analysis of dry-blended cements offers a more the methods.
accurate indication of blend homogeneity than perfor- In the context of well cementing, prejob testing of
mance testing, and knowing the additive content also cement slurries is frequently performed with a mix-
can be useful in explaining performance variations from water sample from the job location. Depending upon the
one blend to another. Many analyses involve some type of test results, the cement-slurry composition may be
separation technique to isolate the material of interest. adjusted to obtain the desired performance. In cases in
Most retarders and dispersants are chemically struc- which problems have occurred during a cement job, mix-
tured so that they absorb ultraviolet radiation; hence, water samples are often analyzed in an effort to deter-
these materials can be selectively dissolved or extracted mine why difficulties were experienced.
and determined by ultraviolet (UV)-absorption spec- The most common chemical variables that are known
trophotometry (ASTM C114, Standard Test Methods for to affect the performance of Portland cement slurries
Chemical Analysis of Hydraulic Cement). Extraction include
techniques have also been applied to fluid-loss additives, ■ pH
although the diverse chemical nature of these materials
■ alkalinity
has necessitated the use of a wide variety of analytical

656 Well Cementing

■ ammonia B-9 Acronym list
■ chloride AAS Atomic absorption spectrometry
■ iron API American Petroleum Institute
■ organic acids BET Brunauer-Emmett-Teller
■ sulfate. BHCT Bottomhole circulating temperatures
These variables can either affect the hydration and BHST Bottomhole static temperature
setting of Portland cement itself or interfere with CHA Cement hydration analyzer
cement additives. DTA Differential thermal analysis
The analytical methods involved in mix-water testing
vary depending upon the analyte. In larger laboratories, FTIR Fourier transform infrared spectroscopy
techniques such as AAS and ICP are commonly used to ICP Inductively coupled plasma spectroscopy
determine metallic species. Organics are determined by ISO International Organization for
Fourier transform infrared spectroscopy (FTIR) or chro- Standardization
matographic methods. Wet chemistry methods are also ISRM International Society for Rock Mechanics
used. In smaller laboratories, or at field locations, ion-
specific electrode meters or portable titration kits are LVDT Linear variable differential transformer
frequently employed to determine inorganic and some PUNDIT Portable ultrasonic nondestructive digital
organic species. indicating tester
RP Recommended Practice
TGA Thermogravimetric analysis
B-8 Summary UCA Ultrasonic cement analyzer
The technology currently available for testing well
UCS Uniaxial compressive strength
cements is sophisticated; nevertheless, there is much
room for improvement. Consequently, energy compa- UV Ultraviolet
nies, service companies, and instrument manufacturers XRD X-ray diffraction
are engaged in research to improve existing techniques XRF X-ray fluorescence
or to develop new procedures and equipment to simulate
well conditions more accurately. Many methods or
devices are proprietary, so their use is often limited to
the company that invented the method.
Table B-4 provides references to procedures that are
presently available to all industry laboratories.

Appendix B Laboratory Testing, Evaluation, and Analysis of Well Cements 657

Table B-4. Summary of Test Procedures for Well Cements
Test Category Equipment or Methods References
Sampling Mechanical splitter ASTM C702
Tube sampler API RP 10B/ISO 10426-2
Diverted flow sampler Gerke et al., 1985

Slurry preparation Two-speed propeller mixer API RP 10B/ISO 10426-2

Thickening time Atmospheric consistometer API RP 10B/ISO 10426-2

Pressurized consistometer

Fluid loss High-pressure fluid-loss cell using 325-mesh screen API RP 10B/ISO 10426-2

Compressive strength A 2-in. × 2-in. curing mold, placed in a API RP 10B/ISO 10426-2
water bath or in a pressurized autoclave
Hydraulic press
UCA Rao et al., 1982

Mechanical properties Curing equipment ASTM C39, ASTM

Compression load frame C496/C496M-04
Axial and radial deformation sensors

Free fluid 250-mL graduated cylinder API RP 10B/ISO 10426-2

Permeability Water permeameter API RP 10B/ISO 10426-2

Rheology Rotational viscometer (couette type) API RP 10B/ISO 10426-2

Rotational viscometer (Searl type) Orban and Parcevaux, 1986

Static gel strength Rotational viscometer (couette type) API RP 10B/ISO 10426-2
API RP 13B

Expansion and shrinkage Bar mold ASTM C151

Cylindrical sleeves ISO 10426-5
Annular ring mold
Membrane test
CHA test API TR 10TR

Spacer, wash, and cement compatibility Rotational viscometer (couette type) with API RP 10B/ISO 10426-2
pressurized consistometer, fluid-loss cell,
pressurized autoclave, and hydraulic press

Portland cement phase analysis Wet chemical methods, XRD, XRF, AA, or ICP Aldridge, 1982
phases, calculated by Bogue equations Bogue, 1929

Determination of gypsum, hemihydrate, Wet chemical methods Simpson, 1988

and free lime in Portland cement
TGA El-Jazairi and Illston, 1977
DTA Midgley, 1979

Particle analysis of Portland cement Blaine permeameter ASTM C204

Wagner turbidimeter ASTM C115
BET Brunauer et al., 1938
Laser light scattering Wertheimer and Wilcock,1976

Specific gravity of Portland cement Le Châtelier flask ASTM C188

Pycnometer ASTM C114

Chemical analysis of dry blends UV-absorption spectrophotometry ASTM C114

XRD Simpson, 1988
XRF Reeves et al., 1983
Optical microscopy

Chemical analysis of mix water Wet chemical method API RP 45

AAS Simpson, 1988
ICP

658 Well Cementing

C-1 Introduction
Calculations are an essential part of a cement job
design, because they determine the properties of a
Cementing Calculations
Thomas J. Griffin, Jr.—Griffin Cement Consulting LLC

Table C-1. API/ISO Cement-Slurry Specifications

for Oilwell Cements†
Cement Mix Water Slurry Density
C Yield
cement system—density, yield, volume of mix water, Class (%BWOC) (lbm/gal [kg/m3]) (ft3/sk [m3/t])
and proportions of additives. In addition, depending A 46 15.6 [1,870] 1.18 [0.355]
upon the type of cement job, calculations are necessary
to determine the slurry-volume requirements and pres- B 46 15.6 [1,870] 1.18 [0.355]
sures exerted by various fluids during the treatment. C 56 14.8 [1,770] 1.32 [0.397]
In this appendix, five categories of calculations are
discussed. D 38 16.45 [1,970] 1.05 [0.316]
■ Cement-slurry properties E 38 16.45 [1,970] 1.05 [0.316]
■ Primary cement job design
F 38 16.45 [1,970] 1.05 [0.316]
■ Squeeze cement job design
G 44 15.8 [1,890] 1.15 [0.346]
■ Cement plug design

■ Foamed cement job design

H 38 16.45 [1,970] 1.05 [0.316]
† Reproduced courtesy of the American Petroleum Institute.
U.S. customary units are used for oilfield calculations
in many areas of the world. Accordingly, the calculations
presented in this appendix are in U.S. units. Conversion The mix-water concentrations are mainly a function
factors to Système Internationale (SI) units appear at of the cements’ surface areas. When cement additives
the end of this appendix. are present in the system, the water concentrations nec-
The proportions of cements and additives are tradi- essary to obtain the desired cement-slurry properties
tionally expressed in the industry as “weights.” “Mass” is will vary. The important cement-slurry properties
the correct term; nevertheless, in this appendix, the tra- affected by the mix-water concentration and the pres-
ditional term will be used. ence of additives include slurry density (for well control
and avoiding lost circulation), free water, sedimentation,
rheology, compressive strength, fluid-loss control, and
C-2 Cement-slurry properties permeability. This appendix presents methods for calcu-
American Petroleum Institute (API) Specification 10A lating the proportions of materials for the various types
(2002, Specifications for Cements and Materials for of cement systems.
Well Cementing, and the corresponding International
Organization for Standardization (ISO) 10426-1 publica- C-2.1 Specific gravity of Portland cement
tion specify an amount of water to be mixed with each
class of oilwell cement (Table C-1). The API water con- The specific gravity of Portland cement varies between
centrations are expressed by weight of cement (BWOC), 3.10 and 3.25, depending upon the raw materials used in
and the corresponding slurry densities and yields are its manufacture. For the cement-slurry calculations to
used for testing to verify that a cement conforms to the be precise, one should measure the specific gravity of
API performance specifications (Chapter 2). The API the cement to be used on a specific job (Appendix B). In
water concentrations are not intended to be a recom- this appendix, a specific gravity of 3.14 is assumed in all
mendation for field application, because the use of addi- calculations.
tives modifies the performance properties and makes it
necessary to adjust the water content.

Well Cementing ■ Appendix C Cementing Calculations 659

C-2.2 Absolute and bulk volumes Table C-3. Absolute Volumes of Common Cementing
The volume of a particulate material is expressed in two Materials
ways. The absolute volume is the volume occupied only Material Absolute Volume Specific
by the material itself and does not include the volume of (gal/lbm [m3/t]) Gravity
air surrounding the particles. The bulk volume is the Barite 0.0278 [0.231] 4.33
sum of the volumes occupied by the material and its sur-
rounding air. Bentonite 0.0454 [0.377] 2.65
In the United States, Portland cement normally has Coal (ground) 0.0925 [0.769] 1.30
a bulk volume of 1 ft3 for 94 lbm, which is commonly
referred to as a sack. The absolute volume occupied by a Gilsonite 0.1123 [0.935] 1.06
94-lbm sack of cement is 3.59 U.S. gal or 0.48 ft3, assum- Hematite 0.0244 [0.202] 4.95
ing a specific gravity of 3.14. Other cements such as com-
mercial lightweight formulations or calcium aluminate Ilmenite 0.0270 [0.225] 4.44
cement have different absolute and bulk volumes. A list
Silica sand 0.0454 [0.377] 2.65
of bulk and absolute volumes of several common
cements is presented in Table C-2. NaCl (above saturation) 0.0556 [0.463] 2.15

Fresh water 0.1198 [1.000] 1.00

Table C-2. Cement Volume Factors
Cement Sack Bulk Absolute
Weight Volume Volume
(lbm) (ft3/sk [m3/t]) (gal/lbm [m3/t]) C-2.3 Concentrations of additives
API Classes A through H 94 1.0 [0.301] 0.0380 [0.317] The concentrations of most solid cement additives are
expressed as a percentage BWOC. This method is also
TXI™ Lightweight† 75 1.0 [0.377] 0.0425 [0.355]
used for water. For example, if 35% (BWOC) silica sand
Ciment Fondu Lafarge™‡ 87.5 1.0 [0.323] 0.0373 [0.312] is used in a cement blend, the amount for each sack of
cement is 94 lbm × 0.35 = 32.9 lbm of silica sand. This
LUMNITE™§ 94 1.0 [0.301] 0.0380 [0.317] results in 94 + 32.9 = 126.9 lbm of total mix. The true
† TXIis a trademark of Texas Industries, Inc.
‡ Ciment Fondu Lafarge is a trademark of Lafarge Aluminates. percent of silica sand in the mix is
§ LUMNITE is a trademark of Heidelberger Calcium Aluminates.

32.9
× 100 = 25.9%.
The absolute and bulk volumes of cement additives 126.9
are available from literature published by the major In the case of cement/fly ash blends, the fly ash is
cementing service companies. A listing of such informa- considered to be part of the cementitious material
tion for some commonly used cementing materials is because it reacts with calcium hydroxide and con-
shown in Table C-3. This table should not be used for tributes C-S-H phase to the set cement matrix (Chapter
slurry design purposes. Cementing companies may 3). Therefore, additive concentrations are expressed by
obtain materials from different sources; consequently, weight of cement plus fly ash, and the abbreviation
the absolute and bulk volumes may vary. “BWOC” is used.
Materials that dissolve in water do not occupy as In some cases, additive concentrations are given by
much space as their dry absolute volumes. For soluble weight of blend (BWOB). This convention is useful when
additives like retarders, dispersants, and fluid-loss addi- the blend is a complex mixture of cement and other bulk
tives, which are added in relatively small amounts, the components.
difference is negligible. Salt (NaCl), however, is usually Salt is a special exception. It is added by weight of
added in much larger concentrations; consequently, the mix water (BWOW). In addition, weighting materials
difference must be taken into account. This point is dis- such as barite are often added on a “pounds per sack
cussed later. (lbm/sk)” basis. This is done for convenience, because it
eliminates the need to convert from percent BWOC to
pounds in the bulk plant.
Liquid additive concentrations are commonly
expressed in gallons per sack of cement (gal/sk) [liters
per SI ton (L/t)] of cementitious material.

660 Well Cementing

C-2.4 Slurry density and yield essary to ensure that enough water is available at the
The slurry density is calculated by adding the weights of wellsite. It is simply the number of gallons calculated as
the cement-system components, and dividing that sum above, multiplied by the number of sacks of cement to be
by the total of their absolute volumes. In other words, to mixed.
determine the slurry density in lbm/gal, divide the total Most additives are handled in the same manner.
weight in pounds of cement, water, and additives by the Often, when the calculations are performed by hand, the
total volume in gallons of cement, water, and additives additives present in minor amounts (less than 1%) are
(Eq. C-1). ignored. Today, most laboratories use computers to cal-
culate the slurry mixes and the density and to determine
mcem + mwater + madd the amounts of additives to use for laboratory testing. All
ρ slurry = , (C-1)
Vcem + Vwater + Vadd additives are taken into consideration.

where Example calculation

m = weight Consider a slurry composed of Class G cement + 35%
V = volume. silica flour + 1% solid fluid-loss additive + 0.2 gal/sk
The yield of a cement system is the volume occupied liquid dispersant + 44% water.
by one unit of cement plus all of the additives and the Component Weight Absolute Volume Volume
mix water. For cement measured in sacks, the yield is (lbm) (gal/lbm) (gal)
expressed in cubic feet per sack (ft3/sk) [or cubic meters
Cement 94 0.0382 3.59
per SI ton (m3/t)]. The yield is then used to calculate the
number of sacks required to achieve the desired fill in Water 41.36 0.1198 4.96
the annulus. Most slurry density calculations are per-
formed on the basis of one sack of cement (94 lbm) (or Silica flour 32.9 0.0454 1.49
SI ton). This simplifies the calculation of the slurry yield. Fluid-loss additive 0.94 0.0932 0.088

Liquid dispersant 1.97 0.1014 0.20

Example calculation
Total 171.17 10.32
Consider a slurry composed of Class G cement plus 44%
water (94 lbm × 0.44 = 41.36 lbm water).
171.17 lbm
Component Weight Absolute Volume Volume ρ slurry = = 16.59 lbm/gal
(lbm) (gal/lbm) (gal) 10.32 gal
Cement 94 0.0382 3.59 10.32 gal/sk
Yslurry = 3
= 1.38 ft3/sk
Water 41.36 0.1198 4.96 7.48 gal/ft
Total 135.36 8.54

C-2.4.1 Special additives

135.36 lbm
ρ slurry = = 15.84 lbm/gal Salt
8.54 gal
As mentioned earlier, NaCl concentration is commonly
The slurry yield, Yslurry, is now determined by dividing expressed as a percentage BWOW. Because it is used at
the total volume of the slurry per 94-lbm sack of cement high concentrations, NaCl must be considered in both
(8.54 gal) by the conversion factor of 7.48 gal/ft3. density and yield calculations. The absolute volume of
NaCl dissolved in water is less than its dry absolute
8.54 gal/sk
Yslurry = 3
= 1.14 ft3/sk volume, and it varies with concentration (Table C-4).
7.48 gal/ft

Another important calculation is the amount of mix

water required for the cementing operation. This is nec-

Appendix C Cementing Calculations 661

Table C-4. Absolute Volume of Salt in Water at 80°F [26.7°C] 150.75 lbm
ρ slurry = = 16.31 lbm/gal
NaCl Concentration Absolute Volume in Water 9.22 gal
(%BWOW) (gal/lbm [m3/t])
9.22 gal/sk
2 0.0371 [0.310] Yslurry = 3
= 1.24 ft3/sk
7.48 gal/ft
4 0.0378 [0.316]
For other concentrations of salt, the corresponding
6 0.0384 [0.321] absolute volume factor for the concentration would be
8 0.0390 [0.326] used in the above calculation.
10 0.0394 [0.329] Fly ash
12 0.0399 [0.333] As discussed in Chapter 3, fly ash is a pozzolanic exten-
der that is often used to replace part of the cement. A
14 0.0403 [0.336]
special convention defined by the API (Recommended
16 0.0407 [0.340] Practice 10B, [Recommended Practice for Testing Well
Cements] Section 17) is used to describe fly ash/cement
18 0.0412 [0.344] blends. Such mixtures are normally written as ratios of
20 0.0416 [0.347] the absolute-volume contributions of the two compo-
nents. A ratio of 35:65 refers to 35% fly ash and 65%
22 0.0420 [0.351] cement (the first number always represents the fly ash
24 0.0424 [0.354] and the second the cement). Other common ratios are
50:50 and 75:25. The quantity of fly ash necessary to pre-
26 0.0428 [0.357] pare 3.59 gal (absolute volume) of cement blend may be
28 0.0430 [0.359] calculated from the following formula.

30 0.0433 [0.361] ⎛ γ flyash ⎞

m flyash = mcemrepl ⎜ ⎟ (C-2)
34 0.0439 [0.366] ⎝ γ cem ⎠
37.2 (saturated) 0.0442 [0.369]
where
m = weight
γ = specific gravity.
Example calculation The weight of the cement replaced, mcemrepl, is 94 lbm
Consider a slurry composed of Class G cement + 37.2% minus the amount of cement remaining. For a 35:65 fly
NaCl (BWOW) + 44% water. What are the slurry density ash/cement blend, this is 94 lbm – (0.65 × 94 lbm) = 32.9
and the yield? lbm. Thus, for a 35:65 fly ash/cement blend, using a
cement with a specific gravity of 3.14 and a fly ash with
94 lbm cement × 0.44 = 41.36 lbm water a specific gravity of 2.46, the required weight of fly ash is
41.36 lbm water × 0.372 = 15.39 lbm NaCl
2.46
Referring to Table C-4, the absolute volume of NaCl at m flyash = 32.9 lbm × = 25.8 lbm.
a concentration of 37.2% BWOW is 0.0442 gal/lbm. Thus, 3.14
the calculation can be completed as follows.
The weight of cement, mcem, is
Component Weight Absolute Volume Volume mcem = 0.65 × 94 lbm = 61.1 lbm.
(lbm) (gal/lbm) (gal)
Cement 94 0.0382 3.59 The mixture of 25.8 lbm of fly ash with 61.1 lbm of
cement weighs 86.9 lbm. This weight is referred to as the
NaCl 15.39 0.0442 0.68 equivalent sack (abbreviated as eq sk), because the
Water 41.36 0.1198 4.96
absolute volume of the blend is 3.59 gal. For such sys-
tems, additive concentrations are calculated as a func-
Total 150.75 9.22 tion of the equivalent sack, not a 94-lbm sack of Portland
cement. In the above case, 1% of an additive (BWOC)
would mean 1% of an 86.9-lbm eq sk.

662 Well Cementing

 Fly ashes vary widely, and the specific gravity must be hydrate bentonite in fresh water, and the efficiency of
determined for each. The water requirements for the dif- the material as an extender is improved up to four times.
ferent fly ashes will vary according to desired perfor- The ability of bentonite to absorb water is known as its
mance properties, fineness, and the chemical composi- yield. Prehydrated bentonite “yields” four times as effi-
tions of both the fly ash and cement. ciently as when it is dry blended. Unfortunately, this
Outside the United States, ratios of fly ash and has resulted in some confusion concerning calculations
cement are referred to using other conventions. Users involving prehydrated bentonite. The improved effi-
should check to ensure consistency of use in communi- ciency of prehydrated bentonite has no bearing on the
cations. slurry density or the yield of a sack of cement. Bentonite
calculations are based on the actual amounts of materi-
als that are present in the blend.
Example calculation In editions before 1989, API Specification 10 recom-
Consider a slurry consisting of a 50:50 fly ash/Class G mended the addition of 5.3% water for each percent of
blend plus 2% bentonite and 54% water (γ flyash = 2.48; bentonite (BWOC). This is only a general guide, because
therefore, absolute volume = 0.0483). the extending efficiency of bentonite varies from batch
mcemrepl = 94 lbm – (0.50 × 94 lbm) = 47.0 lbm to batch. The actual amount of water should be based on
the desired performance properties of the cement (e.g.,
2.48 free water, gel strength, etc.) as determined in the labo-
m flyash = 47.0 lbm × = 37.12 lbm ratory.
3.14
mcem = 0.50 × 94 lbm = 47.0 lbm Example calculation
meqsk = 37.12 lbm + 47.0 = 84.12 lbm For a slurry consisting of Class G cement + 2% prehy-
mwater = 0.54 × 84.12 = 45.47 lbm drated bentonite + 86.4% water, the calculation is per-
formed as follows.
Component Weight Absolute Volume Volume
(lbm) (gal/lbm) (gal)
Component Weight Absolute Volume Volume
Cement 47 0.0382 1.794 (lbm) (gal/lbm) (gal)
Fly ash 37.12 0.0483 1.794 Cement 94 0.0382 3.59
Bentonite 1.68 0.0452 0.076 Bentonite 1.88 0.0452 0.085
Water 45.47 0.1198 5.44 Water 81.22 0.1198 9.73
Total 131.23 9.11 Total 177.10 13.40

131.23 lbm 177.10 lbm

ρ slurry = = 14.41 lbm/gal ρ slurry = = 13.22 lbm/gal
9.11 gal 13.40 gal
9.11 gal 13.40 gal
Yslurry = 3
= 1.22 ft3/eq sk Yslurry = 3
= 1.79 ft3/eq sk
7.45 gal/ft 7.48 gal/ft

Note that in this case, water and bentonite are deter- For 8% dry-blended bentonite at the same density (90.6%
mined based on the equivalent sack and the yield is water), the calculation is shown below.
expressed in cubic feet per equivalent sack (ft3/eq sk) to
show that it is based on both the cement and the fly ash. Component Weight Absolute Volume Volume
(lbm) (gal/lbm) (gal)
Bentonite Cement 94 0.0382 3.59
As discussed in Chapter 3, bentonite and attapulgite are
Bentonite 7.52 0.0452 0.34
clay minerals added to develop gel strength and allow
the addition of extra water to cement slurries. The Water 85.16 0.1198 10.21
resulting slurries are less dense and more economical.
Bentonite can be dry blended with the cement or pre- Total 186.68 14.14
hydrated in the mix water. About 30 min is required to

Appendix C Cementing Calculations 663

 186.68 lbm Water requirement
ρ slurry = =13.2 lbm/gal
14.14 gal It is sometimes necessary to calculate the volume of
water required to achieve a desired density. This is done
14.14 gal with an unknown, similar to the calculation of a weight-
Yslurry = 3
=1.89 ft3/eq sk
7.48 gal/ft ing agent as illustrated above. A variation is that it is
convenient to set the unknown equal to the volume of
For comparison, the yield and mix water for a water.
13.2-lbm/gal slurry are shown below.
Example calculation
Bentonite Mix Water (gal/sk) Yield (ft3/sk) For a slurry consisting of Class G cement plus 8% ben-
Dry blend 10.21 1.89 tonite, how much water is necessary for a density of 13.0
lbm/gal?
Prehydrated 9.73 1.79

Component Weight Absolute Volume Volume

(lbm) (gal/lbm) (gal)
Weighting agents Cement 94 0.0382 3.59
Certain materials are used in cement slurries to increase Bentonite 7.52 0.0452 0.34
the density of the slurry for well-control purposes. When
these materials are used, density must be specified to Water 8.34x x
calculate the amount of weighting agent to achieve a Total (101.52 + 8.34x ) (3.93 + x )
given slurry density.

Example calculation (101.52 + 8.34 x)lbm

ρ slurry = 13.0lbm/gal =
For a slurry composed of Class G cement plus 44% water, (3..93 + x )gal
how much hematite (absolute volume: 0.0242 gal/lbm) x = 10.82 gal
is required to prepare a slurry with a density of 18.5 lbm/
gal? Thus, 10.82 gal of water are required for each sack of
cement.
Component Weight Absolute Volume Volume
(lbm) (gal/lbm) (gal)
Cement 94 0.0382 3.59
C-3 Primary cementing calculations
Calculations for primary cementing determine the fol-
Hematite x 0.0242 0.0242x lowing.
Water 41.36 0.1198 4.96 ■ Cement volume (annular volume) with excess

■ Cement density and yield

Total (135.36 + x) (8.55 + 0.0242x )
■ Cement mix water

■ Displacement to land plug

(135.36 + x )lbm
ρ slurry = 18.5 lbm/gal = ■ Pump pressure to land plug
(8.55 + 0.0242 x)gal
■ Hydrostatic pressure on formation (fracture and pore
x = 40.97 lbm pressures)
■ Pressure to lift casing
Thus, because the calculation was based on one sack
of cement, 40.97 lbm of hematite is required per sack to
increase the density of the Class G slurry to 18.5 lbm/gal.
C-3.1 Annular volumes
This may be expressed in pounds per sack (40.97 lbm/sk)
or in percent (40.97 lbm/94 lbm × 100 = 43.6% BWOC). Annular volumes are calculated to determine the volume
The properties of the slurry are also carefully measured of cement slurry required to fill the annulus to a certain
to ensure that the slurry will perform as desired. level. Often during the design stage, these calculations
Particularly important are the slurry’s rheology, free are based on the bit size plus an excess volume to
water, and tendency for the weighting agent to settle. account for hole enlargement or losses as determined
from experience in a field.

664 Well Cementing

 After reaching the casing point, caliper logs should be given in standard tables. The displacement volume
run and the cement volume adjusted based on the actual should be that between the pumps and the landing
hole size. Even with the caliper, it is common practice to collar. Normally, a small excess volume may be pumped
use an excess volume to ensure fill-up by cement across to allow for the compression of entrained air in the fluid
all zones requiring isolation. The type of caliper can and to account for pump inefficiency.
affect the amount of cement computed and the resulting
fill-up by the cement. Two- and three-arm calipers, with
arms that operate together, may underestimate—or C-3.4 Pump pressure to land the plug
overestimate in the case of the two-arm caliper—the The pump pressure required to land the plug is the
size of the hole. This is especially true for deviated wells, hydrostatic-pressure difference between the fluids in
because the holes tend to be oval. For these situations, the annulus and the pipe. Depending on the pump rate,
four- and six-arm calipers with independently operating additional pressure may be required to overcome fric-
arms provide the most accurate hole-size information. tion pressure. U-tube or placement simulators are the
Most logging companies also offer computerized best way to determine the friction pressure. Such pres-
annular volume calculations that are presented on the sure information is needed to determine the type of
log. The use of caliper logs for cement-job design is dis- pump required, to ensure that the cementing head is
cussed in detail in Chapter 12. adequate, and to avoid bursting the casing.
Once the annular volume is calculated, an excess is
added (normally 10% to 100%, based on experience in C-3.5 Hydrostatic pressure on the formation
the field), and the number of sacks required to fill this
(fracture and pore pressure)
volume is computed based on the yield of the cement.
Excess factors must be based on experience. To ensure the safety of the well, one must determine
Normally, the excess is calculated only for the openhole whether the well will flow or be fractured during or after
section being cemented. This excess accounts for the cementing treatment. A good first approximation
cement slurry that may be lost into the formation, for an can be obtained by calculating the hydrostatic pressures
enlarged hole, or for fluid that may be lost from the at the critical points in the wellbore.
cement into permeable zones. When returns to the sur- Friction pressure is particularly important when
face are desired or required, additional excess volume determining the possibility of fracturing weak forma-
may be used to ensure that this is achieved. tions. Because the hydrostatic and friction pressures are
Care must be taken in using excess. If the well has constantly changing during a primary cement job, the
weak formations that are close to being fractured, excess preferred method is to use computerized placement sim-
cement could raise the top of the cement and may cause ulators. If such simulators are unavailable, the calcula-
the formation to be fractured because of increased tions given below can be used.
hydrostatic and friction pressures. To determine the hydrostatic pressure component,
The final step in calculating the cement-slurry the following equation is used.
volume is to add the volume of slurry that will remain in ph = 0.052ρh, (C-3a)
the shoe joints between the float collar and the shoe.
This is simply the casing volume between the two points. where
h = height of column having density ρ (ft)
C-3.2 Density, yield, and mix water ph = hydrostatic pressure (psi)
ρ = density of the fluid (lbm/gal).
The mix-water volume, slurry density, and yield are
determined as described in Section C-2.4. The cement- Alternatively
ing service company usually calculates the mix-water ph = 0.00982ρh, (C-3b)
volume. This is the sum of the water volumes required
for mixing cement, spacers, and preflushes, as well as for where
displacement. Excess water should also be available at h = height of column having density ρ(m)
the wellsite for other contingencies. ph = hydrostatic pressure (kPa)
ρ = density of the fluid (kg/m3).
C-3.3 Displacement volume to land the plug If there are several fluids in the wellbore, this calcu-
The displacement volume to land the plug is simply a lation must be performed for each and their respective
calculation of the capacity of the pipe. The length of the hydrostatic pressures added. This value should then be
pipe (or segments of pipe if the entire string is not the compared with the pressures of the critical formations,
same size or weight) is multiplied by the pipe capacity as and should lie between their pore and fracturing pres-
sures.

Appendix C Cementing Calculations 665

C-3.6 Example well calculations In open hole:
The following calculations would be performed for the Volume between 95⁄8-in. casing and 121⁄4-in. open hole,
well depicted in Fig. C-1. Additional well information is with 25% excess
given below.
V2 = (4,450 – 1,700)ft × 0.3132ft3/ft × 1.25 = 1,076.6 ft3

13 3⁄ 8 in., 54.50 lbm/ft casing

Total lead cement volume:
95⁄ 8 in., 36.00 lbm/ft casing Vlead = V1 + V2 = 108.8 + 1, 076.6 = 1,185.4 ft3.
V1 1,400 ft (top of cement)
300 ft
1,700 ft Tail cement
Open hole:
V2
2,750 ft Volume between 4,450 ft and 4,950 ft
121⁄4 in. (open hole)
V3 = 500 ft × 0.3132 ft3/ ft × 1.25 = 195.8 ft3.
V4 4,450 ft (top of tail)

4,908 ft (float collar)

Shoe joint:
500 ft V3
4,950 ft (shoe) V4 = 42ft × 0.4341ft3/ ft = 18.2 ft3 .

Fig. C-1. Example well for primary cementing calculations. Total tail cement volume:

Vtotal = V3 + V4 = 195.8 + 18.2 = 214.0 ft3 .

Well information
Displacement volume
Surface casing: 13 3⁄8 in. (54.50 lbm/ft) to 1,700 ft
Openhole size: 121⁄4 in. to 4,950 ft Vdisp = (4,950 – 42)ft × 0.0773 bbl/ft = 379.4 bbl.
Drilling mud: 11.50 lbm/gal
Casing: 9 5⁄8 in. (36.00 lbm/ft) Pump pressure to land the plug
Excess required: 25% (over open hole) Note: When calculating pressure, true vertical depth
Shoe joint: 42 ft (TVD) is used.
For hydrostatic pressure inside casing (ph)i:
Top of cement: 300 ft inside 133⁄8-in. casing
Mud hydrostatic pressure,
Top of tail: 4,450 ft
Lead system: 13.0 lbm/gal (yield: 1.50 ft3/sk) ( ph ) mud = 0.052 × 11.5 lbm/gal × (4,950 – 42) ft
Tail system: 16.4 lbm/gal (yield: 1.05 ft3/sk) = 2,935.0 psi.
Spacer: 12.5 lbm/gal (volume: 40 bbl)
Tail hydrostatic pressure,
Displacement fluid: 11.5 lbm/gal mud
Weak formation: 3,215-psi fracture pressure at 4,320 ft ( ph )tail = 0.052 × 16.4 lbm/gal × 42ft = 35.8 psi.
Highest pore pressure: 3,150 psi at 4,800 ft ( ph ) i = 2,935.0 psi + 35.8 psi = 2,970.8 psi.

For hydrostatic pressure outside casing (ph)o, the sum

Cement volume calculations of hydrostatic pressures of all annular fluids must be
Note: When calculating volume, measured depth is used. calculated.
Mud: The height of the mud is 1,400 ft less than the
Lead cement length of the spacer. The length of the spacer is 40 bbl ×
Inside surface casing: 15.48 ft/bbl (from standard tables) = 619.2 ft. Thus, the
Volume between 9 5⁄8-in. and 133⁄8-in. casings mud height is 1,400 – 619.2 = 780.8 ft.

V1 = 300 ft × 0.3627 ft3/ft = 108.8 ft3 ( ph ) mud = 0.052 × 11.5 lbm/gal × 780.8 ft = 4666.9psi

666 Well Cementing

Spacer: C-3.7 Pressure to lift the casing
( ph)sp = 0.052 × 12.5 lbm/gal × 619.2 ft = 4022.5 psi. During some cementing treatments, there is a danger
that the casing may be pumped or lifted out of the well.
Lead cement: Conditions that favor such an occurrence are
1. lightweight pipe
( ph )lead = 0.052 × 13.0 lbm/gal × (4,450 – 1,400)fft
2. short pipe length
= 2,061.8 psi. 3. large-diameter pipe
Tail cement: 4. high-density cement slurries
5. low-density displacement fluids
( ph )tail = 0.052 × 16.4 lbm/gal × (4,950 – 4,450)fft 6. high annular friction pressures
= 426.4 psi. 7. bridging in the annulus
Total: 8. backpressure.
Conditions 2, 3, and 5 are frequently met when
( ph )o = ( ph ) mud + ( ph ) sp + ( ph )lead + ( ph )tail cementing surface or conductor casings. The lifting pres-
= 466.9 + 402.5 + 2, 061.8 + 426.4 sure is calculated as shown below, using the wellbore
diagram shown in Fig. C-2.
= 3,357.6 psi.
Pressure to land plug, pLP (excluding friction pressure):
pLP = ( ph )o − ( ph ) i = 3,357.6 – 2,970.8 = 386.8 psi.
pp

Hydrostatic pressure on formations

pp
Fracture pressure: 3,215 psi at 4,320 ft
ph at 4,320 ft = (ph ) mud + (ph ) sp + (ph )lead
(at 4,320 ft)
(ph )lead (at 4,320 ft) = 0.052 × 13.0 lbm/gal
× (4,320 – 1,400) ft = 1,973.9 psi

ph = 466.9 + 402.5 + 1,973.9 = 2,843.3 psi.

mcsg mfluid
Therefore, the hydrostatic pressure is 371.7 psi below
the fracture pressure.
Pore pressure: 3,150 psi at 4,800 ft
ph at 4,800 ft = (ph ) mud + (ph ) sp + (ph )lead + (ph )tail
(att 4,800 ft) ph

( ph )tail (at 4,800 ft) = 0.052 × 16.4 lbm/gaal AID

× (4,800 – 4,450) ft = 298.5 psi
AOD
ph = 466.9 + 402.5 + 2,061.8 + 298.5 = 3,2299.7 psi
A
Therefore, the hydrostatic pressure is 79.7 psi above
the pore pressure. Fig. C-2. Wellbore diagram for calculation of pressure to lift casing.

Appendix C Cementing Calculations 667

 Under static conditions, if a check valve (float) is Under static conditions, using Eq. C-4b,
used at the end of the casing, the differential force,
⎪⎧ ⎡ ⎡ 2 ⎤ ⎤⎪ ⎫
ΔF = ⎨( 0.052 × 14.8 lbm/gal × 800 ft ) ⎢ π ⎣(13.375 in.) – (12.515 in.) ⎦ ⎥ ⎬
2
ΔF, is calculated with Eq. C-4a.
⎩⎪ ⎣ 4 ⎦ ⎭⎪
(
Δ F = ⎡⎣( ph )ann × AOD ⎤⎦ − mcsg + m fluid , ) (C-4a) ⎧( 0.052 × 14.8 lbm/gal × 800 ft ) – ( 0.052 × 8.33 lbm/gal × 800 ft ) ⎫
⎪
+⎨ ⎡ (
⎪
π 12.515 in.) ⎤⎥ ⎬
2
where ⎪ × ⎢ ⎪
⎩ ⎣ 4 ⎦ ⎭
AOD = cross-sectional area (in.2)of casing outside – ( 800 ft × 61 lbm/ft )
diameter = 10,766 + 33,110 – 48, 800 = – 4, 923 lbf
mcsg = casing weight (lbm)
mfluid = weight of fluid inside casing (lbm) The negative force indicates that there is not enough
(ph)ann = hydrostatic pressure of annular fluid(s) buoyancy to float the casing under static conditions. The
(lbf/in.2). pump pressure to bump the plug is
If the casing is open-ended (no floats), the calcula- pp = (14.8 lbm/gal – 8.33 lbm/gal) × 0.052 × 800 ft
tion is different (Eq. C-4b), but the result is the same. = 269.2 psi.
Δ F = ⎡ ph
⎣ ( )ann ( AOD − AID )⎤⎦ + This pump pressure balances the hydrostatic-pres-

{ }
sure difference between the fluid in the annulus and the
⎡( p
) − ( ph )csg ⎤⎦⎥( AID )
⎣⎢ h ann
− mcsg , (C-4b) fluid in the casing. Equation C-5b is used to calculate the
maximum allowable pump pressure to avoid lifting the
casing from the well.
where
AID = cross-sectional area (in.2) of casing inside diameter ⎡ ⎤
(ph)csg = hydrostatic pressure of fluid(s) in casing ⎢ 800 ft × 61 lbm/ft ⎥
ppmax =⎢ ⎥
(lbf/in.2).
⎢ (13.375 in.)2 × π ⎥
When pumping, the pump pressure (pp) acting on the ⎢⎣ 4 ⎥⎦
inner-diameter cross-sectional area (AID) must be added
⎡ ⎛ (12.5115 in.)2 ⎞ ⎤
to the above equation. − ⎢(800 ft × 0.052 × 8.33)⎜ 1 – ⎟⎥
⎢⎣ ⎝ (13.375 in.)2 ⎠ ⎥⎦
( ) (
Δ F = pBH AOD − AID + p p × AID − mcsg , ) (C-5a)
= 347.33 – 43.13 = 304.2 lbf/in.2
where
Therefore, for this casing, the pump pressure must be
pBH = effective pressure at the bottom of the hole: maintained below 304.2 lbf/in.2, or other precautions
pBH = (ph)csg + pp. should be taken to avoid having the casing come out of
If Δ F is positive, the casing may come out of the well. the well.
Working this problem backward, the value of pp that This example uses single fluids in the casing and in
gives a Δ F value of zero is the critical pump pressure the annulus. In practice, there may be several fluids in
above which the casing may be pumped from the well. the annulus at the end of the job (e.g., tail, lead cement,
The service crew should ensure that the pump pressure spacer). In such cases, the calculation is worked with
during the treatment never exceeds this value unless the the appropriate contributions from the different fluids.
casing is restrained.

mcsg ⎡ ⎛ A ⎞⎤ C-4 Plug balancing

p p max = − ⎢( ph )csg ⎜ 1 − ID ⎟ ⎥ , (C-5b) Cement plugs, especially sidetrack or “kickoff” plugs, are
AOD ⎢⎣ ⎝ AOD ⎠ ⎦⎥
normally balanced in the borehole (Chapter 14). This
means that the hydrostatic pressures in the annulus and
where ppmax = maximum allowable pump pressure.
in the workstring are equal at the time of placement.
This is done to prevent U-tubing after cement placement
Example calculation and helps to prevent contamination so that a strong plug
Consider a 133⁄8-in., 61-lbm/ft casing set at 800 ft with will be set. In practice, the cement is frequently slightly
14.8-lbm/gal cement and 8.33-lbm/gal water for dis- underdisplaced from the balance point. This allows
placement. Is there danger of the casing coming out of cement to fall while the pipe is being pulled, filling up
the well? the space that was occupied by the pipe. It also allows
the pipe to be pulled without bringing fluids out onto the
rig floor.
668 Well Cementing
 When balancing a plug, the hydrostatic pressures in
the pipe and in the annulus must be equal. To achieve
this, the fluids used to displace the cement must be the
same fluids that are ahead of the cement, but in the
reverse order. The heights of each of the fluids in the
pipe and the annulus must be equal (Fig. C-3).

Mud

Spacer

7,000

Cement L

7,500
(a) (b)
(a) Desired plug (b) At end of displacement
Fig. C-4. Reversing of excess cement during plug cementing.
Fig. C-3. Wellbore diagram for plug cementing calculations.

Volume of spacer behind the cement, Vsp2

To ensure that the top of the plug is placed at the
desired location, sufficient cement is normally run so Vsp1
that the top is well above its desired location. This Vsp2 = × Stub , (C-8)
excess may be left to set and is then removed with the bit Sann
before it has developed its full strength. In some cases
where
the excess cement may be reversed out so that the top of
the cement plug is at the appropriate location (Fig. C-4). Vsp1 = volume of spacer ahead of the cement.
Displacement volume, Vdisp
C-4.1 Equations
Volume of cement, Vcem ⎣ (
Vdisp = Stub × ⎡ D − Lcem + Lsp2 ⎤ ,
⎦ ) (C-9)

Vcem = L × Soh , (C-6) where

where D = depth of workstring (bottom of cement plug)(ft)
Lsp2 = length of spacer behind (ft) = Vsp2 /Stub.
L = length of column of cement in open hole (ft)
Soh = capacity of open hole from standard tables (ft3/ft).
C-4.2 Example calculations
Length of balanced plug, Lcem (with workstring The well in question is shown in Fig. C-5.
in place)
Well data
Vcem
Lcem = , (C-7) Openhole size: 81⁄2 in.
Sann + Stub Hole capacity (Soh) = 0.3941 ft3/ft
Drillpipe: 4 in., 14.0 lbm/ft
where Pipe capacity (Stub) = 0.01084 bbl/ft or 0.06084 ft3/ft
Sann = capacity of annulus between tubing or drillpipe Annular capacity (Sann) = 0.0546 bbl/ft or 0.3068 ft3/ft
and open hole (ft3/ft) Spacer ahead of cement (Vsp1): 10 bbl
Stub = capacity of tubing or drillpipe (ft3/ft).

Appendix C Cementing Calculations 669

Volume of cement ■ pressure to inject into the void (to avoid fracturing
[maximum pressure limit] or to fracture if desired)
Vcem = L × Soh = 500 ft × 0.3941 ft3/ft = 197.1ft3 ■ bottomhole pressure (at various stages of the treat-
ment)—the sum of the hydrostatic pressure and
Length of balanced plug pump pressure, minus the friction pressure†
■ squeeze pressure—an established increment over
Vcem 197.1 ft3
Lcem = = = 536.1 ft injection pressure
Sann + Stub 0.3068 + 0.06084 ft3/ft ■ maximum surface pressure safely applied to the
annulus
Volume of spacer behind the cement ■ forces on the casing at the packer
Vsp1 10 bbl ■ maximum allowable squeeze pressure
Vsp2 = × Stub = × 0.01084 bbl/ft ■
Sann 0.0546 bbl/ft maximum allowable pressure to reverse circulate
fluids from the wellbore.
= 2.0 bbl
When calculating pressures, TVDs are used.
Displacement volume Ideally, fracturing is not required and the final
squeeze pressure is the only required calculation. The
⎣ (
Vdisp = Stub × ⎡ D − Lcem + Lsp2 ⎤
⎦ ) final squeeze pressure calculation is based on the injec-
tion pressure. Normally, it is 500 or 1,000 psi (3,448 or
⎡7,500 ft – (536.1 ft + ⎤ 6,895 kPa) above the injection pressure (Chapter 14). In
= 0.01084 bbl/ftt × ⎢ ⎥ some cases, because of well operations required before
⎣2.0 bbl/0.01084 bbl/ft)⎦ the cement setting, it is necessary to apply a higher
= 0.01084 bbl/ft × 6,779 ft squeeze pressure. Reversing out excess cement is a typ-
= 73.5 bbl ical example. In this case, because of the cement column
in the tubing, the hydrostatic pressure is greater than
Note that the use of the wrong hole size or inaccurate the desired final squeeze pressure. Before beginning
excess volumes may result in an improperly balanced reverse circulation, the squeezed perforations must be
plug, and the top of the plug may be at the wrong depth. tested to the corresponding pressure.

C-5.1 Example calculations

C-5 Squeeze cementing
Most squeeze-cementing calculations involve simple
Squeeze cementing involves two separate sets of calcu- volumetrics or hydrostatics; however, several of the
lations: calculations require further explanation and are demon-
■ volumes during the treatment strated with an example. The example wellbore is illus-
■ pressures at various points in the wellbore during trated in Fig. C-5.
various stages of the treatment.
The volumes calculated for a squeeze treatment are C-5.1.1 Forces on the casing at the packer
■ cement-slurry volume (from the estimated void to be There are pressures on the casing at the packer from two
filled and/or experience) directions. The greater pressure is on the outside of the
■ volume to the end of workstring casing, pext, tending to collapse it (Fig. C-5). This pres-
■
sure is a combination of the hydrostatic pressure plus
casing volume below the workstring
the pumping pressure. If the annulus is not open at the
■ volume to spot water or cement to the perforations surface, or if there is a bridge, the external pressure is
■ volume to spot cement to the tool. calculated by
These are simple pipe-volume calculations that take
into account the various fluids in the pipe. Measured ( ) (
pext = p p + ⎡⎣ 0.052 × D1 + D2 × ρ1 ⎤⎦ − 0.052 × D2 × ρ2 , )
depths are used when calculating volumes. (C-10)
Pressure calculations ensure the safety of the well
and determine the anticipated pressure to squeeze the
well. These include † Because
of low pump rates, the friction pressure is usually negligible during
■ pressure to kill the well a squeeze treatment.

670 Well Cementing

 51⁄ 2 in., 17.0 lbm/ft casing If it is not certain whether there is communication to
Collapse pressure rating: 4,910 psi surface, or if a bridge may be near the surface, the most
Burst pressure rating: 5,320 psi severe case should be assumed and Eq. C-10 should be
Squeeze pressure: 3,500 psi used.
To offset this pressure, it is sometimes necessary to
pp pann
apply pump pressure, pann, to the annulus between the
casing and the workstring. This pressure must be suffi-
cient to prevent the collapse of the casing, which may be
in poor condition if it has been in the well a long time.
Mud = 11.0 lbm/gal pi is the sum of the pressure applied to the annulus,
pann, and the hydrostatic pressure, ph, of the fluid in the
workstring/casing annulus. This pressure must be
D1 greater than or equal to the difference between the
collapse pressure of the casing (adjusted for its condi-
tion) and the external pressure, pext. The applied pres-
sure must not exceed the burst pressure of the casing
pi pext (adjusted for its condition) at any point, the highest
4,200 differential pressure being at the surface.
Cement In the example (Fig. C-5), the required final pump
D2 pressure is expected to be 3,900 psi. The workover fluid
Channel
is 8.5-lbm/gal brine (annular fluid, preflush, and dis-
4,500 psq Perforations placement fluid). The packer is at 4,200 ft. Using
Eq. C-10,

D1 = depth from surface to the packer ( ) (

pext = p p + ⎡⎣ 0.052 × D1 + D2 × ρ1 ⎤⎦ − 0.052 × D2 × ρ2 )
D2 = distance from the perforations to the packer
pann = pressure applied to the annulus = 3,900 + (0.052 × 4,500 × 8.5)
pext = external pressure to which the casing is exposed – (0.052 × 300 × 8.5)
at the packer
pi = internal pressure on the casing at the packer = 5,756 psi.
pp = pump pressure during the squeeze
psq = squeeze pressure experienced by the formation If cement remains in the pipe at the end of the
squeeze, the hydrostatic pressure contribution of the
Fig. C-5. Wellbore diagram for squeezing cement. cement and the brine must be taken into account.
With a collapse pressure rating of 4,910 psi, there
must be at least 846 psi on the inside of the casing at the
where packer to prevent collapse. The hydrostatic pressure of
D1 = depth to the packer the 8.5-lbm/gal brine is 1,856 psi, so the requirement is
D2 = distance from packer to squeeze perforations. exceeded by 1,010 psi and should be safe if the casing is
pp = pump pressure (maximum) new and undamaged. Depending on the condition of the
ρ1 = density of fluid inside the pipe casing, the operator may desire a larger safety margin
ρ2 = density of lightest fluid pumped during the and therefore require that pressure be maintained on
treatment the annulus. In addition, to make the appropriate calcu-
If there are several fluids in the tubing and casing, lation, one must determine the extent to which the
the term (0.052 × [D1 + D2] × ρ1) becomes the sum of collapse pressure rating of the casing has deteriorated.
the hydrostatic pressures of the individual fluids.
If the casing/hole annulus is open to the surface, the
pressure is
pext = (0.052 × D1 × ρ1 ). (C-10a)

Appendix C Cementing Calculations 671

C-5.1.2 Maximum surface pressure safely applied The total hydrostatic pressure in the workstring must
to the annulus be considered. This includes all cement in the work-
As discussed above, in some squeeze treatments, it is string plus any between the workstring and the perfora-
necessary to apply pressure to the annulus between the tions. If this pressure is greater than the final squeeze
workstring and the casing to avoid collapsing the casing pressure, consideration should be given to a higher
at the packer. The burst rating of the casing must be eval- squeeze pressure. This is not necessary for the squeeze
uated—and downgraded for old or used casing—to itself but is necessary to ensure that the squeeze will not
ensure that the casing will not be damaged. Normally, the be damaged by the reversing process. In the case of
worst conditions will be at the surface, with essentially small-diameter workstrings, the friction pressure may be
no external pressure and the internal pressure being the significant. Calculation of friction pressure is best per-
pressure applied to the tubing/casing annulus. However, formed by placement simulators but can be done manu-
the effect of the condition of the casing may be worse ally with a calculator (Chapter 4). In any event, to pro-
deeper in the well if the casing is corroded or worn. tect the casing, a limit should be set on the pump
pressure applied to reverse fluids.

C-5.1.3 Maximum allowable squeeze pressure

Before the squeeze treatment, the maximum allowable C-6 Calculations for foamed cement jobs
squeeze pressure must be determined, considering the Most cementing companies use computer programs to
placement technique selected for the treatment and the design foamed cement jobs; however, it is useful to know
integrity of all the pipe. If the placement technique how to perform the calculations manually. As discussed
involves fracturing the formation before cement place- in Chapter 7, foamed cement jobs can be divided into
ment, the integrity of the pipe is the main concern. If the two types (depending on the method of scheduling the
treatment is performed without fracturing, the formation gas phase): constant nitrogen (or air) ratio and con-
fracturing pressure must not be exceeded. Of course, the stant density. During a constant-nitrogen-ratio job, the
pressure to which the pipe and the formation are nitrogen is added to the base cement slurry at a constant
exposed is the total of the hydrostatic pressure and the rate (standard cubic feet [scf]/bbl of slurry). Because of
pump pressure. Once the maximum allowable bottom- compression, this method results in cement with vari-
hole pressure has been determined, the hydrostatic pres- able density—lightest at the top, heaviest at the bottom.
sure of the displacing fluids (and any cement column in A constant-density job is one in which several stages of
the pipe) must be subtracted to calculate the maximum foamed cement, each with a different volume ratio of
allowable surface pressure during the squeeze. nitrogen to slurry, are used. The nitrogen-to-slurry ratios
are calculated so that each stage will have the same
C-5.1.4 Maximum allowable pressure when average density at its final position in the annulus; how-
reversing out ever, the density varies within each stage because of dif-
ferences in hydrostatic pressure within the stage. In this
The final set of pressure calculations involves deter-
section, the design calculations for the constant-density
mining the pressures to which the well will be exposed
method are presented. The same calculations may be
while reverse circulating fluids out of the well. In some
used to design a constant-rate job. In that case, the
cases, the only fluid reversed out will be the workover
entire interval is considered as one stage.
fluid. In others, cement may be reversed out.
The following items are calculated to design a foamed
The critical areas to be considered during this
cement job.
process are casing at the surface (burst), the tubing
■ Minimum fracture pressure (less a safety factor) and
(collapse), and the squeezed interval itself. If reversing
cement, it is best to assume that no cement was injected location
into the formation and that all the cement is in the ■ Hydrostatic pressure of the fluids above the weak
tubing. The calculations are a simple matter of deter- zone(s) (if the tail slurry is above the weak zone, it
mining the fill of each fluid in the pipe and the hydro- must be included)
static pressures of these fluids in the pipe and in the ■ Allowable average density of the foamed cement (to
annulus. The difference in hydrostatic pressure between meet the fracture-pressure limitation)
the pipe and the annulus is the required pump pressure ■ Number of stages
(not taking into account friction pressure) and is the
■ Hydrostatic pressure at the midpoint of each stage
pressure to which the casing will be exposed. This pres-
sure must be compared to the burst rating of the casing ■ Nitrogen requirement for each stage based on the
(appropriate to its condition). midpoint hydrostatic pressure

672 Well Cementing

■ Foam quality (or gas ratio) for each stage Minimum fracture pressure
■ Yield of the foamed cement for each stage
■ Volume of each stage (from length of stage and pfmin = ( 0.512 psi/ft × 7,700 ft ) − 500 psi (safetty factor)
caliper log) = 3,442 psi
■ Hydrostatic pressure as each successive stage of
foamed cement enters the annulus (this is necessary Hydrostatic pressure of the fluids (mud and
because the gas ratio of the first stages of foamed spacer) ahead of the foamed cement
cement is lower at greater depths owing to gas com-
pression; thus, their density as they pass the shoe will ph = 0.052 × [(1,755 ft × 9.2 lbm/gal)
be higher) + (745 ft × 8.6 lbm/gal)]
■ Pump schedule including = 1,173 psi
– base slurry volume
– nitrogen ratio Allowable average density of the foamed cement
– nitrogen volume The allowable average density of the foamed cement in
– nitrogen pump rate the annulus, ρ1, is calculated as below.
– foamer pump rate
pf min − ph
Once these calculations are made, a series of tables ρ1 =
and graphs should be constructed to show the parame- 0.052 × h
ters for each stage. The following example illustrates 3,442 psi − 1,173 psi
this procedure. =
0.0052 × (7,700 ft − 2,500 ft)
= 8.4 lbm/gal (C-11)

Surface

1,755 ft
9.2 lbm/gal
drilling mud
Design data 745 ft 1,755 Top of
Total depth: 9,000 ft 8.6 lbm/gal cement
Lost circulation: 5,000 to 7,700 ft spacer 2,500 ft
2,500
Fracture gradient: 0.512 psi/ft at 7,700 ft
(equivalent to 9.85 lbm/gal) 1,100 ft 1,413 psi
Full circulation: 9.2 lbm/gal mud at 3,050 ft
Casing: 51⁄2 in., 17.0 lbm/ft 3,600
Hole caliper: 91⁄4 in.
1,100 ft 1,894 psi
Static temperature: 180°F at 9,000 ft
at 4,150 ft
165°F at 7,700 ft 4,700
Circulating temperature: 160°F at 7,450 ft
150°F at 6,350 ft 1,100 ft 2,374 psi 5,500
140°F at 5,250 ft at 5,250 ft
130°F at 4,150 ft 5,800
120°F at 3,050 ft 2,855 psi
Spacer: 745 ft at 8.6 lbm/gal 1,100 ft Weak
at 6,350 ft zone at
Top of tail cement: 8,000 ft 6,900
Top of cement: 2,500 ft 7,700 ft
Base cement density: 14.2 lbm/gal 1,100 ft 3,335 psi
Base cement yield: 1.29 ft3/sk at 7,450 ft
8,000
Tail Top of
cement tail cement

Fig. C-6. Wellbore diagram for staged foamed cementing.

