Ugwu Thesis

CEMENT FATIGUE AND HPHT WELL INTEGRITY WITH
APPLICATION TO LIFE OF WELL PREDICTION
A Thesis
by
IGNATIUS OBINNA UGWU
Submitted to the Office of Graduate Studies of

Texas A&M University
in partial fulfillment of the requirements for the degree of
MASTER OF SCIENCE
December 2008
Major Subject: Mechanical Engineering

CEMENT FATIGUE AND HPHT WELL INTEGRITY WITH
APPLICATION TO LIFE OF WELL PREDICTION
A Thesis
by
IGNATIUS OBINNA UGWU
Submitted to the Office of Graduate Studies of

in partial fulfillment of the requirements for the degree of
MASTER OF SCIENCE
Approved by:
Co-Chairs of Committee, Xin-Lin Gao

Catalin Teodoriu
Committee Members, Anastasia Muliana
Alan Palazzolo
Head of Department, Dennis O’Neal
December 2008
Major Subject: Mechanical Engineering

iii
ABSTRACT
Cement Fatigue and HPHT Well Integrity with Application to Life of Well Prediction.
(December 2008)
Ignatius Obinna Ugwu,
B.Eng., Federal University of Technology, Owerri, Nigeria
Co-Chairs of Advisory Committee, Dr. Xin-Lin Gao

Dr. Catalin Teodoriu
In order to keep up with the world’s energy demands, oil and gas producing companies
have taken the initiative to explore offshore reserves or drill deeper into previously
existing wells. The consequence of this, however, has to deal with the high temperatures
and pressures encountered at increasing depths.
For an oil well to maintain its integrity and be produced effectively and
economically, it is pertinent that a complete zonal isolation is achieved during well
completion. This complete zonal isolation can be compromised due to factors that come
into play when oil well cement experiences cyclic loading conditions which can lead to
fatigue failure as a consequence of extensive degradation of the microstructure of the
cement material depending on stress levels and number of cycles. There have been a lot
of research and experimental investigations on the mechanism of fatigue failure of
concrete structures but the fatigue behavior of oil well cement is still relatively unknown
to engineers.
iv
Research in the area of oil well cement design has led to improved cement
designs and cementing practices but yet many cement integrity problems persist and this
further strengthens the need to understand the mechanism of cement fatigue.
This research seeks to develop a better understanding of the performance of the
casing cement bond under HPHT well conditions that can lead to best practices and a
model to predict well life. An analytical model, which can be used to evaluate stresses in
the cement sheath based on actual wellbore parameters, was developed and combined
effectively with finite element models to evaluate the fatigue and static loading behavior
of a well.
Based on the findings of this investigation, the mechanical properties of the
casing, cement and formation as well loading conditions play a very big role in the static
and fatigue failure of well cement.
Finally, recommendations for future work on this subject were also presented in
order to understand all tenets of cement fatigue and to develop governing equations.
v
DEDICATION
To God who gave me the strength and intellect to venture into this work, this work and
its progress is dedicated to Him.
To my parents who have been shinning beacons and continuous source of
encouragement, even in the face of difficulty, to me, this work is especially dedicated to
you.
To everyone else who helped along the way, you are a great part of this work.
vi
ACKNOWLEDGMENTS
I would love to express my profound gratitude to my committee chairs, Dr. Xin-Lin Gao
and Dr. Catalin Teodoriu, for their untiring efforts in seeing to the success of this work.
Your willingness to devote your time to answer my questions whenever I called and your
continued guidance, understanding and support is something I highly appreciate.
Further appreciation is expressed to Dr. Palazzolo and Dr. Muliana, for making time
to serve on my thesis committee.
I would also not forget to acknowledge the minerals management services (MMS)
for funding this project.
To all my friends here at Texas A&M, especially Jerome Rajnauth, your warmth
and care made my transition through this great school much easier and enjoyable.
vii
NOMENCLATURE
HPHT High Pressure -High Temperature
S-N Diagram Stress-Cycle Diagram
LS Low Strength
HS High Strength
BWOC By Weight of Cement
BVOB By Volume of Blend
BWOW By Weight of Water
FEA Finite Element Analysis
SF Safety Factor
E Young’s Modulus
v Poisson Ratio
σa Stress Amplitude
σm Mean Stress
viii
TABLE OF CONTENTS
Page
ABSTRACT…………………………………………………………………………… iii
DEDICATION…………………………………………………………………………. v
ACKNOWLEDGMENTS……………………………………………………………… vi
NOMENCLATURE…………………………………………………………………… vii
TABLE OF CONTENTS………………………………………………………………. viii
LIST OF FIGURES…………………………………………………………………….. x
LIST OF TABLES……………………………………………………………………… xiv
CHAPTER
I INTRODUCTION: RESEARCH MOTIVATION……………………...................... 1
1.1 Background……………………………………………........................... 2
1.1.1 HPHT Well Integrity……………………………………………... 2
1.1.2 Failure Criteria……………………………………………………. 3
1.1.3 Fatigue…………………………………………………………….. 4
1.1.4 Fatigue in Cement/Concrete………………………………………. 6
1.2