Appendix C Cementing Calculations 673

Number of stages where
Divide into intervals of 1,000 to 1,500 ft ρbs = base slurry density
Total interval is 8,000 – 2,500 ft = 5,500 ft ρfc = foamed cement density
Divide into 5 stages of 1,100 ft each ρN2 = nitrogen density.
Foamed cement yield, Yfc (ft3/sk):
Hydrostatic pressure at the midpoint of each Ybs
stage = Y fc = . (C-14)
1 − Q foam
ph = ph above stage + ph to midpoint of stage
where
Stage No. 1 (top) Ybs = base slurry yield (ft3/sk).
Annular volume, Vann:
ph1 = 1,173 psi
Vann = L × Sann , (C-15)
+ [(Stage number – 0.5) × 1,100 ft × ρ1 × 0.052]
= 1,173 psi + (0.5 × 1,100 ft × 8.4 lbm/gal × 0.052) where
= 1,413 psi L = length
Sann = annular capacity.
Stage No. 2 Cement requirement, C (sk):

ph2 = 1,173 psi + (1.5 × 1,100 ft × 8.4 lbm/gal × 0.052) Vann

C= . (C-16)
= 1,894 psi Y fc

Nitrogen requirement at conditions in the annulus,

Stage No. 3 RN2:
ph3 = 1,173 psi + (2.5 × 1,100 ft × 8.4 lbm/gal × 0.052)
RN 2 = Vann × Q foam . (C-17)
= 2,374 psi
The nitrogen requirement refers to the volume
Stage No. 4 required at circulating temperature and pressure. For
job-design purposes, this value must be converted to the
ph4 = 1,173 psi + (3.5 × 1,100 ft × 8.4 lbm/gal × 0.052) equivalent volume of nitrogen in standard cubic feet (at
standard temperature and pressure [STP]).
= 2,855 psi Nitrogen volume,VN2, at STP (scf):

Stage No. 5 (bottom) VN 2 = RN 2 × KN 2 . (C-18)

ph5 = 1,173 psi + (4.5 × 1,100 ft × 8.4 lbm/gal × 0.052)

Stage No. 1, ph1 = 1,413 psi
= 3,335 psi
( )
ρ N 2 = 1.724 × 10 −3 × K N 2
Nitrogen requirement for each stage based on = 1.724 × 10 −3 × 476 scf/bbl
the midpoint
= 0.8206 lbm/gal
Nitrogen density, ρN2:
The nitrogen volume factor can be calculated based
ρ N 2 = (1.724 × 10 –3 ) × K N 2 . (C-12) upon pressure and bottomhole circulating temperature
where or more easily looked up in standard tables published by
KN2 = nitrogen volume factor (scf/bbl). most cementing companies.
Foamed cement quality, Qfoam: Foamed cement quality, Qfoam:
ρ fc − ρ N 2 8.4 − 0.8206
Q foam = 1 − , (C-13) Q foam = 1 −
ρbs − ρ N 2 14.2 − 0.8206
= 0.4335.

674 Well Cementing

Foamed cement yield, Yfc: Hydrostatic pressure as each stage of foamed
cement enters the annulus
1.29 ft3/sk
Y fc = Actually, hydrostatic pressure should be calculated for
1 − 0.4335 the position of each stage immediately above the weak
= 2.28 ft3/sk. formation(s). Because a job designed for constant den-
sity uses a slurry with the lowest concentration of nitro-
Annular volume, Vann: gen in the first stages, these initial stages will be signifi-
cantly more dense when they pass the weaker
Vann = 1,100 ft × 0.3017 ft3/ft
formations below. Therefore, to ensure the integrity of
= 331.9 ft3. the well, the hydrostatic pressure exerted on the weak
formations must be calculated and compared to the
Cement requirement, C (sk): fracturing pressure of these formations. To do this, the
following steps must be followed.
331.9 ft3
C= 1. Determine the volume occupied by each stage at the
2.28 ft3/sk weak zone.
= 145.6 sk. 2. Calculate its length based on the annular capacity.
Nitrogen requirement, RN2: 3. Calculate the hydrostatic pressure of the fluids in the
annulus above.
RN = 331.9 ft3 × 0.4335 4. Calculate the hydrostatic pressure of the foamed
2
cement stage(s).
= 143.9 ft3 . 5. Add the results of Steps 3 and 4 to obtain the total
Nitrogen volume, VN2, at STP: hydrostatic pressure.
6. Compare this value to the fracturing pressure of the
VN = 143.9 ft3 × 0.178 bbl/ft3 × 476 scf/bbl weak formation.
2
7. If the results indicate a risk of formation fracturing,
= 12,192 scf.
consider a constant-nitrogen-rate or a hybrid job. A
Similarly, the requirements for the other stages are hybrid job involves several stages of foamed cement
calculated, and the following table can be built. with different designed densities.
From the preceding example, this calculation is per-
formed as shown below.
Stage 1 2 3 4 5 1. Determine the volume occupied by each stage at the
Hydrostatic pressure 1,413 1,894 2,374 2,855 3,335 weak zone.
(psi) ■ For Stage No. 1, 145.6 sk or (145.6 × 1.29 ft3/sk
Nitrogen density 0.821 1.069 1.291 1.493 1.677 × 0.178 bbl/ft3) = 33.4 bbl of cement slurry are
(lbm/gal) required.
Foamed cement 0.4335 0.4417 0.4493 0.4564 0.4631 ■
quality
The nitrogen requirement is 12,192 scf.
■ The weak zone is at 7,700 ft.
Foamed cement yield 2.28 2.31 2.34 2.37 2.40
(ft3/sk) ■ The fluids ahead of the cement are 9.2-lbm/gal mud
Annular volume (ft3) 331.9 331.9 331.9 331.9 331.9 (0.4784 psi/ft), and 745 ft of spacer at 8.6 lbm/gal
Cement requirement 145.6 143.7 141.8 140.0 138.3 (0.4472 psi/ft).
(sk)
■ Assuming the foamed cement occupies 850 ft
Nitrogen per sack 83.7 112.6 140.2 166.8 192.5
(scf/sk)
(because of compression), the length of the mud
column is 7,700 – 850 (cement) – 745 (spacer) =
Nitrogen requirement 12,192 16,179 19,878 23,353 26,620 6,105 ft.
(scf)
■ The hydrostatic pressure from the mud is 6,105 ft
Total cement requirement: 943.3 sk (709.4 sk for foamed stages +
233.9 sk for tail) × 0.4784 psi/ft = 2,920 psi, and the hydrostatic
pressure from the spacer is 745 ft × 0.4472 psi/ft
Total nitrogen requirement: 98,222 scf = 333 psi; thus, the total hydrostatic pressure is
2,920 + 333 = 3,253 psi.

Appendix C Cementing Calculations 675

 ■ Reading from the nitrogen tables, the volume occu- 6. Compare this value to the fracturing pressure of the
pied by 12,192 scf of nitrogen is 12.7 bbl at 160°F. weak formation.
Thus, the volume of the foamed cement slurry is The fracturing pressure of the weak formation is
12.7 + 33.4 = 46.1 bbl. 3,942 psi.
2. Calculate its length based on the annular capacity. Note that the pressure calculated in the previous
With an annular capacity of 0.3017 ft3/ft and 46.1 bbl step does not exceed the fracturing pressure of the
of slurry, the fill-up is weak formation but does exceed the safety margin set
in the first step. This is what would be expected based
46.1 bbl on the method for calculating the foamed cement
= 858 ft.
0.3017 ft /ft × 0.178 bbl/ft3
3 density, based on the fracturing pressure less the
safety margin. This calculation should be repeated as
If this result had not been close to the assumed each stage passes the weak formation.
foamed cement length of 850 ft (Step No. 1), the cal-
culation would be repeated using an adjusted length.
Job execution tables
3. Calculate the hydrostatic pressure of the fluids in It is helpful for control of the job to construct tables of
the annulus above. the pumping schedule containing the following informa-
The height of the mud column is 7,700 – 745 – 858 tion.
= 6,097 ft. Therefore, the hydrostatic pressure is ■ Base slurry volume
6,097 ft × 0.4784 psi/ft = 2,916.8 psi. For the spacer,
■ Nitrogen ratio
the hydrostatic pressure is 745 ft × 0.4472 psi/ft =
333.1 psi. ■ Nitrogen volume

■ Nitrogen pump rate

4. Calculate the hydrostatic pressure of the foamed
cement stage(s). ■ Foamer pump rate

ρ N = 1.724 × 10 –3 × nitrogen volume factor (scf/bbl) Job Schedule per Stage

2
Stage 1 2 3 4 5
= 1.724 × 10 –3 × 950 scf/bbl
Base slurry volume (bbl) 33.4 33.0 32.6 32.1 31.8
= 1.638 lbm/gal.
Nitrogen ratio 365 490 610 728 837
The foam quality can be calculated by (scf/bbl base slurry)

VN Nitrogen volume (scf) 12,192 16,179 19,878 23,353 26,620

12.7 bbl
Q foam = 2
= = 0.2755
Vbs + VN 12.7 bbl + 33.4 bbll
2

The previous equation for foamed cement quality

can be rearranged to calculate foamed cement den- Nitrogen and Foamer Rate
sity, Base Slurry Rate 3 4 5 6 7
(bbl/min)
ρ fc = (1 − Q foam )(ρbs − ρ N ) + ρ N
2 2
Nitrogen rate (scf/min)
= (1 – 0.2755)(14.22 – 1.638) + 1.638 = 10.74 lbm/gal. Stage No. 1 1,095 1,460 1,825 2,190 2,555

Therefore, the hydrostatic pressure from foamed Stage No. 2 1,470 1,960 2,450 2,940 3,430
cement is 10.74 lbm/gal × 0.052 psi/ft/lbm/gal × 858 ft Stage No. 3 1,830 2,440 3,050 3,660 4,270
= 479.2 psi.
Stage No. 4 2,184 2,912 3,640 4,368 5,096
5. Add the three hydrostatic pressures to obtain the
total hydrostatic pressure. Stage No. 5 2,511 3,348 4,185 5,022 5,859

2,916.8 psi + 333.1 psi + 479.2 psi = 3,729.1 psi. Foamer rate (gal/min)
All stages 1.31 1.74 2.18 2.61 3.05

676 Well Cementing

Conversion Factors C-7 Acronym list
Customary Unit Multiply by To Obtain SI API American Petroleum Institute
Units
BWOB By weight of blend
in. 2.54 (25.4) cm (mm) BWOC By weight of cement
ft 0.305 m BWOW By weight of water
ft3 0.0283 m3 ISO International Organization for Standardization
SI Système International
bbl 0.159 m3
STP Standard temperature and pressure
U.S. gal 3.785 (3.785 × 10–3) L (m3)
TVD True vertical depth
ft3/ft (capacity) 0.0929 m3/m

bbl/ft (capacity) 0.522 m3/m

C-8 Suggested reading
gal/sk (94 lbm sack) 88.78 L/t
American Petroleum Institute: Specifications for
ft3/sk (94 lbm sack) 0.301 m3/t Cements and Materials for Well Cementing (API
Specification 10A), 23rd edition, American National
lbm 0.454 kg
Standards Institute (ANSI)/API 10A/ISO 10426-1:2001,
lbm/gal 120 kg/m3 Dallas, Texas, USA (April 2002).
lbm/ft (pipe weight) 1.49 kg/m Recommended Practice for Testing Well Cements, ANSI/
API Recommended Practice 10B-2 (formerly 10B), 1st
psi 6.895 (6.895 × 10–3) kPa (MPa) ed., July 2005 (identical to ISO 10426-2:2003, including
psi/ft 22.7 kPa/m ISO 10426-2:2005/FDAM 1:2005).
Schlumberger: i-Handbook, desktop application avail-
°F (°F – 32)/1.8 °C able from http://www.slb.com/content/services/resources/
software/ihandbook.asp (accessed Aug. 31, 2006).

Appendix C Cementing Calculations 677

Symbol
a
Definition
Activity (Chapter 6)
N
Nomenclature

aw Water activity

A Area (Chapters 1, 4, 6, 8, 15; Appendices B, C)

AAPI API surface area

AID Cross-sectional area of casing inside diameter

AOD Cross-sectional area of casing outside diameter

Apipe External pipe surface area

A Signal attenuation (Chapter 15)

b Offset in linear regression (Chapter 4)

B Correction factor (Chapters 1, 4, 15)

B Formation volume factor

Blam Laminar flow correction factor

Bres Formation volume factor

Bturb Turbulent-flow correction factor

c Compressibility (Chapters 6, 8, 9)

C Specific heat (Chapter 8; Appendix B)

C Cement requirement (Appendix C)

Well Cementing ■ Nomenclature 679

Symbol Definition
C Circumference (Chapter 15)

Ccsg Casing circumference

d Diameter (Chapters 4, 5, 6, 7, 8, 9, 11, 15; Appendices A, B)

d* Sample diameter after deformation

dbody Outer diameter of liner hanger body

dcsg Casing diameter

ddisk Disk diameter

dhole Hole diameter

di Inside diameter

do Outside diameter

dpart Particle diameter

dpipe Pipe diameter

dslip Slip OD

dw Pipe diameter at the wall

D Depth (Chapters 6, 9, 11; Appendices B, C)

Dboc Depth at bottom of cement

Dbog Depth at bottom of gas zone

Dpl Depth of permeable layer

Dtoc Depth of top of column

Dtog Depth at top of gas zone

DTV True vertical depth of cement column

E Young’s modulus (Chapters 7, 8, 15; Appendix B)

E Amplitude (Chapters 6, 15)

E1, E2, E3 CBL amplitude

(E1)1 E1 amplitude at Receiver 1

(E1)2 E1 amplitude at Receiver 2
(E1)fp E1 amplitude in free pipe

680 Well Cementing

Symbol Definition
E100%cem Amplitude of 100% cemented pipe

Efp CBL amplitude (E1 peak) for free pipe

Emeas Measured CBL amplitude (E1 peak) for free pipe

Ewave Wave amplitude

E Energy (Chapter 6)

Eact Activation energy

f Frequency (Chapter 15)

f0 Resonant frequency of pipe

f Factor (Chapters 4, 9, 11, 14; Appendix A)

fc Slurry compressibility factor

fform Formation factor

ffr Fanning friction factor

ffr1, ffr2 Fanning friction factor value at critical values of the Reynolds number

ffrict Mud friction factor

fhyd Hydrostatic factor

fspring Normalized spring factor

fγ⋅ Shear rate factor

fτ Shear stress factor

f Fraction (Chapters 6, 7, 9, 14)

fcV Cement volume fraction

ffc Solid volume fraction of the cement cake

fgas Volumetric fraction of gas in the slurry at surface conditions

fpV Packing volume fraction

fsV Solid volume fraction

(fsV)max Maximum solid volume fraction

fwV Water volume fraction

fwV 0 Initial water volume fraction

Nomenclature 681
Symbol Definition
F Force (Chapters 4, 8, 11; Appendices B, C)

F– Resultant force

Ffail Load at failure

Fmax Maximum unsupported load that the casing in which a liner hanger is to be set will withstand

g Acceleration of gravity (Chapters 4, 5, 6, 7, 9, 14; Appendix B)

gz z component of gravity acceleration

g Gradient (Chapter 11)

gfrac Fracture gradient

G Shear modulus (Chapter 8)

h Height or thickness (Chapters 1, 4, 5, 6, 8, 9, 14, 15; Appendices B, C)

–
h Head height of protuberance

hcem Cement slurry height in the annulus

hcomb Combined thickness of cement sheath and casing

hcsg Casing thickness

hdisk Disk thickness

hfc Filtercake thickness

(hfc)cem Cement filtercake thickness

(hfc)mud Mud filtercake thickness

hform Formation thickness

hmud Mud height in the annulus

hnode Height of node building up inside casing

hpl Permeable layer height

hres Reservoir height

hsp Spacer height in the annulus

i Wave intensity (Chapter 15)

682 Well Cementing

Symbol Definition
I Index (Chapters 1, 15)

I%bond Percentage bonding index

Ibond Bond index

IZI Index of zonal isolation

k Permeability (Chapters 1, 6, 8, 9, 14; Appendix B)

k* Equivalent permeability value

kcem Cement permeability

kfc Filtercake permeability

(kfc)cem Cement filtercake permeability

(kfc)mud Mud filtercake permeability

kform Formation permeability

kres Reservoir permeability

k Consistency index (Chapters 1, 4, 5, 7; Appendix A)

kann Annular consistency index

k ′ann Local annular consistency index

kbf Base fluid consistency index

kng Consistency index calculated according to a narrow-gap approximation

kNW Consistency index calculated according to a Newtonian shear rate approximation

kpipe Pipe consistency index

k ′pipe Local pipe consistency index

K Coefficient (Chapters 5, 6, 8, 9, 14, 15; Appendix C)

Kdep Deposition constant

Kdisp Dispersion coefficient

Khyd Coefficient related to degree of cement-phase hydration

Kfp Formation parameter

Kgfp Gas flow potential

Kmpr Maximum pressure restriction

Kmrp Mud removal parameter

KN2 Nitrogen volume factor

Nomenclature 683
Symbol Definition
Kos Osmotic coefficient

Kpdlp Pressure decay limit parameter

Kref Reflection coefficient

Ktrans Transmission coefficient

KT Temperature-related constant

Kx Coefficient related to active matter content x

K Standard spring calibration constant (Chapter 4)

L Length or distance (Chapters 1, 4, 5, 6, 8, 9, 15; Appendices A, B, C)

L* Sample length after deformation

L1 Average distance between pins when sleeve is empty (Appendix B)

L2 Average distance between pins when sleeve is expanded (Appendix B)

L1, L2 Distance between receivers (Chapter 15)

Lcem Cement length

Lecc Eccentering limit

Lf Final distance between pins (Appendix B)

Li Initial distance between pins (Appendix B)

Lmax Maximum distance

Lmin Minimum distance

Lpin Length of pin

Lsp2 Length of spacer behind cement

m Weight or mass (Appendix C)

madd Weight of additives

mcem Weight of cement

mcemrepl Weight of cement replaced

mcsg Casing weight

meqsk Weight in equivalent sacks

mfluid Weight of fluid in casing

mflyash Weight of fly ash

mwater Weight of water

684 Well Cementing

Symbol Definition
M A hardening parameter (Chapter 8)

n Power-law index (Chapters 4, 5, 7; Appendix A)

n′ Local power-law index

ncem Power-law index of cement

nmud Power-law index of mud

n– Stress vector in Fig. 8-1 (Chapter 8)

N Number (Chapters 4, 5, 7, 9, 11; Appendix A)

NBG Bingham number of the displacing fluid

Nconeseg Number of cone segments cut

Ncpads Number of cone pads

NHe Hedström number

(NHe)HB Generalized Hedström number for Herschel-Bulkley fluids

NRe Reynolds number

NRe1, NRe2 Critical values of Reynolds number

(NRe)AN Annular Reynolds number

(NRe)BG Bingham Reynolds number

(NRe)HB Herschel-Bulkley Reynolds number

(NRe)i Reynolds number of fluid i

(NRe)MR Metzner and Reed Reynolds number

(NRe)ecc Reynolds number for eccentric annuli

Nslips Number of slips

Nslipseg Number of slip segments cut

Nsp Slurry performance number

p Pressure (Chapters 1, 4, 5, 6, 8, 9, 11, 14, 15; Appendices B, C)

pann Annular pressure

pAPI API pressure

pback Backpressure

Nomenclature 685
Symbol Definition
pBH Bottomhole pressure

pburst Bursting pressure

pcem Cement slurry pressure

pcol Collapse pressure

pcon Confined pressure value

pe Effective filtration pressure

pext External pressure

pfbh Flowing bottomhole pressure

pfmin Minimum fracture pressure

pfrac Fracturing pressure

pgas Gas pressure

pgauge Gauge pressure

ph Hydrostatic pressure

(ph)ann Hydrostatic pressure of annular fluids

(ph)csg Hydrostatic pressure of fluids inside casing

(ph)i Hydrostatic pressure inside casing

(ph)lead Hydrostatic pressure for lead slurry

(ph)mud Mud hydrostatic pressure

(ph)o Hydrostatic pressure outside casing

(ph)sp Spacer hydrostatic pressure

(ph)tail Hydrostatic pressure for tail slurry

pi Acoustic pressure (Chapter 15)

pi Initial formation pressure (Chapter 1)

pi Iterative pressure change (Chapter 14)

pi Sum of the pressure applied to the annulus and hydrostatic pressure (Appendix C)

pinlet Inlet pressure

pLP Pressure to land plug

pmin Minimum casing collapse resistance

(Δp)min Minimum pressure gradient

pmod Modified pressure

pmud Hydrostatic pressure from mud column

pob Overbalance pressure

poutlet Outlet pressure

poverburden Overburden pressure

686 Well Cementing

Symbol Definition
pp Pump pressure

ppmax Maximum allowable pump pressure

ppore Pore pressure

(ppore)max Maximum pore pressure

pres Reservoir pressure

ps Safe pressure, the lesser of psf or psc

psc Pressure for casing protection

psf Pressure for formation protection

psp Spacer/wash pressure

psq Squeeze pressure to the formation

pwb Wellbore pressure

px Signal pressure at location x

px + L Signal pressure at location x + L

P Pressure gradient (Chapter 5)

q Flow rate (Chapters 1, 4, 5, 8, 9, 14; Appendices A, B, C)

qann Flow rate in the annulus

qAPI API fluid-loss rate

(qAPI)calc Calculated API fluid-loss rate

qcf Crossflow contributions to flow

qgas Gas flow rate

qleak Leakage rate

qres Reservoir lateral flow rate

Q Quality (Chapters 4, 7; Appendix C)

Qfoam Foamed cement quality

r Radius (Chapters 1, 4, 5, 8, 9, 13, 14; Appendices A, B)

r* Sample radius after deformation

r1, r2 Radii in coaxial cylinder viscometer

Nomenclature 687
Symbol Definition
rbreak Breakthrough radius for pipe/annular flow

rcsg Radius of the casing

rext External drainage radius

rhole Hole radius

ri Inner radius

ri + L Inner radius plus length of pin

ro Outer radius

rperf Perforation radius

rres Reservoir radius

rw Pipe or annulus inner radius

rwb Wellbore radius

ry Critical radius beyond which the fluid is unsheared in a coaxial cylinder geometry

R Ratio (Chapters 4, 5)

R2 Density ratio

R3 Effective viscosity ratio

R4 Dimensionless flow ratio

Rb Buoyancy ratio

Rn Power-law index ratio

Rr Radius ratio for a coaxial cylinder geometry

RSTO Standoff ratio

Ry Ratio of yield stress

R Requirement (Appendix C)

RN2 Nitrogen requirement

R Gas constant (Chapter 6)

R Rate (Appendix B)

Rpu Pressure-up rate

688 Well Cementing

Symbol Definition
s Skin factor (Chapters 1, 9, 14)

sres Reservoir skin factor

S Capacity (Appendix C)

Sann Annulus capacity

Soh Open hole capacity

Stub Tubing or drillpipe capacity

S Compressive strength (Chapters 7, 8, 9; Appendix B)

Sgel Static gel strength

Stens Tensile strength

S Sensitivity (Chapter 15)

Sj Receiver sensitivity

t Time (Chapters 4, 5, 6, 8, 9, 14, 15; Appendices A, B)

t* Dimensionless time

t*break Dimensionless breakthrough time

t0 Initial time

t100Bc Time to 100 Bearden units of consistency

t30Bc Time to 30 Bearden units of consistency

tAPI API testing time

tbreak Breakthrough time

tcrit Critical time

tdisp Time to displace leading edge of cement slurry to bottom

tf Upper time boundary

tgel Gelling time

ti Iterative time (difference)

ttr Transit time

t ′tr Stretched transit time

Nomenclature 689
Symbol Definition
T Temperature (Chapters 1, 4, 6, 8, 9, 15; Appendix B)

Tmax Maximum temperature

Tres Reservoir temperature

v Velocity (Chapters 4, 5, 6, 7, 11, 15; Appendix A)

v– Mean fluid velocity

v–lam Laminar mean velocity for a given shear stress at the wall

v–max Maximum mean velocity

v–min Minimum mean velocity

v–′ Mean square velocity fluctuation

v–1, v–2 Values of the mean velocity at critical value of the Reynolds number

v0 Slurry velocity below the permeable formation

v1, v2 Plate velocity

vcem Acoustic velocity through cement

vdf Dimensionless filtration rate

(vfilt)max Maximum admissible filtration velocity

vfilt Filtration velocity into the formation

vfluid Fluid velocity

vmax Maximum fluid velocity

vmud Acoustic velocity through mud

vP, vS Compressional and shear velocity

vpipe Maximum casing velocity to prevent damage to casing or formation; tripping speed of pipe

vpl Velocity of plate wave

vrun Speed at which the casing is run

vs Slurry settling velocity

vslip Slip velocity

vsteel Acoustic velocity through steel

V Volume (Chapters 5, 6, 7, 9, 14; Appendices A, B, C)

Vab Absolute shrinkage

Vadd Additive volume

Vann Annular volume

VAPI API test fluid volume

690 Well Cementing

Symbol Definition
Vbs Base slurry volume

Vbulk Bulk volume

Vcem Cement volume

Vcrit Critical annular volume

Vcum Cumulative volume pumped

Vdisp Displacement volume

Vfc Filtercake volume

Vfilt Filtrate volume

Vfl Fluid-loss volume reduction

Vfoam Total foam volume

Vgas Gas volume

Vhyd Hydration volume reduction

Vlead Volume of lead cement

VN2 Nitrogen volume or nitrogen volume at standard temperature and pressure

Vpipe Pipe volume

Vslurry Slurry volume

Vsolids Volume of solids

Vsp Volume of spacer

Vt Filtrate volume at time t

Vtotal Total volume

Vwater Water volume

V Wellbore volume times porosity (Chapter 14)

w Slot width or a local annular gap (Chapters 1, 4, 5)

wmin Minimum annular gap

w Deposition factor (Chapter 14)

x Normalized distance to the pipe axis or to the plane of symmetry of the slot (Chapter 4; Appendix A)

Nomenclature 691
Symbol Definition
x Active matter content (Chapter 6)

y Relative error (Chapter 4)

ycalculated Calculated relative error

ymeasured Measured relative error

Y Yield force (Chapter 5)

Y Yield (Appendix C)

Ybs Base slurry yield

Yfc Foamed cement yield

Yslurry Slurry yield

Y Value (Chapters 8, 9)

Ycc Cement concentration value

Yco Cohesion value

Yform Formation parameter

z Axis of symmetry of cylindrical geometry (Chapter 4)

Z Gas deviation factor (Chapter 1)

Zres Reservoir gas deviation factor

Z Acoustic impedance (Chapter 15)

Zcem Cement acoustic impedance

Zmth Theoretical mud impedance

Zmud Mud acoustic impedance

692 Well Cementing

Symbol Definition
α Coefficient of linear thermal expansion (Appendix B)

α Annulus diameter ratio (Chapter 4)

α Angle of deviation (Chapter 5)

α Poroelastic constant (Chapter 8)

α Hanger cone angle (Chapter 11)

α Attenuation rate (Chapter 15)

α100%cem CBL attenuation rate for a 100% cemented pipe

α BHC Bottomhole-compensated attenuation rate

αfp CBL attenuation rate for a free pipe

αsig Signal attenuation

β Coefficient of linear thermal expansion (Chapter 8)

χ Torque per unit length (Chapter 4)

Δ Sum of the squares of deviations (Chapter 4)

ε Strain (Preface; Chapter 8; Appendix B)

εa Axial strain

εc(t) Time-dependent deformation

εel Elastic strain

εpl Plastic strain

εr Radial strain

Nomenclature 693
Symbol Definition
εT Linear thermal strain

εx X-axis strain

εy Y-axis strain

εz Z-axis strain

εθ Angular strain

ε Annulus eccentricity (Chapter 4)

φ Porosity (Chapters 6, 7, 8, 14, 15)

φcs Cement slurry porosity

φfc Filtercake porosity

φ Phase spectrum (Chapter 15)

Φ Angle of internal friction (Chapter 8)

γ⋅ Shear rate (Chapters 4, 5; Appendix A)

γ⋅ 1, γ⋅ 2 Shear rate at inner and outer walls of coaxial cylinder geometry

γ⋅ ng Shear rate calculated using a narrow-gap approximation

(γ⋅ ng)i Iterative narrow-gap shear rate

γ⋅ NW Newtonian shear rate at the wall (pipe and annular flow)

γ⋅ w Shear rate at the wall (pipe and annular flow)

γ Strain or shear strain (Chapters 4, 8)

γe Elastic strain

γp Plastic strain

γrz The rz component of rate of strain tensor

γxy Shear strain, xy direction

γxz Shear strain, xz direction

γyz Shear strain, yz direction

694 Well Cementing

Symbol Definition
γ Specific gravity (Chapter 7; Appendix C)

γcem Specific gravity of cement

γflyash Specific gravity of fly ash

γL Secific gravity of liquid

η Efficiency (Chapters 4, 5; Appendix A)

ηcirc Circulation efficiency

ηdisp Displacement efficiency

λ Thermal conductivity (Chapter 8; Appendix B)

λ Normalized minimum distance from the axis of symmetry in a concentric annulus in which the shear stress
is nil (Chapter 4)

λ Wavelength (Chapter 15)

μ Viscosity (Chapters 1, 4, 5, 6, 7, 8, 9, 14; Appendices A, B)

μeff Effective viscosity

μfilt Filtrate viscosity

μgas Gas viscosity

μL Viscosity of liquid medium

μp Plastic viscosity

(μp)c Plastic viscosity of cement

(μp)mud Plastic viscosity of mud

(μp)ng Plastic viscosity calculated using a narrow-gap approximation

(μp)p Plastic viscosity measured in a pipe

μres Reservoir viscosity

μs 0 Viscosity of suspending liquid

μslurry Slurry viscosity

μwater Water viscosity

Nomenclature 695
Symbol Definition
μ Coefficient of friction (Chapter 11)

ν Poisson's ratio (Chapters 4, 8, 15; Appendix B)

θ Angle (Chapters 4, 5, 8, 9, 11, 15)

θhc Hanger cone angle

θ Dial readings (Chapter 4)

ρ Density (Chapters 4, 5, 6, 7, 8, 9, 11, 14, 15; Appendices A, B, C)

ρ0 Initial slurry density

ρa Allowable average density of the foamed cement

ρb Density of blend

ρbs Base slurry density

ρcem Cement density

vdf Density of drilling fluid

ρeq Equivalent density of normal formation pressure gradient

ρfc Foamed cement density

ρL Specific gravity of liquid

ρmud Mud density

ρmw Mix water density

ρN2 Nitrogen density

ρslurry Slurry density

ρsp Spacer density

ρw Water density

σ Stress (Preface; Chapters 8, 9, 14)

σ′ Total applied stress

σ′ Intergranular or effective stress related to gel strength development

σa Axial stress

σh Stress in a horizontal direction

696 Well Cementing

Symbol Definition
σn Stress state on a plane

σr Radial stress

σvert Vertical stress

σx Stress in x direction

σy Stress in y direction

σz Total stress exerted at a given linear depth, z

σθ Tangential stress

σ Strength (Chapter 8)

σc Compressive strength

τ Group delay spectrum (Chapter 15)

τ Shear (Chapter 8)

τxy Shear in the xy direction

τxz Shear in the xz direction

τyz Shear in the yz direction

τ Shear stress (Chapters 4, 5, 7, 8, 9, 10, 14, 15; Appendix A)

τ1 Shear stress at inner walls of coaxial cylinder geometry

τ2 Shear stress at outer walls of coaxial cylinder geometry

τcem Shear stress of cement slurry

τgel Minimum shear stress value to induce flow

τmud Shear stress of mud

τng Shear stress calculated using a narrow-gap approximation

(τng)icalc Calculated iterative narrow-gap yield stress

(τng)idata Measured iterative narrow-gap yield stress

τpdlp Shear stress at wellbore wall that causes pressure to reach critical value for gas entry

τrz rz component of shear stress tensor

τw Shear stress at the wall

Nomenclature 697
Symbol Definition
τy Yield stress

(τy)c Yield stress measured in viscometer

(τy)cem Cement slurry yield stress

(τy)est Estimated yield stress

(τy)mud Mud yield stress

(τy)ng Yield stress calculated using a narrow-gap approximation

(τy)NW Yield stress calculated according to a Newtonian shear rate approximation

(τy)p Yield stress measured in pipe

Ω Rotational speed (Chapter 4)

ω Angular frequency (Chapter 15)

ξ Dimensionless shear rate (Chapters 4, 5)

ξcrit Critical dimensionless shear rate

ψ Reciprocal dimensionless shear stress (Chapters 4, 5)

ψ1 Critical value of reciprocal dimensionless shear stress

Ψ Change of angle (Chapter 8)

698 Well Cementing

Abdel-Mota’al, A.A.: “Detection and Remedy of Behind-
Casing Communication During Well Completion,” paper
SPE 11498 presented at the Middle East Oil Technical
R
References

of Mechanical Engineers (ASME) Petroleum Division

Petroleum Mechanical Engineering Conference, Tulsa,
Oklahoma, USA (September 21–25, 1975).
Conference and Exhibition, Bahrain (March 14–17, 1983). Aldridge, L.P.: “Accuracy and Precision of Phase
Abrams, A.: “Mud Design to Minimize Rock Impairment Analysis in Portland Cement by Bogue, Microscopic and
due to Particle Invasion,” Journal of Petroleum X-ray Diffraction Methods,” Cement and Concrete
Technology (May 1977) 29, No. 5, 586–592; also presented Research (May 1982) 12, No. 3, 381–398.
as paper SPE 5713 at the SPE Symposium on Formation Ali, S. et al. “Filter cakes: Part 1: fluid design impacts
Damage Control, Houston, Texas, USA (January 19–30, drilling and production operations,” Oil & Gas Journal
1976). (January 25, 1999) 97, No. 4, 91–93, 96.
Ahmed, U.: “Fracture-Height Predictions and Post- Ali, S. et al.: “High-Productivity Horizontal Gravel
Treatment Measurements,” Reservoir Stimulation (2nd Packs,” Oilfield Review (Summer 2001) 13, No. 2, 52–73.
ed.), M.J. Economides and K.G. Nolte (eds.), Houston,
Ali, S. et al.: “Chevron Spacer System Successfully
Texas, USA, Schlumberger Educational Services (1989),
Tested,” Offshore Engineer (April 1998) 23, No. 4, 151–152.
10-1–10-13.
Ali, S.A. et al.: “New test identifies completion fluid
Aignesberger, A., and Plank, J.: “Dispersant for Concrete
compatibility problems,” Oil & Gas Journal (August 25,
Mixtures of High Salt Content,” U.S. Patent No. 4,818,288
1997) 95, No. 34, 95–96, 99–101.
(April 4, 1989).
Allen, F.L., Best, G.H., and Lindroth, T.A.: “Welan gum
Al Khayyat, B. et al.: “Successes In Production-Liner
in cement compositions,” U.S. Patent No. 5,004,506
Cementing in Oil-Based Mud: A Case Study,” paper
(April 2, 1991).
SPE 57560 presented at the SPE/IADC Middle East
Drilling Technology Conference held in Abu Dhabi, UAE Allen, S.L., and Wood, M.W.: “Cement Bond Log Quality
(November 8–10, 1999). Control through Simultaneous Recording of Fixed Gate
and Sliding Gate Amplitudes with Transit Times,”
Albert, L.E. et al.: “A Comparison of CBL, RBT and PET
Transactions of the SPWLA 26th Annual Logging
Logs in a Test Well with Induced Channels,” Journal of
Symposium, Dallas, Texas, USA (June 17–20, 1985),
Petroleum Technology (September 1988) 40, No. 9,
paper NN.
1211–1216; also presented as paper SPE 16817 at the
SPE Annual Technical Conference and Exhibition, Dallas, Allen, T.: Particle Size Measurement (3rd ed.), London,
Texas, USA (September 27–30, 1987). UK, Chapman & Hall (1981).
Alberta Energy and Utilities Board: “Statistical Series 57, Allouche, M., Frigaard, I.A., and Sona, G.: “Static wall
Field Surveillance April 1998/March 1999, Provincial layers in the displacement of two visco-plastic fluids
Summaries,” http://www.eub.gov.ab.ca/bbs/products/ in a plane channel,” Journal of Fluid Mechanics
STs/st57-1999.pdf (accessed October 10, 2005). (December 10, 2000) 424, 243–277.
Al-Buraik, K., Al-Abdulqader, K., and Bsaibes, R.: Al-Mashat, Ali Mohsen: “Rheology of Foam Cement,” PhD
“Prevention of shallow gas migration through cement,” dissertation, Colorado School of Mines, Golden, Colorado,
paper SPE 47775 presented at the IADC/SPE Asia Pacific USA (November 23, 1976).
Drilling Technology, Jakarta, Indonesia (September 7–9, Alsdorf, H., and Dittmar, A.: “Material for Sealing Borehole
1998). Walls,” U.S. Patent No. 4,670,056 (June 2, 1987).
Alcorn, I.W., and Bond, D.C.: “Cementing Earth Bores,” Al-Suwaidi, A. et al.: “Light as a Feather, Hard as a Rock,”
U.S. Patent No. 2,469,353 (May 10, 1949). Oilfield Review (Summer 2001) 13, No. 2, 2–15.
Aldrich, C.H., Mitchell, B.J.: “Strength, Permeabilities, Amanullah, Md.: “Experimental determination of adhesive-
and Porosity of Oilwell Foam Cement,” American Society cohesive bond strength (ACBS) and adhesion-cohesion

Well Cementing ■ References 699

modulus (ACM) of mudcakes,” paper SPE 77198 present- presented as paper SPE 7853 at the SPE Deep Drilling
ed at the IADC/SPE Asia Pacific Drilling Technology, and Production Symposium, Amarillo, Texas, USA (April
Jakarta, Indonesia (September 9–11, 2002). 1–3, 1979)
Anderson, G.: Coring and Core Analysis Handbook, Arnold, E.S.: “Cementing: Bridging the Gap from Laboratory
Tulsa, OK, Petroleum Publishing Company (1975). Results to Field Operations,” Journal of Petroleum
Anderson, W.L., and Riddle, G.A.: “Acoustic Amplitude Technology (December 1982) 34, No. 12, 2843–2852; also
Ratio Logging,” Journal of Petroleum Technology presented as paper SPE 9708 at the SPE Permian Basin
(November 1964) 16, No. 1, 1243–1248; also presented Oil and Gas Recovery Symposium, Midland, Texas, USA
as paper SPE 722 at the SPE Annual Meeting, New (March 12–13, 1981).
Orleans, Louisiana, USA (October 6–9, 1963). Aston, M. et al.: “ Towards Zero Fluid Loss Oil Based
Anderson, W.L., and Walker, T.: “Research Predicts Muds,” paper SPE 77446 presented at the SPE Annual
Improved Cement Bond Evaluations with Acoustic Logs,” Technical Conference and Exhibition, San Antonio, Texas,
Journal of Petroleum Technology (November 1961) USA (September 29–October 2, 2002).
13, No. 11, 1093–1097 (paper SPE 196); also presented Audebert, R. et al.: “Cement Composition Containing
at Society of Professional Well Log Analysts Annual Chemically Crosslinked Polyvinyl Alcohol (PVA),” U.S.
Meeting, Dallas, Texas, USA (May 18–19, 1961). Patent No. 5,594,050 (January 14, 1997).
Angstadt, R.L., and Hurley, F.R.: “Hydration of the Alite Audibert, A. et al.: “ Well Cementing Method Using HMHPG
Phase in Portland Cement,” Nature (February 16, 1963) Filtrate Reducer,” U.S. Patent No. 6,257,336 (July 10, 2001).
197, No. 4,868,688. Azouz, I. et al.: “Numerical Simulation of Laminar Flow
Anres, S. et al.: “CATWELL, an Innovative Solution to of Newtonian and Non-Newtonian Fluids in Conduits of
Extend the Reservoir Drainage Area from a Wellhead Arbitrary Cross Sections,” paper SPE 24406 available from
Floating Platform in Deepwater,” paper presented at the SPE, Richardson, Texas, USA (1992).
6th Offshore Mediterranean Conference and Exhibition Babasheikh, A. et al.: “New Shock Resistant Cement
(OMC 2003) Ravenna, Italy (March 26–28, 2003). Reduces the Required Time for Side-Tracking from
Anres, S.: “Catwell Concept Minimizes Deepwater Kick-Off Cement Plugs,” paper SPE 79908 presented
Platforms,” Deepwater Technology (a supplement to at the SPE/International Association of Drilling
Hart’s E&P and Hart PipeLine & Gas Technology) Contractors (IADC) Drilling Conference, Amsterdam,
(July 2003) 26–28. The Netherlands (February 19–21, 2003).
API Mid-Continent District Study Committee on Cementing Backe, K.R. et al.: “Characterising Curing Cement Slurries
Practices and Testing of Oil Well Cements: “Report of by Permeability, Tensile Strength and Shrinkage,” paper
Cooperative Tests on Sulfate Resistance of Cement and SPE 38267 presented at the SPE Western Regional
Additives,” API, Washington, D.C. (1955). Meeting, Long Beach, California, USA (June 25–27, 1997).
API Task Group on Cements for Geothermal Wells: Bade, J.F.: “Cement Bond Logging Techniques—How They
“API Work Group reports field tests of geothermal Compare and Some Variables Affecting Interpretation,”
cements,” Oil & Gas Journal (February 11, 1985) 83, Journal of Petroleum Technology (January 1963) 15,
6, 93–97. No. 1, 17–22; also presented as paper SPE 383 at the
Appleby, S. and Wilson, A.: “Permeability and suction SPE Rocky Mountain Joint Regional Meeting, Billings,
in setting cement,” Chemical Engineering Science Montana, USA (May 24–25, 1962).
(January 1996) 51, No. 2, 251–267. Bailey, L. et al.: “Filter Cake Integrity and Reservoir
Arebrat, P.A.: “Liquid separation device,” U.S. Patent Damage,” paper SPE 39429 presented at the SPE
Application No. 20040112602 (June 17, 2004). International Symposium on Formation Damage Control,
Lafayette, Louisiana, USA (February 18–19, 1998).
Arliguie, G., and Grandet, J.: “Etude par Calorimetrie
de L’Hydratation du Ciment Portland en Presence de Bailey, L. et al.: “New insight into the mechanisms of
Zinc,” Cement and Concrete Research (September shale inhibition using water based silicate,” paper SPE
1985) 15, No. 5, 825–832. 39401 presented at the IADC/SPE Drilling Conference,
Dallas, Texas, USA (March 3–6, 1998).
Armentrout, A.L.: “Material for Recovering Lost
Circulation in Wells,” U.S. Patent No. 2,836,555 (May 27, Bailey, W. et al.: “Water Control,” Oilfield Review (Spring
1958). 2000), 12 (1), 30–51.
Arnold, D.R.: “Planning and Progress on Drilling a Bair, K.A. et al.: “Polymeric Fluid-Loss Additives and
Record-Depth Well in the Rocky Mountains,” Journal Methods of Use Thereof,” U.S. Patent No. 6,465,587
of Petroleum Technology (April 1980) 694–702; also (October 15, 2002).

700 Well Cementing

Bakchoutov, V.S. et al.: “Study of the Grain Composition of at the IADC/SPE Drilling Conference, New Orleans,
Oil-Well Cements,” Proceedings of the 7th Congress on the Louisiana, USA (March 12–15, 1996).
Chemistry of Cement, Paris, France (1980) 5, 202–207. Baret, J.F., Le Roy-Delage, S., and Dargaud, B.:
Baker, W.S., and Harrison, J.J.: “Cement Composition “Cementing compositions and use of such compositions
and Method of Cement Casing in a Well,” U.S. Patent for cementing oil wells or the like,” U.S. Patent No.
No. 4,462,836 (July 31, 1984). 6,458,198 (October 1, 2002).
Banfill, P.F.G., and Kitching, D.R.: “Use of a Control Baret, J.F.: “ Why Cement Fluid Loss Additives Are
Stress Rheometer to Study the Yield Stress of Oilwell Necessary,” paper SPE 17630 presented at the SPE
Cement Slurries,” Proceedings of the International International Meeting on Petroleum Engineering, Tianjin,
Rheology Conference, P.F.G. Banfill (ed.), Liverpool, China (November 1–4, 1988).
England, British Society of Rheology (1990). Barkman, J.H., and Davidson, D.H.: “Measuring Water
Bannister, C.E. et al.: “Critical Design Parameters To Quality and Predicting Well Impairment,” Journal of
Prevent Gas Invasion During Cementing Operations,” Petroleum Technology (July 1972) 24, No. 7, 865–873; also
paper SPE 11982 presented at the SPE Annual Technical presented as paper SPE 3543 at the SPE Annual Meeting,
Conference and Exhibition, San Francisco, California, New Orleans, Louisiana, USA (October 3–6, 1971).
USA (October 5–8, 1983). Barlet-Gouédard, V. et al.: “Cementing Compositions
Bannister, C.E., and Lawson, V.M.: “Role of Cement Fluid and the Application of Such Compositions to Cementing
Loss in Well Bore Completion,” paper SPE 14433 pre- Oil or Analogous Wells,” U.S. Patent No. 6,312,515
sented at the SPE Annual Technical Conference and (November 6, 2001).
Exhibition, Las Vegas, Nevada, USA (September 22–25, Barlet-Gouédard, V. et al.: “Cementing Compositions for
1985). High-Temperature Applications,” WO Patent No. 2003
Bannister, C.E.: “Aqueous Chemical Wash Composition,” 068708 (February 14, 2003).
U.S. Patent No. 4,681,165 (July 21, 1987b). Barlet-Gouédard, V., and Goffé, B.: “Cementing
Bannister, C.E.: “Aqueous Treatment Fluid and Method Compositions and Application of these Compositions to
of Use,” Canadian Patent No. 1,185,777 (1981). Cementing Oil Wells or the Like,” Patent PCT WO 01/70646
Bannister, C.E.: “Aqueous Treatment Fluid and Method (2002).
of Use,” U.S. Patent No. 4,646,834 (March 3, 1987a). Barlet-Gouédard, V., and Vidick, B.: “A Non-Conventional
Bannister, C.E.: “Evaluation of Cement Fluid-Loss Way of Developing Cement Slurries for Geothermal Wells,”
Behavior Under Dynamic Conditions,” paper SPE 7592 Transactions of the Geothermal Resources Council
presented at the SPE Annual Fall Technical Conference Annual Meeting, Davis, California, USA, Geothermal
and Exhibition, Houston, Texas, USA (October 1–3, 1978). Resources Council (2001), 85–91.
Bannister, C.E.: “Rheological Evaluation of Cement Barlet-Gouédard, V., Hendriks, H., and Maroy, P.: “High
Slurries: Methods and Models,” paper SPE 9284 pre- Temperature Retarders for Oil Field Cements, Cement
sented at the SPE Annual Fall Technical Conference and Slurries and Corresponding Cementing Processes,” U.S.
Exhibition, Houston, Texas, USA (September 21–24, 1980). Patent No. 5,503,672 (April 2, 1996).
Bannister, C.E.: “The Role of Cement Fluid-Loss in Barlet-Gouédard, V., Vigneaux, P., and Thiercelin, M.:
Wellbore Completions,” paper SPE 14433 presented at “Methods for Analyzing and Designing Gas-Tight Cement
the SPE Annual Technical Conference and Exhibition, Slurries,” European Patent 21.1010 (October 4, 2001).
Las Vegas, Nevada, USA (September 22–25, 1985). Barnes, B.: “The morphology of the paste-aggregate
Banthia, N.: “Crack Growth Resistance of Hybrid Fiber interface 1,” PhD thesis, Purdue University, 1975.
Reinforced Cement Compositions,” Cement & Concrete Barnes, P.: Structure and Performance of Cements,
Composites (January 2003) 25, No. 1, 3–9. London, England, Applied Science Publishers Ltd.
Barclay, I. et al.: “The Beginning of the End: A Review of (1983).
Abandonment and Decommissioning Practices,” Oilfield Barret, P. et al.: “Thermodynamic and Kinetic Aspects
Review (Winter 2001), 13 (4), 28–41. of C3S Passage in Solution and C-S-H formation,”
Baret, J.F., and Boussouira, B.: “Oil-well cement slurries Proceedings of the 7th International Congress on the
with good fluid-loss control,” European Patent 315,243 Chemistry of Cement (1980b) 2, 279–284.
(May 10, 1989). Barret, P., and Bertrandie, D.: “Fundamental Hydration
Baret, J.F., Garnier, A., and Rashad, B.: “High Performance Kinetic Features of the Major Cement Constituents:
Water Reduced Cement Slurries Prepared With Low Tricalcium Silicate (Ca3SiO5) and Beta-Dicalcium
Cost Optimized Blend,” paper SPE 35088 presented

References 701
Silicate (β-Ca2SiO4),” Journal de Chimie Physique design in horizontal wells,” SPE Drilling Completion
(November–December 1986) 83, No. 11/12, 765–775. (September 2000) 15, No. 3, 191–197.
Barret, P., Bertrandie, D., and Ménétrier, D.: “Comparative Beirute, R.M., and Cheung, P.R.: “Method for Selection of
Study of C-S-H Formation from Supersaturated Solutions Cement Recipes to Control Fluid Invasion After
and C3S Solution Mixtures,” Proceedings of the 7th Cementing,” SPE Production Engineering (November
Congress on the Chemistry of Cement (1980a) 2, 261–266. 1990) 5, No. 4, 433–440; also presented as paper SPE
Barret, P., Ménétrier, D., and Betrandie, D. et al.: 19522 at the SPE Annual Technical Conference and
“Mechanism of C3S dissolution and problem of the con- Exhibition, San Antonio, Texas, USA (October 8–11,
gruency in the very initial period and later on,” Cement 1989).
and Concrete Research (September 1983) 13, No. 5, Beirute, R.M., and Flumerfelt, R.W.: “Mechanics of the
728–738. Displacement Process of Drilling Muds by Cement Slurries
Barret, P.: “Hydration Mechanism of Calcium Silicates Using an Accurate Rheological Model,” paper SPE 6801
(C3S, C2S), Cement Compounds, Through the General presented at the SPE Annual Fall Technical Conference
Concepts of the Reactivity of Solids,” Proceedings of the and Exhibition, Denver, Colorado, USA (October 9–12,
8th International Congress on the Chemistry of Cement 1977).
(1986) 3, 86–92. Beirute, R.M., Sabins, F., and Ravi, K.V.: “Large-Scale
Barton, N.A., Archer, G.L., and Seymour, D.A.: Experiments Show Proper Hole Conditioning: A Critical
“Computational fluid dynamics improves liner cement- Requirement for Successful Cementing Operations,”
ing operation,” Oil & Gas Journal (September 26, 1994) paper SPE 22774 presented at the SPE Annual Technical
92, No. 39, 92–98. Conference and Exhibition, Dallas, Texas, USA (October
6–9, 1991).
Barvinok, M.S., Komokhov, P.S., and Bondareva, N.F.:
“Effect of Temperature and Additives on the Early Beirute, R.M., Wilson, M.A., and Sabins, F.L.: “Attenuation
Hardening Stage,” Proceedings of the 6th International of Casing Cemented with Conventional and Expanding
Conference on the Chemistry of Cement (1976) 2, Cements across Heavy-Oil and Sandstone Formations,”
151–155. SPE Drilling Engineering (September 1992) 7, No. 3,
200–211; also presented as paper SPE 18027 at the SPE
Basile, F. et al.: “Effect of Gypsum State in Industrial
Annual Technical Conference and Exhibition, Houston,
Cements on the Action of Superplasticizers,” Cement and
Texas, USA (October 2–5, 1988).
Concrete Research (September 1987) 17, No. 5, 715–722.
Beirute, R.M.: “A Circulating and Shut-in Well
Baumgarte, C. et al.: “Case Studies of Expanding Cement
Temperatures Profile Simulator,” Journal of Petroleum
To Prevent Microannular Formation,” paper SPE 56535
Technology (September 1991) 43, No. 9, 1140–1146; also
presented at the SPE Annual Technical Conference and
presented as paper SPE 17591 at the SPE International
Exhibition, Houston, Texas, USA (October 3–6, 1999).
Meeting on Petroleum Engineering, Tianjin, China
Baumgarte, C., Thiercelin, M., and Klaus, D.: “Case (November 1–4, 1988).
Studies of Expanding Cement to Prevent Microannular
Beirute, R.M.: “Flow Behavior of an Unset Cement Plug
Formation,” paper SPE 56535 presented at the SPE
In Place,” paper SPE 7589 presented at the SPE Annual
Annual Technical Conference and Exhibition, Houston,
Technical Conference and Exhibition (October 1–3,
Texas, USA (October 3–6, 1999).
1978).
Beach, H.J., and Goins, W.C., Jr.: “A Method of Protecting
Beirute, R.M.: “High-Temperature Cement Mud Spacer,”
Cements Against the Harmful Ef fects of Mud
U.S. Patent No. 4,276,182 (June 30, 1981).
Decontamination,” Transactions, SPE (1957) 210,
148–152. Beirute, R.M.: “API Revises Procedures to Measure
Cement Slurry Rheology,” Oil & Gas Journal
Beach, H.J.: “Consequences of Salting Well Cements,”
(September 1, 1986) 84, No. 38, 36–38.
paper SPE 10032 presented at the International
Petroleum Exhibition and Technical Symposium, Beijing, Beirute, R.M.: “On-Site Diagnosis of Cement Job
China (March 18–26, 1982). Problems: The Concept of Job Signatures,” SPE Drilling
Engineering (December 1988) 3, No. 4, 374–380; also
Bearden, W.G.: “Effect of Temperature and Pressure on
presented as paper SPE 16649 at the SPE Annual
the Physical Properties of Cement,” Oil-Well Cementing
Technical Conference and Exhibition, Dallas, Texas, USA
Practices in the United States, Washington, D.C., USA,
(September 27–30, 1987).
American Petroleum Institute (1959) 49–59.
Beirute, R.M.: “The Phenomenon of Free Fall During
Becker, T.E., and Gardiner, N.H.: “Laboratory modeling
Primary Cementing,” paper SPE 13045 presented at the
of mudcake application and erosion for gravel-pack

702 Well Cementing

SPE Annual Technical Conference and Exhibition, Bentur, A. et al.: “Structural Properties of Calcium
Houston, Texas, USA (September 16–19, 1984). Silicate Pastes: II. Effect of Curing Temperature,”
Bell, D.R., et al: “Evaluation and Procedures for Preparing Journal of the American Ceramic Society (July–August
and Sampling Dry-Blended Cements,” paper SPE 18525 1979) 62, No. 7–8, 362–366.
presented at the SPE Eastern Regional Meeting in Bentz, D.P. et al.: “Cellular Automaton Simulations of
Charleston, West Virginia, USA (November 1–4, 1988). Cement Hydration and Microstructure Development,”
Bell, G., Bensted, J., and Glasser, F.P.: “Characterization Modelling and Simulation in Materials Science and
of Hydrothermally Treated Calcium Silcate and Oilwell Engineering (July 1994) 2, No. 4, 783–808.
Cement Hydration Products,” Advances in Cement Berger, R.L., and McGregor, J.D.: “Influence of
Research (1989) 12, No. 6, 61–72. Admixtures on the Morphology of Calcium Hydroxide
Belousov, G.A. et al.: “Spacer Fluid for Separating Formed During Tricalcium Silicate Hydration,” Cement
Drilling Fluid and Cement Slurry,” Neftyanoe Khozyaisto and Concrete Research (January 1972) 2, No. 1, 43–55.
(August 1987) No. 8, 25–29. Bergeron, H.A., and Grant, W.H.: “Cement Quality
Benabdelkarim, M., and Galiana, C.: “Nonradioactive Control Program Shows Substantial Savings,” SPE
Densitometer for Continuous Monitoring of Cement Production Engineering (August 1990) 5, No. 3,
Mixing Process,” paper SPE 23262 presented at the 231–235; also presented as paper SPE 18621 at the
First International Conference in Health, Safety and SPE/IADC Drilling Conference, New Orleans, Louisiana,
Environment, The Hague, The Netherlands (November USA (February 28–March 3, 1989).
10–14, 1991). Bergren, F.E., and Bradley, D.A.: “Design and Evaluation
Ben-Dor, L., and Perez, D.: “Influence of Admixtures on Of Pump-In Temperature Surveys for Detecting Cement
Strength Development of Portland Cement and on the Channels Above Perforations,” paper SPE 18145 pre-
Microstructure of Tricalcium Silicate,” Journal of sented at the SPE Annual Technical Conference and
Materials Science (February 1976) 11, No. 2, 239–245. Exhibition, Houston, Texas, USA (October 2–5, 1988).
Benge, O.G., and Webster, W.W.: “Evaluation of Blast Berke, N.S.: “The Effects of Calcium Nitrite and Mix
Furnace Slag Slurries for Oilfield Application,” paper Design on the Corrosion Resistance of Steel in Concrete
International Association of Drilling Contractors (Part 1),” paper National Association of Corrosion
(IADC)/SPE 27449 presented at the IADC/SPE Drilling Engineers (NACE) No. 273, presented at NACE
Conference, Dallas, Texas, USA (February 15–18, 1994). Corrosion/85 Meeting, Boston, Massachusetts, USA
(March 25–29, 1985).
Benge, O.G., et al.: “A New Low-Cost Permafrost Cementing
System,” paper SPE 10757 presented at the SPE California Bern, P.A. et al.: “Barite sag: measurement, modeling
Regional Meeting, San Francisco, California, USA and management,” paper SPE 47784 presented at the
(March 24–26, 1982). IADC/SPE Asia Pacific Drilling Technology, Jakarta,
Indonesia (September 7–9,1998).
Bensted, J., and Barnes, P.: Structure and Performance
of Cements (2nd ed.), London, Spon Press (2002). Berra, M. et al.: “Behavior of a Cementing Hydraulic
Binder Under Severe Geothermal Conditions,”
Bensted, J.: “Effect of Accelerator Additives on the Early
Geothermics (1998) 17, No. 5–6, 785–813.
Hydration of Portland Cement,” Il Cemento (January–
March, 1978) 75, No. 1, 13–20. Bertrandie, D., and Barret, P.: “Initial Interfacial Steps
in Hydration of Calcium Aluminates as Cement
Bensted, J.: “Fase ferritica-uno studio spettroscopio
Compounds,” Proceedings of the 8th International
all’infrarosso,” Il Cemento (January–March, 1976) 73,
Conference on the Chemistry of Cement (1986) 3,
No. 1, 45–51.
79–85.
Bensted, J.: “A discussion of the paper ‘alintte—chemical
Biagini, S. et al.: “Sulfonated Polyst yrene as
composition, solid solution and hydration behaviour’ by
Superplasticizer,” Il Cemento (1982) 79, No. 4, 345–354.
J. Neubauer and H. Pöllmann,” Cement and Concrete
Research (1995) 25, No. 8, 1806–1807. Biczok, I.: Concrete Corrosion—Concrete Protection,
New York City, New York, USA, Chemical Publishing Co.,
Bensted, J.: “Discussion Of Chemical Shrinkage Properties
Inc. (1967) 287–298.
Of Oilfield Cements,” SPE Drilling Engineering
(September 1991) 231. Biezen, E., Van der Werff, N., and Ravi, K.: “Experimental
and Numerical Study of Drilling Fluid Removal from a
Bensted, J.: “Special Cements,” Lea’s Chemistry of
Horizontal Wellbore,” SPE 62887 presented at the SPE
Cement and Concrete (4th ed.), P.C. Hewlett (ed.),
Annual Technical Conference and Exhibition, Dallas,
London, UK, Butterworth-Heinemann (2001), 783–840.
Texas, USA (October 1–4, 2000).

References 703
Bigelow, E.L.: “A Practical Approach to the Interpretation Journal of the American Ceramic Society (August
of Cement Bond Logs,” Journal of Petroleum Technology 1963) 46, No. 8, 395–399.
(July 1985) 37, No. 7, 1285–1294; also presented as paper Blomberg, N. et al.: “Boue de ciment hydraulique pour la
SPE 13342 at the SPE California Regional Meeting, cimentation des puits de petrole,” French Patent
Bakersfield, California, USA (March 27–29, 1985). 2,587,988 (1986).
Biles, J.W.: “Selective Plugging Method,” U.S. Patent Blount, C.G., Copoulos, A.E., and Myers, G.D.: “A cement
No. 3,658,131 (1972). channel-detection technique using the pulsed-neutron
Bimaputra, A. et al.: “Problem solving of non productive log,” SPE Formation Evaluation (December 1991) 6,
time of plug and squeeze cementing offshore No. 4, 485–492; previously presented as paper SPE 20042
Kalimantan,” paper IPA99-E-156 presented at the 27th at the SPE California Regional Meeting, Ventura,
Annual Indonesian Petroleum Association Convention, California, USA (April 4–6, 1990).
Jakarta (February 2–3, 2000). Bloys, J.B. et al.: “Dispersant Compositions for
Binkley, G.W., Dumbauld, G.K., and Collins, R.E.: “Factors Subterranean Well Drilling and Completion,” U.S. Patent
Affecting the Rate of Deposition of Cement in Unfractured No. 5,292,367 (March 8, 1994).
Perforations During Squeeze-Cementing Operations,” Bloys, J.B., Wilson, W.N., and Bradshaw, R.D.: “Cementing
Trans., American Institute of Mining, Metallurgical and Oil and Gas Wells,” U.S. Patent No. 5,076,852 (December
Petroleum Engineers (AIME) (1958) 213, 51–58. 31, 1991).
Biot, M.A.: “General Solutions of the Equations of Bogue, R.H.: “Calculation of the Compounds in Portland
Elasticity and Consolidation for a Porous Material,” Cement,” Industrial & Engineering Chemistry Anaysis
Journal of Applied Mechanics (1956) 78, 91–96. Edition (October 1929) 1, No. 4, 192–197.
Biot, M.A.: “General Theory of Three-Dimensional Boisnault, J.-M. et al.: “Concrete developments in
Consolidation,” Journal of Applied Physics (1941) 12, cementing technology,” Oilfield Review (Spring 1999)
155–164. 11, No. 1, 16–29.
Birchall, J.D., Howard, A.J., and Double, D.D.: Cement and Bol, G.M. et al.: “Preventing Fluid Migration Around a Well
Concrete Research (March 1980) 10, No. 2, 145–155. Casing,” European Patent Application 0197609 (1986).
Bird, R.B., Stewart, W.E., and Lightfoot, E.N.: Transport Bol, G.M. et al.: “Borehole stability in shales,” SPE
Phenomena, New York City, New York, USA, John Wiley Drilling & Completion (June 1994) 9, No. 2, 87–94; also
& Sons Incorporated (1960). presented as paper SPE 24975 at the SPE European
Bittleston, S., and Guillot, D.: “Mud Removal: Research Petroleum Conference, Cannes, France (November
Improves Traditional Cementing Guidelines,” Oilfield 16–18, 1994).
Review (April 1991) 3, No. 2, 44–54. Bol, G.M. et al.: “Cementing: How to Achieve Zonal
Bittleston, S., and Hassenger, O.: “Flow of viscoplastic Isolation,” paper presented at the Offshore Mediterranean
fluids in a rotating concentric annulus,” Journal of Non- Conference, Ravenna, Italy (March 19–21, 1997).
Newtonian Fluid Mechanics (March 1992) 42, No. 1–2, Boncan, V.G., and Gandy, R.: “Well Cementing Method
19–36. Using an AM/AMPS Fluid Loss Additive Blend,” U.S.
Bittleston, S., Ferguson, J., and Frigaard, I.A.: “Mud Patent No. 4,632,186 (December 30, 1986).
removal and cement placement during primary cement- Bosma, M.G.R. et al.: “Design Approach to Sealant
ing of an oil well,” Journal of Engineering Mathematics Selection for the Life of the Well,” paper SPE 56536, pre-
(August 2002) 43, Nos. 2–4, 223–253. sented at the SPE Annual Technical Conference and
Bjordal, A., Harris, K.L., and Olaussen, S.R.: “Colloidal Exhibition, Houston, Texas, USA (October 3–6, 1999).
Silica Cement: Description and Use in North Sea Bosma, M.G.R. et al. “Development of a novel silicone
Operations,” paper SPE 26725 presented at the SPE rubber/cement plugging agent for cost effective thru'
Of fshore European Conference, Aberdeen, UK tubing well abandonment,” Proceedings of the 1998
(September 7–10, 1993). IADC/SPE Asia Pacific Drilling Technology Conference,
Blanco, A. et al.: “Effective Pay Zone Isolation of Steam 483–496.
Injection Wells,” paper SPE 53689 presented at the SPE Bosma, M.G.R., Cornelissen, E.K., and Schwing, A.:
Latin American and Caribbean Petroleum Engineering “Improved Experimental Characterisation of
Conference, Caracas, Venezuela (April 21–23, 1999). Cement/Rubber Zonal Isolation Materials,” paper SPE
Blank, B., Rossington, D.R., and Weinland, L.A.: 64395 presented at the SPE Asia Pacific Oil and Gas
“Adsorption of Admixtures on Portland Cement,” Conference and Exhibition, Brisbane, Australia

704 Well Cementing

(October 16–18, 2000); also presented as paper SPE Brice, J.W., and Holmes, B.C.: “Engineered Casing
64762 at the SPE International Oil and Gas Conference, Cementing Programs Utilizing Turbulent Flow
Beijing, China (November 7–10, 2000). Techniques,” Journal of Petroleum Technology
Boukhelifa, L. et al.: “Evaluation of Cement Systems for (May 1964) 16, No. 5, 503–508; also presented as paper
Oil and Gas Well Zonal Isolation in a Full-Scale Annular SPE 742 at the SPE Annual Technical Conference and
Geometry,” paper SPE 87195 presented at the IADC/SPE Exhibition, New Orleans, Louisiana, USA (October 8–9,
Drilling Conference, Dallas, Texas, USA (March 2–4, 1963).
2004). Brookhaven National Laboratory: “Brookhaven Lab,
Bour, D.L., and Vennes, M.R.: “Application of Foam Halliburton Company, and Unocal Corporation Win R&D
Cement in the Williston Basin,” paper SPE 18984 pre- 100 Award for High-Performance Cement,” press release
sented at the SPE Joint Rocky Mountain Regional/Low 00–56 (2000).
Permeability Reservoirs Symposium and Exhibition, Brothers, L.E. et al.: “Cement Compositions and
Denver, Colorado, USA (March 6–8, 1989). Methods for High Temperature Wells Containing Carbon
Bour, D.L., Creel, P., and Kulakofsky, D.S.: “Computer Dioxide,” U.S. Patent No. 6,332,921 (December 25, 2001).
Simulation Improves Cement Squeeze Jobs,” paper Brothers, L.E. et al.: “Light Weight High Temperature
21595 presented at the CIM/SPE International Well Cement Compositions and Methods,” U.S. Patent
Technical Meeting, Calgary, Alberta, Canada (June No. 5,900,053 (May 4, 1999).
10–13, 1990). Brothers, L.E. et al.: “Light Weight High Temperature
Bour, D.L., et al.: “New Expansive Cement System for Well Cement Compositions and Methods,” U.S. Patent
High Temperature,” Proceedings of the 1988 No. 6,143,069 (November 7, 2000).
Southwestern Petroleum Short Course, Lubbock, Texas, Brothers, L.E., and deBlanc, F.X.: “New Cement
USA, Southwestern Petroleum (1988) 1–9. Formulation Helps Solve Deep Cementing Problems,”
Bowman, G. and Sherer, B.: “How To Run and Cement Journal of Petroleum Technology (June 1989) 41, No. 6,
Liners: Pt. 4,” World Oil (July 1988) 207, No.1, 84, 86–88. 611–614; also presented as paper SPE 17181 at the
Boyington, W.H. et al.: “Optimised design, operations IADC/SPE Drilling Conference, Dallas, Texas, USA
procedures and materials improve cementing results in (February 29–March 4, 1988).
the North Sea,” paper SPE 19249 presented at Offshore Brothers, L.E., Onan, D.D., and Morgan, R.L.: “Well
Europe 89, Aberdeen, Scotland (September 5–8, 1989). cement compositions containing rubber particles and
Bozich, M.P., Montman, R.C., and Harms, W.M.: methods of cementing subterranean zones,” U.S. Patent
“Application of Foamed Portland Cement to Deep Well No. 5,779,787 (July 14, 1988).
Conditions in West Texas,” paper SPE 12612 presented at Brothers, L.E.: “Method of Reducing Fluid Loss in
the SPE Deep Drilling and Production Symposium, Cement Compositions Containing Substantial Salt
Amarillo, Texas, USA (April 1–3, 1984). Concentrations,” U.S. Patent No. 4,640,942 (February 3,
Bradford, B. and Reiners, B.: “Analysis gives successful 1987).
cement squeeze,” Oil & Gas Journal (April 1, 1985), 83 Brothers, L.E.: “Set Retarded Ultra Fine Cement
(13), 71–74. Compositions and Methods,” U.S. Patent No. 5,340,397
Brady, S.D. et al.: “Recent technological advances help (August 23, 1994).
solve cement placement problems in the Gulf of Brough, A.R. et al.: “In-situ solid-state NMR studies of
Mexico,” paper SPE 23927 presented at the IADC/SPE Ca3SiO5: hydration at room temperature and at elevated
Drilling Conference, New Orleans, Louisiana, USA, temperatures using 29Si enrichment,” Journal of
(February 18–21, 1992). Materials Science (August 1994) 29, 15, 3926–3940.
Brand, F., Peixinho, J., and Nouar, C.: “A Quantitative Brown, D., and Ferg, T.: “The use of lightweight cement
Investigation of the Laminar-to-Turbulent Transition: slurries and downhole chokes improves the success
Application to Efficient Mud Cleaning,” paper SPE 71375 of primary cementing on air-drilled wells,” paper
presented at the SPE Annual Technical Conference and SPE 84561 presented at the SPE Annual Technical
Exhibition, New Orleans, Louisiana, USA (September 30– Conference and Exhibition held in Denver, Colorado,
October 3, 2001). USA (October 5–8, 2003).
Brandt, W. et al.: “Deepening the Search for Offshore Brown, H.D., Grijalva, V.E., and Raymer, L.L.: “New
Hydrocarbons,” Oilfield Review (Spring 1998) 10, No. 1, Developments in Sonic Wave Train Display and Analysis
2–21. in Cased Holes,” Transactions of the SPWLA 11th
Annual Logging Symposium, (May 3–6, 1970), paper F.

References 705
Brown, M.D., and Gledhill, A.D.: “Expandable Drill Bit Reservoir Engineering (February 1990), 41–46; revised
Provides New Method Of Drilling Increased Diameter for publication from paper SPE 16963 presented at the
Hole,” paper SPE/IADC 79793 presented at the SPE Annual Technical Conference and Exhibition,
SPE / IADC Drilling Conference, Amsterdam, The Dallas, Texas, USA (September 27–30, 1987).
Netherlands (April 19–21, 2003). Bulatov, A.I., Obosin, O.N., and Kuksov, A.K.:
Brown, P.W. et al.: “Analyses of the aqueous phase during “Occurrence of Channels in the Annular Spaces of Wells
early C3S hydration,” Cement and Concrete Research after Cementing,” Gazovaya Prom (February 1970) 15,
(March 1984); 14, No. 2, 257–262. No. 2, 3–6, translated from Russian.
Brown, P.W.: “Kinetics of Tricalcium Aluminate and Bulatov, A.I.: Plugging Materials and the Cementing of
Tetracalcium Aluminoferrite Hydration in the Presence Wells (2nd ed.), Moscow, Russia, Mir Publishers (1985).
of Calcium-Sulfate,” Journal of the American Ceramic Burba, J.L. III, Holman, W.E., and Crabb, C.R.:
Society (December 1993) 76, No. 12, 2971–2976. “Laboratory and Field Evaluations of Novel Inorganic
Bruckdorfer, R.A., Jacobs, W., and Masson, J.-P.: “CBL Drilling Fluid Additive,” paper SPE 17198 presented at
Evaluation of Foam- and Synthetic-Cemented Casings,” the IADC/SPE Drilling Conference, Dallas, Texas, USA
Journal of Petroleum Technology (November 1984) (February 28–March 2, 1988).
36, No. 11, 1917–1921; also presented as paper SPE Burkhalter, J.F., Childs, J.D., and Sutton, D.L.: “Well
11980 at the SPE Annual Technical Conference and Cementing Process and Gasified Cements Useful
Exhibition, San Francisco, California, USA (October Therein,” U.S. Patent 4,450,010 (May 22, 1984).
5–8, 1983).
Burkhardt, J. A.: “Wellbore Pressure Surges Produced by
Bruckdorfer, R.A.: “Carbon Dioxide Corrosion in Oilwell Pipe Movement,” Journal of Petroleum Technology
Cements,” paper SPE 15176 presented at the SPE Rocky (June 1961) 13, No. 6, 595–605.
Mountain Regional Meeting, Billings, Montana, USA
Butler, R.M.: “Application of SAGD, Related Processes
(May 19–21, 1986).
Growing in Canada,” Oil & Gas Journal (May 14, 2001)
Bruere, G.M.: “Set-Retarding Effects of Sugars in 99, No. 20, 74–78.
Portland Cement Pastes,” Nature (October 1966) 212,
Butler, R.M.: “SAGD Comes of Age,” Journal of
No. 5061, 502–503.
Canadian Petroleum Technology (1998) 37, No. 7, 9–12.
Brunauer, S. et al.: “Adsorption of Gases in Multi-molec-
Butler, R.M.: Thermal Recovery of Oil and Bitumen,
ular Layers,” Journal of the American Chemical Society
Englewood Cliffs, New Jersey, USA (1991), Prentice-
(1938) 60, 309–319.
Hall, 285–359.
Bruton, J., Ivan, C., and Bedel, J.P.: “Unique Engineering
Butsch, R.J. et al.: “The Evaluation of Specialised
Software Developed for Lost Circulation Planning and
Cements,” paper SPE 76713 presented at the SPE
Assessment,” paper AADE-04-DF-HO-05 presented at
Western Regional/AAPG Pacific Section Joint Meeting,
the American Association of Drilling Engineers National
Anchorage, Alaska, USA (May 20–22, 2002).
Technical Conference, Houston, Texas, USA (April 6–7,
2004). Butsch, R.J.: “Overcoming Interpretation Problems of
Gas-Contaminated Cement using Ultrasonic Cement
Bryant, G.A.: “Successful Alternatives to Conventional
Logs,” paper SPE 30509 presented at the SPE Annual
Cement Designs in the Williston Basin,” paper SPE
Technical Conference and Exhibition, Dallas (October
12904 presented at the SPE Rocky Mountain Regional
22–25, 1995).
Meeting, Casper, Wyoming, USA (May 21–23, 1984).
Byrdson, R., Richardson, I.G., and Groves, G.W.:
Brydson, R., Richardson, I.G., and Groves, G.W.:
“Determining the local coordination of aluminium in
“Determining the Local Coordination of Aluminium in
cement using electron energy loss near-edge structure,”
Cement Using Electron Energy Loss Near-Edge Structure,”
Mikrochimica Acta (1994) 114–115, 221–229.
Mikrochimica Acta (1994) 114–115, 221–229.
Caliskan, S.: “Aggregate/mortar interface: influence of
Buchan, L., and Little, M.T.S.: “Advanced Techniques
silica fume at the micro- and macro-level,” Cement and
Improve Liner Cementation in North Sea Wells: An
Concrete Composites (May–July 2003) 25, Nos. 4–5,
Operator’s Experience,” SPE Drilling Engineering
557–564.
(September 1988) 3, No. 3, 281–288; also presented as
paper SPE 15896 at the SPE European Petroleum Calloni, G., D’antona, P., and Moroni, N.: “A New
Conference, London, UK (October 20–22, 1986). Rheological Approach Helps Formulation of Gas
Impermeable Cement Slurries,” Cement and Concrete
Buell, R.S., Kazemi, H., and Poettmann, F.H.: “Analyzing
Research (April 1999) 29, No. 4, 523–526.
Injectivity of Polymer Solutions With the Hall Plot,” SPE

706 Well Cementing

Calvert, D.G., and Griffin, T.J. Jr.: “Determination of Carpenter, R.B. et al.: “Well completion spacer fluids and
Temperatures for Cementing Wells Drilled in Deep Water,” methods,” U.S. Patent No. 5,789,352 (August 4, 1998).
paper SPE 39315 presented at the 1998 IADC/SPE Drilling Carré, G. et al.: “High Expectations from Deepwater
Conference, Dallas, Texas, USA (March 3–6, 1998). Wells,” Oilfield Review (Winter 2002) 14, No. 4, 36–51.
Calvert, D.G., Heathman, J.F., and Griffin, J.E.: “Plug Cartalos, U. et al.: “Defining the Rheological Properties
Cementing: Horizontal to Vertical Conditions,” paper of Cement Slurries for the Successful Cementation of an
SPE 30514 presented at the SPE Annual Technical Oil Well,” Proceedings of the 4th European Conference,
Conference and Exhibition, Dallas, Texas (October C. Gallegos (ed.), Sevilla, Spain (September 1994),
22–25, 1995). 275–277.
Campo, D. et al.: “Monodiameter Drilling Liner—From Carter, G., and Smith, D.K.: “Properties of Cementing
Concept to Reality,” paper SPE/IADC 79790 presented Compositions at Elevated Temperatures and Pressures,”
at the SPE/IADC Drilling Conference, Amsterdam, The AIME Petroleum Transactions (1958) 213, 20–27.
Netherlands (February 19–21, 2003).
Carter, L.G., and Evans, G.W.: “A Study of Cement-Pipe
Carathers, K., and Crook, R.: “Surface Pipe Cement Bonding,” Journal of Petroleum Technology (February
Gives High Early Strength with New Cement Additive,” 1964) 16, No. 2, 157–160; also presented as paper SPE
Proceedings of the 1987 Southwestern Petroleum Short 764 at the SPE California Regional Meeting (October
Course, Lubbock, Texas, USA, Southwestern Petroleum 24–26, 1963).
(1987) 12–19.
Carter, L.G. et al.: “Expanding Cements for Primary
Carden, R.S. et al.: “Unique Aspects of Drilling and Cementing,” Journal of Petroleum Technology (May 6,
Completing Hot Dry Rock Geothermal Wells,” SPE 1966) 22, No. 5, 551–558; also presented as paper SPE
Drilling Engineering (May 1985) 37, No. 5, 821–834; also 1235 at the SPE Annual Meeting, Denver, Colorado, USA
presented as paper IADC/SPE 11373 at the IADC/SPE (October 3–6, 1965).
Drilling Conference, New Orleans, Louisiana, USA
Carter, L.G. et al.: “Resilient Cement Decreases
(February 20–23, 1983).
Perforating Damage,” paper presented at API Mid-
Cardno, D.G. et al.: “The use of silicate WBM facilitates Continent District, Division of Production Spring
successful primary cementation operations on the BHP Meeting, Amarillo, Texas, USA (April 3–5, 1968).
Petroleum Liverpool Bay Development,” paper SPE/IADC
Carter, L.G., and Slagle, K.A.: “A Study of Completion
37611 presented at the SPE/IADC Drilling Conference,
Practices to Minimize Gas Communication,” Journal of
Amsterdam, The Netherlands (March 4–6, 1997).
Petroleum Technology (September 1972) 24, No. 9,
Caretta, F., Young, G., and Fox, P.: “A New Cost Effective 1170–1174; also presented as paper SPE 3164 at the SPE
Approach to Comprehensive Gravel Pack Completion,” Central Plains Regional Meeting, Amarillo, Texas, USA
SPE paper 62730 presented at the IADC/SPE Asia (November 16–17, 1970).
Pacific Drilling Technology (Sept. 11–13, 2000).
Carter, L.G., Beecroft, W.H., and Pollock, R.W.:
Carnahan, B.D. et al.: “Fiber Optic Temperature “Cementing Practices for Thermal Wells,” Journal of
Monitoring Technology,” paper SPE 54599 presented at Canadian Petroleum Technology (July–September
the 1999 SPE Western Regional Meeting, Anchorage, 1966) 5, No. 3, 130–134.
Alaska (May 26–28, 1999).
Carter, L.G., Cooke, C., and Snelson, L.: “Cementing
Carneiro, F., and Barcellos, A.: “The Tensile Strength of Research in Directional Gas Well Completions,” paper
Concrete,” Réunion Internationale des Laboratoires et SPE 4313 presented at the SPE Annual European
Experts des Matériaux, Systèmes de Constructions et Meeting, London, England (April 2–3, 1973).
Ouvrages (RILEM) Bulletin (1953) 13, 97–123.
Casabonne, J.M., Jouve, M., and Nelson, E.B.: “High
Carney, L.: “Cement Spacer Fluid,” Journal of Petroleum Temperature Retarders for Oil Field Cements, Cement
Technology (August 1974) 26, 856–858; also presented Slurries and Corresponding Cementing Processes,” U.S.
as paper SPE 4784 at the SPE Formation Damage Patent No. 5,503,671 (April 2, 1996).
Meeting, Houston, Texas, USA, (October 6–9, 1974).
Casson, N.: “Flow Equation for Pigment-Oil Suspensions
Carpenter, R.B., and Johnson, D.L.: “Method and of the Printing Ink Type,” Rheology of Dispersed
Cement-Drilling Fluid Cement Composition for Systems (C.C. Mill, ed.), Oxford, England, Pergamon
Cementing a Wellbore,” U.S. Patent No. 5,874,387 Press (1959), 84–104.
(February 23, 1999).
Catala, G., Stowe, I., and Henry, D.: “A Combination of
Carpenter, R.B.: “Matrix Control Cementing Slurry,” U.S. Acoustic Measurements to Evaluate Cementations,”
Patent No. 4,569,395 (February 11, 1986). paper SPE 13139 presented at the SPE Annual Technical

References 707
Conference and Exhibition, Houston, Texas, USA Chatterji, J., Brake, B.G., and Tinsley, J.M.: “Liquid
(September 16–19, 1984). Water Loss Reducing Additives for Cement Slurries,”
Caughron, D.E. et al.: “Unique Crosslinking Pill in U.S. Patent No. 4,466,837 (August 21, 1984).
Tandem With Fracture Prediction Model Cures Chatterji, S.: “Electron-optical and X-ray Diffraction
Circulation Losses in Deepwater Gulf of Mexico,” paper Investigation of the Effects of Lignosulphonates on the
SPE 74518 presented at the IADC/SPE Drilling Hydration of C3A,” Indian Concrete Journal (April,
Conference, Dallas, Texas (February 26–28, 2002). 1967) 41, No. 4, 151–160.
Cerasi, P. et al.: “Measurement of the Mechanical Cheatham, J.B., and McEver, J.W.: “Behavior of Casing
Properties of Filtercakes,” paper SPE 68948 presented Subjected to Salt Loading,” Journal of Petroleum
at the SPE European Formation Damage Conference, Technology (September 1964) 16, No. 9, 1069–1075; also
The Hague, The Netherlands (May 21–22, 2001). presented as paper SPE 828 at the SPE Rocky Mountain
Chambers, M.R.: “Making Multi-lateral Wells Cost Regional Meeting, Casper, Wyoming, USA (May 25–26,
Effective,” paper SPE 51244 presented at the SPE 1964).
Annual International Conference and Exhibition, Lagos, Chekiri, K.: “Effect of Jet Perforation on Foam Cement,”
Nigeria (August 5–7, 1998). T-2069 PhD thesis, Colorado School of Mines, Golden,
Chan, A.F.: “Tandem spacer fluid system and method for Colorado, USA (1978).
positioning a cement slurry in a wellbore annulus,” U.S. Chen, W.F., and Han, D.J.: Plasticity for Structural
Patent No. 6,283,213 (September 4, 2001). Engineers, New York City, New York, USA, Springer-
Chan, K.S. and Griffin, T.J.: “Low temperature, low rhe- Verlag (1988).
ology synthetic cement,” US Patent No. 5547027 (August Chenevert, M.E., and Jin, L.: “Model for predicting well-
20, 1996). bore pressure in cement columns,” paper SPE 19521
Chan, K.S.: “Injection profile modification with a new presented at the SPE Annual Technical Conference and
non-polymer gelling system," paper CIM 89-40-46 pre- Exhibition, San Antonio, Texas, USA (October 8–11, 1989).
sented at the 40th Technical Meeting of the Petroleum Chenevert, M.E., and Shrestha, B.K.: “Chemical
Society of CIM, Banff, Alberta, Canada (May 28–31, Shrinkage Properties of Oilfield Cements,” SPE Drilling
1989). Engineering (March 1991) 6, No. 1, 37–43; also pre-
Chandra, S., and Björnström, J.: “Influence of Cement sented as paper 16654 at the SPE Annual Technical
and Superplasticizers Type and Dosage on the Fluidity Conference and Exhibition, Dallas, Texas, USA
of Cement Mortars—Part I,” Cement and Concrete (September 27–30, 1987).
Research (October 2002a) 32, No. 10, 1605–1611. Cheung, P.R., and Beirute, R.M.: “Gas Flow in Cements,”
Chandra, S., and Björnström, J.: “Influence of Journal of Petroleum Technology (June 1985); also pre-
Superplasticizer Type and Dosage on the Slump Loss sented as paper SPE 11207 SPE Annual Technical
of Portland Cement Mortars—Part II,” Cement and Conference and Exhibition, New Orleans, Louisiana,
Concrete Research (October 2002b) 32, No. 10, 1613–1619. USA (September 26–29, 1982).
Chatterji, J. et al.: “Cement Compositions and Cheung, P.R., and Myrick, B.D.: “Field Evaluation of an
Biodegradable Dispersants Therefor,” U.S. Patent No. Impermeable Cement System for Controlling Gas
6,019,835 (February 1, 2000). Migration,” paper SPE 11983 presented at the SPE
Annual Technical Conference and Exhibition, San
Chatterji, J. et al.: “Cement Compositions,” U.S. Patent
Francisco, California, USA (October 5–8, 1983).
No. 5,711,801 (January 27, 1998).
Childers, M.A.: “Primary Cementing of Multiple Casing,”
Chatterji, J. et al.: “Well Cementing Methods Using
Journal of Petroleum Technology (July 1968) 20, No. 7,
Compositions Containing Liquid Polymeric Additives,”
751–762; also presented as paper SPE 1919 at the SPE
U.S. Patent No. 6,268,406 (July 31, 2001).
Annual Meeting, Houston, Texas, USA (October 1–4,
Chatterji, J. and Borchardt, J.K.: “Applications of Water- 1967).
Soluble Polymers in the Oilfield,” paper SPE 9288 pre-
Childs, J., Sabins, F., and Taylor, M.J.: “Method of Using
sented at the SPE Annual Technical Conference and
Thixotropic Cements for Combating Lost Circulation,”
Exhibition, Dallas, TX, USA (September 21–24, 1980).
U.S. Patent No. 4,515,216 (May 7, 1985).
Chatterji, J., and Brake, B.G.: “Water Loss Reducing
Childs, J.D. et al.: “Well cementing methods and compo-
Additives for Salt Water Cement Slurries,” British Patent
sitions,” European Patent No. 177308 (1986).
No. GB 2 080 812 (February 10, 1982).

708 Well Cementing

Childs, J.D., Brothers, L.E., and Taylor, M.J.: “Method of (July 1973) 25, No. 7, 775–783; also presented as
Preparing a Light Weight Cement Composition from Sea SPE 4090 at the SPE Annual Meeting, San Antonio,
Water,” U.S. Patent No. 4,450,009 (May 22, 1984). Texas, USA (October 8–11, 1972).
Childs, J.D., Sutton, D.L., and Sabins, F.L.: “Set Delayed Clark, P.E., and Barkat, O.: “Analysis of fluid-loss data,”
Cement Compositions and Methods of Using the Same,” SPE Production Engineering (August 1990) 5, No. 3,
U.S. Patent No. 4,676,832 (June 30, 1987). 306–310.
Chiocchio, G., and Paolini, A.E.: “Optimum Time for Clarke, W.J., and McNally, A.C.: “Ultrafine Cement for
Adding Superplasticizers to Portland Cement Pastes,” Oilwell Cementing,” paper SPE 25868 presented at the
Cement and Concrete Research (September 1985) 15, SPE Rocky Mountain Regional/Low Permeability
No. 5, 901–908. Reservoirs Symposium, Denver, Colorado, USA (April
Chmilowski, W., and Kondratoff, L.B.: “Foamed cement 12–14, 1993).
for Squeeze Cementing Low-Pressure Highly Permeable Clarke, W.J.: “Grouting Composition Comprising Slag,”
Reservoirs: Design and Evaluation,” SPE Drilling U.S. Patent No. 4,761,183 (August 2, 1988).
Engineering (December 1992), 7, (4), 284–290; also Clement, C.C.: “A Scientific Approach to the Use of
presented as paper SPE 20425 at the SPE Annual Thixotropic Cement,” Journal of Petroleum Technology
Technical Conference and Exhibition, New Orleans, (March 1979) 31, No. 3, 344–346.
(September 23–26, 1990).
Cloud, W.B.: “Downhole temperature tool accurately
Cho, H., Shah, S.N., and Osisanya, S.O.: “Effects of Fluid measures well bore profile,” Oil & Gas Journal (July 20,
Flow in a Porous Cuttings-Bed on Cuttings Transport 1992) 90, No. 29, 101–104.
Efficiency and Hydraulics,” paper SPE 71374 presented
Cobb, J.A., and Pace, R.S.: “Elements Affecting the
at the SPE Annual Technical Conference and Exhibition,
Thickening Time of a Cement Blend,” paper SPE 14195
New Orleans, Louisiana, USA (September 30–October 3,
presented at the SPE Annual Technical Conference and
2001).
Exhibition, Las Vegas, Nevada, USA (September 22–25,
Chow, T.W. et al.: “The Rheological Properties of Cement 1985).
Slurries: Effect of Vibration, Hydration Conditions and
Cobb, S., Maki, V., and Sabins, F.: “Method predicts
Additives,” SPE Production Engineering (November
foamed-cement compressive strength under tempera-
1988) 3, No. 4, 543–550 (paper SPE 13936).
ture, pressure,” Oil & Gas Journal (April 15, 2002), 100
Christian, W.W., Chatterji, J., and Ostroot, G.W.: “Gas (15), 48–52.
Leakage in Primary Cementing—A Field Study and
Colavecchio, G.P., and Adamiak, R.: “Foamed Cement
Laboratory Investigation,” Journal of Petroleum
Achieves Predictable Annular Fill in Appalachian
Technology (November 1976) 28, No. 11, 1361–1369; also
Devonian Shale Wells,” paper SPE 17040 presented at the
presented as paper SPE 5517 at the SPE Annual
SPE Eastern Regional Meeting, Pittsburgh, Pennsylvania,
Technical Conference and Exhibition, Dallas, Texas, USA
USA (October 21–23, 1987).
(September 28–October 1, 1975).
Cole, R.C.: “Epoxy Sealant for Combating Well
Chu, C.: “State-of-the-Art Review of Fireflood Field
Corrosion,” paper SPE 7874 presented at the SPE
Projects,” paper SPE/U.S. Department of Energy (DOE)
International Symposium on Oilfield and Geothermal
9772 presented at the SPE/DOE Enhanced Oil Recovery
Chemistry, Houston, Texas, USA (January 22–24, 1979).
Symposium, Tulsa, Oklahoma, USA (April 5–8, 1981).
Collepardi, M. et al.: “Influence of Sulfonated Naphthalene
Chu, C.: “State-of-the-Art Review of Steamflood Field
on the Fluidity of Cement Pastes,” Proceedings of the
Projects,” Journal of Petroleum Technology (October
7th International Congress on the Chemistry of Cement
1985) 37, No. 10, 1887–1902; also presented as paper
(1980), 3, paper VI-20–VI-25.
SPE 11733 at the SPE California Regional Meeting,
Ventura, California, USA (March 23–25, 1983). Collepardi, M. et al.: “Tetracalcium aluminoferrite
hydration in the presence of lime and gypsum,” Cement
Ciach, T.D., and Swenson, E.G.: “Morphology and
and Concrete Research (July 1979) 9, No. 4, 431–437.
Microstructure of Hydrating Portland Cement and Its
Constituents II. Changes in Hydration of Calcium Collepardi, M. et al.: “Tricalcium aluminate hydration in the
Silicates Alone and in the Presence of Triethanolamine presence of lime, gypsum or sodium sulfate,” Cement and
and Calcium Lignosulfonate, Both With and Without Concrete Research (September 1978) 8, No. 5, 571–580.
Gypsum,” Cement and Concrete Research (March 1971) Collepardi, M., and Marchese, B.: “Morphology and
1, No. 2, 159–176. Surface Properties of Hydrated Tricalcium Silicate
Clark, C.R., and L.G. Carter: “Mud Displacement With Pastes,” Cement and Concrete Research (January 1972)
Cement Slurries,” Journal of Petroleum Technology 2, No. 1, 57–65.

References 709
Collepardi, M., and Massidda, L.: “Hydration of Beta Cowan, K.M., Hale, A.H., and Nahm, J.J.W.: “Water-Base
Dicalcium Silicate Alone and in the Presence of CaCl2 or Mud Conversion for High Temperature Cementing,” U.S.
C2H5OH,” Journal of the American Ceramic Society Patent No. 5,515,921 (June 14, 1996).
(April 1973) 56, No. 1, 181–183. Cowan, K.M., Nahm, J.J.W., and Hale, A.H.: “Method to
Collepardi, M.: “Il comportamento reologico delle paste Cement a Wellbore in the Presence of Carbon Dioxide,”
cementizie,” Il Cemento (1971) 68, No. 1, 99–106. U.S. Patent No. 5,307,876 (June 3, 1994a).
Collett, T.S., Lewis, R., and Uchida, T.: “Growing interest Cowan, M., and Sabins, F.: “New correlations improve
in Gas Hydrates,” Oilfield Review (Summer 2000) 12, temperature predictions for cementing and squeezing,”
No. 2, 42–57. Oil & Gas Journal (August 21, 1996) 93, No. 4, 53–57.
Constien, V.G. et al.: “Performance of Fracturing Crammond, N.J., and Currie, R.J.: “Survey of condition of
Materials,” Reservoir Stimulation (3rd ed.), M.J. precast high-alumina cement components in internal
Economides and K.G. Nolte (eds.), Chichester, England, locations in 14 existing buildings,” Magazine of Concrete
John Wiley & Sons (2000) 8-1–8-26. Research (1993) 43, No. 165, 275–279.
Cooke Jr., C.E., Kluck, M.P., and Medrano R.: “Field Crawshaw, J.P., and Frigaard, I.: “Cement plugs: stability
Measurements of Annular Pressure and Temperature and failure by buoyancy-driven mechanism,” paper
During Primary Cementing,” Journal of Petroleum SPE 56959 presented at the SPE Offshore Europe
Technology (August 1983) 35, No. 9, 74–83; also pre- Conference, Aberdeen, Scotland (September 7–10, 1999).
sented as paper SPE 11206 at the SPE Annual Technical Creel, P., and Crook, R.: “Foamed Cement Solves
Conference and Exhibition, New Orleans, Louisiana, Producing, Injection Problems,” Oil & Gas Journal
USA (September 26–29, 1982). (January 1, 1998) 96, No. 2, 41–45.
Cook, C., and Cunningham, W.C.: “Filtrate Control—A Creel, P., and Crook, R.: “Gels, Monomer Solutions Fix
Key in Successful Cementing Practices,” Journal of Pinhole Casing Leaks,” Oil & Gas Journal (October 13,
Petroleum Technology (August 1977) 29, No. 8, 951–956; 1997) 95, No. 41, 44–46.
also presented as paper SPE 5898 at the SPE Rocky
Crema, S.C., Kucera, C.H., and Konrad, G.: “New Fluid-Loss
Mountain Regional Meeting, Cody, Wyoming, USA
Additives for Oilfield Cementing,” paper SPE 18901 pre-
(May 11–12, 1978).
sented at the SPE Production Operations Symposium,
Costa, U., Massazza, F., and Barrilà, A.: “Adsorption of Oklahoma City, Oklahoma, USA (March 13–14, 1989).
Superplasticizers on C3S: Changes in Zeta Potential and
Crinklemeyer, O.W., Puntney, A.W., and Sharpe, J.R.:
Rheology of Pastes,” Il Cemento (1982) 79, No. 4, 323–336.
“Use of Water Base Spacer With Thixotropic Cement
Courtney, G. et al.: “New open-hole displacement proce- Systems Improves Cement Jobs,” paper SPE 6367 pre-
dure dramatically reduces loss of synthetic fluids in sented at the SPE Eastern Regional Meeting, Columbus,
deepwater GOM,” paper SPE/IADC 52813 presented at Ohio, USA (November 18–19, 1976).
the SPE/IADC Drilling Conference, Amsterdam, The
Crook, R.J., Keller, S.R., and Wilson, M.A.: “Deviated
Netherlands (March 9–11, 1999).
wellbore cementing: Part 2—Solutions,” Journal of
Couturier, M. et al.: “Design Rules and Associated Spacer Petroleum Technology (August 1987) 39, No. 8, 961–966;
Properties for Optimal Mud Removal in Eccentric also presented as paper SPE 14198 at the SPE Annual
Annuli,” paper Canadian Institute of Mining, Metallurgy Technical Conference and Exhibition, Las Vegas,
& Petroleum (CIM) 90-112/SPE 21594 presented at the Nevada, USA (September 22–25, 1985).
1990 International Joint Meeting, Calgary, Alberta,
Crump, D.K., and Wilson, D.A.: “Set Retarding Additives
Canada (June 10–13, 1990).
for Cement from Aminomethylenephosphonic Acid
Cowan, K.M., and Hale, A.H.: “Squeeze Cementing,” U.S. Derivatives,” U.S. Patent No. 4,468,252 (August 28, 1984).
Patent No. 5,322,124 (June 21, 1994).
Cumberland, D.J., and Crawford, R.J.: The Packing of
Cowan, K.M., Hale, A.H., and Nahm, J.J.: “Conversion Particles, Elsevier (1987).
of Drilling Fluids to Cements With Blast Furnace
Cunha, J.C. et al.: “Planning Extended Reach Wells for
Slag: Performance Properties and Applications for
Deep Water,” paper SPE 74400 presented at the SPE
Well Cementing,” paper SPE 24575 presented at the
International Petroleum Conference and Exhibition in
SPE Annual Technical Conference and Exhibition,
Mexico, Villahermosa, Mexico (February 10–12, 2002).
Washington, DC, USA (October 4–7, 1992).
Cunningham, W.C., and Smith, D.K.: “Effect of Salt
Cowan, K.M., Hale, A.H., and Nahm, J.J.W.: “Cement
Cement Filtrate on Subsurface Formations,” Journal of
Slurry Composition and Method to Cement Wellbore
Petroleum Technology (March 1968) 20, No. 3, 259–264;
Casings in Salt Formations,” U.S. Patent No. 5,309,999
(June 10, 1994b).

710 Well Cementing

also presented as paper SPE 1920 at the SPE Annual Technical Conference, Manama, Bahrain (March 9–12,
Meeting, Houston, Texas, USA (October 1–4, 1967). 1981).
Curtis, C. et al.: “Heavy-Oil Reservoirs,” Oilfield Review Davies, J.T., and Rideal, E.K.: Interfacial Phenomena,
(Autumn 2002) 14, No. 3, 30–51. New York, New York, USA, Academic Press (1963).
Cutforth, H.G.: “Low Water-Loss Cement Slurry and Davies, S.N. et al.: “Field Studies of Circulating
Method of Cementing a Well Therewith,” U.S. Patent Temperatures Under Cementing Conditions,” SPE
No. 2,598,675 (June 3, 1952). Drilling & Completion (March 1994) 9, No. 1, 12–16.
Cuvillier, G. et al.: “Solving Deepwater Well-Construction Davis, R.E. et al.: “Properties of Cements and Concretes
Problems,” Oilfield Review (Spring 2000) 12, No. 1, 2–17. Containing Fly Ash,” Journal of the American Concrete
Daccord, G., and Baret, J.F.: “How Fluid Loss Influences Institute (May–June 1937) 33, 5, 577–612.
Primary Cementing: Literature Review and Methodology,” Davison, J.M. et al.: “Oil-based muds for reservoir
SPE Drilling & Completion (June 1994) 9, No. 2, 133– drilling: their performance and cleanup characteristics,”
138; revised from paper OTC 25150 presented at the SPE Drilling & Completion (June 2001) 16, No. 2,
Offshore Technology Conference, Houston, Texas, USA 127–134; also presented as paper SPE 72063 at the SPE
(May 4–7, 1992). Formation Damage Control International Symposium,
Daccord, G., de Rozières, J., and Boussouira, B.: “Cement Lafayette, Louisiana, USA (February 23–24, 2000).
slurry behavior during hydration and consequences De Jong, J.G.M., Stein, H.N., and Stevels, J.M.:
for oil-well cementing,” Hydration and Setting of “Hydration of Tricalcium Silicate,” Journal of Applied
Cements—Proceedings of the International RILEM Chemistry (1967) 17, 246–250.
Workshop, A. Nona and J.C. Mutin (eds.), E&FN Spon, De Larrard, F., and Sedran, T.: “Optimization of Ultra-
London (1992) 277–286. High-Performance Concrete by the Use of Packing
Dahua, W. et al.: “Low Density Slag/Mud Mix Slurries Model,” Cement and Concrete Research (1994) 24, No. 6,
Applied in Changqing Oilfield for Sealing Long Annuli,” 997–1009.
paper SPE 50891 presented at the SPE International De Larrard, F.: “Ultrafine particles for the making of very
Oil & Gas Conference and Exhibition, Beijing, China high-strength concrete,” Cement and Concrete Research
(November 2–6, 1998). (March 1989) 19, No. 2, 161–172.
Daigle, C.L. et al.: “Expandable Tubulars: Field De Rozières, J., and Ferrière, R.: “Foamed-Cement
Examples of Application in Well Construction and Characterization Under Downhole Conditions and Its
Remediation,” paper SPE 62958 presented at the SPE Impact on Job Design,” SPE Production Engineering
Annual Technical Conference and Exhibition, Dallas, (August 1991) 6, No. 3, 297–304; also presented as paper
Texas, USA (October 14, 2000). SPE 19935 at the IADC/SPE Drilling Conference,
Damidot, D., Nonat, A., and Mutin C. eds.: Hydration Houston, Texas, USA (February 27–March 2, 1990).
and setting of cement, E&FN Spon, 1992, 23. De Rozières, J., and Sabins, F.: “Shrinkage and
Dao, B., and Vijn, J.P.: “Environmentally Acceptable Expansion of Oil Well Cements,” Activity report of the
Cement Fluid Loss Additive,” paper SPE 74088 presented API work group on shrinkage, API, Washington, DC,
at the SPE International Conference on Health, Safety and USA (February 1995).
Environment in Oil and Gas Exploration and Production, Dean, G.D., and Brennen, M.A.: “A Unique Laboratory
Kuala Lumpur, Malaysia (March 20–22, 2002). Gas Flow Model Reveals Insight to Predict Gas Migration
Darley, H.C.H., and Gray, G.R.: Composition and proper- in Cement,” paper SPE 24049 presented at the SPE
ties of drilling and completion fluids (5th ed.), Gulf Western Regional Meeting, Bakersfield, California, USA
Professional Publishing (Elsevier), Houston, Texas, USA (March 30–April 1, 1992).
(1988). Debacq, M. et al.: “Buoyant mixing of miscible fluids of
Daulton, D.J., et al.: “Field experience with application of varying viscosities in vertical tubes,” Physics of Fluids
blast furnace slag to the drilling and cementing program (December 2003) 15, No. 12, 3846–3885.
in the Stratton field, South Texas,” paper SPE 29472 Debacq, M. et al.: “Self-similar concentration profiles in
presented at the SPE Production Operations Symposium, buoyant mixing of miscible fluids in a vertical tube,”
Oklahoma City, Oklahoma, USA (April 2–4, 1995). Physics of Fluids (November 2001) 13, No. 1, 3097–3100.
Davies, D.R., Hartog, J.J., and Cobbett, J.S.: “Foamed Deeg, W. et al.: “How Foamed Cement Advantages
Cement—A Cement with Many Applications,” paper Extend to Hydraulic Fracturing Operations,” World Oil
SPE 9598 presented at the SPE Middle East Oil Magazine (November 1999) 220, No. 11, 51–53.

References 711
Dees, J.M., and Spradlin, W.N. Jr.: “Successful Deep Chemistry in the Oil Industry, Manchester, UK (April
Openhole Cement Plugs for the Anadarko Basin,” paper 19–20, 1988).
SPE 10957 presented at the SPE Annual Technical Desouky, S.E.M: “A New Laminar-Turbulent Transition
Conference and Exhibition, New Orleans, Louisiana, Criterion for Pseudoplastic Fluids,” Journal of Petroleum
USA (September 26–29, 1982). Science Engineering (April 1991) 5, No. 3, 285–291.
Defossé, C.: “Fluid Loss Additive for Cement,” French Detournay, E., and Cheng, A.: “Fundamentals of
Patent No. 2,540,098 (1985a). Poroelasticity,” Comprehensive Rock Engineering,
Defossé, C: “Cement Slurry Compositions for Cementing John A. Hudson (ed.), Oxford, England, Pergamon Press
Oil Wells, Adapted to Control Free Water and (1993) 2, 113–171.
Corresponding Cementing Process,” French Patent Detroit, W.J.: “Lignosulfonate Derivatives,” U.S. Patent
No. 2,540,097 (1985b). No. 4,219,471 (August 26, 1980).
Degni, C.D., and Marcinkevicius, C.R.: “Primary cementing Di Lullo Arias, G.F.: “Method of Using Construction Grade
optimization: well conditioning procedure and foamed Cement in Oil and Gas Wells,” U.S. Patent No. 5,547,024
spacer use,” paper SPE 53955 presented at the SPE (August 20, 1996).
Latin American and Caribbean Petroleum Engineering
Di Lullo, G., and Rae, P.: “Cements for Long Term
Conference, Caracas, Venezuela (April 21–23, 1999).
Isolation—Design Optimization by Computer Modelling
Degouy, D., and Martin, M.: “Characterization of and Prediction,” paper SPE 62745 presented at the
the Evolution of Cementing Materials After Aging IADC/SPE Asia Pacific Drilling Technology, Kuala
Under Severe Bottomhole Conditions,” SPE Drilling & Lumpur, Malaysia (September 11–13, 2000).
Completion (March 1993) 8, No. 1, 57–63; also present-
Diaz, H. et al.: “Modeling of ECD in Casing Drilling
ed as paper SPE 20904 at the SPE European Petroleum
Operations and Comparison with Experimental and
Conference, The Hague, The Netherlands (October
Field Data,” paper IADC/SPE 87149 presented at the
22–24, 1990).
IADC/SPE Drilling Conference, Dallas, Texas, USA.
Delhommer, H.J., and Walker, C.O.: “Encapsulated Oil (March 2–4, 2004).
Absorbent Polymers as Lost Circulation Additives for
Dick, M.A. et al.: “Optimizing the Selection of Bridging
Oil-base Drilling Fluids,” U.S. Patent No. 4,704,213
Particles for Reservoir Drilling Fluids,” paper SPE 58793
(November 3, 1987).
presented at the SPE International Symposium on
Demong, K., and Rivenbark, M.: “Planning the Well Formation Damage, Lafayette, Louisiana (February
Construction Process for the Use of Solid Expandable 23–24, 2000).
Casing,” paper SPE/IADC 85303 presented at the
Diggins, E.: “Classification provides framework for
SPE/IADC Middle East Drilling Technology Conference
ranking multilateral complexity and well type,” Oil &
& Exhibition, Abu Dhabi, UAE (October 20–22, 2003).
Gas Journal (December 29, 1997) 95, No. 52, 73–75.
Demong, K., Rivenbark, M., and Mason, D.:
Dillenbeck, R.L., and Cooper, D.: “Successful optimization
“Breakthroughs Using Solid Expandable Tubulars to
of primary cement designs enables annular placement
Construct Extended Reach Wells,” paper IADC/SPE
of casing perforating guns for multiple zone comple-
87209 presented at the IADC/SPE Drilling Conference,
tions,” paper SPE 64526 presented at the 2000 SPE Asia
Dallas, Texas, USA. (March 2–4, 2004).
Pacific Oil and Gas Conference and Exhibition,
Denis, J.H., and Guillot, D.J.: “Prediction of Cement Brisbane, Australia (October 16–18, 2000).
Slurry Laminar Pressure Drops by Rotational
Dillenbeck, R.L., and Nelson, S.G.: “New Clay-Controlling
Viscometry,” paper SPE 16137 presented at the
Technology Addresses Problems Encountered in Cement
SPE/IADC Drilling Conference, New Orleans, Louisiana,
Design and Performance,” paper SPE 24383 presented
USA (March 15–18, 1987).
at the SPE Rocky Mountain Regional Meeting, Casper,
Dent-Glasser, L.S. et al.: “A Multi-Method Study of C3S Wyoming, USA (May 18–21, 1992).
Hydration,” Cement and Concrete Research (November
Dillenbeck, R.L., and Simpson, S.: “New down-hole exter-
1978) 8, No. 6, 733–739.
nal casing mud removal technology improves primary
Dent-Glasser, L.S.: “Osmotic pressure and the swelling cement results,” paper SPE 53943 presented at the SPE
of gels,” Cement and Concrete Research (July 1979) 9, Latin American and Caribbean Petroleum Engineering
No. 4, 515–517. Conference, Caracas, Venezuela (April 21–23, 1999).
Desbrières, J.: “Influence of Polymeric Additives on Dillenbeck, R.L., and Smith, J.: “Highly Relaxed Fluid
Cement Filter Cake Permeability,” presented at the Royal Loss, Surfactant Enhanced Cement Improves Results on
Society of London Third International Symposium On Deep Gas Wells,” paper SPE 38599 presented at the SPE

712 Well Cementing

Annual Technical Conference and Exhibition, San SPE Annual Meeting, Houston, Texas, USA (October
Antonio, Texas, USA (October 5–8, 1997). 11–14, 1964).
Dillenbeck, R.L., Mueller, D.T., and Orr, B.R.: “The Effect Dusseault, M.B., Bruno, M.S., and Barrera, J.: “Casing
of Microsilica on the Thermal Stability of Lightweight Shear: Causes, Cases, Cures,” paper SPE 48864 presented
Cement Systems,” paper CIM/SPE 21597 presented at at the SPE International Conference and Exhibition,
the International Technical Meeting jointly hosted by Beijing, China (November 2–6, 1998).
CIM and SPE, Calgary, Alberta, Canada (June 10–13, Dusseault, M.B., Gray, M., and Nawrocki, P.: “Why
1990). Oilwells Leak: Cement Behavior and Long-Term
Dillenbeck, R.L., Strickland, B.D., and Orr, B.R.: “KCl may Consequences,” paper SPE 64733 presented at the SPE
in some cases prevent good casing cement jobs,” Oil & Gas International Oil and Gas Conference and Exhibition,
Journal (November 4, 1991) 89, No. 4, 71–75. Beijing, China (November 7–10, 2000).
Dodge, D.W., and Metzner, A.B.: “Turbulent Flow of Earlougher, R.C.: “Some Practical Considerations in the
Non-Newtonian Systems,” AIChE Journal (June 1959) Design of Steam Injection Wells,” Journal of Petroleum
5, No. 2, 189–204. Technology (January 1969) 21, No. 1, 79–86; also
Donche, A., and Isambourg, P.: “Scleroglucan-based presented as paper SPE 2202 at the SPE Annual
compositions and their use as a cementation spacer,” California Fall Meeting, Bakersfield, California, USA
U.S. Patent No. 5,306,340 (April 26, 1994). (November 7–8, 1968).
Doron, P., and Barnea, D.: “A three-layer model for Eaton, M.L., Barry, A.G., and Schoemaker, S.: “New
solid-liquid flow in horizontal pipes,” International Workover and Completion Technology Utilized in Bass
Journal of Multiphase Flow (December 1993) 19, No. 6, Strait,” paper SPE 64400 presented at the SPE Asia
1029–1043. Pacific Oil and Gas Conference and Exhibition,
Brisbane, Australia (October 16–18, 2000).
Double, D.D., Hellawall, A., and Perry, S.J.: “The
Hydration of Portland Cement,” Proceedings of the Eberhard, J.F., and Park, A.: “Portland Cement-
Royal Society of London, London, England, Royal Vinylidene Chloride Polymer Composition, Method of
Society of London (1978) A.359, 435–451. Making, and Method of Using,” U.S. Patent No. 2,819,239
(January 7, 1958).
Double, D.D.: “New Developments in Understanding
the Chemistry of Cement Hydration,” Philosophical Economides, M.J., and Nolte, K.: Reservoir Stimulation
Transactions of the Royal Society of London, London, (3rd ed.), Chichester, England, UK, John Wiley & Sons
(1983), paper A310, 53–66. (2000).
Downton, G. et al.: “New directions in rotary steerable Economides, M.J. et al.: “Horizontal, Multilateral, and
drilling,” Oilfield Review (Spring 2000) 12, No. 1, 18–29. Multibranch Wells in Petroleum Production Engineering,”
Petroleum Well Construction, M.J. Economides and K.G.
Drecq, P., and Parcevaux, P.A.: “A Single Technique
Nolte (eds.), New York, New York, USA, John Wiley &
Solves Gas Migration Problems Across a Wide Range of
Sons (1998).
Conditions,” paper SPE 17629 presented at the SPE
International Meeting on Petroleum Engineering, Economides, M.J., Watters, L.T., and Dunn-Norman, S.
Tianjin, China (November 1–4, 1988). (eds.).: Petroleum Well Construction, New York, New
York, USA, John Wiley & Sons (1998).
Drochon, B., and Maroy, P.: “Permeable Cement
Composition for Oil and Gas Wells,” French Patent Economides, M.J.: “Implications of Cementing on
No. 2808794 (2000). Well Performance,” Well Cementing (1st ed.), E.B.
Nelson (ed.), Amsterdam, The Netherlands, Elsevier
Duchane, D.V.: “Geothermal Energy,” Kirk-Othmer
(1990).
Encyclopedia of Chemical Technology (1994) Vol. 12,
512–539. Edmondson, T.D., and Benge, O.G.: “A Case Study of
Ultralightweight Cementing Practices in the Northeastern
Dumont, A., Patin, J., and Le Floch, G.: “A Single Tool for
United States,” paper SPE 12318 presented at the SPE
Corrosion and Cement Evaluation,” paper SPE 13140
Eastern Regional Meeting, Champion, Pennsylvania,
presented at the SPE Annual Technical Conference and
USA (November 9–11, 1983).
Exhibition, Houston, Texas, USA (September 16–19,
1984). Edwards, G.C., and Angstadt, R.L.: “The Effect of Some
Soluble Inorganic Admixtures on the Early Hydration of
Dumore, J.M.: “Stability Considerations in Downward
Portland Cement,” Journal of Applied Chemistry (May
Miscible Displacements,” SPE Journal (December
1966) 16, No. 5, 166–168.
1964) 4, No. 4, 356–362; also presented as SPE 961 at the

References 713
Eggemeyer, J.C. et al.: “SubSea MudLift Drilling: Design Elphingstone, E.A., McLaughlin, H.C., and Smith, C.W.:
and Implementation of a Dual Gradient Drilling “Temperature Gelation Activated Aqueous Silicate
System,” paper SPE 71359 presented at the SPE Annual Mixtures and Process of Forming Impermeable Gels,”
Technical Conference and Exhibition, New Orleans, U.S. Patent No. 4,293,440 (October 6, 1981).
Louisiana, USA (September 30–October 3, 2001). Eoff, L.S. et al.: “Methods of sealing pipe strings in dispos-
Ehlig-Economides, C.A., and Ayoub, J.A.: “Vertical al wells,” U.S. Patent No. 6,321,841 (November 21, 2001).
Interference Testing Across a Low-Permeability Zone,” Eoff, L.S., and Loughridge, B.W.: “Method For Control
SPE Formation Evaluation (October 1986) 1, No. 5, Of Gas Migration In Well Cementing,” U.S. Patent
497–510; also presented as paper SPE 13251 at the SPE No. 5,339,903 (August 23, 1994).
Annual Technical Conference and Exhibition, Houston,
Erbil, M.M.: “The role of Taylor vortices in drilling,” MS
Texas, USA (September 16–9, 1984).
thesis, University of Texas, Austin, Texas, USA (1995).
Eilers, L.H., and Nelson, E.B.: “Effect of Silica Particle
Evanoff, J.I., and Cook, C.: “Optimizing Cement Design
Size on Degradation of Silica Stabilized Portland
for Improved Job Results,” paper SPE 17441 presented
Cement,” paper SPE 7875 presented at the SPE
at the SPE California Regional Meeting, Long Beach,
International Symposium on Oilfield and Geothermal
California, USA (March 23–25, 1988).
Chemistry, Houston, Texas, USA (January 22–24, 1979).
Evans, G.W., and Carter, L.G.: “Bonding Studies of
Eilers, L.H., and Root, R.L.: “Long-Term Effects of High
Cementing Compositions to Pipe and Formation,” API
Temperature on Strength Retrogression of Cements,”
Drilling and Production Practice (1961), 72–79.
paper SPE 5028 presented at the SPE Annual Meeting,
Houston, Texas, USA (October 6–9, 1974). Evans, H.P.: “La tecnica efectiva para cementacion pri-
maria en pozos de gas,” Cuarto Congreso Latinoamericano
Eilers, L.H., Nelson, E.B., and Moran, L.K.: “High-
de Perforacion, Caracas, Venezuela (October 1–5, 1984).
Temperature Cement Compositions—Pectolite, Scawtite,
Truscottite, or Xonotlite: Which Do You Want?” Journal Every, R.L., and Jacob, J.T.: “Production of Raw Mix
of Petroleum Technology (July 1983) 35, No. 8, Cement Slurries Having Reduced Water Content,” U.S.
1373–1377 (SPE 9286). Patent No. 4,115,139 (September 19, 1978).
Eilers, L.H.: “Process for Cementing Geothermal Wells,” Ewert, D.P., Almond, S.W., and Bierhaus, W.M.: “Small-
U.S. Patent No. 4,556,109 (December 3, 1985). Particle-Size Cement,” SPE Production Engineering
(May 1991) 6, No. 2, 213–216; also presented as paper
Eldin, N.N. et al.: “Rubber-Tire Particles as Concrete
SPE 20038 at the SPE California Regional Meeting,
Aggregate,” Journal of Materials in Civil Engineering
Ventura, California (April 4–6, 1991).
(1993) 5, No. 4, 478–496.
Farkas, R.F. et al.: “New Cementing Technology Cures
El-Hassan, H.I. et al.: “Using a Novel Fiber Cement
40-Year-Old Squeeze Problems,” paper SPE 56537 pre-
System to Control Lost Circulation: Case Histories from
sented at the SPE Annual Technical Conference and
the Middle East and the Far East,” paper SPE 85324
Exhibition, Houston, Texas, USA (October 3–6, 1999).
presented at the SPE/IADC Middle East Drilling
Technology Conference & Exhibition, Abu Dhabi, UAE Farouq Ali, S.M., and Meldau, R.F.: “Current Steamflood
(October 20–22, 2003). Technology,” Journal of Petroleum Technology (October
1979) 13, 1332–1342 (SPE 7183).
El-Jazairi, B., and Illston, J.M.: “A Simultaneous Semi-
Isothermal Method of Thermogravimetry and Derivative Faul, R. et al.: “Fine-Grind Cement Aids GOM (Gulf of
Thermogravimetry, and its Application to Cement Mexico) Plug-and-Abandon Operations,” Proceedings of
Pastes,” Cement and Concrete Research (May 1977) 7, the First Annual SPE Offshore Technology Conference
No. 3, 247–257. (1999) 3, 311–315; also paper OTC-108963 presented at
the Offshore Technology Conference, Houston, Texas,
Ellis, L. et al.: “Mud gas isotope logging assists in oil and
USA (May 3, 1999).
gas drilling operations,” Oil & Gas Journal (May 26,
2003), 101 (21), 32–41. Fernando, H.J.S.: “Turbulent mixing in stratified fluids,”
Annual Review of Fluid Mechanics (January 1991) 23,
Elmoniem, H.A., Zaki, S., and Al-Arda, H.: “Cementing
455–493.
the Deepest 20-in Casing in Abu Dhabi Using a
Combination of Noval Lightweight Slurry and Fiber,” Fertl, W.H., Pilkington, P.E., and Scott, J.B.: “A Look at
paper ADIPEC-0940 presented at the 9th Abu Dhabi Cement Bond Logs,” Journal of Petroleum Technology
International Petroleum Exhibition and Conference, (June 1974) 26, No. 6, 607–617; also presented as paper
Abu Dhabi, UAE (October 15–18, 2000). SPE 4512 at the SPE Annual Meeting, Las Vegas, Nevada
(September 30–October 3, 1973).

714 Well Cementing

Finlayson, C.G. et al.: “Results of Field Testing the Cement Fontenot, K. et al.: “Casing Drilling Activity Expands
Evaluation Tool,” Transactions of the SPWLA 25th in South Texas,” paper SPE 79862 presented at the
Annual Logging Symposium, New Orleans, Louisiana, SPE / IADC Drilling Conference, Amsterdam, The
USA (June 10–13, 1984), paper H. Netherlands (February 19–21, 2003).
Fitzgerald, D.D. et al.: “Guidelines for 90% Accuracy Ford, R.E. et al.: “Obtaining Quality Primary Cement
in Zone Isolation Decision,” Journal of Petroleum Jobs in the Williston Basin,” paper SPE 10874 presented
Technology (November 1985); also presented as paper at the SPE Rocky Mountain Regional Meeting, Billings,
SPE 12141 at the SPE Annual Technical Conference Montana, USA (May 19–21, 1982).
and Exhibition, San Francisco, California, USA Fordham, E.J. et al.: “Dynamic filtration of bentonite
(October 5–8, 1983). muds under different flow conditions,” paper SPE 18038
Flatt, R.J., and Bowen, P.: “Electrostatic Repulsion presented at the SPE Annual Technical Conference and
Between Particles in Cement Suspensions: Domain of Exhibition, Houston, Texas, USA (October 2–5, 1988).
Validity of Linearized Poisson-Boltzmann Equation for Fordham, E.J., and Ladva, H.K.J.: “Cross-flow filtration
Nonideal Electrolytes,” Cement and Concrete Research of bentonite suspensions II,” Journal of Colloid and
(June 2003) 33, No. 6, 781–791. Interface Science (January 1992) 148, No. 1, 29–34.
Flatt, R.J., and Houst, Y.F.: “A Simplified View on Chemical Fordham, E.J., Bittleston, S.H., and Tehrani, M.A.:
Effects Perturbing the Action of Superplasticizers,” “Viscoplastic Flow in Centered Annuli, Pipes, and Slots,”
Cement and Concrete Research (August 2001) 31, No. 8, I & EC Research (1991) 30, No. 3, 517–524.
1169–1176.
Frank, U. et al.: “Shore-based ERD program taps offshore
Fleckenstein, W., Eustes, A., and Miller, G.: “Burst salt dome—Mittelplate field protected by sensitive
Induced Stresses in Cemented Wellbores,” paper SPE equipment,” Offshore International (September 1998) 58,
62596 presented at the SPE/AAPG Western Regional No. 9, 110–111, 113, 162, 186.
Meeting, Long Beach, California, USA (June 19–23,
Fraser, L. et al.: “Seamless Fluids Programs: A Key to
2000).
Better Well Construction,” Oilfield Review (Summer
Fleming, J.J., Boyd, S.R., and Wolverton, D.M.: 1996) 8, No. 2, 42–56.
“Gravel-pack Isolation Squeeze for Plug-down
Fraser, L.J., and Aragão, A.F.L.: “Successful Application
Recompletion,” paper SPE 63231 presented at the SPE
of MMH Fluid to Drill in Narrow Pressure Window, Ultra
Annual Technical Conference and Exhibition, Dallas,
Deepwater Situation. A Case History,” paper SPE 67734
Texas, USA (October 1–4, 2000).
presented at the SPE/IADC Drilling Conference,
Fletcher, P. et al.: “Predicting Quality and Performance Amsterdam, The Netherlands (February 27–March 1,
of Oilfield Cements with Artificial Neural Networks and 2001).
FTIR Spectroscopy,” Journal of Petroleum Technology
Fredrickson, A.G., and Bird, R.B.: “Non-Newtonian Flow
(February 1995) 47, No. 2, 129–130 (paper SPE 30079).
in Annuli,” Industrial & Engineering Chemistry
Flumerfelt, R.W.: “An Analytical Study of Laminar Non- (March 1958) 50, No. 3, 347–352.
Newtonian Displacement,” SPE Journal (April 1975)
Frey, M., de Capitani, C., and Liou, J.G.: “A New
15, No. 2, 169–180; also presented as paper SPE 4486 at
Petrogenetic Grid for Low-Grade Metabasites,” Journal
the SPE Annual Meeting, Las Vegas, Nevada, USA
of Metamorphic Geolology (1991) 9, 497–509.
(September 30–October 3, 1973).
Frigaard, I.A., Allouche, M., and Gabard-Cuoq, C.:
Folmar, L.W.: “Methods of Detecting Top of Cement
“Setting Rheological Targets for Chemical Solutions
Behind Casing,” Oil-Well Cementing Practices in the
in Mud Removal and Cement Slurry Design,” paper
United States 3, Washington, DC, USA, American
SPE 64998 presented at the SPE International
Petroleum Institute (1959), 132–139.
Symposium on Oilfield Chemistry, Houston, Texas, USA
Fontenot, E.P.: “Drilling a 26-in. Diameter Hole to 12,550 (February 13–16, 2001).
ft: A Case History,” paper SPE 15365 presented at the
Frigione, G., and Marra, S.: “Relationship Between
SPE Annual Technical Conference and Exhibition, New
Particle Size Distribution and Compressive Strength in
Orleans, Louisiana, USA (October 5–8, 1986).
Portland Cement,” Cement and Concrete Research
Fontenot, J.E., and Clark, R.K.: “An Improved Method for (January 1976) 6, No. 1 (113–127).
Calculating Swab and Surge Pressure and Calculating
Frisch, G.J., Graham, L., and Griffith, J.: “A Novel and
Pressures in a Drilling Well,” SPE Journal (October
Economical Processing Technique using Conventional
1974) 451–462 (SPE 4521).
Bond Logs and Ultrasonic Tools for Enhanced Cement

References 715
Evaluation,” Transactions of the SPWLA 41st Annual Garboczi, E.J., and Bentz, D.P.: “Fundamental Computer
Logging Symposium, Dallas (June 4–7, 2000), paper EE. Simulation Models for Cement-Based Materials,”
Frisch, G.J., Graham, W.L., and Griffith, J.: “Assessment Materials Science of Concrete II, J. Skalny and S.
of Foamed-Cement Slurries Using Conventional Cement Mindess (eds.), Westerville, Ohio, USA, The American
Evaluation Methods and Improved Interpretation Ceramic Society, Inc. (1991), 249–277.
Methods,” paper SPE 55649 presented at the SPE Rocky Garcia Jr., H., Maidla, E.E., and Onan, D.L.: “Annular
Mountain Regional Meeting, Gillette, Wyoming, USA Pressure Prediction Throughout Foam Cement
(May 15–18, 1999). Operations,” paper SPE 25439 presented at the SPE
Froelich, B. et al.: “Cement Evaluation Tool: A New Production Operations Symposium, Oklahoma City,
Approach to Cement Evaluation,” Journal of Petroleum Oklahoma, USA (March 21–23, 1993).
Technology (August 1982) 34, No. 8, 1835–1841; also pre- Garcia, J.A. and Clark, C.R.: “An Investigation of Annular
sented as paper SPE 10207 at the SPE Annual Technical Gas Flow Following Cementing Operations,” paper SPE
Conference and Exhibition, San Antonio, Texas, USA 5701 presented at the SPE Symposium on Formation
(October 5–7, 1981). Damage Control, Houston, Texas, USA (January 29–30,
Fry, S.E. et al.: “Method of Reducing Fluid Loss in Cement 1976).
Compositions Which May Contain Substantial Salt Garrison, A.D.: “Surface chemistry of shales and clays,”
Concentrations,” U.S. Patent No. 4,676,317 (June 30, 1987). Trans. AIME (1939) 132, 423–436.
Fuller, M., Littler, M., and Pollock, I.: “Innovative Way Gates, C.F., and Holmes, B.G.: “Thermal Well Completions
to Cement a Liner Utilizing a New Inner String Liner and Operation,” Proceedings of the 7th World Petroleum
Cementing Process,” paper IADC/SPE 39349 presented Congress (1967) PD-11, No. 6, 419–429.
at the IADC/SPE Drilling Conference, Dallas, Texas, USA Gatlin, C., and Nemir, C.: “Some Effects of Size
(March 3–6, 1998). Distribution on Particle Bridging in Lost Circulation and
Gabard, C.: “Déplacement de fluids miscibles non- Filtration Tests,” Journal of Petroleum Technology
Newtoniens en conduite cylindrique verticale,” PhD the- (June 1961) 13, No. 6, 575–578.
sis, Paris VI University, Paris (2000). Gaynor, T.M. et al.: “Tortuosity versus micro-tortuosity—
Gaddis, P.G.: “Method of Making High Viscosity Aqueous Why little things mean a lot,” paper SPE/IADC 67818
Mediums,” U.S. Patent No. 3,909,421 (September 30, 1975). presented at the SPE/IADC Drilling Conference,
Gai, H., and Lockyear, C.: “Cement Bond Logs—A New Amsterdam, The Netherlands (February 27–March 1,
Analysis to Improve Reliability,” SPE Applied 2001).
Technology Series (March 1994) 2, No. 1, 34–42; also Geiker, M., and Knudsen, T.: “Chemical Shrinkage of
presented as paper SPE 23729 at the SPE Latin Portland Cement Pastes,” Cement and Concrete
American Petroleum Engineering Conference, Caracas, Research (September 1982) 12, No. 5, 603–610.
Venezuela (March 8–11, 1992). George, C.R., and Gerke, R.R.: “Oil Field Cements,”
Galloway, G.: “Drilling With Casing Ready For Deepwater European Patent No. 85303475 (1985).
Application,” Drilling Contractor (July–August 2003) Geothermal Education Office: “Geothermal Energy—
59, No. 4, 38–39. A Global View,” brochure (2001).
Gallus, J.P., Pyle, D.E., and Moran, L.K.: “Physical and Gerber, A.H.: “Admixture for hydraulic cement,” U.S.
Chemical Properties of Cement Exposed to Geothermal Patent No. 4,473,405 (September 25, 1984).
Dry Steam,” paper SPE 7876 presented at the SPE
Gerke, R.R., et al.: “A Study of Bulk Cement Handling
International Symposium on Oilfield and Geothermal
and Testing Procedures,” SPE Production Engineering
Chemistry, Houston, Texas, USA (January 22–24, 1979).
(November 1990) 5, No. 4, 425–432; also presented as
Gallus, J.P., Pyle, D.E., and Watters, L.T.: “Performance paper SPE 14196 at the SPE Annual Technical
of Oil-Well Cementing Compositions in Geothermal Conference and Exhibition, Las Vegas, Nevada, USA
Wells,” paper SPE 7591 presented at the SPE Annual (September 22–25, 1985).
Technical Conference and Exhibition, Houston, Texas,
Ghofrani, R., and H. Plack: “CaO- and/or MgO-Swelling
USA (October 1–3, 1978).
Cements: A Key for Providing a Better Annular Sealing?”
Gambino, F. E. et al.: “Experimental Study of Fluid/Rock paper SPE 25697 presented at the SPE/IADC Drilling
interaction Caused by Drilling and Cementing Filtrates Conference, Amsterdam, The Netherlands (February
in Carito Field,” paper SPE 65004 presented at the 23–25, 1993).
SPE International Symposium on Oilfield Chemistry,
Ghosh, K. et al.: “Horizontal Drilling Experience in the
Houston, Texas, USA (February 13–16, 2001).
Panna Field,” paper SPE 57564 presented at the

716 Well Cementing

SPE/IADC Middle East Drilling Technology Conference, presented as paper SPE 2473 at the SPE Permian Basin
Abu Dhabi, UAE (November 8–10, 1999). Oil Recovery Conference, Midland, Texas, USA (May
Ghosh, S.N. (ed.): Advances in Cement Technology, 8–9, 1969).
Oxford, England, Pergamon Press Ltd. (1983). Gouda, V.K., Mourad, W.E., and Mikhail, R. Sh.:
Gibson, D.L., and Kucera, C.H.: “Low Water Loss Aqueous “Additives to Cement Pastes: Simultaneous Effects on
Cement Slurry and Method of Use,” U.S. Patent No. Pore Structure and Corrosion of Steel Reinforcement,”
3,491,049 (January 20, 1970). Journal of Colloid and Interface Science (May 1973) 43,
No. 2, 293–302.
Gilmour, A., Cheung, R., and Bui, A.: “Removal of well-
bore residues,” U.S. Patent No. 6,534,449 (March 18, Govier, G.W., and Aziz, K.: The Flow of Complex Mixtures
2003). in Pipes (1st ed.), Malabar, Florida, USA, R.E. Krieger
Publishing Company (1977).
Glasser, F.P.: “Barrier Performance of Cements and
Concretes in Nuclear Waste Repositories,” Report EUR Grabowski, E., and Gillott, J.E.: “Effect of Replacement
19780 EN, Luxembourg, Commission of the European of Silica Flour with Silica Fume on Engineering
Communities (2001). Properties of Oilwell Cements at Normal and Elevated
Temperatures and Pressures,” Cement and Concrete
Glasser, F.P.: “The Burning of Portland Cement,”
Research (May 1989) 19, No. 3, 333–344.
Lea’s Chemistry of Cement and Concrete (4th ed.),
P.C. Hewlett (ed.), Oxford, England, Butterworth- Grachyov, V.V., and Leonov, E.G.: “Study of Pore and
Heinemann (2001). Skeletal Pressure of Cement Slurry Column During the
Period of Setting,” Burenie, No. 3, pp. 17–21 (1969),
Goins, W.C. Jr.: “How to Combat Circulation Loss,” Oil &
translated from Russian.
Gas Journal (June 9, 1952) 55, No. 5, 71–92.
Graham, H.L.: “Rheology Balanced Cementing Improves
Golapudi, U.K., Ramakrishnan, V., and Islam, M.R.:
Primary Success,” Oil & Gas Journal (December 18,
“Improvement of Properties and Microstructure of
1972) 70, No. 51, 53–60.
Oil-Well Cements,” paper SPE 26312 available from
SPE, Richardson, Texas, USA (1993). Graham, W.L. et al.: “Cements Evaluation and Casing
Inspection with Advanced Ultrasonic Scanning,” paper
Gollwitzer, L.H., and Masson, J.P.: “The Cement Bond
SPE 38651 presented at SPE Annual Technical
Tool,” Transactions of the SPWLA 23rd Annual Logging
Conference and Exhibition, San Antonio, Texas, USA
Symposium, Corpus Christi, Texas, USA (July 6–9,
(October 5–8, 1997).
1982), paper Y.
Graham, W.L. et al.: “Enhanced Statistical Variation
Goodwin, K.J., and Crook, R.J.: “Cement Sheath Stress
Method for Determining Zone Isolation in Complex
Failure,” SPE Drilling Engineering (December 1992) 7,
Cements,” paper SPE 28772 presented at the SPE Asia-
No. 4, 291–296; also presented as paper SPE 20453 at the
Pacific Oil and Gas Conference, Melbourne, Australia
SPE Annual Technical Conference and Exhibition, New
(November 7–10, 1994).
Orleans, Louisiana, USA (September 23–26).
Granberry, V.L., Grant, W.H., and Clarke, J.W.:
Goodwin, K.J., and Phipps, K.: “Salt-Free Cement—An
“Monitoring Blended Cement Quality and Design With a
Alternative to Collapsed Casing in Plastic Salts,”
Mobile Cement Testing Laboratory,” Journal of Petroleum
Journal of Petroleum Technology (February 1984) 36,
Technology (October 1989) 41, No. 10, 1032–1079; also
No. 2, 320–324; also presented as paper SPE 10885 at
presented as paper SPE 17179 at the IADC/SPE Drilling
the SPE Rocky Mountain Regional Meeting, Billings,
Conference, Dallas, Texas, USA (February 28–March 2,
Montana, USA (May 19–21, 1982).
1988).
Goodwin, K.J.: “Principles of Squeeze Cementing,” paper
Grant, W.H. et al.: “Successful Squeezing of Shallow
SPE 12603 presented at the Permian Basin Oil and Gas
and Low-Pressure Formations,” paper SPE/IADC 19937
Recovery Conference, Midland, Texas, USA (March 8–9,
presented at the IADC/SPE Drilling Conference, Houston,
1984).
Texas, USA (February 27–March 2, 1990).
Goodwin, K.J.: “Guidelines for Cement Sheath Evaluation,”
Grant, W.H., and White, R.L.: “The Role of Filtration
SPE Production Engineering (August 1992) 7, No. 3,
Control in Squeeze Cementing,” Chevron Technical
280–284; also presented as paper SPE 19538 at the SPE
Memorandum No. 87-01 (1987).
Annual Technical Conference and Exhibition, San
Antonio, Texas, USA (October 8–11, 1989). Grant, W.H., Rutledge, J.M., and Gardner, C.A.: “Quality
of Bentonite and Its Ef fect on Cement-Slurry
Goolsby, J.L.: “A Proven Squeeze-Cementing Technique
Performance,” SPE Production Engineering (November
in a Dolomite Reservoir,” Journal of Petroleum
1990) 5, No. 4, 411–414; also presented as paper SPE
Technology (October 1969) 21, No. 10, 1341–1346; also

References 717
19940 at the IADC/SPE Drilling Conference, Houston, Guillot, F., Boisnault, J.M., and Hujeux, J.C.: “A
Texas, USA (February 27–March 2, 1990). Cementing Temperature Simulator to Improve Field
Graves, K.S.: “Planning Would Boost Liner Cementing Practices,” paper SPE 25436 presented at the Production
Success,” Oil & Gas Journal (April 15, 1985) 83, No. 15, Operations Symposium, Oklahoma City, Oklahoma,
47–52. USA (March 21–23, 1993).
Graves, W. Gordon, Ott, William K., and Woods, Joe D.: Gulrajani, S.N., and Nolte, K.G.: “Fracture Evaluation
Mature Oil and Gas Wells Downhole Remediation Using Pressure Diagnostics,” Reservoir Stimulation (3rd
Handbook, Houston, Texas, USA, World Oil Magazine ed.), M.J. Economides and K.G. Nolte (eds.), Chichester,
Division of Gulf Publishing Company (2004). England, John Wiley & Sons (2000) 9-1–9-63.
Greer, F.C., and Shryock, S.H.: “New Technique Improves Guyvoronsky, A.A., and Farukshin, L.K.: “Hydrostatic
Steam Stimulation Completions,” paper SPE 1944 Pressure of Cement Slurry,” Neftyanik (October 1963)
presented at the Annual California Regional Fall 10, 30–32, translated from Russian.
Meeting, Los Angeles, California, USA (October 26–27, Haberman, J.P. et al.: “Downhole Fluid Loss Measurements
1967). from Drilling Fluid and Cement Slurries,” Journal of
Greminger, G.K.: “Hydraulic Cement Compositions for Petroleum Technology (August 1992) 44, No. 8, 872–879;
Wells,” U.S. Patent No. 2,844,480 (July 22, 1958). also presented as paper SPE 22552 at the SPE Annual
Technical Conference and Exhibition, Dallas, Texas, USA
Griffin, T.J., and Valkó, P.P.: “Prevent poor cementing
(October 6–9, 1991).
results by the use of bottom cementing plugs,”
Proceedings of the 1997 Southwestern Petroleum Short Haberman, J.P., and Wolhart, S.P.: “Reciprocating
Course, Lubbock, Texas, USA, Southwestern Petroleum Cement Slurries After Placement by Applying Pressure
(1997) 276–290. Pulses in the Annulus,” paper SPE 37619 presented
at the SPE/IADC Drilling Conference, Amsterdam, The
Griffin, T.J., Spangle, L.B., and Nelson, E.B.: “New
Netherlands (March 6, 1997).
Expanding Cement Promotes Better Bonding,” Oil &
Gas Journal (June 25, 1979) 77, No. 26, 143–151. Haciislamoglu, M., and Cartalos, U.: “Practical Pressure
Loss Predictions in Realistic Annular Geometries,”
Griffith, J.E., and Ravi, K.M.: “Monitoring circulatable
paper SPE 28304 presented at the SPE Annual Technical
hole with real-time correction: case histories,” paper
Conference and Exhibition, New Orleans, Louisiana,
SPE 29470 presented at the Production Operations
USA (September 25–28, 1994).
Symposium held in Oklahoma City, Oklahoma, USA
(April 2–4, 1995). Haciislamoglu, M., and Langlinais, J.: “Non-Newtonian
Flow in Eccentric Annuli,” Journal of Energy Resources
Grinrod, M., Vassoy, B., and Dingsoyr, E.O.: “Development
Technology (September 1990) 112, No. 3, 163–169.
and Use of a Gas-Tight Cement,” paper SPE 17258 pre-
sented at the IADC/SPE Drilling Conference, Dallas, Haimoni, A., and Hannant, D.J.: “Developments in the
Texas, USA (February 28–March 2, 1988). Shear Vane Test to Measure Gel Strength of Oilwell
Cement Slurry,” Advances in Cement Research
Grosmangin, M. et al.: “A Sonic Method for Analyzing the
(October 1988) 1, No. 4, 221–229.
Quality of Cementation of Borehole Casings,” Journal
of Petroleum Technology (February 1961) 13, No. 2, Haimoni, A.M.: “Rheology of a Specific Oilwell Cement,”
165–71. MS thesis, Surrey University, Guildford, England
(December 1987).
Growcock, F.B. et al.: “Alternative Aphron-Based Drilling
Fluid,” paper SPE 87134 presented at the IADC/SPE Hale, A.H., and Cowan, K.M.: “Solidification of Water Based
Drilling Conference, Dallas, Texas, USA (March 2–4, Muds,” U.S. Patent No. 5,058,679 (October 22, 1991).
2004). Hale, A.H., and Cowan, K.M.: “Solidification of Water Based
Guillot, D.J., Parcevaux, P., and Jennings, D.B.: “Aqueous Muds,” U.S. Patent No. 5,673,753 (October 7, 1997).
Composition for Universal Spacer and Its Use in the Hale, A.H.: “Drilling and Cementing with Glycoside-Blast
Field of Drilling Wells, Notably Oil and Gas wells,” Furnace Slag-Drilling Fluid,” U.S. Patent No. 5,479,987
European Patent No. 273,471 (1986). (January 2, 1996).
Guillot, D.J., and Denis, J.H.: “Prediction of Laminar and Hale, B.W.: “Well-Cementing Method Using Low
Turbulent Friction Pressures of Cement Slurries in Pipes Fluid-Loss Cement Slurry,” U.S. Patent No. 4,258,790
and Centered Annuli,” paper SPE 18377 presented at the (March 31, 1981).
European Petroleum Conference, London, England Hall, C., and Hughes, T.L.: “Ion chromatography tracer
(October 16–19, 1988). experiments during drilling,” paper SPE 25178 present-

718 Well Cementing

ed at the SPE International Symposium on Oilfield Harestad, K. et al.: “New Tool Improves Success Ratio
Chemistry, New Orleans, Louisiana, USA (March 2–5, for Balanced Plug Cementing,” paper IADC/SPE 35084
1993). presented at the IADC/SPE Drilling Conference, New
Handin, J.: “Strength of Oil Well Cements at Downhole Orleans, Louisiana, USA (March 12–15,1996).
Pressure-Temperature Conditions,” Journal of Petroleum Harmeen, G.J., and Stuve, J.G.: “Cement Composition,”
Technology (December 1965) 5, No. 4, 341–347 (paper U.S. Patent No. 2,822,873 (February 11, 1958).
SPE 1300). Harms, W.M., and Febus, J.S.: “Cementing of Fragile-
Hanehara, S., and Yamada, K.: “Interaction Between Formation Wells With Foamed Cement Slurries,”
Cement and Chemical Admixture From the Point of Journal of Petroleum Technology (June 1985) 37, No. 7,
Cement Hydration, Absorption Behaviour of Admixture, 1049–1057; also presented as paper SPE 12755 at the
and Paste Rheology,” Cement and Concrete Research SPE California Regional Meeting, Long Beach,
(August 1999) 29, No. 8, 1159–1165. California, USA (April 11–13, 1984).
Hanks, R.W., and Larsen, K.M.: “The Flow of Power Harms, W.M., and Sutton, D.L.: “Ultra-Low Density
Law Non-Newtonian Fluids in Concentric Annuli,” Cementing Compositions,” Journal of Petroleum
Industrial & Engineering Chemistry Fundamentals Technology (January 1983) 35, No. 1, 61–69; also present-
(1979) 18, No. 1, 33–35. ed as paper SPE 9597 at the SPE Middle East Oil Technical
Hanks, R.W.: “A Theory of Laminar Flow Stability,” Conference, Manama, Bahrain (March 9–12, 1981).
AIChE Journal (January 1969) 15, No. 1, 25–28. Harness, P.E., Sabins, F.L., and Griffith, J.E.: “New
Hanks, R.W.: “On the Flow of Bingham Plastic Slurries in Technique Provides Better Low-Densit y-Cement
Pipes and Between Parallel Plates,” SPE Journal Evaluation,” paper SPE 24050 presented at the SPE
(December 1967) 7, No. 4, 342–346; also presented as Western Regional Meeting, Bakersfield, California
paper SPE 1682 at the SPE Symposium on Mechanics of (March 30–April 1, 1992).
Rheologically Complex Fluids, Houston, Texas, USA Harris, K., Grayson, G., and Langlinais, J.: “Obtaining
(December 15–16, 1966). Successful Shoe Tests In The Gulf Of Mexico: Critical
Hanks, R.W.: “Principles of Slurry Pipeline Hydraulics,” Cementing Factors,” paper SPE 71388 presented at the
Encyclopedia of Fluid Mechanics: Slurry Flow SPE Annual Technical Conference and Exhibition, New
Technology, N.P. Cheremisinoff (ed.) Houston, Texas, Orleans, Louisiana, USA (September 30–October 3, 2001).
USA, Gulf Publishing Company (1986) 5, 213–276. Harris, K.L.: “New Lightweight Technology for the
Hanks, R.W.: “The Axial Laminar Flow of Yield- Primary Cementing of Oilfield Casings in Cold
Pseudoplastic Fluids in a Concentric Annulus,” Environments,” paper SPE 22065 presented at the SPE
Industrial & Engineering Chemistry Process Design International Arctic Technology Conference, Anchorage,
Development (1979) 18, No. 3, 488–493. Alaska, USA (May 29–31, 1991).
Hanks, R.W.: “The Laminar-Turbulent Transition for Harris, P.C., and Reidenbach, V.G.: “High-Temperature
Flow in Pipes, Concentric Annuli and Parallel Plates,” Rheological Study of Foam Fracturing Fluids,” Journal
AIChE Journal (January 1963a) 9, No. 1, 45–48. of Petroleum Technology (May 1987), 39, No. 5, 613–619;
also presented as paper SPE 13177 at the SPE Annual
Hanks, R.W.: “The Laminar-Turbulent Transition for
Technical Conference and Exhibition, Houston, Texas,
Fluids with a Yield Stress,” AIChE Journal (May 1963b)
USA (September 16–19, 1984).
9, No. 3, 306–309.
Harris, P.C.: “Effects of Texture on Rheology of Foam
Hannant, D.J., and Keating, J.: “Equipment for Assessing
Fracturing Fluids,” SPE Production Engineering
the Development of Structure in Fresh Cement Pastes by
(August 1989) 4, No. 3, 249–257; also presented as paper
the Measurement of Shear Modulus,” Cement and
SPE 14257 at the SPE Annual Technical Conference and
Concrete Research (July 1985), 15, No. 4, 605–612.
Exhibition, Las Vegas, Nevada, USA (September 22–25,
Harder, C., Carpenter, R., and Wilson, W.: 1985).
“Surfactant/cement blends improve plugging operations
Harrison, H.T.: “Aqueous Cementing Composition
in oil-based muds,” paper SPE 23928 presented at the
Adaptable to High Turbulent Flow and Method of
IADC/SPE Drilling Conference, New Orleans, Louisiana,
Cementing a Well Using Same,” U.S. Patent No. 3,409,080
USA (February 18–21, 1992).
(November 5, 1968).
Harder, C.A.: “Optimization of oil-based mud chemistry
Harrison, T.W., and Blount, C.G.: “Coiled Tubing Cement
for cementing,” paper SPE 25183 presented at the SPE
Squeeze Technique at Prudhoe Bay, Alaska,” paper SPE
International Symposium on Oilfield Chemistry, New
15104 presented at the SPE California Regional Meeting,
Orleans, Louisiana, USA (March 2–5, 1993).
Oakland, California, USA (April 12–14, 1986).

References 719
Hartog, J.J., Davies, D.R., and Stewart, R.B.: “An Heathman, J., Wilson, J.M., and Cantrell, J.J.: “New test
Integrated Approach for Successful Primary procedures optimize surfactant blends,” Oil & Gas
Cementations,” Journal of Petroleum Technology Journal (October 4, 1999) 97, No. 40, 71–74.
(September 1983) 35, No. 10, 1600–1610; also presented Heathman, J.F. et al.: “Acid-Resistant Microfine Squeeze
as paper SPE 9599 at the SPE Middle East Technical Cement: From Conception to Viable Technology,”
Conference, Manama, Bahrain (March 9–12, 1981). paper SPE 26571 presented at the 68th SPE Annual
Haut, R.C., and Crook, R.J.: “Laboratory Investigation of Technical Conference and Exhibition, Houston, Texas,
Lightweight, Low-Viscosity Cementing Spacer Fluids,” USA (October 3–6, 1993).
Journal of Petroleum Technology (August 1982) 34, No. Heathman, J.F. et al.: “Quality Management Alliance
8, 1828–1834; also presented as paper SPE 10305 at the Eliminates Plug Failures,” paper SPE 28321 presented at
SPE Annual Technical Conference and Exhibition, San the SPE Annual Technical Conference and Exhibition,
Antonio, Texas, USA (October 5–7, 1981). New Orleans, Louisiana, USA (September 25–28, 1994).
Haut, R.C., and Crook, R.J.: “Primary Cementing: The Heathman, J.F., and East, L.E. Jr.: “Case Histories
Mud Displacement Process,” paper SPE 8253 presented Regarding the Application of Microfine Cements,” paper
at the SPE Annual Technical Conference and Exhibition, SPE 23926 presented at the IADC/SPE Drilling
Las Vegas, Nevada, USA (September 23–26, 1979). Conference, New Orleans, Louisiana, USA (February
Havira, M.: “Ultrasonic Bond Evaluation in Multilayered 18–21, 1992).
Media,” Journal of the Acoustical Society of America Heathman, J.F.: “Advances In Cement-Plug Procedures,”
(Fall 1979) 66, Supplement 1, S41. Journal of Petroleum Technology (September 1996) 48,
Havira, M.: “Ultrasonic Cement Bond Evaluation,” No. 9, 825–826, 828, 831 (SPE 36351).
Transactions of the 23rd Annual SPWLA Logging Hedenquest, J.W., and Stewart, M.K.: “Natural CO2-rich
Symposium, Corpus Christi, Texas, USA (July 6–9, steam-heated waters in the Broadlands-Okaaki geother-
1982) paper N. mal system, New Zealand: Their chemistry distribution
Hayman, A.J., Gai, H., and Toma, I.: “A Comparison of and corrosive nature,” Transactions of the Geothermal
Cementation Logging Tools in a Full-Scale Simulator,’’ Resources Council International Symposium on
paper SPE 22779 presented at the SPE Annual Technical Geothermal Energy, Davis, California, USA, Geothermal
Conference and Exhibition, Dallas, Texas (October 6–9, Resources Council (1985).
1991). Hedström, B.O.A.: “Flow of Plastics Materials in Pipes,”
Hayman, A.J., Hutin, R., and Wright, P.V.: “High- Industrial & Engineering Chemistry (March 1952) 44,
Resolution Cementation and Corrosion Imaging by No. 3, 651–656.
Ultrasound,’’ Transactions of the SPWLA 32nd Annual Heindl, R.A., and Post, Z.A.: “Refractory Castables II:
Symposium, Midland, Texas, USA (June 16–19, 1991) Some Properties and Effects of Heat Treatment,”
paper KK. Journal of the American Ceramic Society (May 1954)
Hayman, A.J.: “Ultrasonic Properties of Oil-Well Drilling 37, No. 5, 206–216.
Fluids,” Proceedings of the Institute of Electrical and Heller, John P., and Kuntamukkula, Murty S.: “Critical
Electronics Engineers 1989 Ultrasonics Symposium Review of the Foam Rheology Literature,” Industrial
(paper PE-1), New York, Institute of Electrical and and Engineering Chemistry Research (February 1987)
Electronics Engineers (1989), 327–332. 26, No. 2, 318–325.
Head, P. et al.: “A Tool that Controls U-Tubing during Oil Hengst, R.R., and Tressler, R.E.: “Fracture of Foamed
Well Cementing,” paper CADE/CAODC 95-1008 present- Portland Cements,” Cement and Concrete Research
ed at the Spring Drilling Conference, Calgary, Alberta, (January 1983) 13, No. 1, 127–134.
Canada (April 19–21, 1995).
Herschel, W.H., and Bulkley, R.: “Konsistenzmessungen
Heathman, J. et al.: “Removing Subjective Judgment von gummi-benzöllösungen,” Kolloid-Z (1926) 39,
From Wettability Analysis Aids Displacement,” paper 291–303; also “Measurement of Consistency as Applied
SPE 59135 presented at the IADC/SPE Drilling to Rubber-Benzene Solutions,” Proceedings of the ASTM
Conference, New Orleans, Louisiana, USA (February (June 1926) 26, No. 2, 621–633.
23–25, 2000).
Hewitt, C.H.: “Analytical Techniques for Recognizing
Heathman, J., Tare, U., and Ravi, K.: “Understanding Water-Sensitive Reservoir Rocks,” Journal of Petroleum
Formation (In)Stability During Cementing,” paper Technology (August 1963) 15, No. 8, 813–818; also pre-
SPE/IADC 79913 presented at the SPE/IADC Drilling sented as paper SPE 594 at the SPE Rocky Mountain
Conference, Amsterdam, The Netherlands (February Regional Meeting, Denver, Colorado, USA (May 27–28,
19–21, 2003). 1968).

720 Well Cementing

Hewlett, P.C. (ed.): Lea’s Chemistry of Cement and SPE Rocky Mountain Regional Meeting, to be held in
Concrete (4th ed.), London, England, Butterworth- Denver, Colorado, USA (May 25–27, 1969).
Heinemann (2001). Hook, F.E., Morris, E.F., and Rosene, R.B.: “Silica-Lime
Hibbeler, J., Graham, T., and Ehlinger, J.: “New float collar Systems for High Temperature Cementing Applications,”
design to eliminate wet shoe tracks,” paper SPE 62751 pre- paper SPE 3447 presented at the SPE Annual Meeting,
sented at the IADC/SPE Asia Pacific Drilling Technology, New Orleans, Louisiana, USA (October 3–6, 1971).
Kuala Lumpur, Malaysia (September 11–13, 2000). Hook, F.E.: “Aqueous Cement Slurry; and Method of Use,”
Hill, D.G.: “Clay Stabilization—Criteria for Best U.S. Patent No. 3,483,007 (December 9, 1969).
Performance,” paper SPE 10656 presented at the SPE Hoskins, J. et al.: “A Predictive Model of Mud Removal
Formation Damage Control Symposium, Lafayette, Efficiency Using Historical Well Data,” SPE 26221,
Louisiana, USA (March 24–25, 1982). SPE/PCC 8th Petroleum Computer Conference, New
Hill, K., Von Rybinski, W., and Stoll, G. (eds.): Alkyl Orleans, Louisiana, USA (July 11–13, 1993).
Polyglycosides—Technology, Properties and Applications, Hou, J., and Liu, Z.: “Synthesizing Dispersant for MTC
Weinheim, Germany, Wiley-VCH Verlag GmbH (1997). Design and Its Effect on Slurry Rheology,” SPE Drilling
Hill, R.: The Mathematical Theory of Plasticity, Oxford, & Completion (March 2000) 15, No. 1, 31–36 (paper SPE
England, Clarendon Press (1951). 62095).
Hille, M. et al.: “Cement Slurries for Deep Holes, with a Howard, G.C., and Clark, J.B.: “Factors to be Considered
Copolymer Content for Reducing the Water Loss,” in Obtaining Proper Cementing of Casing,” API Drilling
Canadian Patent No. 1,228,373 (1987). and Production Practices, Washington, D.C., USA,
Hirljac, J., Wu, Z.Q., and Young, J.F.: “Silicate American Petroleum Institute (1948), 257–272.
Polymerization During the Hydration of Alite,” Cement and Howard, G.C., and Scott, P.P. Jr.: “An Analysis and Control
Concrete Research (November 1983) 13, No. 6, 877–886. of Lost Circulation,” Trans., AIME (1951) 192, 171–182.
Hodgkinson, E., and Hughes, S.: “The Mineralogy and Huddleston, D.A., Gabel, R.K., and Williamson, C.D.:
Geochemistry of Cement/Rock Reactions: High- “Method For Reducing Fluid Loss From Oilfield Cement
Resolution Studies of Experimental And Analogue Slurries Using Vinyl Grafted Wattle Tannin,” U.S. Patent
Materials,” Chemical Containment of Waste in the No. 5,147,964 (September 15, 1992).
Geosphere, R. Metcalfe and C. A. Rochelle (eds.), Huddleston, D.A.: “Set Retarding Additive for Cement
London, England, Geological Society of London (1999), Slurries,” U.S. Patent No. 5,417,759 (May 23, 1995).
195–212.
Hughes, C.R., Hill, P.J., and Isaac, K.P.: “An Experimental
Hodne, H., and Saasen, A.: “The Effect of the Zeta Evaluation of the Formation Damage Potential of
Potential and Slurry Conductivity on the Consistency Cement Filtrates and Leachates,” SPE Drilling &
of Oilwell Cement Slurries,” Cement and Concrete Completion (June 1997) 12, No. 2, 99–104; also present-
Research (November 2000) 30, No. 11, 1767–1772. ed as paper SPE 30087 at the SPE European Formation
Hogg, C.: “Level 6 Multilateral Numbers Increase,” Oil & Damage Conference, The Hague, The Netherlands (May
Gas Journal (September 30, 2002) 100, No. 40, 63–64, 15–16, 1995).
66–67. Hughes, T.L. et al.: “Rapid cement quality control
Holhjem, A., Berk, G.W., and Hoch, R.S.: “Reciprocation method for improved oilfield cementing,” paper OTC
of Casing While Cementing From a Floating Drilling 7582 presented at the Offshore Technology Conference,
Unit,” paper EUR 364 presented at the SPE European Houston, Texas, USA (May 2–5, 1994b).
Petroleum Conference, London, England (October Hughes, T.L. et al.: “The use of Fourier transform infrared
25–28, 1982). spectroscopy to characterise cement powders, cement
Honore, R.S. Jr. et al.: “Cementing Temperature hydration and the role of additives,” Proceedings of the
Predictions Based on Downhole Measurements and Royal Society of Chemistry 5th International Symposium
Computer Predictions: A Case History,” paper SPE 25436 on Chemistry in the Oil Industry (1994a) 99–115.
presented at the Production Operations Symposium, Humphrey, H.C.: “Casing Failures Caused by Thermal
Oklahoma City, Oklahoma, USA (March 21–23, 1993). Expansion,” World Oil (November 1960) 151, No. 6,
Hook, F.E., and Ernst, E.A: “The Effect of Low-Water- 105–108.
Loss Additives, Squeeze Pressure, and Formation Hunt, L.P., and Elspass, C.W.: “Particle-Size Properties of
Permeability on the Dehydration Rate of a Squeeze Oilwell Cements,” Cement and Concrete Research
Cementing Slurry,” paper SPE 2455 presented at the (November 1986) 16, No. 6, 805–812.

References 721
Hunt, L.P.: “Prediction of Thickening Time of Well (4th ed.), P.C. Hewlett (ed.), London, UK, Butterworth-
Cements from Blaine Air Permeability,” Cement and Heinemann (2001).
Concrete Research (March 1986) 16, No. 2, 190–198. Jain, B. et al.: “Using Particle-Size Distribution
Hurlbut, C.S.: Dana’s Manual of Mineralogy, John Wiley Technology for Designing High-Densit y, High-
& Sons, New York, 1971. Performance Cement Slurries in Demanding Frontier
Hutchinson, M., and Rezmer-Cooper, I.: “Using downhole Exploration Wells in South Oman,” paper SPE 59134
annular pressure measurements to anticipate drilling presented at the IADC/SPE Drilling Conference, New
problems,” paper SPE 49114 presented at the SPE Orleans, Louisiana, USA (February 23–25, 2000).
Annual Technical Conference and Exhibition, New Jakobsen, J. et al.: “Displacements in Eccentric Annuli
Orleans, Louisiana, USA (September 27–30, 1998). During Primary Cementing in Deviated Wells,” paper
Iljas, R.: “Lost Circulation and Control in Reefal SPE 21686 presented at the SPE Production Operations
Limestone Depositions,” Proceedings of the Indonesian Symposium, Oklahoma City, Oklahoma, USA (April 7–9,
Petroleum Association Annual Convention (May 1984) 1991).
2, 1–10; first presented at the Indonesian Petroleum Jamot, A.: “Déplacement de la boue par le laitier de
Association Annual Convention (June 7–8, 1983). ciment dans l’espace annulaire tubage-paroi d’un puits,”
International Society for Rock Mechanics: “Suggested Revue de l’Association Français Technique Petroliére
methods for determining tensile strength of rock mate- (March–April 1974), No. 224, 27–37.
rials,” International Journal of Rock Mechanics, Jawed, I., and Skalny, J.: “Alkalies in Cement: A Review.
Mineral Science and Geomechanics Abstracts II. Effects of Alkalies on the Hydration and Performance
(December 1978); 15, No. 6, 99–103. of Portland Cement,” Cement and Concrete Research
Irani, C.A.: “Method for Plugging Gas Migration (January 1978) 8, No. 1, 37–51.
Channels in the Cement Annulus of a Wellbore Using Jawed, I., Klemm, W.A., and Skalny, J.: “Hydration of
High Viscosity Polymers,” US Patent No. 5,950,727 Cement-Lignosulfonate-Alkali Carbonate System,”
(September 14, 1999). Journal of the American Ceramic Society (September–
Ismail, S.A.A., and Khalaf, F.: “Effectiveness of Low-Salt October, 1979) 62, No. 9–10, 461–464.
Cement Opposite Salt Bodies,” paper SPE 25542 pre- Jefferies, N., Tweed, C., and Wiseby, S.: “The Effects
sented at the SPE Middle East Oil Technical Conference of Changes in pH Within a Clay Surrounding a
and Exhibition, Bahrain (April 3–6, 1993). Cementitious Repository,” Scientific basis for nuclear
Israelachvili, J.: Intermolecular and Surface Forces waste management, Materials Research Society (1988)
(2nd ed.), Academic Press (1992). 11,43–52.
Italcementi S.p.a.: “GEOTERM,” customer brochure Jennings, A.R.: “Cementing efficiency in horizontal well-
(1977). bores via dual density fluids and cements,” U.S. Patent
No. 5,452,764 (September 26, 1995).
Ivan, D.C. et al.: “Making a Case for Rethinking Lost
Circulation Treatments in Induced Fractures,” paper Jennings, H.M. et al.: “Interpretation of the Effects
SPE 77353 presented at the SPE Annual Technical of Retarding Admixtures on Pastes of C3S, C3A plus gyp-
Conference and Exhibition, San Antonio, Texas sum, and Portland Cement,” Proceedings of the 8th
(September 29–October 2, 2002). International Congress on the Chemistry of Cement
(1986) 239–243.
Iyoho, A.W., and Azar, J.J.: “An Accurate Slot Flow Model
for Non-Newtonian Fluid Flow Through Eccentric Jeong, S.Y., and Wagh, A.S.: “Cementing the Gap
Annuli,” SPE Journal (October 1981) 21, No. 5, Between Ceramics and Cements,” Materials Technology
565–572; also presented as paper SPE 9447 at the SPE (September 2003) 18, No. 3, 162–168.
Annual Technology Conference and Exhibition, Dallas, Jestes, C. et al.: “Monitoring Cementing Component
Texas, USA (September 21–24, 1980). Temperature Yields Downhole: Cement Slurry
Jackson, P.B., and Murphey, C.E.: “Effect of Casing Excellence and Job Quality—A Case Study,” paper
Pressure on Gas Flow Through a Sheath of Set Cement,” SPE 84840 presented at the SPE Eastern
paper SPE 25698 presented at the 1993 SPE/IADC Regional/AAPG Eastern Section Joint Meeting,
Drilling Conference, Amsterdam, The Netherlands Pittsburgh, Pennsylvania, USA (September 6–10, 2003).
(February 23–25, 1993). Jiang, S., Kim, B.-G., and Aïtcin, P.-C.: “Importance
Jackson, P.J.: “Portland Cement: Classification and of Adequate Soluble Alkali Content to Ensure
Manufacture,” Lea’s Chemistry of Cement and Concrete Cement/Superplasticizer Compatibility,” Cement and
Concrete Research (January 1999) 29, No. 1, 71–78.

722 Well Cementing

Johannessen, R., Karlstad, S., and Piot, B.: “Gravel-Pack Conference and Exhibition, New Orleans, Louisiana,
Shutoff Using Deep Penetrating Squeeze Cement USA (October 5–8, 1986).
System in the Gullfaks Field,” paper SPE 61184 pre- Justnes, H. et al.: “An Evaluation Of Methods for
sented at the SPE International Conference on Measuring Chemical Shrinkage of Cementitious Pastes,”
Health, Safety, and the Environment in Oil and Gas Nordic Concrete Research (1994) Publication No. 14,
Exploration and Production, Stavanger, Norway 45–61.
(June 26–28, 2000).
Justnes, H. et al.: “Chemical Shrinkage of Oil Well
Johnson, M.H, Ashton, J.P., and Nguyen, H.: “The Effects Cement Slurries,” Advances in Cement Research (April
of Erosion Velocity on Filter Cake Stability During 1995) 7, No. 26, 85–90.
Gravel Placement of Openhole Horizontal Gravel-Pack
Justnes, H. et al.: “Total and External Chemical Shrinkage
Completions,” paper SPE 23773 presented at the SPE
of Low W/C Ratio Cement Pastes,” Advances in Cement
International Symposium on Formation Damage Control,
Research (July 1996) 8, No. 31, 121–126.
Lafayette, Louisiana, USA (February 26–27, 1992).
Justnes, H., and Nygaard, E.C.: “Technical Calcium
Johnson, R., Jellison, M.J., and Klementich, E.F.:
Nitrate as Set Accelerator for Cement at Low
“Triaxial-Load-Capacity Diagrams Provide a New
Temperatures,” Cement and Concrete Research
Approach to Casing and Tubing Design Analysis,” SPE
(December 1995) 25, No. 8, 1766–1774.
Drilling Engineering (September 1987) 2, No. 3,
268–274. Jutten, J.J., and Corrigall, E.: “Studies with Narrow
Cement Thicknesses Lead to Improved CBL in
Johnston, N., and Miller, H.: “The effect of pH on the sta-
Concentric Casings,” Journal of Petroleum Technology
bility of smectite,” Mississauga, Ontario, Atomic Energy
(November 1989) 41, No. 11, 1158–1159, 1186–1192; also
of Canada Ltd. Report No. AECL 8366 (1984).
presented as paper SPE 18028 at the SPE Annual
Johnston, N.C., and Senese, M.: “New Approach to High- Technical Conference and Exhibition, Houston, Texas,
Density Cement Slurries for Cementing High-Pressure, USA (October 2–5, 1988).
High-Temperature Wells,” paper SPE 24976 presented at
Jutten, J.J., and Hayman, A.J.: “Microannulus Effect on
the SPE European Petroleum Conference, Cannes,
Cement Logs: Experiments and Case Histories,” paper
France (November 16–18, 1992).
SPE 25377 presented at the SPE Asia Pacific Oil and Gas
Jones, F.O.: “Influence of Chemical Composition of Conference, Singapore (February 8–10, 1993).
Water on Clay Blocking of Permeability,” Journal of
Jutten, J.J., and Parcevaux, P.: “Relationship between
Petroleum Technology (April 1964) 16, No. 4, 441–446.
Cement Bond Log Output and Borehole Geometrical
Jones, P.H., and Berdine, D.: “Oil Well Cementing: Parameters,” paper SPE 16139 presented at the
Factors Influencing Bond Between Cement and SPE/IADC Drilling Conference, New Orleans, Louisiana,
Formation,” API Drilling and Production Practices, USA (March 15–18, 1987).
Washington, D.C., USA, American Petroleum Institute
Jutten, J.J., Guillot, D., and Parcevaux, P.A.:
(March 1940), 45–63.
“Relationship Between Cement Slurry Composition,
Jones, R. and Watters, L.: “Remedial Cementing,” Mechanical Properties, and Cement-Bond-Log Output,”
Petroleum Well Construction, M.J. Economides and K.G. SPE Production Engineering (February 1989) 4, No. 1,
Nolte (eds.), New York, New York, USA, John Wiley & 75–82; also presented as paper SPE 16652 at the SPE
Sons (1998), 320–347. Annual Technical Conference and Exhibition, Dallas,
Jones, R.R., and Carpenter, R.B.: “New Latex, Expanding Texas, USA (September 27–30, 1987).
Thixotropic Cement Systems Improve Job Performance Juvkam-Wold, H.C., and Wu, J.: “Casing Deflection and
and Reduce Costs,” paper SPE 21010 presented at the Centralizer Spacing Calculations,” SPE Drilling
SPE International Symposium on Oilfield Chemistry, Engineering (December 1992) 7, No. 4, 268–274; also
Anaheim, California, USA (February 20–22, 1991). presented as paper SPE 21282 at the SPE Eastern
Jones, R.R., Carpenter, R.B., and Conway, M.W.: “A study Regional Meeting, Columbus, Ohio, USA (October 31–
of formation damage potential during cementing opera- November 2, 1990).
tions,” paper SPE 22777 presented at the SPE Annual Kaeuffer, M.: “Determination de l’optimum de rem-
Technical Conference and Exhibition, Dallas, Texas, USA plissage granulométrique et quelques proprietés s’y
(October 6–9, 1991). rattachant,” paper presented at Congrés International
Jones, R.R.: “A Novel Economical Approach for Accurate de l’A.F.T.P.V., Rouen, France (October 1973).
Real-Time Measurement of Wellbore Temperatures,” Kalousek, G.L., and Chow, S.Y.: “Research on Cements
paper SPE 15577 presented at the SPE Annual Technical for Geothermal and Deep Oil Wells,” SPE Journal

References 723
(December 1976) 16, No. 6, 307–309; also presented Kelm, C.H., and Faul, R.R.: “A Best Practices Approach
as paper SPE 5940 at the SPE Annual Technical Can Reduce Environmental Risk,” paper SPE 54344 pre-
Conference and Exhibition, New Orleans, Louisiana, sented at the SPE Asia Pacific Oil and Gas Conference
USA (October 3–6, 1976). and Exhibition, Jakarta, Indonesia (April 20–22, 1999).
Kalousek, G.L., and Nelson, E.B.: “Hydrothermal Kieffer, J., and Rae, P.: “How Gelation Affects Oilwell
Reactions of Dicalcium Silicate and Silica,” Cement and Cements,” Petroleum Engineering International (May
Concrete Research (May 1978) 8, No. 3, 283–290. 1987) 59, No. 5, 46–48.
Kalousek, G.L.: “Development of Expansive Cements,” Kilne, W.E., Kocian, E.M., and Smith, W.E.: “Evaluation of
Proceedings of the Klein Symposium on Expansive Cementing Practices by Quantitative Radiotracer
Cement Concretes, American Concrete Institute Measurements,” paper SPE 14778 presented at the
Publication SP-38, Farmington Hills, Michigan, American SPE/IADC Drilling Conference, Dallas (February 9–12,
Concrete Institute (1973). 1986).
Kalousek, G.L.: “The Reactions of Cement Hydration Kim, B.-G. et al.: “The Adsorption Behavior of PNS
at Elevated Temperatures,” Proceedings of Third Superplasticizer and its Relation to Fluidity of Cement
International Symposium on the Chemistry of Cement Paste,” Cement and Concrete Research (June 2000) 30,
(1952), 334–354. No. 6, 887–893.
Kar, K.K.: “Metallo-Organo Aluminates as Lubricant Kim, J.H., and Robertson, R.: “Effects of polyvinyl alco-
Additives,” U.S. Patent No. 4,610,797 (September 9, 1986). hol on aggregate paste bond strength and the interfacial
Karimov, N.H. et al.: “Research Goal: Improved Rheological transition zone,” Advanced Cement-Based Materials
Properties of Cement Slurries and Improve Compressive (September 1998) 8, No. 2, 66–76.
Strength,” Soviet Union Patent No. 1323699 (1985). Kimura, K. et al.: “Custom-blending foamed cement for
Kastrop, J.E. (ed.): Fundamentals of Thermal Oil multiple challenges,” paper SPE/IADC 57585 presented
Recovery, Dallas, Texas, USA, Petroleum Engineer at the SPE/IADC Middle East Drilling Technology
Publishing Company (1965). Conference, Abu Dhabi, UAE (November 8–10, 1999).
Keese, R. et al.: “Novel Cementing Slurries to Remediate King, I., Trenty, L., and Vit, C.: “How the 3D modeling
Steam Breakthrough in the World’s Largest Steam could help hole-cleaning optimization,” paper SPE 63276
Flood,” paper IPA99-E-145, Proceedings of the Annual presented at the SPE Annual Technical Conference and
Indonesian Petroleum Association Convention, Exhibition, Dallas, Texas, USA (October 1–4, 2000).
Jakarta, Indonesia (February 1–3, 2000) 2, 295–304. King, T.G.: “The Application of Permanent Completion
Kelessidis, V.C. et al.: “Field data demonstrate improved Techniques to Offshore Operations,” Journal of
mud removal techniques lead to successful cement jobs,” Petroleum Technology (September 1966) 18, No. 9,
SPE Advanced Technology Series (1995) 4, No. 1, 53–58. 1031–1040; also presented as paper SPE 1444 at the SPE
Symposium on Offshore Technology and Operations,
Kelessidis, V.C. et al.: “Simulator models ‘U-tubing’ to
New Orleans, Louisiana, USA (May 23–24, 1966).
improve primary cementing,” Oil & Gas Journal (March
9, 1994) 92, No. 10, 72–80. Kinzel, H., and Martens, J.G.: “The Application of
New Centralizer Types to Improve Zone Isolation in
Keller, S.R. et al.: “Problems Associated with Deviated
Horizontal Wells,” paper SPE 50438 presented at the
Wellbore Cementing,” Journal of Petroleum Technology
SPE International Oil and Gas Conference and
(August 1987) 39, No. 8, 955–960; also presented as
Exhibition in China, Beijing, China (November 2–6,
paper SPE 11979 at the SPE Annual Technical
1998).
Conference and Exhibition, San Francisco, California,
USA (October 5–8, 1983). Klein, A., and Troxell, G.E.: “Studies of Calcium
Sulfoaluminate Admixtures for Expansive Cements,”
Kellingray, D.S., Greaves, C., and Dallimer, R.P.: “High
Proceedings of the American Society for Testing and
Temperature and High Pressure Rheology of Oilwell
Materials (1958) 58, 986–1008.
Cement Slurries,” Proceedings of the International
Rheology Conference, P.F.G. Banfill (ed.), Liverpool, Klementich, E.F., and Jellison, M.J.: “A Service Life
England, British Society of Rheology (March 1990). Model for Casing Strings,” SPE Drilling Engineering
(April 1986) 1, No. 2, 141–152.
Kelly, E.B. et al.: “SlurryMinder: A Rational Oil Well
Completion Design Module,” Proceedings of the Fourth Klotz, J.A.: “To Determine Herschel-Bulkley Coefficients,”
Conference on Innovative Applications of Artificial Journal of Petroleum Technology (November 1998) 50,
Intelligence, C. Scott and P. Klahr (eds.), Menlo Park, 11, 80–81.
California, USA, American Association for Artificial Kolthoff, K.W., and Scales, G.H.: “Improved Liner
Intelligence (1992). Cementing Techniques for Alaska’s Prudhoe Bay Field,”

724 Well Cementing

paper SPE 10756 presented at the SPE California über die Hydratation der Calciumsilikate 3CaO SiO2 und
Regional Meeting, San Francisco, California, USA β-2CaO · SiO2 und den Einfluβ von Calciumchlorid und
(March 24–26, 1982). Gips auf den Hydrationsvorgang,” Tonindustrie-Zeitung
Kondo, R. et al.: “Influence of organic salts on the hydra- und Keramische Rundschau (1960) 24, 585–597.
tion of tricalcium silicate,” Journal of Applied Chemistry Kuru, E. et al.: “New Directions in Foam and Aerated
and Biotechnology (1977), 27, 191–197. Mud Research and Development,” paper SPE 53963
Koomareni, S., and White, W.: “Hydrothermal Reactions presented at the SPE Latin American and Caribbean
of Clay Minerals and Shales with Cesium Phases from Petroleum Engineering Conference, Caracas, Venezuela
Spent Fuel Elements,” Clays and Clay Minerals (April 21–23, 1999).
(August 1981) 29, No. 4, 299–308. Kutasov, I.M.: “Method corrects API bottomhole circu-
Krause, D.C.: “Solids-free brines: efficient displacement lating-temperature correlations,” Oil & Gas Journal
techniques can save you money,” paper SPE 14830 pre- (July 2002) 100, No. 28, 47–50.
sented at the SPE Symposium on Formation Damage Kuzel, H.-J., and Pöllmann, H.: “Hydration of C3A in
Control, Lafayette, Louisiana, USA (February 26–27, 1986). the Presence of CH, Gypsum and Calcium Carbonate,”
Krause, R.E., and Reem, D.C.: “New Coiled-Tubing Cement and Concrete Research (September 1991) 21,
Cementing Techniques at Prudhoe Developed to No. 5, 885–895.
Withstand Higher Differential Pressure,” paper SPE Labat, C. et al.: “3D Azimuthal LWD Caliper,” paper
24052 presented at the SPE Western Regional Meeting, SPE 77526 presented at the SPE Annual Technical
Bakersfield, California, USA (March 30–April 1, 1992). Conference and Exhibition, San Antonio, Texas, USA
Krause, R.E.: “Lost Circulation Material Usage in Coiled (September 29–October 2, 2002).
Tubing Remedial Cementing at Prudhoe Bay,” paper SPE Lach, V., and Bures, J.: “Phase Composition and
22067 presented at the International Arctic Technology Microstructure of Cement Paste Hydrated at Elevated
Conference, Anchorage, Alaska, USA (May 29–31, 1991). Temperatures,” Proceedings of the 6th International
Kroken, W., Sjaholm, A.J., and Olsen, A.S.: “Tide flow: Conference on the Chemistry of Cement (1976) 2, 129–135.
a low rate density driven cementing technique for Ladva, H.K.J. et al.: “The Cement-to-Formation Interface
highly deviated wells,” paper SPE 35082 presented at the in Zonal Isolation,” paper SPE 88016 presented at
IADC/SPE Drilling Conference, New Orleans, Louisiana, the IADC/SPE Asia Pacif ic Drilling Technology
USA (March 12–15, 1995). Conference and Exhibition, Kuala Lumpur, Malaysia
Kuchuk, F., and Sengul, M.: “The Challenge Of Water (September 13–15, 2004).
Control,” Middle East Well Evaluation Review (1999) Ladva, H.K.J. et al.: “Multiphase Flow and Drilling
No. 22, 25–43. Fluid Filtrate Effects on the Onset of Production,”
Kucyn, P. V. et al.: “Prevention des manifestations du SPE Journal (December 2001) 6, No. 4, 425–432
gaz entre le tubage et les parois du puits et de l’erup- (SPE 74706); revised for publication from paper
tion incontrolee au cours du forage,” Gazovaja SPE 58795, first presented at the SPE International
Promyshlennost (1977) 2, 48 (translated from Russian). Symposium on Formation Damage Control, Lafayette,
Louisiana, USA (February 23–24, 2000).
Kuhlmann, L.A.: “Latex Modified Concrete for the
Repair and Rehabilitation of Bridges,” International Ladva, H.K.J. et al.: “Use of Oil-Based Reservoir
Journal of Cement Composites and Lightweight Drilling Fluids in Open-Hole Horizontal Gravel-Packed
Concrete (November 1985) 7, No. 4, 241–247. Completions: Damage Mechanisms and How to Avoid
Them,” paper SPE 68959 presented at the SPE European
Kulakofsky, D., and Creel, P.: “Computer Simulation
Formation Damage Conference, The Hague, The
Improves Remedial Cementing Success,” World Oil
Netherlands (May 21–22, 2001).
(November 1992) 213, No. 11, 59–63.
Lai, Q.J. et al.: “Gel-Cement Combination Squeezes for
Kumar, A.: “Diffusion of Cs+ and Cl– Through Sealing
Gas Shutoff,” paper SPE 54596 presented at the SPE
Materials,” Cement and Concrete Research (January
Western Regional Meeting, Anchorage, Alaska, USA
1987) 17, No. 1, 153–160.
(May 26–28, 1999).
Kunze, K.R.: “Obtaining and Verifying Quality Cement
Lajeunesse, E. et al.: “Miscible displacement in a Hele-
Blends,” paper SPE 15576 presented at the SPE Annual
Shaw cell at high angles,” Journal of Fluid Mechanics
Technical Conference and Exhibition, New Orleans,
(November 10, 1999) 398, 299–319.
Louisiana, USA (October 5–8, 1986).
Lama, R.D., and Vutukuri, V.S.: “Dynamic Elastic
Kurczyk, H.G., and Schwiete, H.E.: “Elektronen-
Constants of Rock,” Handbook on Mechanical
mikroskopische un thermochemische Untersuchungen

References 725
Properties of Rocks—Volume II, Testing Techniques and Leimkuhler, J.M. et al.: “Downhole Performance
Results, Trans Tech Publications (1978) 231–236. Evaluation of Blast Furnace Slag-Based Cements:
Landrum, W.R., Porter, J.E., and Turner, R.D.: “Rotating Onshore and Offshore Field Applications,” paper SPE
Liners During Cementing in the Grand Isle and West 28474 presented at the SPE Annual Technical
Delta Areas, Louisiana,” Journal of Petroleum Conference and Exhibition, New Orleans, Louisiana,
Technology (July 1985) 37, No. 8, 1263–1266. USA (September 25–28, 1994).
Langton, C.A. et al.: “High Temperature Cements with Lenn, C., Bamforth, S., and Jariwala, H.: “Flow Diagnosis
Geothermal Applications,” Proceedings of the 7th in an Extended Reach Well at the Wytch Farm Oilfield
International Congress on the Chemistry of Cement, Using a New Toolstring Combination Incorporating
Paris, France (1980), 145–51. Novel Production Technology,” paper SPE 36580 pre-
sented at the SPE Annual Technical Conference and
Lapasin, R., Papo, A., and Rajgelj, S.: “Flow Behavior of
Exhibition, Denver, Colorado, USA (October 6–9,1996).
Fresh Cement Pastes: A Comparison of Different
Rheological Instruments and Techniques,” Cement and Lerch, W.: “The Influence of Gypsum on the Hydration
Concrete Research (May 1983) 13, No. 3, 349–356. and Properties of Portland Cement Pastes,” Portland
Cement Research Laboratory Bulletin (March 1946) 12.
Last, N. et al.: “Casing Deformation in a Tectonic Setting:
Evaluation, Impact and Management,” paper SPE 74560 Leslie, H.D., de Selliers, J., and Pittman, D.J.: “Coupling
presented at the SPE/IADC Drilling Conference, Dallas, and Attenuation: A New Pair in Cement Bond Logging,”
Texas, USA (February 26–28, 2002). paper SPE 16207 presented at the SPE Production
Operations Symposium, Oklahoma City, Oklahoma, USA
Lavaure, G., and Galiana, G.: “A Safer Approach to
(March 8–10, 1987).
controlling the Cementing Operations at the Wellhead,”
paper SPE 23261 presented at the First International Lester, R.A.: “The Segmented Bond Tool: A Pad-type
Conference on Health, Safety, and Environment, The Cement Bond Device,” Transactions of the 12th Annual
Hague, The Netherlands (November 10–14, 1991). Canadian Well Logging Society Formation Evaluation
Symposium, Calgary (September 26–29, 1989), paper A.
Lawrence, C.D.: “The Constitution and Specification of
Portland Cements,” Lea’s Chemistry of Cement and Leutwyler, K.: “Casing Temperature Studies in Steam
Concrete (4th ed.), P.C. Hewlett (ed.), Oxford, England, Injection Wells,” Journal of Petroleum Technology
Butterworth-Heinemann (2001). (September 1966) 18, No. 9, 1157–1162 (SPE 1264).
Le Roy-Delage, S. et al.: “New Cement Systems for Levine, D.C. et al.: “Annular Gas Flow After Cementing:
Durable Zonal Isolation,” paper SPE 59132 presented at A Look at Practical Solutions,” paper SPE 8255 present-
the IADC/SPE Drilling Conference, New Orleans, ed at the SPE Annual Fall Technical Conference and
Louisiana, USA (February 23–25, 2000). Exhibition, Las Vegas, Nevada, USA (September 23–26,
1979).
Le Roy-Delage, S., Dargaud, B., and Thiercelin, M.:
“Cementing Compositions and Application of Such Lewis, W.J., et al.: “Salt Cements for Improved Hydraulic
Compositions for Cementing Oil Wells or the Like,” Isolation and Reduced Gas Channeling,” paper SPE 16386
Patent WO 00/20387 (2000). presented at the SPE California Regional Meeting,
Ventura, California, USA (April 8–10, 1987).
Le Roy-Delage, S., Dargaud, B., Baret, J.F., and
Thiercelin, M.: “Cementing Compositions and the Use of Lichtman, I., and Gammon, T.: “Defoamers,” Kirk-
Such Compositions for Cementing Oil Wells or the Like,” Othmer Encyclopedia of Chemical Technology 7,
Patent WO 00/20350 (2000). Grayson M. and Eckroth D. (eds.) New York, John Wiley
& Sons (1979), 430–438.
Lea, F.M.: The Chemistry of Cement and Concrete, New
York, New York, USA, Chemical Publishing Co., Inc. Lindner, A., Coussot, P., and Bonn, D.: “Viscous
(1971). Fingering in a Yield Stress Fluid,” Physical Review
Letters (July 2000) 85 (2), 314–317.
Lea, P.J., and Fisher, H.B.: "Low Water-Loss Cement and
Method of Cementing," U.S. Patent No. 2,614,634 Lindstrom, G. et al.: “Shear-Sensitive Fluid Cures
(October 21, 1952). Canadian Lost Circulation,” Oil & Gas Journal
(September 27, 1999) 97, No. 39, 83–85.
Lea, P.J.: “Low Water-Loss Cement Slurry,” U.S. Patent
No. 2,614,998 (October 21, 1952). Lizak, K.F., Zeltmann, T.A., and Crook, R.J.: “Permian
Basin Operators Seal Casing Leaks With Small-Particle
LeChatelier, H.: “Recherches Experimentales sur la
Cement,” paper SPE 23985 presented at the SPE
Constitution des Mortiers Hydrauliques,” 2ème Edition,
Permian Basin Oil and Gas Recovery Conference,
Paris, Dunod, 1887.
Midland, Texas, USA (March 18–20, 1992).

726 Well Cementing

Locher, F.W., Richartz, W., and Sprung, S.: “Setting of Lyman, J., and Fleming, R.H.: “Composition of Sea Water,”
Cement. Part II. Effect of Adding Calcium Sulfate,” Journal of Marine Research (1940) 2, No. 3, 134–146.
Zement-Kalk-Gips (June 1980) 33, No. 6, 271–277. Maberry, J. et al.: “Well cementing compositions for cold
Lockyear, C.F., and Hibbert, A.P.: “Integrated Primary environment,” U.S. Patent Application No. 10/029,333
Cementing Study Defines Key Factors for Field (filed December 21, 2001).
Success,” Journal of Petroleum Technology (December MacKenzie, A., and Hogg, C.: “Multilateral Classification
1989), 41, No. 12, 1320–1325; also presented as paper System With Example Applications,” World Oil (January
SPE 18376 at the SPE European Petroleum Conference, 1999) 220, No. 1, 55–61.
London (October 16–19, 1988).
Mackenzie, R.J. Gordon, and Patterson, Andrew E.:
Lockyear, C.F., Ryan, D.F., and Gunningham, M.M.: “Hydraulics As It Affects Our Tools—The Inflatable
“Cement Channeling: How to Predict and Prevent,” SPE Paradox,” paper SPE 89537 presented at the SPE/ICoTA
Drilling Engineering (September 1990) 5, No. 3, Coiled Tubing Conference and Exhibition, Houston,
201–208; also presented as paper SPE 19865 at the Texas, USA (March 23–24, 2004).
SPE Annual Technical Conference and Exhibition,
Mackenzie, R.J. Gordon, and Plante, Mark E.: “Through-
San Antonio, Texas, USA (October 8–11, 1989).
Tubing Inflatables: Isolation and Guidelines for Coiled-
Lopez, H., and Renshaw, D.: “Squeeze Cement Matrix Tubing Applications,” paper SPE 54476 presented at the
Improves Communication, Increases Efficiency, and SPE/ICoTA Coiled Tubing Roundtable, Houston, Texas,
Reduces Administrative Costs for Alliance Partners,” USA (May 25–26, 1999).
paper SPE 39767 presented at the SPE Permian Basin
Maeso, C., and Tribe, I.: “Hole Shape from Ultrasonic
Oil and Gas Recovery Conference, Midland, Texas, USA
Calipers and Density While Drilling—A Tool for
(March 25–27, 1998).
Drillers,” paper SPE 71395 presented at the SPE Annual
Love, A.E.H.: A Treatise on the Mathematical Theory of Technical Conference and Exhibition, New Orleans,
Elasticity (4th ed.), New York City, New York, USA, Louisiana, USA (September 30–October 3, 2001).
Dover Publications (1927).
Maier, L.F. et al.: “Cementing Materials for Cold
Loveland, K. R., and Bond, A. J.: “Recent Applications of Environments,” Journal of Petroleum Technology
Coiled Tubing in Remedial Wellwork at Prudhoe Bay,” (October 1971) 23, No. 10, 1215–1220.
paper SPE 35587 presented at the SPE Western Regional
Malhotra, V.M., and Malanka, D.: Super-plasticizers in
Meeting, Anchorage, Alaska, USA (May 22–24, 1996).
Concrete, Detroit, Michigan, USA, American Concrete
Low, N., Daccord, G., and Bedel, J.-F.: “Designing Institute (1979).
Fibered Cement Slurries for Lost Circulation
Manev, E.D., Scheludko, A., and Exerowa, D.: “Effect of
Applications: Case Histories,” paper SPE 84617 present-
Surfactant Concentration on the Critical Thickness of
ed at the SPE Annual Technical Conference and
Liquid Films,” Colloid and Polymer Science (July–
Exhibition, Denver, Colorado, USA (October 5–8, 2003).
August 1974) 252, Nos. 7–8, 586–593.
Lubinski, A.: “Influence of Tension and Compression on
Mannheimer, R.J.: “Effect of Slip on Flow Properties of
Straightness and Buckling of Tubular Goods in Oil Wells,”
Cement Slurries Can Flaw Resistance Calculations,” Oil
API Proceedings (Production Bulletin No. 237) (1951).
& Gas Journal (December 5, 1983) 81, No. 49, 144–147.
Ludwig, N.C.: “Effects of Sodium Chloride on Setting
Mannheimer, R.J.: “Laminar and Turbulent Flow of
Properties of Oil Well Cements,” Drilling and Production
Cement Slurries in Large Diameter Pipe,” Report No.
Practices, Washington, DC, USA, API (1951), 20–27.
SwRI 8983, prepared for American Petroleum Institute
Luke, K. et al.: “New Techniques for Monitoring Cement (September 1988).
Hydration Under Simulated Well Conditions,” paper
Mannheimer, R.J.: “Rheological Evaluation of Cement
SPE 28958 presented at the SPE International
Slurries,” Report No. SwRI 6836, prepared for American
Symposium on Oilfield Chemistry, San Antonio, Texas,
Petroleum Institute (May 1982).
USA (February 14–17, 1995).
Marangos, C.: “The Effect of Fluid Loss on Fracture
Lukkien, H.B.: “Subsea Shallow Gas Presents Unique
Initiation During Squeeze Cementing Operations,”
Problems and Solutions,” Oil & Gas Journal (August 2,
paper SPE 19384 (unsolicited) (1989).
1982) 80, No. 31, 120–122.
Maranuk, C.: “Development of an MWD Hole Caliper for
Lummus, J.L., and Azar, J.J.: Drilling Fluids
Drilling and Formation Evaluation,” paper SPE 38585
Optimization, Tulsa, Oklahoma, USA, PennWell
presented at the SPE Annual Technical Conference and
Publishing Co. (1986).
Exhibition, San Antonio, Texas, USA (October 5–8,
1997).

References 727
Maravilla, S.: “A Hydrothermal Setting Cement for paper SPE 10623 presented at the SPE Sixth
Cementing Ultradeep, Hot Wells,” Journal of Petroleum International Symposium on Oilfield and Geothermal
Technology (October 1974) 26, No. 10, 1087–1094 (SPE Chemistry, Dallas, Texas, USA (January 25–27, 1982).
4089). McKenzie, L.F.: “Polyamine Fluid Loss Additive for Oil Well
Maroy, P., and Baret, J.F.: “Oil-wells cement slurries, Cements,” U.S. Patent No. 4,482,383 (November 13, 1984).
their preparation and their use in well cementing McKeon, D.C. et al.: “Improved Oxygen-Activation
operation,” European Patent No. 0621247 (1994). Method for Determining Water Flow Behind Casing,”
Maroy, P.: “Method of cementing a well in geological SPE Formation Evaluation (September 1991) 6, No. 3,
zones containing swelling clays or mud residues 334–342; also presented as paper SPE 20586 at the SPE
containing clays,” U.S. Patent 6,390,197 (May 21, 2002). Annual Technical Conference and Exhibition, New
Marrast, J. et al.: “Well Cementing Process,” U.S. Patent Orleans, Louisiana, USA (September 23–26, 1990).
3,926,257 (December 16, 1975). McKinley, R.M., and Bower, F.M.: “Specialized
Martin, M, Latil, M., and Vetter, P.: “Mud Displacement Applications of Noise Logging,” Journal of Petroleum
by Slurry During Primary Cementing Jobs. Predicting Technology (November 1979) 1387–1395; also presented
Optimum Conditions,” paper SPE 7590 presented at the as paper SPE 6784 at the SPE Annual Technical
SPE Annual Technical Conference and Exhibition, Conference and Exhibition (Oct. 9–12, 1977).
Houston, Texas, USA (October 1–3, 1978). McKinley, R.M., Bower, F.M., and Rumble, R.C.: “The
Martin, R.C.: “Cement Composition,” U.S. Patent No. Structure and Interpretation of Noise from Flow Behind
3,234,154 (February 8, 1966). Cemented Casing,” Journal of Petroleum Technology
(March 1973) 25, No. 3, 329–338; also presented as
Mason, C.J. et al.: “Casing Running Challenges for
paper SPE 3999 at the SPE Annual Meeting, San
Extended-Reach Wells,” paper SPE 84447 presented at
Antonio, Texas, USA (October 8–11, 1972).
the SPE Annual Technical Conference and Exhibition,
Denver (October 5–8, 2003). McLean, R.H., Manry, C.W., and Whitaker, W.W.:
“Displacement Mechanics in Primary Cementing,”
Mason, C.J. et al.: “Casing Running Milestones for
Journal of Petroleum Technology (February 1967) 19,
Extended-Reach Wells,” paper SPE 52842 presented at
No. 2, 251–260.
the SPE/IADC Drilling Conference, Amsterdam, The
Netherlands (March 9–11, 1999). McPherson, S.A.: “Cementation of horizontal wellbores,”
paper SPE 62893 presented at the SPE Annual Technical
Mathis, S.P. et al.: “Horizontal open hole well displace-
Conference and Exhibition held in Dallas, Texas, USA
ment practices: effect of various techniques on well
(October 1–4, 2000).
productivity, operational complexity, and overall project
economics,” paper SPE 62520 presented at the Meek, J.W., and Harris, K.L.: “Repairing Casing Leaks
SPE/AAPG Western Regional Meeting held in Long Using Small-Particle-Size Cement,” paper SPE/IADC
Beach, California, USA (June 19–23, 2000). 21972 presented at the SPE/IADC Drilling Conference,
Amsterdam, The Netherlands (March 11–14, 1991).
Matthews, S.M., and Copeland, J.C.: “Control of Annular
Gas Flow in the Deep Anadarko Basin,” paper SPE 14980 Meeten, G.H., and Pafitis, D.: “Granular Fluids
presented at the SPE Deep Drilling and Production Rheometry,” Proceedings of the Materials Research
Symposium, Amarillo, Texas, USA (April 6–8, 1986). Society Symposium, Warrendale, Pennsylvania, USA,
Materials Research Society (1993) 289, 239–244.
McElfresh, Paul M., and Go Boncan, Virgilio C.:
“Applications of Foam Cement,” paper SPE 11203 pre- Mehta, P.K., and Monteiro, P.J.M.: “Effect of aggregate,
sented at the SPE Annual Technical Conference and cement and mineral admixtures on the microstructures
Exhibition, New Orleans, Louisiana, USA (September of the transition zone,” Proceedings of the Materials
26–29, 1982). Research Society Symposium (1988) 114, 65–74; also
presented at Materials Research Society Symposium,
McGhee, B.F., and Vacca, H.L.: “Guidelines for Improved
Boston, Massachusetts, USA, (November 30 –
Monitoring of Cementing Operations,” Transactions of
December 5, 1987).
the SPWLA 21st Annual Logging Symposium,
Lafayette, Louisiana, USA (July 8–11, 1980), paper V. Mehta, P.K., and Monteiro, P.J.M.: “Interaction between
carbonate rock and cement paste,” Cement and
McIlhenny, W.F., and Zeitoun, M.A.: “A Chemical
Concrete Research (March 1986) 16, No. 2, 127–134.
Engineer’s Guide to Sea Water, Parts 1 and 2,” Chemical
Engineering (1969) 24, 81–86; 25, 251–256. Mehta, S., Jones, R., and Caveney, W.: “Cryogenics with
cement microscopy redefines cement behavior,” Oil &
McKenzie, L.F., and McElfresh, P.M.: “Acrylamide-
Gas Journal (October 3, 1994) 92, No. 40, 47–53.
Acrylic Acid Copolymers for Cement Fluid-Loss Control,”

728 Well Cementing

Meller, N. et al.: “The paste hydration of brownmillerite Michaux, M., and Defossè, C.: “Oil Well Cement Slurries
with and without gypsum: a time resolved synchrotron I. Microstructural Approach of Their Rheology,” Cement
diffraction study at 30, 70, 100 and 150°C,” Journal of and Concrete Research (January 1986) 16, No. 1, 23–30.
Materials Chemistry (2004) 14, No. 3, 428–435. Michaux, M., and Nelson, E.B.: “Flash-set Behaviour of
Menetrier, D.: “Contribution to the Kinetic Study of the Oil Well Cement Slurries Containing Lignosulphonates,”
Initial Period of the Hydration of Tricalcium Silicate, as Proceedings of the 9th International Congress on the
a Major Constituent of Portland Cement,” PhD thesis (in Chemistry of Cement (1992) 4, 584–590.
French), University of Burgundy, Dijon, France (1977). Michaux, M., and Oberste-Padtberg, R.: “The Effect of
Mensa-Wilmot, G., and Stacey, B.: “PDC Bits and Rotary Mix Water Salinity on Cement Slurries Containing
Steerable Tools—Functional Challenges, Stabilization Polynaphthalene Sulfonate,” Proceedings of the 8th
Requirements and the Establishment of Operational International Congress on the Chemistry of Cement
Compatibility,” paper SPE 72280 presented at the (1986), 4, 239–243.
IADC/SPE Middle East Drilling Technology, Bahrain Michaux, M., Fletcher, P., and Vidick, B.: “Evolution at
(October 22–24), 2001. early hydration times of the chemical composition of liq-
Menzel, C.A.: “Studies of High Pressure Steam Curing of uid phase of oilwell cement pastes with and without
Tamped Hollow Concrete Block,” Journal of American additives Part I. Additive free cement pastes,” Cement
Concrete Institute (1935) 7, 64. and Concrete Research (May 1989a) 19, No. 3, 443–456.
Merlo, A. et al.: “Temperature Field Measurements and Michaux, M., Fletcher, P., and Vidick, B.: “Evolution at
Computer Program Predictions Under Cementing early hydration times of the chemical composition of liq-
Operation Conditions,” paper SPE 28822 presented at uid phase of oilwell cement pastes with and without
the SPE European Petroleum Conference, London, UK additives Part II. Cement pastes containing additives,”
(October 25–27, 1994). Cement and Concrete Research (July 1989b) 19, No. 4,
Messenger, J.: Lost Circulation, A practical approach 567–578.
to preventing, assessing, and solving lost circulation Michaux, M., Garnier A., and Revil, P.: “A cementing
problems, Tulsa, Oklahoma, USA, PennWell Publishing composition and application to cementing oil wells or
Co. (1981). the like,” U.K. Patent No. GB 2 362 881 (1998).
Messenger, J.U.: “Cementing Against Evaporites,” U.S. Michaux, M., Menetrier, D., and Barret, P.: “Determination
Patent No. 4,089,376 (May 16, 1978). by Trimethylsilylation of the Degree of Silicate Ion
Messenger, J.U.: “Process of treating a well using a light- Polymerization in Calcium Hydrosilicate,” Comptes
weight cement,” U.S. Patent No. 3,804,058 (April 16, Rendus de l’Académie des Sciences Serie II Paris
1974). (February 1983) 296, No. 2, 104–346 (in French).
Messenger, J.U.: “Treating wells to mitigate flow-after- Michaux, M., Oberste-Padtberg, R., and Defossé, C.:
cementing,” U.S. Patent No. 4,235,291 (Nov. 25, 1980). “Oil Well Cement Slurries II. Adsorption Behaviour
of Dispersants,” Cement and Concrete Research
Messier, E., Stiles, D., and Morgan, R.: “Controlling
(November 1986) 16, No. 6, 921–930.
Cement Tops Through Use of Fiber-Based Slurries
Reduces Drilling Costs,” paper CIPC 2002-085 presented Midgley, H.G.: “The Determination of Calcium Hydroxide
at the 53rd Annual Petroleum CIM Technical Meeting, in Set Portland Cements,” Cement and Concrete
Calgary, Alberta, Canada (June 11–13, 2002). Research (January 1979) 9, No. 1, 77–82.
Metcalf, A.S., and Dresher, T.D.: “How Pressure Affects Milestone, N.B. et al.: “Carbonation of Geothermal
the Set Properties of Various Cement Systems,” paper Grouts—Part 1: CO2 Attack at 150°C,” Cement and
SPE 7186 presented at the SPE Rocky Mountain Concrete Research (November 1986) 16, No. 6, 941–950.
Regional Meeting, Cody, Wyoming, USA (May 17–19, Milestone, N.B. et al.: “Carbonation of Geothermal
1978). Grouts—Part 2: CO2 Attack at 250°C,” Cement and
Method of Test for Thermal Conductivity of Cellular Concrete Research (January 1987) 17, No. 1, 37–46.
Plastics by Means of a Probe (ASTM D2326-70), ASTM Milestone, N.B. et al.: “Carbonation of Geothermal
International, West Conshohocken, Pennsylvania, USA Grouts—Part 3: CO2 Attack on Grouts Containing
(1970). Bentonite,” Cement and Concrete Research (March 1987)
Metzner, A.D., and Reed, J.C.: “Flow of Non-Newtonian 17, No. 2, 295–306.
Fluids—Correlation of the Laminar and Turbulent Flow Milestone, N.B., Kukacka, L.E., and Carciello, N.: “CO2
Regions,” AIChE Journal (December 1955) 1, No. 4, Corrosion of Geothermal Cement Grouts,” Proceedings
434–440.

References 729
of the 8th International Congress on the Chemistry of Montman, R. et al.: “Low-Density Foamed Portland
Cement (1986). Cements Fill Variety of Needs,” Oil & Gas Journal
Milestone, N.B.: “Hydration of Tricalcium Silicate in the (July 26, 1982) 80, No. 30, 209–216.
Presence of Lignosulphonates, Glucose, and Sodium Moon, J., and Wang, S.: “Acoustic Method for Determining
Gluconate,” Journal of the American Ceramic Society the Static Gel Strength of Slurries,” paper SPE 55650
(July–August, 1979) 62, No. 7–8, 321–324. presented at the SPE Rocky Mountain Regional Meeting,
Milestone, N.B.: “The Effect of Lignosulphonate Gillette, Wyoming, USA (May 15–18, 1999).
Fractions on the Hydration of Tricalcium Aluminate,” Moon, J.J. et al.: “Consistency and Static Gel Strength
Cement and Concrete Research (January 1976) 6, No. 1, Measuring Device and Method,” U.S. Patent 4,622,846
89–102. (November 18, 1986).
Mindess, S., and Young, J.F.: Concrete, Prentice-Hall, Inc. Mooney, M.: “Explicit Formulas for Slip and Fluidity,”
(1981), 86–87. Journal of Rheology (1931) 2, 210–222.
Minear, J.W., and Goodwin, K.J.: “Cement-Sheath Mooney, M.: “The Viscosity of Concentrated Solutions of
Evaluation,” Petroleum Well Construction, M.J. Spherical Particles,” Journal of Colloids (April 1951) 6,
Economides and K.G. Nolte (eds.), New York, New York, 162–170.
USA, John Wiley & Sons (1998). Moore, P.: Drilling Practices Manual, Tulsa, Oklahoma,
Minear, J.W., Heysse, D.R., and Boonen, P.: “Initial USA, Petroleum Publication Co. (1974).
Results from an Acoustic Logging-While-Drilling Tool,” Moran, L.K. et al.: “Cement Expansion: A Laboratory
paper SPE 36543 presented at the SPE Annual Technical Investigation,” paper SPE 21685 presented at the SPE
Conference and Exhibition, Denver, Colorado, USA Production Operations Symposium, Oklahoma City,
(October 6–9, 1996). Oklahoma, USA (April 7–9, 1991).
Mishra, P., and Tripathi, G.: “Transition from Laminar to Moran, L.K., and Moran, L.L.: “Composition and Method
Turbulent Flow of Purely Viscous Non-Newtonian Fluids to Control Cement Slurry Loss and Viscosity,” U.S.
in Tubes,” Chemical Engineering Science (June 1971) Patent No. 5,728,210 (March 17, 1998a).
26, No. 6, 915–921.
Moran, L.K., and Moran, L.L.: “Composition and method
Mitchell, J.K.: Fundamentals of Soil Behavior (2nd to control cement slurry loss and viscosity,” U.S. Patent
ed.), John Wiley & Sons (1993), 113–130. No. 5,850,880 (December 22, 1998b).
Mitchell, R.F., and Wedelich, H.F. III: “Prediction of Moran, L.K., Murray, T.R., and Moyer, W.R.: “Cement
Downhole Temperatures Can Be a Key for Optimal Expansion: A Laboratory Investigation,” paper SPE 21685
Wellbore Design,” paper SPE 18900 presented at the presented at the Production Operations Symposium,
SPE Production Operations Symposium, Oklahoma City, Oklahoma City, Oklahoma, USA (April 7–9, 1991).
Oklahoma, USA (March 13–14, 1989).
Moran, L.K.: “A review of the use of fluid loss additives
Mitchell, R.F.: “Dynamic Surge/Swab Pressure in cementing operations in the Permian basin,”
Predictions,” SPE Drilling Engineering (September Proceedings of the 1986 Southwestern Petroleum Short
1988) 3, No. 3, 325–333; also presented as SPE 16156 Course, Lubbock, Texas, USA, Southwestern Petroleum
at the SPE/IADC Drilling Conference, New Orleans, (1986), 23–29.
Louisiana, USA (March 15–18, 1987).
Moranville-Regourd, M.: “Cements Made From
Mody, F.K. et al.: “Development of Novel Membrane Blastfurnace Slag,” Lea’s Chemistry of Cement and
Efficient Water-Based Drilling Fluids Through Concrete (4th ed.), P.C. Hewlett (ed.), London, England,
Fundamental Understanding of Osmotic Membrane Butterworth-Heinemann (2001), 637–678.
Generation in Shales,” paper SPE 77447 presented at the
Morgan, B.E., and Dumbauld, G.K.: “Bentonite Cement
SPE Annual Technical Conference and Exhibition, San
Proving Successful in Permanent-Type Squeeze
Antonio, Texas, USA (September 29–October 2, 2002).
Operations,” World Oil (November 1954) 139, No. 6,
Mondshine, T.C.: “Method of simultaneously strengthen- 220–233.
ing the surface of borehole and bonding cement thereto
Morgan, B.E., and Dumbauld, G.K.: “Measurement of the
and method of forming cementicious pilings,” U.S.
Permeability of Set Cement,” Journal of Petroleum
Patent No. 4,014,174 (March 29, 1977).
Technology (June 1952) 4, No. 6, 16.
Monsalve, A., and Schechter, R.: “The Stability of Foams:
Morgan, D.R.: “Field Measurement of Strain and
Dependence of Observation on the Bubble Size
Temperature While Running and Cementing Casing,”
Distribution,” Journal of Colloid and Interface Science
paper SPE 19552 presented at the SPE Annual Technical
(February 1984) 97, No. 2, 327–335.

730 Well Cementing

Conference and Exhibition, San Antonio, Texas, USA Mueller, D.T., and Dillenbeck, R.L.: “The Versatility of
(October 8–11, 1989). Silica Fume as an Oilwell Cement Admixture,” paper
Moritis, G.: “TAML [Technical Advancement of SPE 21688 presented at the SPE Production Operations
MultiLaterals] Refocuses On Educating Industry On Symposium, Oklahoma City, Oklahoma (April 7–9, 1991).
Multilaterals,” Oil & Gas Journal (February 10, 2003) Mueller, D.T., Franklin, V.L., and Daulton, D.J.: “The
101, No. 6, 39–40, 42–44. Determination of the Static and Dynamic Properties of
Moroni, N., Calloni, G., and Marcotullio, A.: “Gas Nitrified Cements,” paper SPE 20116 presented at the
Impermeable Carbon Black Cements: Analysis of Field SPE Permian Basin Oil and Gas Recovery Conference,
Performances,” Offshore Mediterranean Conference, Midland, Texas, USA (March 8–9, 1990).
Ravenna, Italy (March 19–21, 1997). Mueller, D.T.: “Performance Characteristics of
Morris, C.W., Zanutto, S.L., and Dacres, W.G. Jr.: Vinylsulfonate-Based Cement Fluid-Loss Additives,”
“Lightweight Cement Evaluation Using Ultrasonic paper SPE 24380 presented at the SPE Rocky Mountain
Measurements,” paper SPE 62517 presented at the Regional Meeting, Casper, Wyoming, USA (May 18–21,
SPE/AAPG Western Regional Meeting, Long Beach, 1992).
California, USA (June 19–22, 2002). Mungan, N.: “Permeability Reduction Through Change
Morris, E.F.: “Evaluation of Cement Systems for in pH and Salinity,” Journal of Petroleum Technology
Permafrost,” paper SPE 2824 presented at the SPE Annual (December 1965) 17, No. 12, 1449–1453; also presented
Meeting, Denver, Colorado, USA (February 15–19, 1970). as paper SPE 1283 at the SPE Annual Meeting, Denver,
Colorado, USA (October 3–6, 1965).
Motley, H.R., Morris, E.F., and Pavlich, J.P.: “Use of a
Spacer Composition in Well Cementing,” U.S. Patent No. Murphey, J.R.: “Rapidly Dissolvable Silicates and
3,820,602 (June 28, 1974). Methods of Using the Same,” U.S. Patent No. 4,391,643
(July 5, 1983).
Moulin, E. et al.: “Using Concrete Technology to Improve
the Performance of Lightweight Cements,” paper SPE Muskhelishvili, N.I.: Some Basic Problems of the
39276, presented at the SPE/IADC Middle East Drilling Mathematical Theory of Elasticity, Kluwer Academic
Technology Conference, Bahrain (November 23–25, 1997). Publishers (April 1977).
Moulin, E.: “Fluid Loss Control Agents and Compositions Nägele, E.: “The Zeta-Potential of Cement,” Cement and
for Cementing Oil Wells Comprising Said Fluid Loss Concrete Research (May 1985) 15, No. 3, 453–462.
Control Agent,” U.S. Patent No. 6,180,689 (January 30, Nägele, E.: “The Zeta-Potential of Cement—Part II:
2001). Effect of pH-Value,” Cement and Concrete Research
Moulin, E.M., and Broyer, V.: “Effect of Superplasticizer (November 1986) 16, No. 6, 853–863.
Type on the Fluidity Retention of Portland Cement Nägele, E.: “The Zeta-Potential of Cement—Part III:
Mortars as a Function of the C3A Level and the Nature of The Non-Equilibrium Double Layer on Cement,” Cement
Added Calcium Sulfates,” paper presented at the 11th and Concrete Research (July 1987) 17, No. 4, 573–580.
International Congress on the Chemistry of Cement Nagelhout, A.C.G. et al.: “Laboratory and Field Validation
(2003). of a Sealant System for Critical Plug and Abandon
Mueller, D.T. et al.: “Portland Cement–Blast Furnace Situations,” paper SPE 97347 presented at the
Slag Blends in Oilwell Cementing Applications,” paper SPE/IADC Middle East Drilling Technology Conference
SPE 30513 presented at the SPE Annual Technical and Exhibition, Dubai, UAE (September 12–14, 2005).
Conference & Exhibition, Dallas, Texas, USA (October Nahm, J.J. et al.: “Interfacial Sealing Properties of
22–25, 1995). Slag Mix (Mud-to-Cement Conversion Technology):
Mueller, D.T., and Bray, W.S.: “Characterization of Laboratory and Field Evaluation,” paper SPE 29407
Surfactant-Enhanced Cement Fluid-Loss Additives,” presented at the SPE / IADC Drilling Conference,
paper SPE 25442 presented at the SPE Production Amsterdam, The Netherlands (February 28–March 2,
Operations Symposium, Oklahoma City, Oklahoma, USA 1995).
(March 21–23, 1993). Nahm, J.J. et al.: “Slag Mix Mud Conversion Cementing
Mueller, D.T., and Dickerson, J.P.: “Blast Furnace Slag Technology: Reduction of Mud Disposal Volumes
Technology: Features, Limitations, and Practical and Management of Rig-Site Drilling Wastes,” paper
Applications,” paper SPE 28475 presented at the SPE SPE 25988 presented at the SPE /Environmental
Annual Technical Conference and Exhibition, New Protection Agency (EPA) Exploration and Production
Orleans, Louisiana, USA (September 25–28, 1994). Environmental Conference, San Antonio, Texas, USA
(March 7–10, 1993).

References 731
Nahm, J.J.W., and Wyant, R.E.: “Method for Conversion Geothermal Environment,” Cement and Concrete
of Oil-base Mud to Oil Mud-cement,” U.S. Patent No. Research (May 1981) 11, No. 3, 371–381.
5,213,160 (May 25, 1993). Nelson, E.B.: “Improved cement slurry designed for ther-
Nataraja, M.C., Dhang, N., and Gupta, A.P.: “Toughness mal EOR wells,” Oil & Gas Journal (December 1, 1986)
Characterization of Steel Fiber-Reinforced Concrete by 84, No. 48, 39–44.
JSCE Approach,” Cement and Concrete Research (2000) Nelson, E.B.: “Portland Cements Characterized, Evaluated,”
30, No. 4, 593–597. Oil & Gas Journal (February 7, 1983) 81, No. 5, 73–77.
Nauman, E.B., and Buf fham, B.A.: Mixing in Nelson, E.B.: “Pumpable Thixotropic Cement Slurries
Continuous Flow Systems, New York, New York, USA, For Use in Cementing Pipes in a Well,” U.S. Patent
John Wiley & Sons (1983). No. 4,415,367 (November 15, 1983).
Navarrete, R.C., Seheult, J.M., and Coffey, M.D.: “New Nelson, E.B.: “Sulfonated Poly(Vinyl Aromatics) As
bio-polymers for drilling, drill-in, completion, spacer Fluid-Loss Additives for Salt Cement Slurries,” U.S.
fluids and coiled tubing applications,” paper SPE 62790 Patent No. 4,601,758 (July 22, 1986).
presented at the IADC/SPE Asia Pacific Drilling
Nelson, E.B.: “Well Treating Process and Composition,”
Technology, Kuala Lumpur, Malaysia (September 11–13,
Canadian Patent No. 1,216,742 (1987).
2000).
Nelson, E.B.: “Well Treating Composition,” U.K. Patent
Nayberg, T.M., and Linafelter, R.L.: “Controlling Cement
Application No. GB 2,157,279A (1984).
Circulation Loss to Both High-Permeability and
Fractured Formations,” paper SPE 12905 presented at Nelson, E.B.: “Improved Cement Slurry Design for
the SPE Rocky Mountain Regional Meeting, Casper, Thermal EOR Wells,” Oil & Gas Journal (December 1,
Wyoming, USA (May 21–23, 1984). 1986) 84, No. 48, 39–44.
Nayberg, T.M., and Petty, B.R., “Laboratory Study of Lost Ness, L., and Gatti, A.: “Gas Migration Remediation
Circulation Materials for Use in Both Oil-Based and Process Development,” paper 95-87 presented at the
Water-Based Drilling Muds,” SPE Drilling Engineering 46th Annual Technical Meeting of the Petroleum Society
(September 1987); also presented as paper SPE 14723 at of CIM, Banff, Canada (May 14–17, 1995).
the IADC/SPE Drilling Conference, Dallas, Texas, USA Neubauer, C.M., Yang, M., and Jennings, H.M.:
(February 10–12, 1986). “Interparticle Potential and Sedimentation Behavior
Nayfeh, T.H., Wheelis, W.B. Jr., and Leslie, H.D.: “The of Cement Suspensions: Effects of Admixtures,”
Fluid-Compensated Cement Bond Log,” SPE Formation Advanced Cement Based Materials (July 1998) 8,
Evaluation (August 1986) 1, No. 4, 335–341; also pre- No. 1, 17–27.
sented as paper SPE 13044 at the SPE Annual Technical Newlove, J.C. et al.: “Fluid Loss Control in Oil Field
Conference and Exhibition, Houston, Texas, USA Cements,” U.S. Patent No. 4,480,693 (November 6, 1984).
(September 16–19, 1984). Newman, K., Wojtanowicx, A., and Gahan, B.: “Improving
Nelson, E.B., and Casabonne, J.M.: “New Method for Gas Well Cement Jobs with Cement Pulsation,” GasTIPS
Better Control of Cement Performance in High- (Fall 2001) 7, No. 3, 29–33.
Temperature Wells,” paper SPE 24556 presented at the Newton, L.E. Jr., and Hausler, R.H. (eds.): “Carbon
SPE Annual Technical Conference and Exhibition, Dioxide Corrosion in Oil and Gas Production,” Selected
Washington, DC (October 4–7, 1992). Papers, Abstracts, and References, Houston, Texas, USA,
Nelson, E.B., and Eilers, L.H.: “Cementing Steamflood National Association of Corrosion Engineers (1984).
and Fireflood Wells—Slurry Design,” Journal of Ng, R.C., and Adisa, O.L.: “Coiled Tubing Resin Squeeze
Canadian Petroleum Technology (September–October to Mitigate Water Production in Offshore Gravel-Pack
1985) 24, No. 15, 58–63. Wells,” paper SPE 38836 presented at the SPE Annual
Nelson, E.B., and Eilers, L.H.: “High Temperature Technical Conference and Exhibition, San Antonio,
Expanding Cement Composition and Use,” U.S. Patent Texas, USA (October 5–8, 1997).
No. 4,328,036 (May 4, 1982). Nguyen, D., and Rahman, S.S.: “A computational algo-
Nelson, E.B., and Kalousek, G.L.: “Effects of Na2O on rithm for calculating displacement efficiency in hori-
Calcium Silicate Hydrates at Elevated Temperatures,” zontal annuli—Eccentric case,” Chemical Engineering
Cement and Concrete Research (November 1977) 7, Communications (May 2002) 189, No. 5, 695–709.
No. 6, 687–694. Nguyen, D., Kagan, M., and Rahman, S.S.: “Evaluation of
Nelson, E.B., Eilers, L.H., and Kalousek, G.L.: “Formation drilling fluid removal by cement slurry from horizontal
and Behavior of Calcium Silicate Hydrates in a wells with the use of an accurate mathematical model,”

732 Well Cementing

Journal of Petroleum Science and Engineering Odler, I., and Skalny, J.: “Hydration of Tricalcium Silicate
(October 1992) 8, No. 3, 191–204. at Elevated Temperatures,” Journal of Applied
Nguyen, D.: “Mathematical models of cuttings transport Chemistry and Biotechnology (September 1973) 23,
and drilling fluid displacement by cement slurry in hor- No. 9, 661–667.
izontal wells,” PhD thesis, New South Wales University, Odler, I., and Skalny, J.: “Influence of Calcium Chloride
Sydney, Australia (1997). on Paste Hydration of Tricalcium Silicate,” Journal of
Nilsson, F.: “Alkylglucosides’ Physical-Chemical the American Ceramic Society (July 1971a) 54, No. 7,
Properties,” INFORM (May 1996) 7, No. 5, 490–497. 362–363.
Nishikawa, S., and Wojtanowicz, A.: “Transient Pressure Odler, I., and Skalny, J.: “Pore Structure of Hydrated
Unloading—A Model of Hydrostatic Pressure Loss in Wells Calcium Silicates II. Influence of Calcium Chloride on
after Cement Placement,” paper SPE 77754 presented at the Pore Structure of β-Dicalcium Silicate,” Journal of
the SPE Annual Technical Conference and Exhibition, San Colloid and Interface Science (July 1971b) 36, No. 3,
Antonio, Texas, USA (September 29–October 2, 2002). 293–297.
NIST: Virtual Cement and Concrete Testing Laboratory, Odler, I., Duckstein, U., and Becker, Th.: “On the
http://vcctl.cbt.nist.gov (2001) (accessed October 10, Combined Effect of Water Soluble Lignosulfonates and
2005). Carbonates on Portland Cement and Clinker Pastes 1.
Physical Properties,” Cement and Concrete Research
Noik, C., Rivereau, A., and Vernet, C.: “Novel Cement
(July 1978) 8, No. 4, 469–479.
Materials for High-Pressure/High-Temperature Wells,”
paper SPE 50589 presented at the SPE European Ohama, Y.: “Principle of Latex Modification and Some
Petroleum Conference, The Hague, The Netherlands Typical Properties of Latex-Modified Mortars and
(October 20–22, 1998). Concretes,” ACI Materials Journal (November–
December 1987) 84, No.6, 511–518.
Noles, J. et al.: “Small-Capacity Cement Procedure
Reduces Failure Potential,” World Oil (May 1996) 217, Oil-Well Cementing Practices in the United States, API,
No. 5, 53–55. Washington, D.C., USA (1959).
Nolte, K.G.: “Determination of fracturing fluid loss rate Oliver, J.E., and Singer, A.M.: “Improved Well Cementing
from pressure decline curve,” U.S. Patent 4,398,416 Process,” European Patent No. 238,675 (1986).
(August 16, 1983). Onan, D.D. et al.: “Utilizing Drilling Fluid in Well
Nouri, J.M., Umur, H., and Whitelaw, J.H.: “Flow of Cementing Operations,” U.S. Patent No. 5,327,968
Newtonian and Non-Newtonian Fluids in Concentric and (July 12, 1994).
Eccentric Annuli,” Journal of Fluid Mechanics (August Onan, D.D., Goodwin, K.J., and McPherson, T.W.:
1993) 253, 617–641. “Elastomeric Composites for Use in Well Cementing
Novak, L.H.: “Drilling Mud Composition Which May Be Operations,” paper SPE 26572 presented at the SPE
Converted to Cement Upon Irradiation,” U.S. Patent Annual Technical Conference and Exhibition, Houston,
No. 4,547,298 (October 15, 1985). Texas, USA (October 3–6, 1993).
Novak, L.H.: “Method for Primary Cementing a Well Onan, D.D.: “Effects of Supercritical Carbon Dioxide on
Using a Drilling Mud Composition Which May Be Well Cements,” paper SPE 12593 presented at the SPE
Converted to Cement Upon Irradiation,” U.S. Patent Permian Basin Oil and Gas Recovery Conference,
No. 4,768,593 (September 6, 1988). Midland, Texas, USA (March 8–9, 1984).
Nowak, T.W. et al.: “Shut-Off of a Geopressured Water Orban, J., and Parcevaux, P.: “Viscometers Evaluated
Channel Behind Casing via Coiled Tubing Utilizing a Dual for Accurate Determination of Cement Slurry Rheology,”
Slurry Cement System: A Case History,” paper SPE 29474 Oil & Gas Journal (June 30, 1986) 84, No. 26, 94–100.
presented at the Production Operations Symposium, Orban, J.J. et al.: “New Ultrasonic Caliper for MWD
Oklahoma City, Oklahoma, USA (April 2–4, 1995). Operations,” paper SPE/IADC 21947 presented at the
Nowak, T.W., Patout, T.S., and Buzarde, C.B.: “Rigless SPE / IADC Drilling Conference, Amsterdam, The
Multizone Recompletion Using a Cement Packer Placed Netherlands (March 11–14, 1991).
with Coiled Tubing: A Case History,” paper SPE 35613 Ortenzi, L., Evans, L., and Maeso, C.J.: “An Integrated
presented at the Gas Technology Conference, Calgary, Caliper from Neutron, Density and Ultrasonic Azimuthal
Alberta, Canada (April 28–May 1, 1996). LWD Data,” paper SPE 77479 presented at the SPE
Odler, I., and Dorr, H.: “Early hydration of tricalcium Annual Technical Conference and Exhibition, San
silicate II. The induction period,” Cement and Concrete Antonio, Texas, USA (September 29–October 2, 2002).
Research (May 1979); 9, No. 3, 277–284.

References 733
Osisanya, S.O., and Griffith, J.: “Evaluation of cement (Influence of Curing Temperature and Pressure),”
slurry quality using filter cake permability and thick- Cement and Concrete Research (May 1984b) 14, No. 3,
ness,” paper CIM 97–136 presented at the CIM 419–430.
Petroleum Society Technology Meeting, Calgary, Canada Pardue, G.H., Morris, R.L., and Gollwitzer, L.H.: “Cement
(June 8–11, 1997). Bond Log—A Study of Cement and Casing Variables,”
Ost, B.W.: “Optimum Sulfate Content of Portland Journal of Petroleum Technology (May 1963), 545–554;
Cements,” American Ceramic Society Bulletin (August also presented as paper SPE 453 at the SPE Annual
1974) 53, No. 8, 579–580. Meeting, Los Angeles (October 7–10, 1962).
Oswald, R. et al.: “Well Cementing Aids,” U.S. Patent Park, S.B., Yoon, E.S., and Lee, B.I.: “Effects of
No. 6,277,900 (August 21, 2001). Processing and Materials Variations on Mechanical
Oswald, R.J. et al.: “Water-Soluble Copolymers And Their Properties of Lightweight Cement Composites,” Cement
Use For Exploration And Production Of Petroleum And and Concrete Research (1999) 29, No. 2, 193–200.
Natural Gas,” U.S. Patent No. 6,395,853 (May 28, 2002). Parker, D.G.: “Microsilica Concrete. Part 1: The
Outmans, H.D.: “Mechanics of Static and Dynamic Material,” Current Practice Sheet No. 104, Concrete
Filtration in the Borehole,” SPE Journal (September Society (UK) (October 1985).
1963) 3, No. 3, 236–244; also presented as paper SPE 491 Parker, P.N. et al.: “An Evaluation of a Primary
at the University of Texas-SPE Drilling and Rock Cementing Technique Using Low Displacement Rates,”
Mechanics Conference, Austin, Texas, USA (January paper SPE 1234 presented at the SPE Annual Meeting,
23–24, 1963). Denver, Colorado, USA (October 3–6, 1965).
Pace, R.S. et al.: “Improved Bulk Blending Techniques Parlar, M. et al.: “Filtercake Cleanup Techniques for
for Accurate and Uniform Cement Blends,” paper SPE Openhole Water Injectors with Sand Control: Lessons
13041 presented at the SPE Annual Technical from Laboratory Experiments and Recommendations for
Conference and Exhibition, Houston, Texas, USA Field Practices,” paper SPE 77449 presented at the SPE
(September 16–19, 1984). Annual Technical Conference and Exhibition, San
Pappas, J.M., Creel, P.G., and Crook, R.J.: “Identifying Antonio, Texas, USA (September 29–October 2, 2002).
Water-Flow Problems,” Journal of Petroleum Paske, W.C. et al.: “Theory and Implementation of a
Technology (August 1995) 47, No. 8, 699 (SPE 30231). Borehole Caliper Measurement Made While Drilling,” SPE
Parcevaux, P., and Jennings, D.B.: “Aqueous spacer com- Formation Evaluation (June 1992) 7, No. 2, 145–150.
position compatible with drilling muds and cement slur- Patel, A.D., and Salandanan, C.S.: “Thermally Stable
ries,” U.S. Patent 5,101,902 (April 7, 1992). Polymeric Gellant for Oil-Base Drilling Fluids,” paper
Parcevaux, P.A. et al.: “Annular Gas Flow, a Hazard SPE 13560 presented at the SPE International
Free Solution,” Petroleum Information (July 15, 1983) Symposium on Oilfield and Geothermal Chemistry,
No. 1589, 34–38. Phoenix, Arizona, USA (April 9–11, 1985).
Parcevaux, P.A. et al.: “Cement Compositions for Patel, A.D.: “Reversible Invert Emulsion Drilling
Cementing Wells, Allowing Pressure Gas-Channeling in Fluids—A Quantum Leap in Technology,” paper SPE
the Cemented Annulus to Be Controlled,” U.S. Patent 47772 presented at the IADC/SPE Asia Pacific Drilling
No. 4,537,918 (August 27, 1985). Technology Conference, Jakarta, Indonesia (September
7–9, 1998).
Parcevaux, P.A., and Sault, P.H.: “Cement Shrinkage and
Elasticity: A New Approach for a Good Zonal Isolation,” Patel, V.C., and Head, M.R.: “Some Observations on Skin
paper SPE 13176 presented at the SPE Annual Technical Friction and Velocity Profiles in Fully Developed Pipe
Conference and Exhibition, Houston, Texas, USA and Channel Flows,” Journal of Fluid Mechanics (1969)
(September 16–19, 1984). 38, No. 1, 181–201.
Parcevaux, P.A.: “Gas Migration and GASBLOK ™ Tech- Pattillo, P.D., and Kristiansen, T.G.: “Analysis of
nology,” Drilling & Pumping Journal (August 1987) No. Horizontal Casing Integrity in the Valhall Field,” paper
3, 1122. SPE/International Society for Rock Mechanics (ISRM)
78204 presented at the SPE/ISRM Rock Mechanics
Parcevaux, P.A.: “Mechanisms of Gas Channeling During
Conference, Irvine, Texas, USA (October 20–23, 2002).
Primary Cementation—Methods for Prevention and
Repair,” Chemische Produkte in der Erdolgewinnung, Pauri, M. et al.: “Effect of Triethanolamine on the
Clausthal – Zellerfeld (September 6, 1984a). Tricalcium Silicate Hydration,” Proceedings of the 8th
International Symposium on the Chemistry of Cement
Parcevaux, P.A.: “Pore Size Distribution of Portland
(1986) 3, 125–129.
Cement Slurries at Very Early Stage of Hydration

734 Well Cementing

Peiffer, D.G. et al.: “Fluid Loss Control in Oil Field Pettitt, R.A.: “Completion of Hot Dry Rock Geothermal
Cements,” U.S. Patent No. 4,626,285 (December 2, 1986). Well Systems,” paper SPE 8267 SPE Annual Technical
Peiffer, D.G. et al.: “Fluid Loss Control in Oil Field Conference and Exhibition, Las Vegas, Nevada, USA
Cements,” U.S. Patent No. 4,683,952 (August 4, 1987). (September 23–26, 1979).
Pelipenko, S., and Frigaard, I.A.: “Mud Removal and Philip, Z., Sharma, M.M., and Chenevert, M.E.: “The Role
Cement Placement during Primary Cementing of an Oil Of Taylor Vortices In The Transport Of Drill Cuttings,”
Well,” Journal of Engineering Mathematics (January paper SPE 39504 presented at the SPE India Oil and Gas
2004) 48, No. 1, 1–26. Conference, New Delhi, India (April 7–9, 1998) 77–83.
Pelipenko, S., and Frigaard, I.A.: “On steady state dis- Philippacopoulos, A.J., and Berndt, M.L.: “Mechanical
placements in primary cementing of an oil well,” Journal Response and Characterization of Well Cements,” paper
of Engineering Mathematics (2004) 48, No. 1, 1–26. SPE 77755 presented at the SPE Annual Technical
Conference and Exhibition, San Antonio, Texas, USA
Peralta, M.: “Control de la invasion de gas durante las
(September 29–October 2, 2002).
operaciones de cementacion,” Petroleo Internacional
(July–August 1984) 42, No. 4, 30–37. Pickett, G.R.: “Acoustic Character Logs and Their
Applications in Formation Evaluation,” Journal of
Persinski, L.J., Cook, M.M., and Adams, S.L.: “Low Fluid
Petroleum Technology (June 1963) 15, No. 6, 659–667;
Loss Cementing Compositions Containing Hydrolyzed
also presented as paper SPE 452 at the SPE Annual
Acrylamide/2-Acrylamido-2-Methylpropane Sulfonic
Meeting, Los Angeles, California, USA (October 7–10,
Acid Derivative Copolymers and Their Use,” U.S. Patent
1962).
No. 4,015,991 (April 5, 1977).
Pilkington, P.E.: “Pressure needed to reduce microannu-
Pessin, J-L., and Boyle, B.W.: “Accuracy and Reliability
lus effect on CBL,” Oil & Gas Journal (May 30, 1988) 86,
of Coiled Tubing Depth Measurement,” paper SPE 38422
No. 22, 68–74.
presented at the SPE North American Coiled Tubing
Roundtable, Montgomery, Texas, USA (April 1–3, 1997). Piot, B. et al.: “West Africa Deepwater Wells Benefit from
Low-Temperature Cements,” paper SPE/IADC 67774 pre-
Petit, H.J-M., Valentin, E.P., and Desmarquest, J-P.:
sented at the SPE/IADC Drilling Conference, Amsterdam,
“Air /O 2 Fireflood: Comparison of Field-Scale
The Netherlands (February 27–March 1, 2001).
Performance and Economics,” paper SPE 24082 pre-
sented at the SPE Western Regional Meeting, Piot, B., and Loizzo, M.: “Reviving the Job Signature
Bakersfield, California, USA (March 30–April 1, 1992). Concept for Better Quality Cement Jobs,” paper
lADC/SPE 39350 presented at the lADC/SPE Drilling
Petroleum and natural gas industries - Cements and
Conference, Dallas, Texas, USA (March 3–6, 1998).
materials for well cementing - Part 1: Specification
(ISO 10426-1) (2000) with Addendum 1 (ISO 10426-1 Piot, B., Ferriere, R., and Fraboulet, B.: “A unique
AMD 1) (2002), ISO, Geneva, Switzerland. experience with foamed cement,” paper SPE 28820
presented at the SPE European Petroleum Conference,
Petroleum and natural gas industries - Cements and
London, UK (October 25–27, 1994).
materials for well cementing - Part 2: Testing of well
cements (ISO 10426-2), ISO, Geneva, Switzerland (2003). Pittet, C., and Lemaître, J.: “Mechanical Characterization
of Brushite Cements: A Mohr Circles’ Approach,”
Petroleum and natural gas industries - Cements and
Journal of Biomedical Materials Research (November
materials for well cementing - Part 3: Testing of deep-
2000) 53, No. 6, 769–780.
water well cement formulations (ISO 10426-3), ISO,
Geneva, Switzerland (2003). Plett, E.A.: “Cleaning of fouled surfaces,” Fouling and
cleaning in food processing, C. Sandu, D.B. Lund, and
Petroleum and natural gas industries - Cements and
E.A. Plett (eds.) University of Wisconsin, Madison,
materials for well cementing - Part 4: Preparation
Wisconsin (1985) 286–311.
and testing of foamed cement slurries at atmospheric
pressure (ISO 10426-4), ISO, Geneva, Switzerland (2004). Plona, T.J., and Cook, J.M.: “Effect of Stress Cycles on
Static and Dynamic Young’s Moduli in Castlegate
Petroleum and natural gas industries - Cements and
Sandstone,” paper presented at the 35th Rock
materials for well cementing - Part 5: Test methods for
Mechanics Symposium, University of Nevada, Reno,
determination of shrinkage and expansion of well
Nevada, USA (June 5–7, 1995).
cement formulations at atmospheric pressure (ISO
10426-5), ISO, Geneva, Switzerland (2004). Pokhriyal, J., Gaudlip, T., and Suter, W.R.: “Use of
Concrete Technology for High-Density Cement Systems
Pettijohn, F.J.: Sedimentary Rocks, New York City, New
in South Texas,” paper SPE 67259 presented at the SPE
York, USA, Harper and Row (1975), 260–299.

References 735
Production and Operations Symposium, Oklahoma City, the SPE Western Regional Meeting, Long Beach,
Oklahoma, USA (March 24–27, 2001). California, USA (March 23–25, 1994).
Portland Cements, Skokie, Illinois, USA, Portland Prohaska, M., Ogbe, D., and Economides, M.:
Cement Association (2002). “Determining Wellbore Pressures in Cement Columns,”
Postler, D.P.: “Pressure Integrity Test Interpretation,” SPE 26070 presented at the SPE Western Regional
paper SPE 37589 presented at the SPE/IADC Drilling Meeting, Anchorage, Alaska, USA (May 26–28, 1993).
Conference, Amsterdam, The Netherlands (March 4–6, Purvis, D.L., and Smith, D.D.: “Real-Time Monitoring
1997). Provides Insight into Flow Dynamics during Foam
Powell, J.W. et al.: “Thixotropic, Crosslinking Cementing,” SPE Drilling & Completion (June 1994)
Polymer/Borate/Salt Plug: Development and Application,” 38, No. 1, 124–132; also presented as paper SPE 24570 at
paper SPE 22068 presented at the SPE International the SPE Annual Technical Conference and Exhibition,
Arctic Technology Conference, Anchorage, Alaska, USA Washington, D.C. (October 4–7, 1992).
(May 29–31, 1991). Queirós, J.G.R. et al.: “Through Tubing Rotary Drilling
Powell, J.W., and Seheult, J.M.: “Xanthan and Welan: The and its Associated Cement Challenges: A North Sea
Effect of Critical Polymer Concentration on Rheology Experience,” paper SPE 83955 presented at the 2003
and Fluid Performance,” paper SPE 22066 presented at Offshore Europe Oil and Gas Exhibition and Conference,
the SPE International Arctic Technology Conference, Aberdeen, UK (September 2–4, 2003).
Anchorage, Alaska, USA (May 29–31, 1991). Quemada, D., “Rheological modeling of complex fluids.
Power, D.J. et al.: “Drilling practices and sweep selection I. The concept of effective volume fraction revisited,”
for efficient hole cleaning in deviated wellbores,” paper The European Physical Journal—Applied Physics
SPE 62794 presented at the IADC/SPE Asia Pacific (January 1998) 1, No. 1, 119–127.
Drilling Technology Conference, Kuala Lumpur, Quinn, D., Sunde, E., and Baret, J.F.: “Mechanism of a
Malaysia (September 11–13, 2000). Novel Shear-Sensitive Plugging Fluid To Cure Lost
Powers, C.A., Smith, R.C., and Holman, G.B.: “Low Circulation,” paper SPE 50722 presented at the SPE
Density Cement Slurry and Its Use,” U.S. Patent No. International Symposium on Oilfield Chemistry,
4,370,166 (January 25, 1983). Houston, Texas, USA (February 16–19, 1999).
Powers, T.C.: “Adsorption of Water by Portland Cement Quon, D.H.H., and Malhotra, V.M.: “Performance of High
Paste During the Hardening Process,” Industrial and Alumina Cement Concrete at Elevated Temperature,”
Engineering Chemistry (1935) 27, 790–794. Journal of the Canadian Ceramic Society (1979) 48,
7–16.
Powers, T.C.: “Some Aspects of the Hydration of Portland
Cement,” Journal of Portland Cement Association Rabinowitsch, B.: “über die Viskosität und Elastizität
Research and Development Lab (1961) 3, 47–56. von Solen,” Zeitschrift fur Physikalische Chemie Series
A (1929) 145, 1–26.
Powers, T.C.: “Structure and Physical Properties of
Hardened Cement Paste,” Journal of the American Radenti, G., and Ghiringelli, G.: “Cementing materials
Ceramic Society (1958) 41, No. 1, 1–6. for geothermal wells,” Geothermics (September 1972) 1,
No. 3, 119–123.
Previte, R.W.: “Some Insights on the Mechanism of
Saccharide Set Retardation of Portland Cement,” Rae, G. et al.: “A Case Study in the Design and Execution
Cement and Concrete Research (May 1971) 1, No. 3, of Subsea Production Development Wells in the Captain
301–316. Field,” paper SPE 81029 presented at the SPE Latin
American and Carribbean Petroleum Engineering
Prince, W., Espagne, M., and Aïtcin, P.-C.: “Ettringite
Conference, Port of Spain, Trinidad and Tobago
Formation: a Crucial Step in Cement Superplasticizer
(April 27–30, 2003).
Compatibility,” Cement and Concrete Research (May
2003) 33, No. 5, 635–641. Rae, G., Williams, H., and Hamilton, J.: “Selective
Flotation of Casing From a Floating Vessel,” SPE
Princen, H.M.: “Rheology of Foams and Highly
Drilling & Completion (June 2004) 19, No. 2, 94–103.
Concentrated Emulsions,” Journal of Colloid and
Interface Science (1982) 91, No. 1, 160–175. Rae, P., and Brown, E.: “New Materials Improves the
Cementation of Salt Formations,” Proceedings of the
Prohaska, M., Fruhwirth, R.K., and Economides, M.J.:
1988 Southwestern Petroleum Short Course, Lubbock,
“Modeling Early-Time Gas Migration Through Cement
Texas, USA, Southwestern Petroleum (1988) 38–48.
Slurries,” SPE Drilling & Completion (September 1995)
10, No. 3, 178–185; also presented as paper SPE 27878 at Rae, P., and Johnston, N.: “Liquid Cement Premix for
Improved Abandonment and Workover Operations,”

736 Well Cementing

paper SPE 36477 presented at the SPE Annual Technical (November 1982) 34, No. 11, 2611–2616; also presented
Conference and Exhibition, Denver, Colorado, USA as paper SPE 9283 at the SPE Annual Technical
(October 6–9, 1996b). Conference and Exhibition, Dallas, Texas, USA
Rae, P., and Johnston, N.: “Liquid Cement Premix (September 21–24, 1980).
Introduces New Solutions to Conventional Cementing Rao, S.P., and Burkhalter, J.F.: “Oil Field Cementing
Problems,” paper SPE 35086 presented at the IADC/SPE Methods and Compositions,” European Patent
Drilling Conference, New Orleans, Louisiana, USA Application No. 0 192 447 (1986).
(March 12–15, 1996a). Ravi, K., Bosma, M., and Gastebled, O.: “Improve the
Rae, P., and Johnston, N.: “Storable liquid cementitious Economics of Oil and Gas Wells by Reducing the Risk of
slurries for cementing oil and gas wells,” U.S. Patent No. Cement Failure,” paper SPE 74497 presented at the
5,447,197 (September 5, 1995). IADC/SPE Drilling Conference, Dallas, Texas, USA
Rae, P., Wilkins, D., and Free, D.: “A New Approach for (February 26–28, 2002).
Predicting Gas Flow After Cementing,” paper SPE 18622 Ravi, K.M., and Beirute, R.M.: “Methods and apparatus
presented at the SPE/IADC Drilling Conference, New for measuring the erodability of drilling fluid deposits,”
Orleans, Louisiana, USA (February 28–March 3, 1989). U.S. Patent No. 5,309,761 (May 10, 1994).
Rae, P.: “Cementing: Preventing Gas Migration During Ravi, K.M., and Sutton, D.L.: “New Rheological
Setting,” Noroil (March 1987) 15, No. 3, 31. Correlation for Cement Slurries as a Function of
Rakhimkulov, R.Sh., and Strugovets, E.T.: “Use of Temperature,” paper SPE 20449 presented at the SPE
Vibrations for Improving the Quality of Cementing Technical Conference and Exhibition, New Orleans,
Casing,” Neftyanoe Khozyaisto (June 1990) No. 6, Louisiana, USA (September 23–26, 1990).
35–38. Ravi, K.M., Beirute, R.M., and Covington, R.L.:
Ramachandran, V.S., and Beaudoin, J.J.: “Hydration of “Erodability of partially dehydrated gelled drilling fluid
C4AF + Gypsum: Study of Various Factors,” Proceedings and filter cake,” paper SPE 24571 presented at the SPE
of the 7th International Congress on the Chemistry of Annual Technical Conference and Exhibition,
Cement (1980), 2, 25–30. Washington, DC, USA (October 4–7, 1992).
Ramachandran, V.S.: “Action of Triethanolamine on Ravi, K.M., Beirute, R.M., and Covington, R.L.: “Improve
the Hydration of Tricalcium Aluminate,” Cement and primary cementing by continuous monitoring of circu-
Concrete Research (January 1973) 3, No. 1, 41–54. latable hole,” paper SPE 26574 presented at the SPE
Annual Technical Conference and Exhibition, Houston,
Ramachandran, V.S.: “Admixture and Addition
Texas, USA (October 3–6, 1993).
Interactions in the Cement-Water System,” Il Cemento
(January 1986) 83, No. 1, 13–38. Ray, W.J., and Hines, D.G.: “Method of cleaning a well
bore prior to cementing,” U.S. Patent No. 5,904,208
Ramachandran, V.S.: “Hydration of Cement—Role of
(May 18, 1999).
Triethanolamine,” Cement and Concrete Research
(September 1976a) 6, No. 5, 623–631. Raymondeau, M. et al.: “Three years with no squeeze
at ELF Nigeria: key factors for successful primary
Ramachandran, V.S.: “Interaction of Calcium
cementing,” paper SPENC 9103 presented at the SPE
Lignosulfonate with Tricalcium Silicate, Hydrated
Nigerian Council Conference, Port Harcourt, Nigeria
Tricalcium Silicate, and Calcium Hydroxide,” Cement
(August 28–30, 1991).
and Concrete Research (March 1972) 2, No. 2, 179–194.
Read, D. et al.: “Mineralogical and Microstructural
Ramachandran, V.S.: Calcium Chloride in Concrete, UK,
Changes Accompanying the Interaction of Boom Clay
Applied Science Publishers (1976b).
with Ordinary Portland Cement,” Advances in Cement
Randall, C.A., and Stanke, F.E.: “Mathematical model for Research (January 2001) 13, No. 4, 175–183.
internal ultrasonic inspection of cylindrically layered
Recommended Practice for Analysis of Oilfield Waters
structures,” Journal of the Acoustical Society of
(API RP 45) (3rd ed.), API, Washington, D.C., USA (1998).
America (April 1988) 83, 4, 1295–1305.
Recommended Practice for Cementing Shallow Water
Rang, C.L.: “Evaluation of Gas Flows in Cement,” paper
Flow Zones in Deepwater Wells (API RP 65) (1st ed.,
SPE 16385 presented at the SPE California Regional
2002) (Errata 2003), API, Washington, D.C., USA.
Meeting, Ventura, California, USA (April 8–10, 1987).
Recommended Practice for Field Testing Water-Based
Rao, P.P., et al.: “An Ultrasonic Device for Nondestructive
Drilling Fluids (API RP 13B-1) (3rd ed.), API,
Testing of Oilwell Cements at Elevated Temperatures
Washington, D.C., USA, American Petroleum Institute
and Pressures,” Journal of Petroleum Technology
(2003).

References 737
Recommended Practice for Performance Testing of 1982) 34, No. 1, 37–45; also presented as paper SPE 9755
Cementing Float Equipment (API RP 10F) (3rd ed.) at the SPE Production Operations Symposium,
API, Washington, D.C., USA (2002). Oklahoma City, Oklahoma, USA (March 1–3, 1981).
Recommended Practice for Testing Well Cements (API Rike, J.L.: “Obtaining Successful Squeeze-Cementing
RP 10B) (22nd ed.) (1997) with Addendum 1 (1999), Results,” paper SPE 4608 presented at the SPE
API, Washington, D.C., USA. Annual Meeting, Las Vegas, Nevada, USA (September 30–
Recommended Practice on Testing of Deepwater Well October 3, 1973).
Cement Formulations (API RP 10B-3) (1st ed.) API, Ripley, H.E. et al.: “Ultra-Low Density Cementing
Washington, D.C., USA, American Petroleum Institute Compositions,” paper 80-31-19 presented at the 31st
(2004). Annual Petroleum Society of CIM Technical Meeting,
Reed, T.D., and Pilehvari, A.A.: “A New Model for Calgary, Alberta, Canada (May 25–28, 1980).
Laminar, Transitional, and Turbulent Flow of Drilling Rixom, M.R.: “Development of an Admixture to Produce
Muds,” paper SPE 25456 presented at the SPE Flowing or Self-Compacting Concrete,” Precast Concrete
Production Operations Symposium, Oklahoma City, (November 1974) 5, No. 11, 633–637.
Oklahoma, USA (March 21–23, 1993). Roark, D.N., Nugent, A. Jr., and Bandlish, B.K.: “Fluid
Reese, D.W., Pace, R.S., and McKenzie, L.F.: “Non- Loss Control and Compositions for Use Therein,”
Retarding Fluid Loss Additives for Well Cementing European Patent Application No. 0 201 355 (1986).
Compositions,” U.S. Patent No. 4,610,306 (September 9, Roark, D.N., Nugent, A. Jr., and Bandlish, B.K.: “Fluid
1986). Loss Control in Well Cement Slurries,” U.S. Patent
Reeves, N.K., Bailey, D.E., and Caveny, W.J.: “Microscopic No. 4,657,948 (April 14, 1987a).
Analysis of Dry Cement Blends,” paper SPE 11820 pre- Roark, D.N., Nugent, A. Jr., and Bandlish, B.K.: “Fluid
sented at the SPE Rocky Mountain Regional Meeting, Loss Control in Well Cement Slurries,” U.S. Patent
Salt Lake City, Utah (May 23–25, 1983). No. 4,698,380 (October 6, 1987b).
Reidenbach, V.G. et al.: “Rheological Study of Foam Roark, D.N., Nugent, A. Jr., and Bandlish, B.K.: “Fluid
Fracturing Fluids using Nitrogen and Carbon Dioxide,” Loss Control In Well Cement Slurries,” U.S. Patent
SPE Production Engineering (January 1986), 1, No. 1, No. 4,706,755 (November 17, 1987c).
31–41; also presented as paper SPE 12026 at the
Robertson, R.E., and Stiff, H.A.: “An Improved
SPE Annual Technical Conference and Exhibition,
Mathematical Model for Relating Shear Stress to
San Francisco, California, USA (October 5–8, 1983).
Shear Rate in Drilling Fluids and Cement Slurries,”
Revil, P., and Jain, B.: “A New Approach to Designing SPE Journal (February 1976) 16, No. 1, 31–36; also
High-performance Lightweight Cement Slurries for presented as paper SPE 5533 at the SPE Rocky
Improved Zonal Isolation in Challenging Situations,” Mountain Regional Meeting, Denver, Colorado, USA
paper SPE 47830 presented at the IADC/SPE Asia (April 7–9, 1975).
Pacif ic Drilling Technology, Jakarta, Indonesia
Robson, T.D.: High-Alumina Cements and Concretes,
(September 7–9, 1998).
New York City, New York, USA, John Wiley & Sons, Inc.
Rich, K. et al.: “Carbon Isotope Characterization of (1962), 184–185.
Migrating Gas in the Heavy Oil Fields of Alberta, Canada,”
Rodrigues, K.A., and Lindsey, D.W.: “Set Retarded Ultra
paper SPE 30265 presented at the International Heavy Oil
Fine Cement Compositions and Methods,” U.S. Patent
Symposium in Calgary, Alberta, Canada (June 19–21,
No. 5,398,759 (March 21, 1995).
1995).
Rodriguez, W.J., Fleckstein, W.W, and Eustes, A.W.:
Richardson, E.A.: “Nitrogen Gas Stabilized Cement and a
“Simulation of Collapse Loads on Cemented Casing
Process for Making and Using it,” U.S. Patent 4,333,764
Using Finite Element Analysis,” paper SPE 85466 pre-
(June 8, 1982).
sented at the SPE Annual Technical Conference and
Richardson, J.F., and Zaki, W.N.: “Sedimentation and Exhibition, Denver (October 5–8, 2003).
Fluidization,” Transactions of the Institution of Chemical
Roegiers, J.-C.: “Elements of Rock Mechanics,”
Engineers (1954) 32, No. 1, 35–53.
Reservoir Stimulation (2nd ed.), M.J. Economides and
Rickard, W.M.: “Foam Cement for Geothermal Wells,” K.G. Nolte (eds.), Houston, Texas, USA, Schlumberger
Transactions of the Geothermal Resources Council Educational Services (1989), 2-1 to 2-22.
(1985) 9, Part 1, 147–152.
Rollins, J.T., and Davidson, R.D.: “New Latex Cement
Rike, J.L., and Rike, E.R.: “Squeeze Cementing: State of Solves Special Well Problems,” Petroleum Engineer
the Art,” Journal of Petroleum Technology (January (February 1957) 29, No. 2, B48–51.

738 Well Cementing

Romero, J., and Touboul, E.: “Temperature Prediction for Rueda, F., Heathman, J., and Serrano, M.: “Hole Cleaning
Deepwater Wells: A Field Validated Methodology,” paper and Cement Design for Specific Formation Types,”
SPE 49056 presented at the SPE Annual Technical paper SPE 84560 presented at the SPE Annual Technical
Conference and Exhibition, New Orleans, Louisiana, Conference and Exhibition, Denver, Colorado, USA
USA (September 27–30, 1998). (October 5–8, 2003).
Root, R.L., and Calvert, D.G.: “The Real Story of Cement Runyan, E.E.: “Cementing of Well Casings for Pollution
Expansion,” paper SPE 3346 presented at the SPE Rocky Control,” paper SPE 4812 presented at the SPE Permian
Mountain Regional Meeting, Billings, Montana, USA Basin Oil Recovery Conference, Midland, Texas, USA
(June 2–4, 1971). (March 11–12, 1974).
Rosenburg, A.M. et al.: “A Corrosion Inhibitor Rusch, D.W., and Ellis, B.C.: “Use of Pressure-Activated
Formulated with Calcium Nitrite for Use in Reinforced Sealants to Cure Sources of Casing Pressure,” paper SPE
Concrete,” STP-629, Annual Book of ASTM Standards 58 55996 presented at the SPE Western Regional Meeting,
Philadelphia, Pennsylvania, USA, ASTM (1977), 228–240. Anchorage, Alaska, USA (May 26–28, 1999).
Ross, S.: “Bubbles and Foam,” Industrial & Engineering Russell, J.: “Remedial Cementing in a Low Pressure
Chemistry (1969) 61, No. 10, 48–57. Formation,” Drilling (May 1983) 44, No. 8, 82–83.
Rossington, D.R., and Runk, E.J.: “Adsorption of Rust, C.F., and Wood, W.D.: “Laboratory Evaluations and
Admixtures on Portland Cement Hydration Products,” Field Testing of Silica-CMHEC-Cement Mixtures,”
Journal of the American Ceramic Society (January Journal of Petroleum Technology (November 1966) 18,
1968) 51, No. 1, 46–50. No. 11, 25–29.
Rosskopf, P.A., Linton, F.J., and Peppler, P.B.: “Effect of Ryan, D.F., Browne, S.V., and Burnham, M.P.: “Mud
Various Accelerating Chemical Admixtures on Setting Clean-Up in Horizontal Wells: A Major Joint Industry
and Strength Development of Concrete,” Journal of Study,” paper SPE 30528 presented at the SPE Annual
Testing and Evaluation (July 1975) 3, No. 4, 322–330. Technical Conference & Exhibition, Dallas, Texas, USA
Rosskopf, P.A.: “Additive for hydraulic cement mixes,” (October 22–25, 1995).
U.S. Patent No. 4,373,956 (February 15, 1983). Ryan, D.F., Kellingray, D.S., and Lockyear, C.F.:
Roy, D.M. et al.: “New High Temperature Cementing “Improved Cement Placement on North Sea Wells Using
Materials for Geothermal Wells: Stabilit y and a Cement Placement Simulator,” paper SPE 24977 pre-
Properties,” Brookhaven National Laboratory Final sented at the SPE European Petroleum Conference,
Report BNL 51249, New York City, New York, USA, Cannes, France (November 16–18, 1992).
Brookhaven National Laboratory (1980). Ryan, N.W., and Johnson, M.M.: “Transition from
Roy, D.M. et al.: “Potential New High Temperature Laminar to Turbulent Flow in Pipes,” AIChE Journal
Cements for Geothermal Wells,” paper SPE 7877 (December 1959) 5, No. 4, 433–435.
presented at the SPE International Symposium on Saasen, A. et al.: “The Use of Blast Furnace Slag in
Oilfield and Geothermal Chemistry, Houston, Texas, USA North Sea Cementing Applications,” paper SPE 28821
(January 22–24, 1979). presented at the SPE European Petroleum Conference,
Roy, D.M., and Asaga, K.: “Rheological Properties of London, U.K., (October 25–27, 1994).
Cement Mixes: III. The Effects of Mixing Procedures Saasen, A.: “Hole Cleaning During Deviated Drilling—
on Viscometric Properties of Mixes Containing The Effects of Pump Rate and Rheology,” paper SPE
Superplasticizers,” Cement and Concrete Research 50582 presented at the SPE European Petroleum
(November 1979) 9, No. 6, 731–739. Conference, The Hague, The Netherlands (October
Rubiandini, R.S.: “New Additive for Improving 20–22, 1998).
Shearbond Strength in High Temperature and Pressure Sabins, F., Faul, R., and Sutton, D.L.: “Preventing
Cement,” paper SPE 62750 presented at the IADC/SPE Annular Gas Flow, Part 1: Annular Gas Flow Theory and
Asia Pacific Drilling Technology, Kuala Lumpur, Prevention Method Described,” Oil & Gas Journal
Malaysia (September 11–13, 2000). (December 10, 1984a) 82, No. 50, 84–86, 91–92.
Rucker, R., Sauvageau, K., and Randolph, T.: “A New Sabins, F., Faul, R., and Sutton, D.L.: “Preventing
Approach to an Old Problem—Lightweight Liquid Annular Gas Flow, Part 2: New Evaluation for Annular
Concrete,” paper SPE 67193 presented at the SPE Gas-flow Potential,” Oil & Gas Journal (December 17,
Production and Operations Symposium, Oklahoma City, 1984b) 82, No. 51, 109–112.
Oklahoma, USA (March 24–27, 2001).

References 739
Sabins, F.L. et al.: “Critical Evaluation of Blast Furnace Santarelli, F.J., Dardeau, C., and Zurdo, C.: “Drilling
Slag Mud Converted to Cement,” paper SPE 35085 Through Highly Fractured Formations: A Problem, a
presented at the IADC/SPE Drilling Conference, New Model, and a Cure,” paper SPE 24592 presented at the
Orleans, Louisiana, USA (March 12–15, 1996). SPE Annual Technical Conference and Exhibition,
Sabins, F.L., and Maki, V., “Acoustic method for deter- Washington, DC, USA (October 4–7, 1992).
mining the static gel strength of a cement slurry,” Sasaki, S., Kobayashi, W., and Okabayashi, S.: “Strength
U.S. Patent 5,992,223 (November 30, 1999). Development of 2CaO · SiO2—Silica Cement under
Sabins, F.L., and Sutton, D.L.: “Interrelationship High-Temperature and High-Pressure Condition,” SPE
Between Critical Cement Properties and Volume Production Engineering (January 1986) 1, No. 1, 42–48
Changes During Cement Setting,” SPE Drilling (SPE 13406).
Engineering (June 1991) 6, No. 2, 88–94 (SPE 22952). Satava, V., and Veprek, O.: “Thermal Decomposition of
Sabins, F.L., Sutton, D.L., and Cook, C. Jr.: “The Effect of Ettringite under Hydrothermal Conditions,” Journal of
Excessive Retardation on the Physical Properties of the American Ceramic Society (July–August, 1975) 58,
Cement Slurries,” Journal of Petroleum Technology No. 7–8, 357–359.
(August 1984) 36, No. 8, 1357–1365; also presented as Sauer, C.W.: “Mud Displacement During the Cementing
paper SPE 10221 at the SPE Annual Technical Operation: A State of the Art,” Journal of Petroleum
Conference and Exhibition, San Antonio, Texas, USA Technology (September 1987) Vol. 39, No. 9, 1091–1101
(October 5–8, 1981). (paper SPE 14197).
Sabins, F.L., Tinsley, J.M., and Sutton, D.L.: “Transition Sault, P.H. et al.: “Cement Composition for Cementing
Time of Cement Slurries Between the Fluid and Set Wells Enabling Gas Channeling in the Cemented
State,” SPE Journal (December 1982) 22, No. 6, Annulus to be Inhibited by Right-Angle Setting,”
875–882; also presented as paper SPE 9285 at the SPE European Patent No. 0,189,950 (1986).
Annual Technical Conference and Exhibition, Dallas, Savage, R., and Fowler, H.: “Taking New Materials
Texas, USA (September 21–24, 1980). Downhole—The Composite Bridge Plug,” paper 27
Sabins, F.L., Tinsley, J.M., and Sutton, D.L.: “Transition presented at the 6th Annual Petroleum Network
Time of Cement Slurries Between the Fluid and Set Educational Conference (PNEC) Horizontal Well and
State,” SPE Journal (December 1982) 22, No. 6, Emerging Technological International Conference,
875–882; also presented as paper SPE 9285 at the SPE Houston, Texas, USA (October 24–26, 1994).
Annual Technical Conference and Exhibition, Dallas, Savins, J.G., and Roper, W.F.: “A Direct-Indicating
Texas, USA (September 21–24, 1980). Viscometer for Drilling Fluids,” Drilling and Production
Saltel, J.L. et al.: “In Situ Polymerization of an Inflatable Practices, Washington, D.C., USA, American Petroleum
Composite Sleeve to Reline Damaged Tubing and Shut Institute (1954) 7–22.
Off Perforations,” SPE Drilling & Completion, (June Savoly, A. et al.: “Fluid Loss Agents for Oil Well
1999) 14, No. 2, 115–122; also presented as paper SPE Cementing Compositions,” U.S. Patent No. 4,674,574
56867 at the Offshore Technology Conference, Houston, (June 23, 1987).
Texas, USA (May 6–9, 1996).
Schaaf, S., Mallary, C.R., and Pafitis, D.: “Point-the-bit
Samsuri, A., Junin, R., and Osman, A.M.: “The Utilization rotary steerable system: theory and field results,” paper
of Malaysian Local Bentonite as an Extender and Free SPE 63247 presented at the SPE Annual Technical
Water Controller in Oil-Well Cement Technology,” paper Conference and Exhibition, Dallas, Texas (October 1–4,
SPE 68674 presented at the SPE Asia Pacific Oil and 2000).
Gas Conference and Exhibition, Jakarta, Indonesia
Schlemmer, R. et al.: “Chemical Osmosis, Shale and
(April 17–19, 2001).
Drilling Fluids,” SPE Drilling & Completion
Sanchez, J.M.: “Surfactant Effects on the two-phase (December 2003) 18, No. 4, 318–331 (paper SPE 86912);
Flow of Steam/Water and Nitrogen/Water through an revised for publication from paper SPE 74457 presented
Unconsolidated Permeable Medium,” PhD dissertation, at the IADC/SPE Drilling Conference, Dallas, Texas, USA
University of Texas, Austin, Texas, USA (1987). (February 26–28, 2002).
Sanchez, R.A. et al.: “The effect of drillpipe rotation on Schlichting, H.: Boundary Layer Theory, New York City,
hole cleaning during directional well drilling,” SPE New York, USA, McGraw-Hill Book Company, Inc. (1979).
Journal (June 1999) 4, No. 2, 101–108; also presented as
Schneider, U., and Chen, S.: “The Interface Zone Around
paper SPE 56406 at the SPE/IADC Drilling Conference,
Expanded Shale Grain in Hardened Cement Paste,”
Amsterdam, The Netherlands (March 4–6, 1997).

740 Well Cementing

RILEM Proceedings, Interfaces in Cementitious at the SPE International Symposium on Oilfield and
Composites (1992), No. 18, 149–156. Geothermal Chemistry, Dallas, Texas, USA (January
Schuh, F.J.: “Computer Makes Surge-Pressure Calculations 25–27, 1982).
Useful,” Oil & Gas Journal (August 3, 1964) 62, No. 31, Sharp, L.G.: “Sealing of Deep Permeable Earth
69–98. Formations,” U.S. Patent No. 3,244,230 (April 5, 1966).
Schumacher, J.P. et al.: “Planning and Preparing for the Shaw, D.J.: Introduction to Colloid and Surface
First Subsea Field Test of a Full-Scale Dual-Gradient Chemistry, London, UK, Butterworth & Co., Ltd. (1980).
Drilling System,” SPE Drilling & Completions Shaw, D.J.: Introduction to Colloid and Surface
(December 2002) 17, No. 4, 194–199 (paper SPE 80615); Chemistry, London, UK, Butterworth & Co., Ltd. (1970).
revised for publication from paper SPE 71358 presented
Sheives, T.C. et al.: “A Comparison of New Ultrasonic
at the SPE Annual Technical Conference and Exhibition,
Cement and Casing Evaluation Logs with Standard
New Orleans, Louisiana, USA (September 30–October 3,
Cement Bond Logs,” paper SPE 15436 presented at the
2001).
SPE Annual Technical Conference and Exhibition, New
Scott, L.J. et al.: “Cementing Composition and Well Orleans, Louisiana, USA (October 5–8, 1986).
Cementing Process Employing Same,” U.S. Patent
Shell, F.J., and Wynne, R.A.: “Application of Low-Water-
No. 3,511,314 (May 12, 1970).
Loss Cement Slurries,” paper 875-12-I presented at the
Scrivener, K.L., and Capmas, A.: “Calcium Aluminate API Rocky Mountain District, Division of Production
Cements,” Lea’s Chemistry of Cement and Concrete, Spring Meeting, Denver, Colorado, USA (1958).
P.C. Hewlett (ed.), Oxford, England, Butterworth-
Shelley, R.F., and Sciullo, F.C.: “Catalyzed Silicate Process
Heinemann (2001).
Replaces Cement on Well-Deepening Operations,” paper
Sedillo, L.P., Newlove, J.C., and Portnoy, R.C.: “Fluid Loss SPE 30983 presented at the SPE Eastern Regional
Control in Oil Field Cements,” U.S. Patent No. 4,659,750 Meeting, Morgantown, West Virginia, USA (September
(April 21, 1987). 17–21, 1995).
Sehault, J.M., and Grebe, E.L.: “Spacer Fluid,” European Shen, J.C., and Pyle, D.S.: “Effects of CO2 Attack on
Patent No. 0243067 (1987). Cement in High-Temperature Applications,” paper
Seidel, F.A., and Greene, T.G.: “Use of Expanding Cement SPE/IADC 18618 presented at the SPE/IADC Drilling
Improves Bonding and Aids in Eliminating Annular Gas Conference, New Orleans, Louisiana, USA (February 28–
Migration in Hobbs Grayburg-San Andres Wells,” paper March 3, 1989).
SPE 14434 presented at the SPE Annual Technical Shepard, S.F., Reiley, R.H., and Warren, T.: “Casing
Conference and Exhibition, Las Vegas, Nevada, USA Drilling: An Emerging Technology,” SPE Drilling &
(September 22–25, 1985). Completion (March 2002) 17, No. 1, 4–14 (paper SPE
Sepos, D.J., and Cart, B.W.: “New Quick Setting Cement 76640); revised for publication from paper SPE 67731
Solves Shallow Gas Migration Problems and Reduces presented at the SPE/IADC Drilling Conference,
WOC Time,” paper SPE 14500 presented at the SPE Amsterdam, The Netherlands (February 27–March 1,
Eastern Regional Meeting, Morgantown, West Virginia, 2001).
USA (November 6–8, 1985). Sherman, S. et al.: “Epoxy Resins,” Kirk-Othmer
Setter, N., and Roy, D.M.: “Mechanical Features of Encyclopedia of Chemical Technology, New York City,
Chemical Shrinkage of Cement Paste,” Cement and New York, USA, John Wiley & Sons (1980) Vol. 9, 267–290.
Concrete Research, (1978) 8, No. 5, 623–634. Sherwood, J.D., and Bailey, L.: “Swelling of Shale Around
Sewell, M., and Billingsley, J.: “An Effective Approach to a Cylindrical Wellbore,” Proceedings of the Royal
Keeping the Hole Clean in High-Angle Wells,” World Oil Society of London, Series A (January 8, 1994) 444,
(October 2002) 223, No. 10, 35–39. No. 1920, 161–184.
Shah, S.N., and Sutton, D.L.: “New Friction Correlation Sherwood, J.D.: “Formation of a Cement Filter Cake
for Cements From Pipe and Rotational Viscometer Above a Compactible Mud Cake,” Chemical Engineering
Data,” SPE Production Engineering (November 1989) Science (August 1993) 48, No.15, 2767–2775.
5, No. 4, 415–425. Shi, G., Wenjun, L., and Yonghui, G.: “Modifying Set
Sharma, M.K., Shah, D.O., and Brigham, W.E.: “The Cement Performance for Improving Cementing Job
Chain Length Compatibility and Surface Properties of Quality,” paper SPE 29924 presented at the SPE
Foaming Solutions in Relation to Fluid Displacement International Meeting on Petroleum Engineering,
Efficiency in Porous Media,” SPE Reservoir Engineering Beijing, China (November 14–17, 1995).
(May 1986), 253–260; also presented as paper SPE 10612

References 741
Shideler, J.J.: “Calcium Chloride in Concrete,” Journal Skalny, J., Odler, I., and Hagymassy, J. Jr.: “Pore
of the American Concrete Institute (March, 1952) 23, Structure of Hydrated Calcium Silicates. I. Influence of
No. 7, 537–559. Calcium Chloride on the Pore Structure of Hydrated
Short, A., Struct, M.I., and Kinniburgh, W.: “The Tricalcium Silicate,” Journal of Colloid and Interface
Structural Use of Aerated Concrete,” The Structural Science (March 1971) 35, No. 3, 434–440.
Engineer (January 1961) 39, No. 1, 1–16. Skibsted, J., Henderson, E., and Jakobsen, H.J.:
Shryock, S.H., and Slagle, K.A.: “Problems Related to “Characterization of calcium aluminate phases in
Squeeze Cementing,” Journal of Petroleum Technology cements by 27Al MAS NMR spectroscopy,” Inorganic
(August 1968) 20, No. 8, 801–807; also presented as Chemistry (March 17, 1993) 32, No. 6, 1013–1027.
paper SPE 1993 at the SPE California Regional Meeting, Skinazi, E. et al.: “Development of a Casing/Drilling
Los Angeles, California, USA (October 26–27, 1967). System Improves the Drilling Process,” paper 62780
Shryock, S.H.: “Geothermal Well Cementing Technology,” presented at the 2000 IADC/SPE Asia Pacific Drilling
paper SPE 12454 presented at the SPE Offshore South Technology, Kuala Lumpur (September 11–13, 2000).
East Asia, Singapore (February 21–24, 1984). Slagle, K.A., and Carter, L.G.: “Gilsonite—A Unique
Silk, I.M.: “Exposure to Moisture Alters Well Cement,” Additive for Oil-Well Cements,” Drilling and Production
Petroleum Engineering International (March 1986) 58, Practice, American Petroleum Institute, Washington,
No. 3, 45–49. D.C., USA (1959), 318–328.
Silva, M.G.P. et al.: “Designing fluid velocity profiles Slagle, K.A., and Smith, D.K.: “Salt Cement for Shale and
for optimal primary cementing,” paper SPE 36136 Bentonitic Sands,” Journal of Petroleum Technology
presented at the SPE Fourth Latin American and (February 1963) 15, No. 2, 187–194; also presented as
Caribbean Petroleum Engineering Conference, Port of paper SPE 411 at the SPE Annual Meeting, Los Angeles,
Spain, Trinidad and Tobago (April 23–26, 1996). California, USA (October 7–10, 1962.
Silva, M.G.P. et al.: “Slag Cementing Versus Slater, H.J., Stiles, D.A., and Chmilowski, W.: “Successful
Conventional Cementing: Comparative Bond Results,” Sealing of Vent Flows with Ultra-Low-Rate Cement
paper SPE 39005 presented at the SPE Fifth Latin Squeeze Technique,” paper SPE 67775 presented at
American and Caribbean Petroleum Engineering the SPE/IADC Drilling Conference, Amsterdam, The
Conference and Exhibition, Rio de Janeiro, Brazil Netherlands (February 27–March 1, 2001).
(August 30–September 3, 1997). Slaton, M.: “Foamed Cement—Ultra Light and More,”
Simpson, B.E.: “Analytical Chemistry of Portland Drilling (December 1981) 41, No. 8, 51–52, 15.
Cement and its Oil-Field Admixtures,” SPE Production Slatter, P.T.: “The Laminar/ Turbulent Transition
Engineering (May 1988) 3, No. 2, 158–166; also present- Prediction for Non-Newtonian Slurries,” Proceedings of
ed as paper SPE 14095 at the SPE International Meeting the International Conference on Problems in Fluid
on Petroleum Engineering, Beijing, China (March Mechanics and Hydrology, Prague, Czech Republic
17–20, 1986). (June 1999).
Simpson, J.P., Salisbury, D.P., and Jewell, R.A.: “How to Smith, D.K.: Cementing, Monograph No. 4 in Henry L.
Combat Oil-Base Mud Losses,” World Oil (January Doherty Series, SPE, Dallas, Texas, USA (1976).
1988), 30. Smith, D.K.: Cementing, Monograph No. 4 in Henry L.
Singh, N.B., and Agha, Km.: “Effect of Calcium Formate Doherty Series, SPE, Dallas, Texas, USA (1987).
on the Hydration of Tricalcium Silicate,” Cement and Smith, F.M.: “Thermal Expansion of Cemented Casing,”
Concrete Research (September 1983) 13, No. 5, 619–625. Proceedings of the 25th Pennsylvania State University
Skaggs, B., Dial, H., and Rakitsky, W.: “Stable suspension Petroleum Production Conference, State College,
of hydrocolloids and superplasticizer,” U.S. Patent Pennsylvania, USA, (October 19–21, 1966) 249–273.
No. 6,309,455 (October 30, 2001). Smith, R.C., and Calvert, D.G.: “The Use of Sea Water in
Skalny, J., and Maycock, J.N.: “Mechanisms of Well Cementing,” Journal of Petroleum Technology
Acceleration by Calcium Chloride: A Review,” (June 1975) 27, No. 6, 759–764; also presented as paper
Journal of Testing and Evaluation (July 1975) 3, No. 4, SPE 5030 at the SPE Annual Meeting, Houston, Texas,
303–311. USA (October 6–9, 1974).
Skalny, J., and Young, J.F.: “Mechanisms of Portland Smith, R.C., and Ravi, K.M.: “Investigation of drilling
Cement Hydration,” Proceedings of the 7th International fluid properties to maximize cement displacement effi-
Congress on the Chemistry of Cement (1980) 1, 1–52. ciency,” paper SPE 22775 presented at the SPE Annual

742 Well Cementing

Technical Conference and Exhibition, Dallas, Texas, USA Song, M., Wang, W., and Ma, K.: “Slag MTC Techniques
(October 6–9, 1991). Solve Cementing Problems in Complex Wells,” paper
Smith, R.C., Beirute, R.M., and Holman, G.B.: “Improved SPE 64758 presented at the SPE International Oil
Methods of Setting Successful Whipstock Cement and Gas Conference and Exhibition, Beijing, China
Plugs,” Journal of Petroleum Technology (November (November 7–10, 2000).
1984) 36, No. 11, 1897–1904; also presented as paper Spangle, L.B., and Calvert, D.G.: “Improved Primary and
SPE 11415 at the IADC/SPE Drilling Conference, New Remedial Cementing With Thixotropic Cement Systems,”
Orleans, Louisiana, USA (February 20–23, 1983). paper SPE 3833 presented at the SPE Rocky Mountain
Smith, R.C., Beirute, R.M., and Holmart, G.B.: Regional Meeting, Denver, Colorado (April 10–12, 1972).
“Postanalysis of Abnormal Cementing Jobs With a Spangle, L.B.: “Apparatus and Method for Measuring the
Cementing Simulator,” SPE Production Engineering Expansion Properties of a Cement Composition,” U.S.
(August 1987) 2, No. 3,157–164; also presented as paper Patent No. 4,408,489 (October 11, 1983).
SPE 14201 at the SPE Annual Technical Conference and Spangle, L.B.: “Expandable Cement Composition,”
Exhibition, Las Vegas, Nevada, USA (September 22–25, European Patent No. 254,342, 27 (January 1988).
2001).
Specification for Bow-Spring Casing Centralizers
Smith, R.C., Powers, C.A., and Dobkins, T.A.: “A New (API Spec 10D) (6th ed.) API, Washington, D.C., USA,
Ultra-Lightweight Cement With Super Strength,” American Petroleum Institute (2002).
Journal of Petroleum Technology (August 1980) 32,
Specification for Cements and Materials for Well
1438–1444; also presented as paper SPE 8256 at the SPE
Cementing (API Spec 10A) (23rd ed.), API, Washington,
Annual Fall Technical Conference and Exhibition, Las
D.C., USA (2002).
Vegas, Nevada, USA (September 23–26, 1979).
Specification for Threading, Gauging, and Thread
Smith, R.C.: “Successful Primary Cementing Can Be a
Inspection of Casing, Tubing, and Line Pipe Threads
Reality,” Journal of Petroleum Technology (November
(API Spec 5B) (14th ed.), API, Washington, D.C., USA
1984) 36, No. 12, 1851–1858; also presented as paper
(1996).
SPE 13498 at the SPE California Regional Meeting, Long
Beach, California, USA (March 23–25, 1988). Speers, R.A. et al.: “Drilling Fluid Shear Stress Overshoot
Behavior,” Rheologica Acta (September–October 1987)
Smith, R.C.: “Checklist Aids Successful Primary
26, No. 5, 447–452.
Cementing,” Oil & Gas Journal (November 1, 1982) 80,
No. 44, 72–75. Sprung, S., Kuhlmann, K., and Ellerbrock, H.G.: “Particle
Size Distribution and Properties of Cement—Part II:
Smith, T.R., and Crook, R.J.: “Investigation of Cement
Water Demand of Portland Cement,” Zement-Kalk-Gips
Preflushes for a KCl-Polymer Mud,” paper CIM 82-33-71
(November 1985) 38, No. 11, 275–281.
presented at the Annual CIM Petroleum Society
Technical Meeting, Calgary, Canada (June 6–9, 1982). Stadelmann, C., and Wieker, W.: “On the Influence of
Inorganic Salts on the Hydration of Tricalcium Silicate,”
Smith, T.R., et al.: “Foamed Cement Application in
Il Cemento (1985) 82, No. 4, 203–210.
Canada,” Petroleum Society of CIM paper 83-24-31 pre-
sented at the Annual CIM Petroleum Society Technical Standard Practice for Reducing Samples of Aggregate to
Meeting, Banff, Canada (May 10–13, 1983). Testing Size (ASTM C702-98 [2003]), ASTM International,
West Conshohocken, Pennsylvania, USA (1998).
Smith, T.R.: “Cementing Displacement Practices: Field
Applications,” Journal of Petroleum Technology (May Standard Practice for Sampling and the Amount of
1990), 564–566, 628–629 (paper SPE/IADC 18617, 1989.) Testing of Hydraulic Cement (ASTM C183-02), ASTM
International, West Conshohocken, Pennsylvania, USA
Smith, T., Lukay, R., and Delorey, J.: “Light, Strong
(2002).
Foamed Cement: A New Tool for Problem Wells,” World
Oil (May 1984) 198, No. 6, 135–144. Standard Specification for Portland Cement (ASTM
C 150), ASTM International, West Conshohocken,
Smith, W.H.: “Gelling Aqueous Silicate Compositions,”
Pennsylvania, USA (2002).
European Patent Application No. 0 230 725 A1
(November 1986). Standard Specifications for Processing Additions for
Use in the Manufacture of Portland Cement (ASTM
Sommer, F.S., and Jenkins, D.P.: “Channel Detection
C465), ASTM International, West Conshohocken,
Using Pulsed Neutron Logging in a Borax Solution,”
Pennsylvania, USA (2005).
paper SPE 25383 presented at the SPE Asia Pacific Oil
and Gas Conference, Singapore (February 8–10, 1993). Standard Test Method for Autoclave Expansion of
Portland Cement (ASTM C151-00), ASTM International,
West Conshohocken, Pennsylvania, USA (2000).

References 743
Standard Test Method for Compressive Strength of SPE/CIM/Canadian Heavy Oil Association (CHOA)
Cylindrical Concrete Specimens (ASTM C39/C39M-04), 78950 presented at the International Thermal
ASTM International, West Conshohocken, Pennsylvania, Operations and Heavy Oil Symposium and International
USA (2004). Horizontal Well Technology Conference, Calgary,
Standard Test Method for Density of Hydraulic Cement Alberta, Canada (November 4–7, 2002).
(ASTM C188-95 [2003]), ASTM International, West Stiles, D.A., and Baret, J.F.: “Sedimentation and Free
Conshohocken, Pennsylvania, USA (2003). Water of Cement Slurries: Mathematical Models and
Standard Test Method for Density, Relative Density Practical Solutions,” paper SPE 25866 presented at the
(Specific Gravity), and Absorption of Fine Aggregate SPE Rocky Mountain Regional/Low Permeability
(ASTM C128-04), ASTM International, West Reservoirs Symposium, Denver, Colorado, USA (April
Conshohocken, Pennsylvania, USA (2004). 12–14, 1993).
Standard Test Method for Fineness of Hydraulic Cement Stiles, D.: “Effects of Long-Term Exposure to Ultrahigh
by Air Permeability Apparatus (ASTM C204-00), ASTM Temperature on the Mechanical Parameters of Cement,”
International, West Conshohocken, Pennsylvania, USA paper SPE 98896 presented at the IADC/SPE Drilling
(2000). Conference, Miami, Florida, USA (February 21–23,
2006).
Standard Test Method for Fineness of Portland Cement
by the Turbidimeter (ASTM C115-96a [2003]), ASTM Stiles, D.A.: “Successful Cementing in Areas Prone to
International, West Conshohocken, Pennsylvania, USA Shallow Saltwater Flows in Deep-Water Gulf of Mexico,”
(2003). paper OTC 8305 presented at the Offshore Technology
Conference, Houston, Texas, USA (May 5–8, 1997).
Standard Test Method for Splitting Tensile Strength
of Cylindrical Concrete Specimens (ASTM C496/ Stone, W.H., and Christian, W.W.: “The Inability of Unset
C496M-04), ASTM International, West Conshohocken, Cement to Control Formation Pressure,” paper SPE 4783
Pennsylvania, USA (2004). presented at the SPE Symposium on Formation Damage
Control, New Orleans, Louisiana, USA (February 7–8,
Standard Test Method for Static Modulus of Elasticity
1974).
and Poisson’s Ratio of Concrete in Compression (ASTM
C469-02), ASTM International, West Conshohocken, Stringer, R.A., and Barrett, N.D.: “Fluid Caliper: A Useful
Pennsylvania, USA (2002). Tool to Improve Primary Cementing,” Proceedings of the
1990 Southwestern Petroleum Short Course, Lubbock,
Standard Test Methods for Chemical Analysis of
Texas, USA, Southwestern Petroleum (1990) 54–62.
Hydraulic Cement (ASTM C114-04a), ASTM
International, West Conshohocken, Pennsylvania, USA Sudakas, L.G. et al.: “Alkalies, Microstructure and
(2004). Activity of Industrial-Grade Cement Clinkers,” Tsement
(1978) 12, 11–12.
Stehle, D. et al.: “Conoco Stops Annular Gas Flow with
Special Cement,” Petroleum Engineering International Sugama, T. et al.: “Mullite Microsphere-Filled Lightweight
(April 1985), 57, No. 4, 21–24. Calcium Phosphate Cement Slurries for Geothermal
Wells: Setting and Properties,” Cement and Concrete
Stein, H.N.: “Influence of Some Additives on the Hydration
Research (August 1995) 25, No. 6, 1305–1310.
Reactions of Portland Cement. II. Electrolytes,” Journal
of Applied Chemistry (December 1961a) 11, 12, 482–492. Sugama, T., and Carciello, N.R.: “Carbonation of calcium
phosphate cements after long-term exposure to Na2CO3-
Stein, H.N.: “Influence of Some Additives on the
laden water at 250°C,” Cement and Concrete
Hydration Reactions of Portland Cement. I. Non-ionic
Research (November 1993) 23, No. 6, 1409–1417.
Organic Additives,” Journal of Applied Chemistry
(December 1961b) 11, 12, 474–482. Sugama, T., and Carciello, N.R.: “Carbonation of
hydrothermally treated phosphate-bonded calcium alu-
Stephens, M.: “Fluid Loss Additives For Well Cementing
minate cements,” Cement and Concrete Research
Compositions,” U.S. Patent No. 5,294,651 (March 15, 1994).
(September 1992) 22, No. 5, 783–792.
Stewart, R.B., and Schouten, F.C.: “Gas Invasion and
Sugama, T., and Carciello, N.R.: “Sodium Phosphate-
Migration in Cemented Annuli: Causes and Cures,” SPE
Derived Calcium Phosphate Cements,” Cement and
Drilling Engineering (March 1988) 3, No. 1, 77–82; also
Concrete Research (January 1995) 25, No. 1, 91–101.
presented as paper SPE 14779 at the IADC/SPE Drilling
Conference, Dallas, Texas, USA (February 10–12, 1986). Sugama, T., Kukacka, L.E., and Galen, B.G.: “Advanced
High-Temperature Lightweight Foamed Cements for
Stiles, D., and Hollies, D.: “Implementation of Advanced
Geothermal Well Completions,” Brookhaven National
Cementing Techniques to Improve Zonal Isolation in
Laboratory Report No. BNL-38087-DE87005263, New York
Steam Assisted Gravit y Drainage Wells,” paper

744 Well Cementing

City, New York, USA, Brookhaven National Laboratory SPE Annual Technical Conference and Exhibition,
(1986). Dallas, Texas, USA (September 27–30, 1987).
Suman, G.O., and Ellis, R.C.: World Oil’s Cementing oil Szabo, P., and Hassager, O.: “Displacement of one
and gas wells… including casing handling procedures, Newtonian fluid by another: density effects in axial
Houston, Texas, USA, Gulf Publishing Company (1977). annular flow,” International Journal of Multiphase
Suman, G.O.: “Well Completion Method,” U.S. Patent Flow (February 1997) 23, No. 1, 113–129.
4,440,226 (April 3, 1984). Tadros, M.E., Skalny, J., and Kalyoncu, R.S.: “Early
Sutton, D.L., and Faul, R.: “Annular Gas Flow Theory Hydration of Tricalcium Silicate,” Journal of the American
and Prevention Methods Described,” Oil & Gas Journal Ceramic Society (July–August, 1976) 59, No. 7–8, 344–347.
(December 10, 1984) 82, No. 50,84–86. Tanner, C.H., and Harms, W.M.: “Unique Ultra Light-
Sutton, D.L., and Prather, D.A.: “New Expansion Weight Cement Slurry Compositions for Use in Unique
Additive Gives Good Results With Low C3A Cements,” Well Conditions, Laboratory Evaluation, and Field
Proceedings of the Southwestern Petroleum Short Performance,” paper SPE 11486 presented at the SPE
Course, Lubbock, Texas, USA, 39–48 (1986). Middle East Oil Technical Conference, Manama, Bahrain
(March 14–17, 1983).
Sutton, D.L., and Ravi, K.M.: “Low-rate pipe movement
during cement gelation to control gas migration and Taplin, J.H.: Discussion of “Some Chemical Additions
improve cement bond,” paper SPE 22776 presented at and Admixtures in Cement Paste and Concrete by H.E.
the SPE Annual Technical Conference and Exhibition, Vivian,” Proceedings of the 4th International Congress
Dallas, Texas, USA (October 6–9, 1991). on the Chemistry of Cement, Washington, D.C., USA
(October 2–7, 1960) 2, 924.
Sutton, D.L., Sabins, F.L., and Love, R.: “Well Cementing
Methods and Compositions,” European Pat. Appl. Tare, U.A.: “Utilizing granulated quick quenched blast
No. 853069268, 1985. furnace slag in water based improved productivity drilling,”
MS thesis, Tulsa University, Tulsa, Oklahoma, USA (1997).
Sutton, D.L., Sabins, F.L., and Paul, R.: “New Evaluation
for Annular Gas-Flow Potential,” Oil & Gas Journal Tattersall, G.H.: “Present Problems Associated With
(December 17, 1984) 82, No. 51, 109–112. the Study of Cement Paste Rheology,” Proceedings of
the RILEM Seminar, Leeds, England (March 1973)
Sutton, D.L.: “Well Cementing Process and Gasified
1, 2.3-1–2.3-18.
Cements Useful Therein,” U.S. Patent 4,340,427 (July 20,
1982). Taylor, G.I.: “The Dispersion of Matter in Turbulent Flow
through a Pipe,” Proceedings of the Royal Society A
Swartzen-Allen, S., and Matijevic, E.: “Surface and
(1954) 223, No. 11155, 446–468.
Colloid Chemistry of Clays,” Chemical Reviews (1974)
74, 385–400. Taylor, H.F.W. (ed.): The Chemistry of Cements, London,
England, Academic Press Inc. Ltd. (1964).
Swayze, M.A.: “Effects of High Temperatures and
Pressures on Strengths of Oil Well Cements,” Drilling Taylor, H.F.W.: “Modification of the Bogue Calculation,”
and Production Practices, API, Washington, D.C., USA Advances in Cement Research (1989) 2, No. 6, 73–77.
(1954) 72. Taylor, H.F.W.: Cement Chemistry, London, England,
Sweatman, R.E. et al.: “First High-Temperature Thomas Telford Publishing, (1997).
Applications of Anti-Gas Migration Slag Cement and Technical Report on Cement Sheath Evaluation (API
Settable Oil-Mud Removal Spacers in Deep South Texas TR 10TR1) (1st ed.), API, Washington, D.C., USA (1996).
Gas Wells,” paper SPE 30512 presented at the SPE Technical Report on Shrinkage and Expansion in
Annual Technical Conference & Exhibition, Dallas, Oilwell Cements (API TR 10TR2) (1st ed.), API,
Texas, USA (October 22–25, 1995). Washington, D.C., USA (1997).
Sweatman, R.E., and Scoggins, W.C.: “Acid-Soluble Technical Report on Temperatures for API Cement
Magnesia Cement: New Applications in Completion and Operating Thickening Time Tests (API TR 10TR3) (1st
Workover Operations,” SPE Production Engineering ed.), API, Washington, D.C., USA (1999).
(November 1990) 5, No. 4, 441–447; also presented as
Tehrani, A., Bittleston, S.H., and Long, P.J.G. “Flow
paper SPE 18031 at the SPE Annual Technical
instabilities during annular displacement of one
Conference and Exhibition, Houston, Texas, USA
non-Newtonian fluid by another,” Experiments in
(October 2–5, 1988).
Fluids (February 1993) 14, No. 4, 246–256.
Sykes, R.L., and Logan, J.L.: “New Technology in Gas
Tehrani, A., Ferguson, J., and Bittleston, S.H.: “Laminar
Migration Control,” paper SPE 16653 presented at the
Displacement in Annuli: A Combined Experimental and

References 745
Theoretical Study,” paper SPE 24569 presented at the Thomas, N.L., and Birchall, J.D.: “The Retarding Action
SPE Annual Technical Conference and Exhibition, of Sugars on Cement Hydration,” Cement and Concrete
Washington, DC, USA (October 4–7, 1992). Research (November 1983) 13, No. 6, 830–842.
Tenoutasse, N.: “The Hydration Mechanism of C3A and Thomas, N.L., and Double, D.D.: “Calcium and Silicon
C3S in the Presence of Calcium Chloride and Calcium Concentrations in Solution During the Early Hydration
Sulphate,” Proceedings of the 5th International of Portland Cement and Tricalcium Silicate,” Cement
Congress on the Chemistry of Cement, 2, 372–378. and Concrete Research (September–November 1981)
Teplitz, A.J., and Hassebroek, W.E.: “An Investigation of 11, Nos. 5–6, 675–687.
Oil-Well Cementing,” API Drilling and Production Thorvaldson, W.M.: “Low Temperature Cementing,”
Practices, Washington, D.C., USA, American Petroleum paper presented at CIM Annual Meeting, Calgary,
Institute (1946). Alberta, Canada (May 2–4, 1962).
Terry, A.J.: “Not all technology is necessary for cost- Timoshenko, S.P., and Goodier, J.N.: Theory of Elasticity
effective cementing,” paper SPE 26319 presented at the (3rd ed.), New York City, New York, USA, McGraw-Hill
SPE Annual Technical Conference and Exhibition, Book Company (1970).
Houston, Texas, USA (October 3–6, 1993). Tinsley, John M. et al.: “Study of Factors Causing
Terry, D.T. et al.: “Converting Drilling Fluids to Annular Gas Flow Following Primary Cement
Cementitious Compositions,” U.S. Patent No. 5,295,543 Placement,” Journal of Petroleum Technology (August
(March 22, 1994). 1980) 32, No. 8, 1427–1437; also presented as SPE paper
Tessari, R.M., and Madell, G.: “Casing Drilling—A 8257 at the SPE Annual Technical Conference and
Revolutionary Approach to Reducing Well Costs,” paper Exhibition, Las Vegas, Nevada, USA (September 23–25,
SPE 52789 presented at the SPE/IADC Drilling 1979).
Conference, Amsterdam (March 9–11, 1999). Tixier, M.P., Alger, R.P., and Doh, C.A.: “New
Texas Railroad Commission: Plugging, Rule 3.14 (2002). Developments in Induction and Sonic Logging,” Journal
of Petroleum Technology (November 1960) 12, No. 11,
Théron, B.E., Bodin, D., and Fleming, J.: “Optimization
79–87.
of Spacer Rheology Using Neural Network Technology,”
paper SPE 74498 presented at the IADC/SPE Drilling Tokay, V.: “Additive composition for Portland cement
Conference, Dallas, Texas, USA (February 26–28, 2002). materials,” U.S. Patent No. 4,337,094 (June 29, 1982).
Thiercelin, M. et al.: “Cement Design Based on Cement Tong, L., and Yang, N.: “Hydration products of calcium
Mechanical Response,” SPE Drilling & Completion aluminoferrite in the presence of gypsum,” Cement and
(December 1998) 13, No. 4, 266–273; revised for publi- Concrete Research (1994) 24, No. 1, 150–158.
cation from paper SPE 38598, first presented at the SPE Toor, I.A.: “Problems in Squeeze Cementing,” paper
Annual Technical Conference and Exhibition, San SPE 11499 presented at the SPE Middle East Oil
Antonio, Texas, USA (October 5–8, 1997). Technical Conference and Exhibition, Manama, Bahrain
Thiercelin, M., and Roegiers, J.C.: “Formation (March 14–17, 1983).
Characterization—Rock Mechanics,” Reservoir Trabelsi, A.M.S., and Al-Samarraie, L.S.: “Fiber Content
Stimulation (3rd ed.), M.J. Economides and K.G. Nolte Affects Porosity, Permeability, and Strength of Cement,”
(eds.), Chichester, England, John Wiley & Sons (2000), Oil & Gas Journal (May 3, 1999) 97, No.18, 108, 110,
3-1–3-35. 112, 114.
Thiercelin, M., Baumgarte, C., and Guillot, D.: “A Soil Traetteberg, A., and Grattan-Bellew, P.E.: “Hydration of
Mechanics Approach to Predict Cement Sheath Behavior,” 3CaO · Al2O3 and 3CaO · Al2O3 + Gypsum With and
paper SPE 47375 presented at the SPE/International Without CaCl2,” Journal of the American Ceramic
Society for Rock Mechanics (ISRM) Eurock 98 Society (May–June, 1975) 58, No. 5-6, 221–227.
Conference, Trondheim, Norway (July 8–10, 1998). Traetteberg, A., Ramachandran, V.S., and Grattan-Bellew,
Thiercelin, M. et al.: “Cement Design Based on Cement P.E.: “A Study of the Microstructure of Tricalcium Silicate
Mechanical Response,” paper SPE 38598 presented at in the Presence of Calcium Chloride,” Cement and
the SPE Annual Technical Conference and Exhibition, Concrete Research (March 1974) 4, No. 2, 203–221.
San Antonio, Texas, USA (October 5–8, 1997). Tubman, K.M. et al.: “Synthetic Full-Waveform Acoustic
Thomas, D.C.: “A Spacer System Useful in Brine Logs in Cased Boreholes, II—Poorly Bonded Casing,”
Completion of Wellbores,” U.K. Patent No. 2073284A Geophysics (April 1986) 151, No.4, 902–913.
(1981).

746 Well Cementing

Tucker, G.R.: “Concrete and Hydraulic Cement,” U.S. lem,” paper SPE 30511 presented at the SPE Annual
Patent No. 2,141,569 (December 27, 1938). Technical Conference & Exhibition, Dallas, Texas, USA
Tumidajski, P.J., and Thomson, L.: “Influence of cadmi- (October 22–25, 1995).
um on the hydration of C3A,” Cement and Concrete Veeken, C.A.M., Wahleitmer, J.P., and Keedy, C.R.:
Research (1994); 24, No. 7, 1359–1372. “Experimental Modelling of Casing Deformation in a
Turki, W.H., and Mackay, A.S.: “Primary Cementing Compacting Reservoir,” paper SPE 28090 presented
Across Massive Lost Circulation Zones,” paper SPE at the SPE / ISRM Rock Mechanics in Petroleum
11490 presented at the Middle East Oil Technical Engineering Conference, Delft, The Netherlands
Conference, Manama, Bahrain (March 14–17, 1983). (August 29–31, 1994).
Uchikawa, H. et al.: “Effect of Admixture on Hydration of Vefring, E.H. et al.: “Optimization of displacement effi-
Cement Adsorptive Behavior of Admixture and Fluidity ciency during primary cementing,” paper SPE 39009 pre-
and Setting of Fresh Cement Paste,” Cement and sented at the SPE Fifth Latin American and Caribbean
Concrete Research (November 1992) 22, No. 6, Petroleum Engineering Conference and Exhibition, Rio
1115–1129. de Janeiro, Brazil (August 30–September 3, 1997).
Uchikawa, H., Hanehara, S., and Sawaki, D.: “The Role of Vert, J.W., and Angman, P.G.: “Method for Drilling and
Steric Repulsive Force in the Dispersion of Cement Casing a Wellbore with a Pump Down Cement Float,”
Particles in Fresh Paste Prepared with Organic Patent WO 0194739 (2001).
Admixture,” Cement and Concrete Research (January Vidick, B. et al.: “Selective Surface Determination of the
1987) 27, No. 1, 37–50. Silicate Phases in Portland Cement Powders Using
Uchikawa, H., Sawaki, D., and Hanehara, S.: “Influence Alkyltrichlorosilane,” Cement and Concrete Research
of Kind and Added Timing of Organic Admixture on the (July 1987) 17, No. 4, 624–632.
Composition, Structure and Property of Fresh Cement Vidick, B., and Reid, P.: “Could Formation Damage
Paste,” Cement and Concrete Research (February 1995) Minimization Provide a Cost-Effective Integrated
25, No. 2, 353–364. Approach for the Design of Drilling and Cementing
Udarbe, R.G., and Hancock-Grossi, K.: “Method For Fluids?” paper SPE 38040 presented at the SPE Asia
Control Of Fluid Loss And Gas Migration In Well Pacific Oil and Gas Conference, Kuala Lumpur, Malaysia
Cementing,” U.S. Patent No. 6,136,935 (October 24, (April 14–16,1997).
2000). Vidick, B., Yearwood, J.A., and Perthuis, H.: “How to
Van Breugel, K.: “HYMOSTRUC: A Computer-Based Solve Lost Circulation Problems,” paper SPE 17811
Simulation Model for Hydration and Formation of presented at the SPE International Meeting on
Structure in Cement-Based Materials,” RILEM Petroleum Engineering, Tianjin, China (November 1–4,
Proceedings (1992) No. 16. 1988).
Van Kleef, R.P.A.R.: “Optimized Slurry Design for Salt Vidovskii, A.I. et al.: “Change in Pressure of a Column of
Zone Cementations,” paper SPE 18620 presented at the Cement Slurry Behind the Casing in a Well During Time
SPE/IADC Drilling Conference, New Orleans, Louisiana, of Setting and Hardening,” Burenie (1971) No. 9, pp.
USA (February 28–March 3, 1989). 27–29.
Van Oort, E. et al.: “Silicate-Based Drilling Fluids: Vigneaux, P.G., Rondeau, J., and Pathak, R.: “Mixing
Competent, Cost-Effective and Benign Solutions to Cement By Solids Fraction Instead Of Density,” paper
Wellbore Stability Problems,” paper SPE 35059 pre- SPE 84584 presented at the SPE Annual Technical
sented at the IADC/SPE Drilling Conference, New Conference and Exhibition, Denver (October 5–8, 2003).
Orleans, Louisiana, USA (March 12–15,1996). Vijn, J.P., Dao, B., and Melbouci, M.: “Environmentally
Van Terzaghi, K.: “Die Berechnung der Durchlassigkeitsziffer Acceptable Well Cement Fluid Loss Control Additives,
des Tones aus dem Verlauf der Hydrodynamischen Compositions and Methods,” U.S. Patent No. 6,405,801
Spannungserscheinungen,” Sitzungsber. Akad. Wiss., (June 18, 2002).
Wien Math. Naturwiss. Kl. (1923) 132, 105–124. Vijn, P., and Fraboulet, B.: “Improved Cement
Van Vliet, J.P.M. et al.: “Development and Field Use of Formulations for Well Abandonment,” paper 077 pre-
Fiber-Containing Cement,” paper OTC 7889 presented at sented at the Offshore Mediterranean Conference,
the Offshore Technology Conference, Houston, Texas, Ravenna, Italy (March 28–30, 2001).
USA (May 14, 1995). Villar, J. et al.: “Cementing compositions and applica-
Van Vliet, J.P.M., Bour, D.L., and Griffith, J.E.: “Slimhole tions of such compositions to cementing oil (or similar)
cementing—Analysis and possible solution of the prob- wells,” U.S. Patent No. 6,060,535 (May 9, 2000).

References 747
Vinson, E.F., Totten, P.L., and Middaugh, R.L.: “Acid Walton, I.C., and Bittleston, S.H.: “The Flow of a
Removable Cement System Helps Lost Circulation in Bingham Plastic Fluid in a Narrow Eccentric Annulus,”
Productive Zones,” paper SPE 23929 presented at the Journal of Fluid Mechanics (January 1991) 222, 39–60.
IADC/SPE Drilling Conference, New Orleans, Louisiana, Wamba Fosso, S. et al.: “Viscous Pill Design Methodology
USA (February 18–21, 1992). Leads to Increased Cement Plug Success Rates:
Vlachou, P.V.: “Rhéologie des Laitiers de Ciment Application and Case Studies from Southern Algeria,”
Pétrolier,” PhD thesis, Institut National Polytechnique paper SPE 62752 presented at the IADC/SPE Asia
de Grenoble, France (1996) (in French). Pacific Drilling Technology Conference, Kuala Lumpur,
Vlachou, V., and Piau, J.M.: “A New Tool for the Malaysia (September 11–13, 2000).
Rheometric Study of Oil Well Cement Slurries and Other Wang, Q.Z., and Xing, L.: “Determination of fracture
Settling Suspensions,” Cement and Concrete Research toughness KIC by using the flattened Brazilian disc spec-
(October 2000) 30, No. 10, 1551–1557. imen for rocks,” Engineering Fracture Mechanics
Volpert, E.: “A Cementing Composition Including a (September 1999) 64, No. 2, 193–201.
Dispersant Agent for Cementing Operation in Oil Wells,” Ward, M. et al.: “A Joint Industry Project to Validate
European Patent No. EP 1 193 230 A1 (2002). Numerical Simulators for Deep Water Temperature
Vutukuri, V.S., Lama, R.D., and Saluja, S.S.: “Compressive Predictions,” SPE Drilling & Completion (June 2003)
Strength of Rock,” Handbook on Mechanical Properties 18, No. 2, 133–137; also presented as paper SPE 71364 at
of Rocks—Volume I, Testing Techniques and Results, the SPE Annual Technical Conference and Exhibition,
Trans Tech Publications (1974), 43–44. New Orleans, Louisiana, USA (September 30–October 3,
2001).
Vyalov, S.S.: Rheological Fundamentals of Soil
Mechanics, Elsevier (1986) 267–283. Warembourg, P.A., Kirksey, J.M., and Bannister, C.E.:
“Improving Cement Bond in the Rocky Mountain Area by
Wagh, A.S., and Brown, D.: “Development of Chemically
the Use of Spacer, Wash and Thixotropic Cement,” paper
Bonded Ceramic Borehole Sealants,” presented at
SPE 9031 presented at the SPE Rocky Mountain Regional
Natural Gas and Oil Technology Partnership Meeting,
Meeting, Casper, Wyoming, USA (May 14–16, 1980).
Houston, Texas, USA, (November 1999).
Warren, T., Houtchens, B., and Madell, G.: “Directional
Wahl, W.W., and Dever, C.D.: “Water Loss Control of
Drilling with Casing,” paper SPE 79914 presented
Aqueous Cement Slurries by Addition of Quaternary
at the SPE / IADC Drilling Conference, Amsterdam
Ammonium Polymers or Sulfonium Polymers,” U.S.
(February 19–21, 2003).
Patent No. 3,094,501 (June 18, 1963).
Warren, T., Houtchens, B., and Portas, W.R. Jr.: “Casing
Wahl, W.W., Dever, C.D., and Ryan, R.F.: “Low Water-Loss
Drilling With Directional Steering In The U.S. Gulf Of
Cement Composition,” U.S. Patent No. 3,086,588 (April
Mexico: Pt.1,” Offshore International (January 2001a)
23, 1963).
61, No. 1, 50, 52–53.
Wahlmeier, M., and Lam, S.: “Mathematical Algorithm
Warren, T., Houtchens, B., and Portas, W.R. Jr.: “Casing
Aids Analysis of ‘U-Tubing’ During Slurry Placement,”
Drilling With Directional Steering In The U.S. Gulf Of
Oil & Gas Journal (January 7, 1985) 83, No. 1, 80–86.
Mexico: Pt.2,” Offshore International (February 2001b)
Walker, C.O.: “Encapsulated Water Absorbent Polymers 61, No. 2, 40–42.
as Lost Circulation Additives for Aqueous Drilling Fluids,”
Warren, T., Angman, P., and Houtchens, B.: “Casing
U.S. Patent No. 4,664,816 (May 12, 1987).
Drilling Application Design Considerations,” paper SPE
Walker, E.J., Gantt, L., and Crow, W.: “Coiled Tubing 59179 presented at the 2000 IADC/SPE Drilling
Operations and Services,” World Oil (June 1992) 213, Conference, New Orleans, Louisiana, USA (February
No. 6, 69–76. 23–25, 2000).
Walker, M.L.: “Cement Fluid Loss Additive,” U.S. Patent Waters, G.A., and Wray, B.L.: “Narrow Annulus
No. 6,448,311 (September 10, 2002). Cementing in the Deep Anadarko Basin,” SPE Drilling
Walker, T.: “A Full-Wave Display of Acoustic Signal in & Completion (March 1995) 10, No. 1, 53–60; also pre-
Cased Hole,” Journal of Petroleum Technology (August sented as paper SPE 27925 at the SPE Mid-Continent
1968) 20, No. 8, 811–824; also presented as paper SPE Gas Symposium, Amarillo, Texas, USA (May 22–24, 1994).
1751 at the SPE Symposium on Mechanical Engineering Watters, L., and Beirute, R.: “Formation-Fluid Migration
Aspects of Drilling and Production, Fort Worth, Texas, After Cementing,” Petroleum Well Construction, M.J.
USA (March 5–7, 1967). Economides and K.G. Nolte (eds.), New York, New York,
USA, John Wiley & Sons (1998).

748 Well Cementing

Watters, L.T., and Sabins, F.L.: “Field Evaluation of 2556 presented at the SPE Annual Meeting, Denver,
Method to Control Gas Flow Following Cementing,” Colorado, USA (September 28–October 1, 1969).
paper SPE 9287 presented at the SPE Fall Technical Wieland, D.R., and Woods, B.L.: “Cement Preflush
Conference and Exhibition, Dallas, Texas, USA Method,” U.S. Patent No. 3,878,895 (April 22, 1975).
(September 21–24, 1980).
Williams, D. et al.: “Annular Gas Migration Can Be
Way, P.W., Howard, P., and Johnson, A.: “Jetting Clean-Up Controlled,” Oil & Gas Journal (January 31, 1983) 81,
Of Formation Damage Caused By The Drilling Process,” No. 4, 146–151.
paper SPE 62799 presented at the Asia Pacific Drilling
Wilson, S. et al.: “Inflatable Packers in Extreme
Technology Conference, Kuala Lumpur, Malaysia
Environments,” paper SPE 89529 presented at the
(September 11–13, 2000).
SPE/International Coiled Tubing Association Coiled
Weaver, C.E.: Developments in Sedimentology 44: Clays, Tubing Conference, Houston, Texas, USA (March 23–24,
Muds, and Shales, Amsterdam, The Netherlands, 2004).
Elsevier (1989).
Wilson, W.N. et al.: “Cementing Oil and Gas Wells Using
Weber, L. et al.: “The Application of a New Corrosion Converted Drilling Fluid,” U.S. Patent No. 4,883,125
Resistant Cement in Geothermal Wells,” Transactions of (November 28, 1989).
the Geothermal Resources Council, Davis, California,
Wilson, W.N., Carpenter, R.B., and Bradshaw, R.D.:
USA, Geothermal Resources Council (1998).
“Conversion of Mud to Cement,” paper SPE 20452 pre-
Webster, W.W., and Eikerts, J.V.: “Flow After sented at the SPE Annual Technical Conference and
Cementing—A Field and Laboratory Study,” paper SPE Exhibition, New Orleans, Louisiana, USA (September
8259 presented at the SPE Fall Technical Conference 23–26, 1990).
and Exhibition, Las Vegas, Nevada, USA (September
Wilson, W.N.: “Incremental density cementing spacers,”
23–26, 1979).
U.S. Patent No. 5,027,900 (July 2, 1991).
Weigand, W.A., and Totten, P.L.: “Fluid Spacer
Winn, R.H., Anderson, T.O., and Carter, L.G.: “A
Composition for Use in Well Cementing,” U.S. Patent No.
Preliminary Study of Factors Influencing Cement Bond
4,588,032 (May 13, 1986).
Logs,” Journal of Petroleum Technology (April 1962)
Welch, J.L., Mason, C.M., and Pavlich, J.P.: “Thixotropic 14, No. 4, 369–372; also presented as paper SPE 132 at
Cementing with Coiled Tubing at Prudhoe Bay,” paper the SPE Annual Meeting, Dallas, Texas, USA (October
SPE 20906 presented at SPE Europec 90, The Hague, 8–11, 1961).
The Netherlands (October 22–24, 1990).
Woodard, G.W., and Merkle, G.H.: “Composition of
Wertheimer, A.L., and Wilcock, W.L.: “Light Scattering Hydraulic Cement and Polyvinyl Acetate and Use
Measurements of Particle Distributions,” Applied Optics Thereof,” U.S. Patent No. 3,158,520 (November 24,
(June 1976) 15, No. 6, 1616–1620. 1964).
Whisonant, B.J. et al.: “New Materials Improve the Wooley, G.R. et al.: “Cementing Temperatures for Deep
Cementation of Salt Formations in the Williston Basin,” Well Production Liners,” paper SPE 13046 presented at
paper SPE 17512 presented at the SPE Rocky Mountain the SPE Annual Technical Conference and Exhibition,
Regional Meeting, Casper, Wyoming, USA (May 11–13, Houston, Texas, USA (September 16–19, 1984).
1988).
Wu, Z.Q., and Young, J.F.: “Formation of Calcium
Whistler, R.L., and BeMiller, J.N.: Industrial gums— Hydroxide from Aqueous Suspensions of Tricalcium
polysaccharides and their derivatives (3rd ed.), Silicate,” Journal of the American Ceramic Society
Academic Press, 1993. (January 1984) 67, No. 1, 48–51.
Whitfill, D.L., and Whitebay, L.E.: “Method for Xenakis, H., Vijn, P., and Covington, R.: “A New Sealant
Restricting Gas Influx During Well Cementing,” U.S. System for Multilateral Junction Geometries,” paper
Patent No. 5,012,869 (May 7,1990). SPE 38495 presented at the SPE Offshore Europe
Whittaker, A. (ed.): Theory and Application of Drilling Conference, Aberdeen, Scotland (September 9–10,
Fluid Hydraulics, EXLOG Series of Petroleum Geology 1997).
and Engineering Handbooks, Boston, Massachusetts, Xu, R., and Wojtanowicz, A.K.: “Diagnosis of Sustained
USA, Springer (1985). Casing Pressure from Bleedof f/Buildup Testing
Whorlow, R.W.: Rheological Techniques (1st ed.), Patterns,” paper SPE 67194 presented at the SPE
Chichester, England, Ellis Horwood Ltd. (1980). Production and Operations Symposium, Oklahoma City,
Wieland, D.R. et al.: “Design of Special Cement Systems Oklahoma, USA (March 24–27, 2001).
For Areas With Low Fracture Gradients,” paper SPE

References 749
Yang, X.M., and Sharma, M.M.: “Formation Damage Journal of Petroleum Technology (June 1969) 21, No. 6,
Caused by Cement Filtrates in Sandstone Cores,” paper 762–774; also presented as paper SPE 2402 at the SPE
SPE 19305, SPE Production Engineering (November Annual Meeting, Houston, Texas, USA (September
1991) 6, No. 4, 399–405. 29–October 2, 1968).
Yearwood, J., et al.: “Cementing Across Massive Salt Zhou, D., and Wojtanowicz, A.: “Estimation of Leakoff
Formations,” CIM paper No. 88-39-104 presented at the Pressure Gradient from Cementing and Casing Data,”
Annual CIM Petroleum Society Technical Meeting, paper SPE 56760 presented at the SPE Annual Technical
Calgary, Canada (June 12–16, 1988). Conference and Exhibition, Houston, Texas, USA
Yearwood, J.A., Vidick, B., and Boissier, J.C.: “A New (October 3–6, 1999).
Technique for Solving Lost Circulation Problems and Zuiderwijk, J.J.M.: “Mud Displacement in Primary
Zone Plugging,” paper CIM 88-39-105 presented at the Cementation,” paper SPE 4830 presented at the SPE
Canadian Petroleum Society Meeting, Calgary, Canada European Spring Meeting, Amsterdam, The Netherlands
(June 1988). (May 29–30, 1974).
Yoshimura, A., and Prud’homme R.K.: “Wall Slip
Corrections for Couette and Parallel Disk Viscometers,”
Journal of Rheology (1988) 32, No. 1, 53–67.
Yoshioka, K. et al.: “Adsorption Characteristics of
Superplasticizers on Cement Component Minerals,”
Cement and Concrete Research (October 2002) 32, No.
10, 1507–1513.
Yoshioka, K. et al.: “Role of Steric Hindrance in the
Performance of Superplasticizers for Concrete,”
Journal of the American Ceramic Society (October
1997) 80, No. 10, 2667–2671.
Young, J.F., Berger, R.L., and Lawrence, F.V. Jr.: “Studies
on the Hydration of Tricalcium Silicate Pastes III.
Influence of Admixtures on Hydration and Strength
Development,” Cement and Concrete Research
(November 1973) 3, No. 6, 689–700.
Young, J.F.: “Influence of Tricalcium Aluminate on
the Hydration of Calcium Silicates,” Journal of the
American Ceramic Society (January 1969) 52, No. 1,
44–46.
Young, R.A., ed.: The Rietveld Method, International
Union of Crystallography Monographs 5, Oxford,
England, Oxford University Press (1995).
Yun, Q., and Qingli, C.: “Studies of the Interfacial Bond
Between Cement Pastes and Expanded Shale,” RILEM
Proceedings, Interfaces in Cementitious Composites
(1992) No. 18, 71–77.
Zamora, M. et al.: “Major advancement in true real-time
hydraulics,” paper SPE 62960 presented at the SPE
Annual Technical Conference and Exhibition, Dallas,
Texas, USA (October 1–4, 2000).
Zeldin, A.N., and Kukacka, L.E.: “Polymer Cement
Geothermal Well Completion Materials,” Brookhaven
National Laboratory Report BNL 51287, New York City,
New York, USA, Brookhaven National Laboratory (1980).
Zemanek, J., and Caldwell, R.L.: “The Borehole
Televiewer—A New Logging Concept for Fracture
Location and Other Types of Borehole Inspection,”

750 Well Cementing

A
Abandonment plugs, 521
Absolute volume, 37, 660
Acoustic properties
Subject Index

of cements, 558–559
of fluids, 559–560
S
Acceleration period, 33–34 of formations, 557–558
Accelerators, 49, 50–54, 91 Acoustics, 556–560
calcium chloride, 50, 51–53 ACPs. See Annular casing packers
chloride-free accelerators, 53–54 Acrylamide copolymers, as fluid-loss control agents, 85
sodium chloride, 53–54 2-Acrylamido-2-methyl propane sulfonic acid (AMPS),
sodium silicate, 51 as fluid-loss control agent, 85, 86
Acid-soluble cements, 265, 267 Additives. See Cement additives
Acoustic devices Adsorption theory of retardation, 54
CBL-VDL tool, 562 AFm phase, 35, 36, 42, 58
for gas migration testing, 316 AFt phase, 35
pad-type sonic tools, 583–584 Aging, effects of, 38–39
ultrasonic pulse-echo technology, 585–597 Akermanite, 246
Acoustic impedance, 557–561 Albite, 324
Acoustic logging, 542, 555–611 Alite, 29
acoustic impedance, 557–561 Alkylene phosphonic acid, as retarder, 56, 57
acoustic properties of cements, 558–559 Alpha dicalcium silicate hydrate, 319
acoustic properties of fluids, 559–560 Alpha quartz, as extender, 64
acoustic properties of formations, 557–558 Aluminate phase, 34–36, 43, 58
bond index, 569, 600, 612 Aluminoferrite monosulfate phase. See AFm phase
borehole-compensated, 561–562, 578 Aluminoferrite trisulfate phase. See AFt phase
calibration, 556 Aluminum nose shoes, 364
CBL/VDL logs, 574–575, 577–583, 599–607 Aluminum powder, 237
cement bond log (CBL), 551, 562–584 Aluminum sulfate, 235
critical angle, 562 Aluminum sulfate/iron (II) sulfate system, 235
cycle skipping, 561, 575 AMPS. See 2-Acrylamido-2-methyl propane sulfonic acid
eccentering, 574, 575 Anionic synthetic polymers, as fluid-loss control agents,
fast formation, 557, 567, 580 85–87, 309
fixed gate, 568, 578 Annular backpressure, 305
inclinometer data, 595 Annular casing packers (ACPs), 386–388
microannulus, 582, 605 Annular expansion ring test, 641–642
of foamed cements, 561, 571, 585, 589 Annular flow slot approximation, equations, 613–625
quality control in, 555–556, 597–611 Annular fluid migration, 10, 289–317
repeat section, 555, 556 Annular packoff equipment, 387–388
sliding gate, 568 Annular volume, calculation, 664–665
slowness, 557 Anorthite, 323, 324
sonic cement evaluation techniques, 561–584 Anorthite cements, 323–324
stretching effect, 574–575 Antifoam agents, 89, 91
transit time, 567 Antileft-hand rotation feature, 414
variable-density log (VDL), 562, 563–567 Antisettling agents, 79–80
Aphrons, 214

Well Cementing ■ Subject Index 751

API/ISO classification systems, 45–48 Bingham-Reynolds number, 122, 131
Argillaceous materials, 24 Bingham yield stress, 97
Artificial-left operations, water and gas influx and, 19 Biopolymers, in spacers, 185
Artificial neural networks, 41 Bisphenol-A, 244
Asphaltic materials, for lost circulation prevention, 211 Blaine method, 39
ASTM classification, 44, 47 Blaine permeameter, 654
Atomic absorption spectrometry, 654 Blast-furnace slag (BFS), 232, 242, 245–246
Atomic force microscopy, 74 Blending tank, 345
Atmospheric consistometer, 631 Block squeeze, 508
Atmospheric silos, 345 Blowout preventer, 460, 461, 462
Atmospheric tank, 347 Bogue equations, 44, 45
Attapulgite, 60 Bogue’s method, 44
Autofill collar, 362 Bond index, 569, 600, 612
Autofill float equipment valves, 374 Bonding
Autofill valves, 372 cement-casing bonding, 220, 221
Automatic fill-up equipment, 371–374 cement-formation bonding, 219–231
hydraulic bond, 219
shear bond, 219
B Borax, as retarder aid, 58
Borehole-compensated logs, 561–562, 578
Balanced plug, 513, 518–519
Borehole quality, 145–146
Ball milling process, 30
Boric acid and salts, as retarders, 58
Ball valves, 368, 370
Boron-pulsed neutron log, 510, 554
Barite, as weighting agent, 69, 70
Borosilicate glass microspheres, as extenders, 67
Basic-slot model, 139
Bottom plugs, 379
Basket shoes, 389
Bottom wiper plug, 469
Batch-blending system, 344–345, 354
Bottomhole assembly (BHA), 506
Bearden unit, 631
Bottomhole circulating temperature (BHCT), 4, 5, 441,
Belite, 29
496
Belite-silica cement (BSC), 323
Bottomhole static temperature (BHST), 4, 441, 442, 646
Bentonite
Bradenhead squeeze, 525, 527, 529, 536, 540
calculations involving, 663–664
Brazilian test, 650–651
encapsulated particles as bridging agent, 209
Breakthrough time, 129, 622
as antisettling agent, 79
Bridge plugs, 516, 530
as extender, 59–60
cased hole bridge plugs, 424–425
as fluid-loss control agents, 80
inflatable, 428
in thixotropic cements, 234
through-tubing inflatable retrievable bridge plug
Bentonitic clay, 240
(TTRBP), 432
BET gas-adsorption method. See Brunauer-Emmett-
through-tubing permanent bridge plug (TTPBP), 432
Teller (BET) gas-adsorption method
Bridging agents, 88–89, 206–209
Beveled nose shoe, 365
Brittle failure, 277
BFS. See Blast-furnace slag
Brownmillerite, 36
BFS cement systems, 245–251, 267
Brucite, 237
BFS-Portland cement blends, 245, 246–247
Brunauer-Emmett-Teller (BET) gas-adsorption method,
BHA. See Bottomhole assembly
39, 655
BHCT. See Bottomhole circulating temperature
BSC. See Belite-silica cement
BHST. See Bottomhole static temperature
Buckingham-Reiner equation, 122
Bingham number, 122
Bulk expansion, 260
Bingham plastic fluids, 97, 103, 104, 105, 107, 108, 124, 133
Bulk shrinkage, 279
circulation efficiency, 152, 154
Bulk tank, 347
flow equations, 122–123, 125, 127, 619–620, 622,
Bulk trailers, 347
623–625
Bulk volume, 37, 660
friction pressure calculations, 131
Buoyant plume, 165
velocity and shear-rate profiles, 126
Bypass plug, 383

752 Well Cementing

C Carob, as fluid-loss control agent, 84, 85
C-S-H. See Calcium silicate hydrate Cased hole bridge plugs, 424–425
C-S-H phase, 31, 33, 34, 37, 41, 43, 52, 319, 320 Cased hole-log interpretation, 557
Calcareous materials, 24 Cased hole squeeze packers, 421–423
Calcination, 26 Casing
Calcined magnesium oxide, 237 calculating pressure to lift, 667–668
Calciochondrodite, 321 cement-to-casing bonding, 220–221
Calcium aluminate-based CBPCs, 265 operational sequence, 498–500
Calcium aluminate cement, 238, 322, 339 Casing buckling, 286
Calcium aluminate hydrates, 35 Casing centralization, 143, 445–446
Calcium aluminoferrite, 36 Casing centralizers, 390–395
Calcium aluminosilicate cement systems, 323–324, 334, Casing deformation, 14, 15
339 Casing drilling, 493–495
Calcium chloride, as accelerator, 50, 51–53 Casing hardware, 10, 182, 362–397
Calcium formate, 53 annular packoff equipment, 387–388
Calcium hydroxide, 30 automatic fill-up equipment, 371–374
Calcium nitrate, as accelerator, 54 casing insert equipment, 372
Calcium nitrite, 53 cement displacement enhancement tools, 389–397
Calcium phosphate cement systems, 334 cementing baskets, 395–397
Calcium silicate hydrate (C-S-H), 30, 31–34 cementing plugs, 378–380
Calcium sulfate, 38 centralizers, 390–395
Calcium sulfate-based systems, 234–235 differential fill equipment, 374–375
Calcium sulfate hemihydrate, 30, 38, 234 float shoes, 362–365
Calcium trisulfoaluminate hydrate, 35, 36 flotation equipment, 365–367
Calculations, 659–677 good cementing practices, 164
absolute and bulk volume, 660 guide shoes, 362–365
annular volume, 664–665 inner string cementing equipment, 375–377
cement additives, 660–664 scratchers, 370, 372, 395, 396
cement-slurry properties, 659–667 shoe and collar shell design, 377–378
conversion factors for units, 677 stage equipment, 381–386
displacement, 495 surge pressure and float equipment, 370–371
displacement volume, 665, 666 turbolizers, 393
flow rate, 15–16 valves, 368–371
foamed cements, 672–677 Casing leaks, 509
fracture pressure, 665 Casing movement, 497–498
friction pressure, 130, 131 good cementing practices, 164
hydrostatic pressure, 665, 666, 673–675 mud circulation, 155
mix water, 665 mud displacement and, 182–183
plug balancing, 668–671 Casing plugs, 416
pore pressure, 665 Casing pressure, 509
pressure to lift casing, 667–668 Casing reciprocation, 182
primary cementing, 495 Casing shearing, 286
pump pressure, 665, 666 Casing strings, 459–466
slurry volume, 495 Casing support, 285
squeeze cementing, 670–672 Casing thickness, 593
well calculations, 666–667 Casson model, 98, 106
Calipers, 436–439 Cast-iron bridge plugs, 426
Carbon dioxide, cements for enhanced oil recovery Cast-iron cement retainers, 422
by flooding with, 245 CAST-V. See Circumferential Acoustic Scanning Tool—
Carbon dioxide corrosion, 245 Visualization
Carbon dioxide-resistant cements, 247 Cation exchange capacity, 225
Carboxymethylhydroxyethylcellulose (CMHEC) Cationic synthetic polymers, as fluid-loss control agents,
as fluid-loss control agent, 83 87
as retarder, 56 Catoosa shale, 229

Subject Index 753

Cavernous formations, 204 Cement hardness, 542
CBL/VDL logs, 574–575, 577, 599 “Cement head,” 358
CBLs. See Cement bond logs Cement hydration, 227
CBPCs. See Chemical bonded phosphate ceramics See also Portland cement hydration
Cellophane flakes Cement hydration analyzer (CHA), 313, 314
as bridging agent, 88 Cement hydration detector, 551–552
as LCMs for cement slurries, 217 Cement job design, 435–458
Cellulose derivatives example of procedure, 448–457
as fluid-loss control agents, 83, 309 placement mechanics, 445–446
as retarders, 56, 57 preparing for the job, 457
Cement, 343, 503 problem analysis, 435–443
acoustic properties, 558–559 slurry selection, 443–444
design guidelines, 287 well control, 446–448
neat cement characterization, 654 Cement job evaluation
Cement additives, 3, 4, 9, 49–91 acoustic logs, 542, 555–611
accelerators, 49, 50–54, 91 CBL/VDL logs, 574–575, 577–583, 599–607
antifoam agents, 89, 91 cement bond log (CBL), 551, 562–584
antisettling agents, 79–80 cement evaluation tools, 7
bridging materials, 88–89 hydraulic testing, 549–551
calculations involving, 660–664 inflow testing, 551
delivery, storage, and mixing of, 344–346 nuclear logging, 554
dispersants, 49, 71–80, 91 oxygen activation logs, 511, 554–555
dry additives, 343, 344–346, 629 of plug cementing, 521
for EPS cement systems, 254 production logs, 511
extenders, 49, 58–69, 91 of squeeze cementing, 541–542
fluid-loss control agents, 49, 80–88, 91 temperature logs, 510, 541, 551–555
for foam cements, 258 ultrasonic pulse-echo technology, 585–597
liquid additives, 344 variable-density log (VDL), 562, 563–567
lost-circulation prevention agents, 49, 88–89, 91 of well construction, 507
mud decontaminants, 90 See also Laboratory testing; Testing
radioactive tracers, 90, 91, 161–162, 542, 554 Cement kilns, 27–28
retarders, 49, 54–58, 91 Cement mix water, 343, 656–657, 658
slurry selection, 444 calculating requirements, 664
specialty additives, 49, 661–662 calculations involving, 665
strengthening agents, 89, 91 Cement mixing, 351–355
variability of responses to, 49–50 cement and dry additive blending, 344–346, 629
weighting agents, 49, 69–71, 91 control devices, 356–358
Cement bond logs (CBLs), 551, 562–584 liquid additive storage and mixing, 349–350
Cement-casing bonding, 220–221 metering of water, 348–349
Cement channeling, 143 See also Cement slurry preparation
Cement displacement enhancement tools, 389–397 Cement mixing units, 354–355
Cement Evaluation Tool (CET), 585, 586 Cement placement, 445–446, 466–474
Cement-formation bonding, 219–231 continuous two-stage cementing, 470, 483
cement-to-casing, 220–221 deepwater cementing, 487
clay minerals, 224–225 immediate gas migration, 291–292
limestone and sandstone cores, 222–224 multistage cementing, 470–474
of sandstone and limestone, 222–224 plug cementing, 515–520
shales and, 226 problems during, 187–188
studies outside the petroleum industry, 230–231 single-stage cementing, 469–470
Cement-formation interactions, 191–232 stab-in cementing, 375–376, 377, 466–467
cement-formation bonding, 219–231 three-stage cementing, 470, 474
cementing across evaporite zones, 231–232 top-up cementing, 468–469
fluid loss, 191–202 two-stage cementing, 470, 472
lost circulation, 202–219

754 Well Cementing

Cement retainers, 422, 423 shear-rate range in wellbore, 98–100
inflatable, 429 spacers, 119
through-tubing cement retainer (TTCR), 433 squeeze cementing, 533–534
Cement sheath evaluation, 7 temperature dependency, 115–117
Cement sheath mechanical failure, 299 thermal cements in deep wells, 326
Cement slurries, 503 time dependency under shear, 117–118
lost-circulation control with, 216–217 turbulent flow, 94
Cement slurry density, 4, 356 wall slip and, 110–112
calculation, 661 See also Rheology
cement system design, 443 Cement slurry stability, EPS systems, 253–254
of EPS systems, 253 Cement slurry viscosification, fluid loss and, 197–198
of flexible cements, 262 Cement slurry volume
fluid loss and, 195, 196–197 calculations, 495
laboratory testing, 630 fluid loss and, 195, 196
squeeze cementing and, 533–534 Cement slurry yield, calculation, 661
thermal cements in deep wells, 326 Cement testing, 5–6
Cement slurry design, 444 Cement transfer systems, 345–347
plug cementing, 520–521 Cementing baskets, 395–397
squeeze cementing, 532–535 Cementing equipment, 10
Cement slurry gelation, 89, 309 Cementing heads, 399, 400, 416
Cement slurry preparation, 346 Cementing mandrel, 468
cement and dry additive blending, 344–346 Cementing materials, 343–344
cement mixing, 351–355 Cementing plugs, 379, 380
controls and instruments, 356–358 Cementing process, flowchart, 344
conventional jet mixer, 351–352 Cementing truck, 359, 360
flow chart, 344 Cementing units, 359–362
high-pressure pumps, 355–356 Cementitious drilling fluids, 250
laboratory testing, 628–630, 658 Cenospheres, as extenders, 67–69
liquid additive storage and mixing, 349–350 Centipoise (unit), 95
metering of water, 348–349 Centralizer sub, 394
surge tanks, 350–351 Centralizers, 362, 370, 445, 475
transportation of bulk materials to wellsite, 346–347 casing centralizers, 390–395
wellsite storage of cement or blends, 347–348 Ceramic microspheres, 336
Cement-slurry properties, calculations, 659–667 as extenders, 67–69
Cement slurry rheology, 9–10, 71, 93, 215, 357 CET. See Cement Evaluation Tool
calcium chloride and, 53 CFD. See Computational fluid dynamics
circulation efficiency, 129–130 CHA test, 642
coaxial cylinder viscometer, 100–101, 104, 105–106, Check valves, 368–371
639 Chemical bonded phosphate ceramics (CBPCs),
dispersants, effect of on, 73 265–266, 267
EPS systems, 253 Chemical resistance, squeeze cementing and, 534
experimental procedures, 101–104 Chemical shrinkage, 279, 298
flow equations, 119–129 Chemical washes, 184–185, 445
flow in wellbore, 119–129 Chemical waste disposal wells, 244
foamed cement and, 113–114 Chloride-free accelerators, 51
friction pressure calculations, 130–141 Chloride salts, as accelerators, 50
gel strength measurement, 118–119, 315, 316, CHP. See Critical hydration period
640–641, 658 Chromic acid and salts, as retarders, 58
laminar flow, 93–94 Circulatable mud, 161
large particles, effect of, 113 Circulating squeeze, 508
limitations of field practices, 104–115 Circulating valves, 428
particle migration and, 112 Circulation efficiency, 129–130, 146, 147, 151–154
pressure dependency, 117 equations, 622
rheological models, 9, 95–100, 105–106, 108–109 field measurement, 161–162

Subject Index 755

Circumferential Acoustic Scanning Tool—Visualization API/ISO requirements, 47
(CAST-V), 585, 586 bentonite and, 59
Citric acid calcium chloride and, 53
as dispersant, 71 cement impurities and, 25, 39
as retarder, 55, 56 cement system design, 443
Class A Portland cement, 45, 47 foamed cement, 258–259
Class B Portland cement, 45, 47 laboratory testing, 636–637, 658
Class C Portland cement, 45, 47 measurement, 357–358
Class D Portland cement, 47, 48 pozzolans and, 62
Class E Portland cement, 47, 48 seawater and, 240
Class F Portland cement, 47, 48 sodium chloride and, 51, 52
Class G Portland cement, 47, 48, 320, 321 sodium silicates and, 61
Class H Portland cement, 47, 48 triaxial, 272, 650
Class J cement, 322–323 uniaxial, 269
Classification of Portland cement. See Portland cement Computational fluid dynamics (CFD), modeling fluid
classification displacement, 172–173
Clay-base systems, 234 Computer-based monitoring system, 357–358
Clay-bearing formations, 222 Computer design, U-tubing, 448
Clay minerals, 27, 224 Computer modeling
Clays cement hydration, 40–42
as extender, 59–60 of cement sheath using thermoelasticity, 280–281
properties of, 224–226 mud circulation, 148–163
as spacers, 185 mud displacement, 165
Clinging constants, 370 mud removal in eccentric annuli, 169–174
Clinker, 23–30 Portland cement hydration, 43–44
Clinkering temperature, 27 Computer simulators, 446
Close-tolerance centralizers, 393 Computer software, 10
Closing plug, 383 for inflatable tools, 430
CMHEC. See Carboxymethylhydroxyethylcellulose lost circulation, 217–218
Coal optimal displacement volume, 519
as bridging agent, 88 through-tubing inflate operations, 433, 434
as extender, 66 Concentric annulus
as LCM for cement slurries, 216 flow equations, 124–129, 125–129
Coaxial cylinder viscometer, 100–101, 104, 105–106, 639 friction factor, 135–137
Coefficient of linear thermal expansion, laboratory laminar flow, 149–151
testing, 653 mud displacement, 165–169
Coiled tubing running tool, 401 turbulent flow, 151
Coiled tubing setting method, 422, 531 Concentric casings, 567
Collar cases, 377 Condensed silica fume, as extender, 53
Collar shells, 377 Conductor pipe, 460–461
Collar stock, 377 Conduction calorimetry, 31, 37
Colloidal silica dispersions, as extender, 64 Cone, 408
Combination casing strings, 465–466 Confined triaxial compression test, 650
Commercial lightweight cements, 63 Confining pressure, 278
Communication indicator, 552–553 Consistency index, 149
Complexation theory of retardation, 54 Consistometers, 631
Composite nose guide shoes, 364 Constant spacing centralizers, 390
Compressibility, 275 Constitutive equation, 271
Compressible cements, 307 Construction cements, treatment of, 249
Compression-set packers, 421 Contamination, of cement plugs, 516–520
Compression-set service packers, 421 Continuous microannulus, 18
Compressional wave, 556 Continuous two-stage cementing, 470, 473
Compressive strength, 8, 285 Control devices, 356–358
alkalis and, 39 Conventional jet mixer, 351–352

756 Well Cementing

Cooling, of clinker, 29 Dimensionless shear rate, 125
Coriolis effect mass flowmeter, 356 Directional drilling, well construction, 506, 521
Corrosive environments, special cement systems for, Dispersants, 49, 71–80, 91, 444
244–245, 267 cement slurry rheology and, 73
Couette viscometer, 357, 639 chemical composition, 71–73
Cracking, 14 free water and, 78–80
Creep, 276 with high-alumina cement, 322
Critical hydration period (CHP), 304, 305 mechanisms of action, 74–77
Crossflow, 15 response to cements, 77–78
Crosslinked cellulose polymer systems, 235 sedimentation and, 78–80
Crosslinked pills, lost circulation prevention with, 209 in spacers, 186
Crosslinked polyvinyl alcohol (PVA) microgels, as in washes, 184–185
fluid-loss control agent, 80, 81 Displacement, calculations, 495
Crushed coal, as LCMs for cement slurries, 216 Displacement efficiency, 146, 147, 166, 167
Crushed firebrick, 322 Displacement number, 174
Cube-crushing test, 269 Displacement plugs, 416–418
Cup seal system, 402 Displacement procedures, 469–470
Cycle skipping, 561, 575 Displacement tank, 348, 349
Cylindrical sleeve test, 642 Displacement volume, 495
for balanced plug, 518–519
calculations, 665, 666
D squeeze cementing, 536
Diutan gum, as spacer, 186
Darcy’s law, 16, 198, 522
Diverter tool, 517
Dart launcher, 486
Double bow centralizers, 393
Data gathering, for design, 435–436
Double grip liner hangers, 413–414
Debonding, 14
“Down jet” guide shoe, 364
Deceleration period, 33–34
Downhole-mixed systems, for lost-circulation
Deep wells, thermal cements for, 325–328
prevention, 212–213
Deepwater cementing, 487
Downhole pressure reduction, 215
Deepwater drilling, 6
Drag, 175
Deepwater wells, conductor casings, 461
Drag-reduction centralizers, 394
Dehydroxylation of clay minerals, 27
Drill-down liner hangers, 414
Depleted formations, 204
Drillable bridge plugs, 425–426, 516
Deposition factor, 522
Drillable cement retainers, 423
Depth and dimensional data, 435
Drillable packoff bushing system, 402
Depth data, 435
Drillable squeeze packers, 422
Design
Drilling, of geothermal wells, 329
cement guidelines, 287
Drilling “blind,” 214
cement slurry design, 444
Drilling fluids
liner hangers, 407
cementitious drilling fluids, 250
wellbore, 440
for massive loss remediation, 214
See also Cement job design
drilling liners, 475
Diatomaceous earth, as extender, 62
Drilling mud, hydraulic-bond strength and, 221
Dicalcium silicate, 30–34, 45
Drillpipe, 517
Diesel engines, 362
Drillpipe dart, 416, 417, 418
Differential fill equipment, 374–375
Drillship, 500
Differential pressure, 536
Drillstem test (DST), 551
Differential thermal analysis (DTA), 654
Dry additives, delivery, storage, and mixing of, 344–346,
Diffuse layer, 73
629
Diffusion period, 34
Dry-blended cements, laboratory analysis, 656, 658
Dilatancy, 272
Dry cement additives, 343
Dilution, 175
Dry microannulus, 607
Dimensionless breakthrough time, 129, 622
Dry process, 25
Dimensionless flow rate, 167

Subject Index 757

Dry testing, 551 ettringite systems, 236
DST. See Drillstem test salt cements, 231, 236, 239–242, 267
DTA. See Differential thermal analysis Extended-reach wells, 487–489
Dual liner wiper plug system, 418 Extenders, 49, 58–69, 91
Ductile failure, 277 clays, 59–60
Dump bailer, 513–514 commercial lightweight cements, 63
Dynamic compressive strength, 651 diatomaceous earth, 62
Dynamic elastic constants, 651–652 fly ash, 63
Dynamic filtration, fluid loss, 193 lightweight particles, 64–69
Dynamic placement pressure, fluid loss and, 198 nitrogen, 69
pozzolans, 61–64
silica, 64
E sodium silicates, 59, 60–61
External casing packers (ECP), 215, 216, 305, 306
Eccentric annulus, mud removal in, 169–174
Eccentric composite-nose float shoe, 364
Eccentricity, 137–138, 170
ECD. See Equivalent circulating density F
ECP. See External casing packers Failure, 18, 277, 286
Eddies, 177 brittle failure, 277
Effective stress component, 274 causes of, 13
Elastomeric composites, 264 cement sheath, 299
Electric wireline setting tool (EWST), 429 ductile failure, 277
Electrical conductivity, of foamed cement, 261 postfailure behavior, 278–279
Electrical double-layer theory, 33 postfailure region, 272
Encapsulated particles, 209 set-cement failure, 277
Engineered particle-size (EPS) distribution cements, shear failure, 14
251–256, 263, 267, 308 squeeze cementing, 542–543
Epidote, 323 stage tools, 384
Epoxy-based polymer cement systems, 244, 334, 521, 534 statistics, 13
EPS distribution cements. See Engineered particle size False set, 38, 39
distribution cements Fanning friction factor, 130, 131, 613
Equivalent circulating density (ECD), 488 Fast formations, 557, 567, 580
Equivalent permeability, 17 Fiber cements, 491
Equivalent sack, 62 Fibers, in flexible cements, 264
Erodability index, 178 Fibrous lost circulation materials (LCMs), 207, 208
Erosion, 165, 175, 193 Fineness, 39, 654, 655
Erosion index, 177 Fireflood wells, 338
Ethoxylated HEC, as fluid-loss control agent, 84 Fixed gate, 568, 578
Ethylene diamine carboxymethylcellulose, as fluid-loss Flapper valves, 368–369, 370, 372
control agent, 83 Flash set, 30, 38
Ettringite, 35, 36, 51, 234 Flexible cements, 262–264, 267, 310
Ettringite cement systems, 236 Flexible particles, 262–263
Evaluation. See Cement job evaluation Flexural strength, 8
Evaporite zones, 231–232, 240 Float collar, 362
EWST. See Electric wireline setting tool Float shoes, 362–365, 377
Expanded perlite, as extender, 64, 65 “Floating mudcap” method, 214
Expandable tubulars, 489, 490 Floating rig, primary cementing techniques, 500
Expanding cements, gas migration and, 310 Flotation equipment, 365–367
Expansion, 279–280 Flow
laboratory testing, 641–643, 658 after cementing, 289
Expansive cement systems, 235–237, 267 laminar flow, 93–94
aluminum powder, 237 turbulent flow, 94
calcined magnesium oxide, 237 Flow equations, 613–625
commercial systems, 236 annular flow, 124–129, 613–625

758 Well Cementing

 generic, 119–123 Foamed muds, for massive loss remediation, 214
pipe flow, 120–124, 613–625 Formation creep, 285–286
Flow rate Formation-damage mechanisms, 199–200
calculation, 15–16 Formation factor, 300
effect on displacement efficiency, 167 Formation parameter, 302
Flowmeter, 348, 356 Fourier transform infrared spectroscopy (FTIR), 40, 41,
Fluid, cement slurry design for squeeze cementing, 657
532–535 Four-arm calipers, 438
Fluid caliper, 161, 162, 438 Four-plug system, 386
Fluid compatibility testing, 644–646, 658 FPM plots. See Fluid properties measurement (FPM)
Fluid loss, 10, 191–202 plots
annular restriction and bridging, 198–199 Fracture pressure, calculations, 665
criteria, 194–195 Fracturing, 285
during primary cementing, 191, 194–201 Free fall tubing, 448
during remedial cementing, 201 Free-fluid control, gas migration and, 309
dynamic filtration, 193 Free water (free fluid)
effect on cement-slurry properties, 195–198 dispersants and, 78–80
field measurements, 201–202 gas migration and, 298
of foamed cement, 260 laboratory testing, 638–639, 658
formation damage and formation invasion depth, Freeze-protected cements, 237–239, 267
199–200 Fresh water, 343
gas migration and, 292–294, 309 Freshwater cement system, 231
laboratory testing, 635–636, 658 Friction factor, 615
mud removal and, 199 Friction pressure calculations, 130–141
pressure transmission during WOC, 200–201 FTIR. See Fourier transform infrared spectoscopy
static filtration, 192–193 Furfuryl alcohol-base cement system, 334
Fluid-loss control, 10
annular fluid migration, 292, 293
slurry selection, 444, 533 G
thermal cements in deep wells, 326
Galactomannans, as fluid-loss control agents, 84
Fluid-loss control agents, 49, 80–88, 91, 192
Gas channeling, 278
particulate materials, 80–81
Gas communication, 289
and spacers, 186
Gas flow potential (GFP), 300
water-soluble polymers, 81–88
Gas flow rate, 15–16, 17
Fluid properties measurement (FPM) plots, 595
Gas hydrates, 232
Fluoric acid and salts, as formation retarders, 58
Gas influx, 19
Fly ash
Gas invasion, 289
calculations involving, 662–663
Gas leakage, 289
classification, 63
Gas microannulus, 607
as extender, 59, 63
Gas migration, 289–317
Foamed cement, 4, 248, 256–262, 336, 337, 534
categories of, 290–292
bubble-size distribution, 257
causes of, 290, 306
calcium aluminate cements, 339
consequences of, 289–290
calculations, 672–677
controlling, 304–310
design, 258–262
immediate gas migration, 291–292
as extender, 59
laboratory testing, 310–316, 643
foam stability and structure, 256–257
long-term gas migration, 292, 293
as LCMs for cement slurries, 217
physical processes, 290–299
for massive loss remediation, 214
postplacement gas migration, 292
mixing, 629
predicting short-term migration, 299–304
properties, 258–262
short-term gas migration, 292, 293, 299–304
rheological properties and, 113–114
Gas migration cell, 313
ultralow-density, 332
Gas migration testers, 310–316

Subject Index 759

Gas wells, thermal cements for deep wells, 325–328 Herschel-Bulkley number, for annular flow, 125
Gaseous extenders, 58 Hesitation squeeze method, 525, 526, 528–529
Gel, 59 HF. See Hydrostatic factor
Gel strength, 118 High Activity Disposal Experimental Site (HADES), 230
gas migration and, 293, 294–296 High-alumina cement, 322
laboratory testing, 315, 316, 640–641, 658 High-density EPS systems, 255–256
measurement, 101–102, 118–119 High-fluid-loss slurry, 209
Gelhenite, 246 High-pressure high-temperature (HPHT) gas wells,
Gelled mud, removal during circulation, 158–161 247–248
Gelling agent, 210 High-pressure high-temperature (HPHT) viscometer,
Geothermal well cementing, 328–334 112
GFP. See Gas flow potential High-pressure pumps, 355–356
Gilsonite High-pressure squeeze, 525, 526, 527–528
as bridging agent, 88 High sulfate-resistant grade (HSR), 46
as extender, 64 High-temperature wells, 247–248, 319
as LCM for cement slurry, 216 See also Thermal cements
Glass fibers, as bridging agent, 88 Highly permeable formations, 204
Glass microspheres, 67, 256 HLB. See Hydrophilic-lipophilic balance
Glucoheptonate salts, as retarders, 55 Hole size, 436
Glucoheptonic acid, as dispersant, 71 Horizontal wells, 487–489
Gluconate, as retarder, 55 Hot dry rock design, 328
Gluconic acid, as dispersant, 71 HPC. See Hydroxypropyl cellulose
Good cementing practices, 164 HPG. See Hydroxy propylated guar
Granular coal, as bridging agent, 88 HPHT gas wells. See High-pressure high-temperature
Granular lost circulation materials (LCMs), 206–207, gas wells
208 HPHT viscometer. See High-pressure high-temperature
Gravity instabilities, in laminar flow, 176 viscometer
Greenhouse gases, 26 HRRT. See Hydraulic release running tool
Ground granulated blast-furnace slag, 246 HSR. See High sulfate-resistant grade
Grossular, 323 Hydraulic bond, 219
Grouting, 468–469 Hydraulic cement, 23
Guar gum, 84, 85, 186 Hydraulic fracturing, 20, 21
Guide shoes, 362–365 Hydraulic press, 636, 637
Gunk plugs, 212 Hydraulic release running tool (HRRT), 428
Gypsum, 38, 42 Hydraulic running tool, 401
Gypsum/Portland cement blends, 238 Hydraulic stage collars, 471
Gyrolite, 320 Hydraulic stage tools, 384–385
Hydraulic testing, 549–551
Hydraulically set liner hanger, 408, 409–410
H Hydrogarnet phase, 36, 42
Hydrophilic-lipophilic balance (HLB), 185
HADES. See High Activity Disposal Experimental Site
Hydrostatic cementing technique, 215
Heat of hydration, 52
Hydrostatic factor (HF), 301
Heat treatment, 26–28
Hydrostatic pressure
Heavy muds, 593
annular fluid migration and, 290–298, 300–304, 309
Heavy oil, 334, 335
calculations, 665, 666, 673–675
HEC. See Hydroxyethylcellulose
fluid loss and, 197
Hedström number, 122, 133, 135, 136
Hydroxyapatite, 324
Helicopter units, 360
Hydroxy propylated guar (HPG), as fluid-loss control
Hematite, as weighting agent, 69, 70
agent, 84
Herschel-Bulkley fluids, 97–98, 106, 108, 109, 116
Hydroxycarboxylic acids
flow equations, 123, 125, 127, 621, 622
as dispersants, 71
friction pressure calculations, 131
as retarder, 55–56

760 Well Cementing

Hydroxyethylcellulose (HEC) J
as antisettling agent, 79 Jackup rig, primary cement placement, 498–499
as fluid-loss control agent, 83, 84 Jet mixers, 351–352
Hydroxypropyl cellulose (HPC) Job monitoring
as fluid-loss control agent, 84 computer-based monitoring system, 357–358
in spacer fluids, 286 primary cementing, 498
Hysteresis, 272 squeeze cementing, 540
Hysteresis loop, 101

I K
Kickoff plugs, 506, 516
IBOP. See Internal blowout preventer Kilchoanite, 321
ICP spectroscopy. See Inductively coupled plasma Kilns, 27–28
spectroscopy
Ideal packing theory (IPT), 207
Ilmenite, as weighting agent, 69, 70
Immediate gas migration, 291–292 L
Immobile mud, 145, 179–181 Laboratory testing, 627–658
Impedance thresholds, 593 cement characterization and analysis, 653–657
Impermeable hydrate layer theory, 32 cement mix water, 343, 656–657, 658
Index of zonal isolation, 9, 15–19 coefficient of linear thermal expansion, 653
Induced fractures, 204 compressive strength, 636–637, 658
Induction period, 32–33 dry-blended cements, 656, 658
Inductively coupled plasma spectroscopy (ICP), 654, 657 expansion and shrinkage, 641–643, 658
Inflatable bridge plugs, 428 fluid loss, 635–636, 658
Inflatable cement retainer, 429 free fluid, 638–639, 658
Inflatable packers, 514 gas migration, 310–316, 643
Inflatable service packers, 427 mechanical properties, 646–653, 658
Inflatable service tools, 427–430 permeability, 643–644, 658
Inflow testing, 551 primary cementing, 632–634
Injection test, 537–538 rheological properties, 639–640, 658
Injectivity test, 511 sample preparation, 628–630, 658
Inner-string cementing, 375–377, 466–467 sampling, 658
Inorganic compounds, as retarders, 58 sedimentation, 638–639
Insert equipment, 372 slurry density, 630
In situ combustion wells, 338, 339 sonic strength, 637–638
Inside blowout preventer, 515 spacers and chemical washes, 644–646
Interface stability, cement plugs, 517–518 squeeze cementing, 634
Intermediate casing, 463–464 standards, 627–628
Intermediate liners, 475 static gel strength, 118–119, 315, 316, 640–641, 658
Intermixing, 188 strength, 636–637
Internal blowout preventer (IBOP), 515 thermal conductivity, 653
Invert emulsion muds, 185 thermal properties, 652–653
Invert-emulsion muds, 184 thickening time, 4, 630–634, 658
Iodine, as radioactive tracer, 90, 542 See also Cement job evaluation
IPT. See Ideal packing theory Lamellar lost circulation materials (LCMs), 207, 208
Iridium, as radioactive tracer, 90, 542 Laminar flow, 93–94, 132–133
Iron (II) sulfate, 235 circulation efficiency, 149
ISO standards, 5 equations, 119–123, 613–621
Land rig, primary cement placement, 498–499
Landing collar, 415, 418
Large-bore autofill float equipment, 372–373
Large-bore subsea wiper plug system, 380

Subject Index 761

Latch-in system, 377 Lost circulation, 10, 202–219
Latex additives, to prevent gas migration, 308, 309 bridging agents to prevent, 206–209
Latex-modified cements, 89, 242–243, 267 consequences of, 202–203
as fluid-loss control agents, 80, 81 controlling massive lost circulation, 214
for wells, 243, 267, 521 design software, 217–218
LCMs. See Lost circulation materials during cementing, 215–217
Lead oxide, as retarder, 58 severity of losses, 203
Lead pumpdown plug, 418 treatment decision tree, 205
Leakage rate, 16 well construction, 505–506
Le Châtelier flask, 656 while drilling, 204–213
“Letting-in tool,” 400 zones, 203–204
Lift equipment, 19 Lost circulation materials (LCMs), 206–209, 216–217,
Lightweight-particle extenders, 64–69 218
Lignosulfonates Lost-circulation plugs, 506, 516, 521
as dispersants, 71 Lost-circulation prevention agents, 49, 88–89, 91
as retarders, 54–55 downhole-mixed systems, 212–213
Limestone, bonding of cement to, 222–224 preflushes, 216
Linear elasticity, 272–273 surface-mixed systems, 209–210
Linear strain, 269 Lost returns, 202
Liner cementing procedure, 476–484 Low-density aggregates, as extenders, 58
Liner hangers, 405–414, 475 Low-density EPS systems, 254–255
Liner placement technique, 405, 406 Low-permeability cement slurries, 308
Liner setting sleeve, 403–405 Low-pressure squeeze, 525, 526, 527, 536
Liner setting tool assembly, 400–403 Low-temperature applications, freeze-protected
Liner stage tools, 386 cements, 237–239, 267
Liner top packers, 403, 404, 405 LWD caliper, 439
Liner wiper plug, 416, 417, 418
Liners, 397–420, 475–484
cementing procedure, 476–484 M
components of, 398–415
M-DOB plugs. See Mud-diesel-oil-bentonite (M-DOB)
displacement plugs for, 416–418
plugs
liner setting tool assembly, 400–403
M-DOB2C plugs. See Mud-diesel-oil-bentonite-cement
planned squeeze, 482
(M-DOB2C) plugs
reasons for installing, 398
Magnesium oxide, 237
running a liner, 475–476, 481
Magnesium oxychloride cement, 265
selection, 419–420
Manganese tetraoxide, as weighting agent, 69, 70–71
shoe track assembly, 415
Matrix simulation, 20
surface cementing equipment, 399–400
Maximum pressure restriction (MPR), 300
tieback liners, 475, 482–483
Maximum tensile stress criterion, 277
types, 397, 475
Mean flow, 187
Liquid additives
Mean velocity, 120
delivery, storage, and mixing of, 344
Mechanical plugs, 187, 517
metering system, 349
Mechanical properties, 8, 269–288
mixing, 349
Brazilian test, 650–651
Local-power-law method, 134–136, 623–625
of cement, 271–280, 647–651
Logging, 20
of cement cased wellbore, 280–287
See also Acoustic logging; Boron-pulsed neutron log;
cement system design, 443–444
CBL/VDL logs; Cement bond logs; Noise logging;
confined triaxial compression test, 650
Nuclear logging; Oxygen activation logs; Production
dynamic properties, 651–652
logs; Pump-in temperature log; Temperature logs;
of EPS cements containing flexible particles, 263
Variable-density logs
of foamed cement, 260
Long-term gas migration, 292, 293
laboratory testing, 646–653, 658

762 Well Cementing

 static properties, 647–651 Mud circulation
uniaxial tensile strength, 650 casing movement, 155
uniaxial unconfined compression test, 647–650 circulation efficiency, 151–154
Mechanical stage collars, 471 laminar flow in concentric annulus, 149–151
Mechanical strength, squeeze cementing and, 536 measuring mud-circulation efficiency in the field,
Mechanical tools, 381, 382 161–162
Mechanically set liner hanger, 408–409 modeling, 148–163
Melilite, 246 string eccentricity, 151–154
Membrane test, 642 turbulent flow in concentric annulus, 151
Merwinite, 246 Mud clinging constant, 370
Metallic microribbons, as strengthening agents, 89 Mud-completion fluid displacement, 183
Metering unit, 349 Mud conditioning, 143, 148, 164, 469
Methylenephosphonic acid derivatives, as retarder, Mud contamination, cement performance and, 516–520
57–58 Mud decontaminants, 90
Microannulus, 14, 16, 17 Mud-diesel-oil-bentonite (M-DOB) plugs, 212, 213
CBL and, 582, 605 Mud-diesel-oil-bentonite-cement (M-DOB2C) plugs, 213
gas migration and, 299 Mud displacement, 143, 144, 145, 163–183
ultrasonic testing, 590–592 casing movement and, 182–183
USI and, 605 concentric annulus, 165–169
Microbubbles, 214 eccentric annulus, 169–174
Microdebonding, 606–608 field studies, 182
Microfine cements, 247, 265, 267 of immobile mud, 179–181
Microsilica, as extender, 64 laminar criteria, 174–175
Microsphere-extended cement systems, 4, 57, 67–69, objectives, 187
256, 332, 336, 629 turbulent flow and, 175–179
Mineral fibers, for lost circulation prevention, 211 Mud filtercake
Minimum fracture pressure, calculation, 673 development, 533
Mixed lost circulation materials (LCMs), 207, 209 removal, 159–160, 161
Mixed metal hydroxides (MMH), as drilling fluid, 214 Mud mobility factor, 177
Mixed metal oxide and hydroxide (MMOH) systems, 214 Mud-mud displacement, 183
Mixed metal oxide (MMO) systems, 214 Mud removal, 10, 143–189, 445
Mixing, 176 cleaning-fluid sequence, 187
Mixing unit, 354–355 drilling fluids, 184
MMH systems. See Mixed metal hydroxides in eccentric annuli, 169–174
MMO systems. See Mixed metal oxide (MMO) systems fluid loss and, 199
MMOH systems. See Mixed metal oxide and hydroxide gas migration and, 299
(MMOH) systems gelled mud, 158–161
Mobile mud, 145 guidelines, 301
Modeling. See Computer modeling of immobile mud, 179–181
Moderate sulfate-resistant grade (MSR), 46 liner cementing, 476–484
Modified lignosulfonates, as dispersants, 71 measuring mud-circulation efficiency in the field,
Modified Sorel cement, 265 161–162
Mohr-Coulomb criterion, 277 mud displacement, 143, 144, 145, 163–183
Monitoring mud filtercake, 158–161
computer-based monitoring system, 357–358 plug cementing, 520
Portland cement hydration, 40–42 preflushes, 183, 184–186
primary cementing, 498 solids bed, 160–161
squeeze cementing, 540 well preparation, 143, 145–163
Monocalcium aluminate, 322 Mud removal factor (MRF), 301
Montmorillonite, 225 Mud removal parameter (MRP), 302, 303
MPR. See Maximum pressure restriction Mud-to-cement systems, 250–251
MRF. See Mud removal factor Multilaterals, 491–493
MRP. See Mud removal parameter Multistage cementing, 470–474
MSR. See Moderate sulfate-resistant grade

Subject Index 763

N P
Narrow concentric annular flow, flow equations and, Packer-bore receptacle, 473
124–126 Packer leaks, 509
Narrow-gap approximation, 107, 108, 128, 129, 139 Packer-setting dog sub, 401
Natural polymers, as fluid-loss control agents, 83–84 Packer shoes and collars, 389
Neat cement, physical characterization, 654 Packers
Negative pressure test, 541 inflatable, 514
Neutral fractures or fissures, 204 retrieval, 529
Newtonian fluids, 95, 105 through-tubing inflatable retrievable packer
flow equations, 121, 127, 616, 622 (TTIRP), 431
friction pressure calculations, 130 Packing element bridge plugs, 425, 426
Nitrogen, as extender, 69 Packing volume fraction (PVF), 251–252
Noise logging, 555 Pad-type sonic tools, 583–584
Non-Newtonian fluids, 96, 131 Paraformaldehyde, as mud decontaminant, 90
Nonionic synthetic polymers, as fluid-loss control Partially hydrolyzed polyacrylamide, as fluid-loss control
agents, 84–85 agent, 85
Nonpyruvylated xanthan gum, as spacers, 185 Particle migration, rheological measurements and, 112
Normal stress, 269 Particle-size analysis, 654–656
Nuclear logging, 554 Particulate materials, as fluid-loss control agents, 80–81
Nucleation of C-S-H theory, 33 Particulated rubber, as strengthening agent, 89
Nucleation of calcium hydroxide theory, 33 Pascal (unit), 95
Nucleation theory of retardation, 54 Pascal-second (unit), 95
Nut shells, as LCMs for cement slurries, 216–217 Patchy bond, 607
Nylon fibers PBR system. See Polished-bore receptacle (PBR)
in flexible cements, 264 packoff system
as LCMs for cement slurries, 217 PCP. See Polymer-based crosslinked pills
as strengthening agents, 89 PDLP. See Pressure decay limit parameter
Pectolite, 321
Perforation, 285, 542, 551
O Perforation plugging theory, 522–525
Perforation washing tool, 530–531, 538
OBM gelling agent, for lost circulation prevention, 211
Performance testing, prejob, 41
Offshore cementing, 239, 484–487
Periclase, 237
diesel engines and, 362
Perlite, as extender, 64, 65
transportation of bulk materials to wellsite, 347
Permafrost, 237
See also Salt cement; Seawater
See also Freeze-protected cements
Oil flow rate, 16
Permeability
Openhole size, 436–438
calcium chloride and, 53
Optical microscopy, 656
of foamed cement, 258–260
Optimal displacement volume, 519
gas migration and, 298
Ordinary Portland cement (OPC), 23
influence on zonal isolation, 16, 17
See also entries under Portland cement
laboratory testing, 643–644, 658
Organophilic clays, for lost circulation prevention, 211
Permeameters, 643–644, 654
Organophosphonates, as retarders, 56–58
PET. See Pulse Echo Tool
Organosiloxane polymer cement, 334
n-phosphonomethyl iminodiacetic acid, as retarder, 56
Orifice fill valves, 372
Phosphoric acid and salts, as retarders, 58
Overbalance pressure, 300
Pills, lost circulation prevention with, 209
Oxford clay, 229
Pipe coatings, 593
Oxford shale, 229
Pipe flow, 99
Oxygen activation logs, 511, 554–555
eccentricity and flow, 137–138
equations, 613–625
friction pressure calculations, 130–135
laminar flow equations, 120–124, 614
power-law parameter equations, 615

764 Well Cementing

Pipe reciprocation, 182 PNS. See Polynaphthalene sulfonate
Pipe removal, plug cementing, 520 “Point of departure,” 632
Pipe standoff ratio, 137 Poisson’s ratio, 273, 651
PIT. See Pressure-integrity test Polished-bore receptacle (PBR) packoff system, 402, 403
Plastic behavior, 275 Polyacrylamides, as fluid-loss control agents, 85
Plastic viscosity, 97 Polyallylamine, as fluid-loss control agent, 87
Plasticity number, 122 Polycarboxylate-based dispersants, 71, 73, 76–77
Plasticizers, as dispersant, 71 Polyethyleneimine, as fluid-loss control agent, 87
Plate wave, 557 Polyglycol ethers, as antifoam agents, 89
Plug balancing, calculations, 668–671 Polymelamine sulfonate (PMS), as dispersant, 71, 72, 75
Plug cementing, 2, 3, 11, 378–380, 503, 512–521 Polymer-based crosslinked systems, lost circulation
balanced plug, 513 prevention with, 209–210
benefits of, 505 Polymer-based crosslinked pills (PCP), 210
cement slurry design, 520–521 Polymeric microgels, 308
cleanup, 520 Polymeric oil gellant, for lost circulation prevention, 211
coiled tubing, 515 Polymeric plugs, lost circulation prevention with, 209
defined, 503 Polynaphthalene sulfonate (PNS), as dispersant, 71–72,
dump bailer, 513–514 75, 77
evaluation, 521 Polypropylene fibers, as LCMs for cement slurries, 217
inflatable packers, 514 Polystyrene sulfonate (PSS), as dispersant, 71, 72, 73
inside blowout preventer, 515 Polyvinyl acetate, as fluid-loss control agents, 80
mud contamination, 516–520 Polyvinyl acetate latex, 243
mud removal, 520 Polyvinyl alcohol (PVA), as fluid-loss control agent, 80,
placement, 515–520 81, 84–85
plug location, 515–516 Polyvinylpyrrolidone (PVP), as fluid-loss control agent,
two-plug method, 514 80, 84–85
umbrella-shaped membranes, 514–515 Poppet valves, 369, 374
Plug flow, 104 Pore pressure, calculations, 665
Plugs, 230, 378–380 Poroelastic constant, 274
abandonment plugs, 521 Poroelasticity, 274–275
balancing calculations, 668–671 Port collars, 381, 386
bottom plugs, 379 Portable ultrasonic nondestructive digital indicating
bridge plugs. See Bridge plugs tester (PUNDIT), 651, 652
cased hole bridge plugs, 424–425 Portland cement, 9, 23–48, 343
casing plugs, 416 BFS-Portland cement blends, 245, 246–247
cementing. See Plug cementing chemical characterization, 654, 655, 658
contamination of, 516–520 geothermal well cements based on, 330–331
good cementing practices, 164 high-temperature chemistry, 319–322
gunk plugs, 212 particle analysis, 658
inflatable bridge plugs, 528 properties, 23, 24
interface stability, 517–518 specific gravity, 658, 659
kickoff plugs, 506, 516 See also Ordinary Portland cement; Portland cement
for lost circulation prevention, 209 chemistry; Portland cement classification; Portland
mechanical plugs, 187, 517 cement hydration; Portland cement manufacture
mud-diesel-oil-bentonite (M-DOB) plugs, 212 Portland cement chemistry
placement, 187 alkalis, 39
polymeric plugs, 209 C-S-H phase, 31, 33, 34, 37, 41, 43, 52, 319, 320
protective plugs for well construction, 507 hydration. See Portland cement hydration notation,
subsea plug systems, 484–486 23
top plugs, 378–379 temperature, effect of, 37–38
whipstock plugs, 506, 516, 521 tricalcium aluminate, 35
PMS. See Polymelamine sulfonate Portland cement classification, 44–48
Pneumatic blending tank, 345 API/ISO system, 45–48
Pneumatic silos, 345 ASTM system, 44, 47

Subject Index 765

 potential phase composition, 44–45 Prehnite, 323
Portland cement hydration, 26, 30–43, 227 Preinduction period, 32, 33
acceleration and deceleration periods, 33–34 Prejob performance testing, 41
aging, effects of, 38–39 Premix method, 349
alkali and, 39 Pressure
aluminate phase hydration, 34–36 downhole pressure, 497
calcium chloride, effect on, 51–52, 52–53 fracturing pressure, 291, 440, 487, 497, 508, 520, 525,
computer modeling, 43–44 536, 537, 539
diffusion period, 34 influence on thickening time, 326
flash set and false set, 38 pore pressure, 290, 291, 440, 443, 487, 497, 520, 536
at high temperatures, 319–321 rheological properties and, 117
induction period, 32–33 for squeeze cementing, 536
mechanism of action, 74–77 Pressure decay limit parameter (PDLP), 302, 303–304
monitoring and modeling techniques, 40–42 Pressure drop tubes, 315
as multicomponent system, 36–40 Pressure-integrity test (PIT), 549
preinduction period, 32, 33 Pressure surges, 370–371
silicate phase hydration, 30–34 Pressure testing, 549–550
sulfate resistance and, 40 Pressurized mud balance, 356, 630
surface area and, 39 Pressurized tanks, 347, 348
temperature, effect of, 37–38 Primary cementing, 9, 11, 459
thermogram, 37 calculations, 495
volume changes, 37 fluid loss during, 191, 194–201
Portland cement manufacture, 23 laboratory testing, 632–634
cooling, 29 monitoring, 498
grinding, 30 techniques. See Primary cementing techniques
heat treatment, 26–28 thickening time, 630–634
impurities, 25 Primary cementing techniques, 1–2, 459–500
raw material preparation, 25–26 casing drilling, 493–495
raw materials, 24–30 casing strings, 459–466
storage, 30 cement placement, 445–446, 466–474
Portland slag cements, 245 extended-reach wells, 487–489
Portlandite, 30, 34 floating rig, 500
Positive pressure test, 541 hole condition, 495–498
Postfailure behavior, 278–279 liners, 397–420, 475–484
Postfailure region, 272 multilaterals, 491–493
Postplacement gas migration, 292 offshore cementing, 484–487
Potassium silicate, as plugging fluid, 213 solid expandable tubulars, 489–490
Potassium sulfate, 39 stationary rig, 498–500
Potential phase composition, 44–45 Principal planes, 270
Powder XRD, 45 Principal stresses, 270
Powdered aluminum, 237 Production casing, 465
Powdered coal, as extender, 66 Production liners, 475
Power-law fluids, 96–97, 103, 105, 106, 108, 133, 134 Production logs, 511
circulation efficiency, 152 Protective plugs, well construction, 507
flow equations, 121–122, 125, 127, 615, 617–618, 622 Pseudoplastic fluids, 96
local power law method, 136 PSS. See Polystyrene sulfonate
velocity and shear rate profiles, 126 Pull tests, 407
Pozzolans, as extenders, 61–64 Pulse Echo Tool (PET), 585
Precipitation theory of retardation, 54 Pump-in temperature log, 552–553
Predisplacement, 143 Pump pressure, calculations, 665, 666
Preflushes, 183, 184–186 Pumping pressure, 356
gas migration and, 309 PUNDIT. See Portable ultrasonic nondestructive digital
lost-circulation control with, 216 indicating tester

766 Well Cementing

PVA. See Polyvinyl alcohol Rheological models, 9, 95–100, 105–106, 108–109
PVF. See Packing volume fraction Rheological properties, laboratory testing, 639–640, 658
PVP. See Polyvinylpyrrolidone Rheology, 93–95
Pycnometer, 656 Bingham plastic fluids, 97, 103, 104, 105, 107, 108, 124
Casson model, 98, 106
coaxial cylinder viscometer, 100–101, 104, 105–106
Q conversion factors, 625
equations, 613–625
Quality control, 497, 555–556, 597–611
experimental procedures, 101–104
See also Cement job evaluation
of foamed cement, 262
Quartz
Herschel-Bulkley fluids, 97–98, 106, 108, 116
as extender, 64
laboratory testing of rheological properties, 639–640,
in thermal cements, 320, 322–323, 330
658
Quick set, 38
laminar flow, 93–94
models, 9, 95–100, 105–106, 108–109
Newtonian fluids, 95, 105
R non-Newtonian fluids, 96
Radioactive densitometer, 356, 357 power-law fluids, 96–97, 103, 105, 106, 108
Radioactive tracers Robertson and Stiff model, 98, 106
as cement additives, 90, 91, 542, 554 turbulent flow, 94
in squeeze cementing, 542 See also Cement slurry rheology
RAS cements. See Right-angle-set (RAS) cements Rietveld method of cement analysis, 45
Reciprocating plunger pump, 355 Right-angle-set (RAS) cements, 309
Reciprocation, 497 Rigid centralizers, 390–391, 392
Recirculation jet mixer, 352, 353 Robertson and Stiff model, 98
Recirculation mixer without conventional jets, 352–353 Roller-equipped centralizers, 395
Remedial cementing, 2, 11, 503–547 Rotating liner hangers, 411–412
best-practices guidelines, 545–546 Rotating-paddle devices, 315, 316
coiled tube placement, 515, 531 Rotation, 497
equipment. See Remedial cementing equipment Rotational viscometer, 640
fluid loss during, 201 Rotor-and-bob geometry, 100
injection test, 537–538 Rugosity, 593
job design, 535–537 Running squeeze method, 525, 526, 528
noncementing solutions, 543
objectives, 549
plug cementing, 2, 3, 11, 503, 505, 512–521 S
production/injection, 509–511
Saccharide compounds, as retarders, 56
squeeze cementing, 2, 3, 11, 247, 503, 505, 521–543
SAGD method. See Steam-assisted gravity drainage
well construction, 505–509
(SAGD) method
Remedial cementing equipment, 421–430
Safety, cementing units, 360–361
cased hole bridge plugs, 424–425
Salicylic acid, as dispersant, 71
cased hole squeeze packers, 421–423
Salt, calculations involving, 661–662
inflatable service tools, 427–430
Salt cements, 231, 236, 239–242, 267
tubing testers and unloaders, 426–427
Salt domes, 240
Repeat section, 555, 556
Salt formations, 567
Retarders, 49, 54–58, 91, 444
Salt-free cement, 231, 241
Retrievable bridge plugs, 424–425
Salt-zone cementing, 247
Retrievable drilling systems, 493
Sandstone, bonding of cement to, 222–224
Retrievable packers, 431, 529
Sawtooth pattern guide shoe, 365
Retrievable seal system, 403
Scab liner, 376, 397, 543
Retrievable squeeze packers, 422
Scab tieback liner, 397
Reversing out, plug cementing, 520
Scanning electron microscope, 654
Reyerite, 321
Scavenger slurry, 175
Reynolds number, 130, 134, 136, 613, 615

Subject Index 767

Scawtite, 321 Silica flour, 244, 330
Scleroglucan, as spacers, 185 Silica “fume,” 331
SCP. See Sustained casing pressure Silica sand, 330
Scratchers, 370, 372, 395, 396 “Silica-stabilized” Portland cements, 320
Screw-in stinger, 376, 377 Silicate extenders, 59, 60–61
Sealing plug, 383 Silicate phase, 58
Searl-type rotational viscometer, 640 Silicate phase hydration, 30–34
Seawater, 51, 79, 239–240, 343 Silicate system, for lost circulation prevention, 211–212
Secondary gypsum, 38 Silicates, 79
Sedimentation Silicones, as antifoam agents, 89
dispersants and, 78–79 Silos, 345
laboratory testing, 638–639 Siloxane sheet, 225
squeeze cementing and, 533 Simulation techniques. See Computer modeling
Seepage losses, 203 Single-arm calipers, 437
Semirigid centralizers, 392 Single-pump cementing unit, 360
Semisaturated cement systems, 241 Single-stage cementing, 469–470
Semisubmersible rig, 500 Single-stage subsea cementing system, 485
Semitrailer-mounted cementing unit, 360, 361 Sintering zone, 27
Semitrailer-mounted units, 360 Six-arm caliber tool, 438
Service packers, inflatable, 427 Skid-mounted units, 359
Set-cement expansion, laboratory testing, 641–643 Slag cements, 245, 246
Set-cement failure, 277 Sliding gate, 568
Set-cement shrinkage, laboratory testing, 641–643 Sliding sleeve valves, 422
Set-cement strength, 277 Slimhole landing collar, 419
Setting period, 33 Slip, 408
Shale, 240 Slip velocity, 110, 111
Shales, 225–226 Slow formation, 572, 578
interaction with drilling fluids, 226 Slow-flow technique, 168
interactions between cement slurries and shales, Slowness, 557
227–230 Slurry. See entries under Cement slurry
Shear band activation, 286 Slurry compressibility factor, 295
Shear bond, 219, 260 Slurry performance number (SPN), 301
Shear failure, 14 Small-particle-size cement systems, 247
Shear instability, in laminar flow, 176–177 Small wellbore simulator (SWBS), 229
Shear plane, 73 Smectite, 59, 225
Shear rate, 95, 125 Soaps, for lost circulation prevention, 211
Shear stress, 95, 99 Sodium calcium silicate hydrate, 321
Shear thickening, 96 Sodium chloride
Shear thinning, 96 as accelerators, 50–51
Shear wave, 556 as retarder, 58
Shoe cases, 377 Sodium chromate, as mud decontaminant, 90
Shoe joint, 362 Sodium metasilicate, as extender, 60–61
Shoe track assembly, 415 Sodium silicate
Short-term gas migration, 292, 293, 299–304 as accelerator, 51
Shrinkage, 279–280 as extenders, 59, 60–61
bulk shrinkage, 279 as plugging fluid, 213
calcium chloride and, 53 Sodium sulfate, 77
chemical shrinkage, 37, 279, 298 Sodium tetraborate decahydrate, as retarder, 58
gas migration and, 296–298 Sodium trichloroacetate, 210
laboratory testing, 641–643, 658 Software. See Computer software
Sidetrack plug, 506 Solid expandable tubulars, 489–490
Sidetracking, well construction, 506 Solid volume fraction (SVF), 71, 251
Silica, as extender, 64 Solids bed, removal, 160–161
Silica-base fibers, as LCMs for cement slurries, 217 Solids fraction monitoring equipment, 357

768 Well Cementing

Sonic cement evaluation techniques, 561–584 preparing for, 543, 544
history, 561–562 running squeeze method, 525, 526, 528
pad-type sonic tools, 583–584 spacers and washes, 535–536
ultrasonic pulse-echo technology, 585–597 squeeze tools and placement techniques, 525, 527,
See also Acoustic logging 529–530
Sonic strength, laboratory testing, 637–638 temperature log, 541
Sorel cement, 265 thickening time, 634
Spacers, 120, 185–186, 445 tools and techniques, 525–532
good cementing practices, 164 Squeeze packers, 421, 530
testing for compatibility, 644–646, 658 Squeeze-tool technique, 525, 527, 529–530
Spearhead fluid, 538 Stab-in cementing, 375–376, 377, 466–467
Special cement systems, 233–268 Stab-in stinger, 376
acid-soluble cements, 265, 267 Stabilizers, for foam cements, 258
chemical bonded phosphate ceramics (CBPCs), Stage collars, 215, 381
265–266, 267 Stage tools, 381–386, 383, 384
for corrosive environments, 244–245, 267 Standardization procedures, 5–6
engineered particle-size (EPS) distribution cements, Standoff ratio, 137–141, 152–154, 156, 157, 463, 164,
251–256, 267 170–172, 179, 180
expansive cement systems, 235–237, 267 Static filtration, fluid loss, 192–193
flexible cements, 262–264, 267 Static fluid pressure, 536
freeze-protected cements, 237–239, 267 Static gel strength, laboratory testing, 315, 316,
latex-modified cement systems, 80, 81, 89, 242–243, 640–641, 658
267 Static heated filterpress cell, 635
microfine cements, 247, 265, 267 Stationary rig, primary cement placement, 498–499
salt cements, 231, 236, 239–242, 267 Statistical models, 41
salt-free cement, 231, 241 Steam-assisted gravity drainage (SAGD) method, 335
storable cement slurries, 266, 267 Steam injection, 248
thixotropic cements, 89, 209, 217, 233–235, 267 Steam recovery wells, 335–338
ultralow-density cement systems, 256–262, 267 Steamflood wells, 335
Specialty additives, 49, 661–662 Steel flow hoses, 358
Specific gravity, Portland cement, 658, 659 Stern layer, 73
SPN. See Slurry performance number Stinger, 517
Spring-bow centralizers, 392, 393, 394 Stokes law, 254
SPVT. See Sulfonated polyvinyltoluene Stop collars, 389–390
Squeeze cementing, 2, 3, 11, 247, 503, 505, 521–543 Storable cement slurries, 266, 267
benefits of, 505 Storage, 345, 347–348
Bradenhead squeeze, 525, 527, 529, 536, 540 in hot regions, 39
calculations, 670–672 of Portland cement, 30
cement slurry design, 532–535 at wellsite, 347
cement slurry rheology, 533–534 Storage tanks, 349
cleanup, 540 Strain, 271
coiled tube placement, 531–532 Strain hardening, 276
defined, 503–504 Strain softening, 276
economics, 535 Strength retrogression, 319, 320
evaluation, 541–542 Strength testing, 636–637
failure, 542–543 Strengthening agents, 89, 91
hesitation squeeze method, 525, 526, 528–529 Stress, 280–282
high-pressure squeeze, 525, 526, 527–528 Stress-strain curve, 271–272
job design, 535–537 Stress vector, 269
laboratory testing, 634 Stretching effect, 574–575
low-pressure squeeze, 525, 526, 527, 536 String eccentricity, mud circulation, 151–154
monitoring, 540 Styrene-butadiene latex, as fluid-loss control agent, 80
perforation plugging theory, 522–525 Styrene butadiene latex systems, 243
perforation washing tool, 530–531, 538 Subsea plug systems, 484–486

Subject Index 769

Subsidence, 285–286 Testing
Sugar compounds. See Saccharide compounds acoustic logs, 542, 555–611
Sulfate resistance, 40 annular expansion ring test, 641–642
calcium chloride and, 53 boron-pulsed neutron log, 510, 554
Sulfonated polymers, as dispersant, 71 Brazilian test, 650–651
Sulfonated polyvinyl aromatics, 87 cased hole-log interpretation, 557
Sulfonated polyvinyltoluene (SPVT), 87 CBL/VDL logs, 574–575, 577–583, 599–607
Superheated formation water, 328 of cement, 5–6
Superplasticizers, as dispersants, 71, 74 cement bond logs (CBLs), 551, 562–584
Surface area, hydration and, 39 cement hydration detector, 551–552
Surface casing, 461–463 CHA test, 642
Surface cementing equipment, 399–400 communication indicator, 552–553
Surfactants, 185, 186, 309 confined triaxial compression test, 650
Surge pressure, 370–371 cube-crushing test, 269
Surge tanks, 350–351 cylindrical sleeve test, 642
Surge tool, 530, 531 drillstem test (DST), 551
Sustained casing pressure (SCP), 13, 504, 509 dry testing, 551
SVF. See Solid volume fraction injection test, 537–538
SWBS. See Small wellbore simulator injectivity test, 511
Synchrotron energy dispersive XRD, 42 membrane test, 642
Syngenite, 39 noise logging, 555
Synthetic cements, 244, 334, 534 nuclear logging, 554
Synthetic polymers oxygen activation logs, 511, 554–555
as bridging agent, 88 positive and negative pressure tests, 541
as fluid-loss control agents, 84–88, 309 prejob performance testing, 41
pressure testing, 549–550
production logs, 511
T pull tests, 407
pump-in temperature log, 552–553
Tailpipe, 517
temperature logs, 510, 541, 551–555
TAML. See Technology Advancement for Multilaterals
tests through perforations, 551
Tandem rise centralizers, 393
ultrasonic tools, 585–597, 651
Tapered casing string, 466
uniaxial tensile strength, 650
Tartaric acid
uniaxial unconfined compression test, 647–650
as dispersant, 71
variable-density log (VDL), 562, 563–567
as retarder, 55
See also Cement job evaluation; Laboratory testing
Taylor vortices, 155, 156
Tetracalcium aluminoferrite, 45
Technology Advancement for Multilaterals (TAML), 491
Texas pattern guide shoe, 365
Tectonic stresses, 285
TGA. See Thermogravimetric analysis
Temperature
Theory of elasticity, 270
rheological properties and, 115–117
Theory of elastoplasticity, 270
squeeze cementing and, 534
Thermal cements, 10, 319–341
Temperature logging
calcium aluminosilicate systems, 323–324
cement hydration detector, 551–552
calcium phosphate cement system, 324–325
communication indicator, 552–553
Class J cement, 322–323
pump-in-temperature log, 552–553
for deep oil and gas wells, 325–328
Temperature logs, 510, 541, 551–555
for geothermal well cementing, 328–334
Temperature modeling, 4–5
high-alumina cement, 322
Temperature regime, cement system design, 441
Portland cements, high-temperature chemistry,
Tensile strength, 8, 650
319–322
Tension packers, 421
for thermal recovery wells, 334–340
Tension-set service packers, 422
Thermal conductivity
Terzaghi’s law, 295
of foamed cement, 261
Test anchor, 507
laboratory testing, 653

770 Well Cementing

Thermal cycling, effect on cement properties, 334, 337, Tracers
338 mud-circulation efficiency measured by, 161–162
Thermal fracturing, 14 radioactive tracers as cement additives, 90, 91, 542,
Thermal properties, of set cement, 652–653 554
Thermal recovery wells, 10, 334–340 Transition time, 295, 301
Thermoelasticity, 273–274, 652 Transit time, 567
Thermogravimetric analysis (TGA), 654 “Treating iron,” 358
Thermoplastics, in flexible cements, 262–264 Tresca criterion, 277
Thickening time Tricalcium aluminate, 35, 45
fluid loss and, 197 Tricalcium silicate, 30–34
of foamed cement, 260 Triethanolamine, as accelerator, 54
laboratory testing, 4, 630–634, 658 Triplex pumps, 355
squeeze cementing, 534 Truck-mounted units, 359
Thief zone, 405 Truscottite, 321
Thiocyanate salts, as accelerator, 54 TTBRP. See Through-tubing inflatable retrievable
Thixotropic cements, 89, 233–235, 267 bridge plug
aluminum sulfate/iron (II) sulfate system, 235 TTCR. See Through-tubing cement retainer
calcium sulfate-based systems, 234–235 TTIRP. See Through-tubing inflatable retrievable packer
as cement plugs, 209 TTPBP. See Through-tubing permanent bridge plug
clay-base systems, 234 Tubing testers, 426
crosslinked cellulose polymer systems, 235 Tubing unloaders, 426, 427
gas migration and, 309 Turbidimeter, 655
as LCMs for cement slurries, 217 Turbolizers, 393
Thixotropy, 101, 209, 233 Turbulent flow, 94, 188
Three-arm calipers, 437 concentric annulus, 151
Three-dimensional cement hydration simulation, 43, 44 equations, 615–625
Three-stage cementing, 470, 474 mud displacement and, 175–179
Through-drillpipe stab-in cementing, 467 Two-arm calipers, 437
Through-tubing cement retainer (TTCR), 433 Two-dimensional cement hydration simulation, 43
Through-tubing electric wireline setting tool, 429 Two-plug cementing method, 1–2
Through-tubing inflatable products, 431–433 Two-plug method, plug cementing, 514
Through-tubing inflatable retrievable bridge plug Two-stage cementing, 470, 472
(TTRBP), 432 TXI Lightweight, 63
Through-tubing inflatable retrievable packer (TTIRP), Type K cement, 236
431 Type M cement, 236
Through-tubing permanent bridge plug (TTPBP), 432 Type S cement, 236
Tieback casings, 483
Tieback liners, 475, 482–483
Tieback receptacle, 483 U
Tieback sealing nipple, 483
U-tubing, 448
Tieback sleeve, 483
UCA. See Ultrasonic cement analyzer
Titus Assembly, 468
Ultrafine Portland cements, 238–239
Titus Ring, 468
Ultralow-density cement systems, 256–262, 267, 332
Tobermorite, 320, 332
Ultrasonic caliper, 439
Tool eccentering, 593
Ultrasonic cement analyzer (UCA), 269, 637, 638
Top-drive subsea cement heads, 485
UltraSonic Imager (USI), 585, 586, 595, 599–611
Top of liner squeeze, 508–509
Ultrasonic pulse-echo technology, 585–597
Top plugs, 378, 379
Ultrasonic tools, 585, 651
Top-up cementing, 468–469
logs, 595–597
Topdrive drilling system, 399
measurements and processing, 586–593
Torque-reduction centralizers, 394
principles, 586, 587
Total chemical shrinkage, 279
Ultraviolet absorption spectrophotometry, 656
Toughness, 278

Subject Index 771

Umbrella-shaped membranes, 514–515 Water influx, 19
Unconsolidated formations, 204 Water metering, 348
Uniaxial compressive strength, 269 Water requirement, calculations involving, 664
Uniaxial tensile strength, 650 Water-soluble polymers, as fluid-loss control agents,
Uniaxial unconfined compression test, 647–650 81–88
Units, conversion factors for, 677 Weighting agents, 49, 69–71, 91
Unsoundness, 236 calculations involving, 664
Uploading sub, 426 and spacers, 186
USI. See UltraSonic Imager Welan gum
as antisettling agent, 79
as spacer, 185–186
V Well abandonment, 503, 512
Well construction, 505–509
Valves, 368–371
block squeeze, 508
Vane rheometers, 315, 316
circulating squeeze, 508
Variable-density log (VDL), 562, 563–567
directional drilling, 506
CBL/VDL logs, 574–575, 577–583, 599–607
evaluation, 507
Variable hole shape, 436
lost circulation, 505–506
Variable spacing centralizers, 390
protective plugs, 507
VDL. See Variable-density log
sidetracking, 506
Velocity-profile equations, 122
Well control, 446–448
“Vertical” well, 435, 436
Well preparation
Vinylidene chloride, as fluid-loss control agents, 80
borehole, 145–146
Vinylidene chloride latex, 243
circulation and displacement efficiency, 146–147
Viscometer, 357
modeling mud circulation, 148–163
coaxial cylinder viscometer, 100–101, 104, 105–106,
mud conditioning, 143, 148
639
Well temperature, slurry selection, 444
couette viscometer, 357, 639
Wellbore
high-pressure high-temperature (HPHT) viscometer,
design, 440
112
fluid properties, 576
rotational viscometer, 640
mechanical pressures, 282–285
Viscosification, fluid loss and, 197–198
Wellbore wipers, 395
Viscosifiers, 185
Wells
Viscosity, 95
calculations involving, 666–667
Viscous fingering, 165
chemical waste disposal wells, 244
Void volume, 537
deep wells, thermal cement for, 325–328
Volumetric flow rate, 120
extended-reach wells, 487–489
Volumetric strain, 272
fireflood wells, 338
Vugular formations, 204, 524
geothermal wells, 329
in situ combustion wells, 338–340
steam recovery wells, 335–338
W steamflood wells, 335
Wagner method, 655 thermal wells, 10
Waiting on cement (WOC), 470, 482 See also Well construction; Well control; Well
fluid loss and, 200–201 preparation
Wall slip, 110–112 Wet cell XRD techniques, 42
Washes, 184–185 Wet chemical methods for cement analysis, 654
good cementing practices, 164 Wet process, 26
squeeze cementing, 535–536 Whipstock plugs, 506, 516, 521
testing for compatibility, 644–646, 658 Wiper plugs, 379, 380
Water, for cementing, 343, 656–657 Wireline caliper tools, 438
Water extenders, 58 WOC. See Waiting on cement
Water glass, 60, 61

772 Well Cementing

X
X-ray diffraction (XRD), 42, 654
X-ray fluorescence (XRF), 654
Xanthan gum, in spacers, 185
Xonotlite, 321, 332
XRD. See X-ray fluorescence

Y
Yield number, 122
Yield point, 272
Young’s modulus, 272, 273, 282, 299, 651

Z
Zeta potential, 73, 75, 76
Zinc oxide, as retarder, 58
Zonal isolation, 9, 14–19, 15, 219, 465

Subject Index 773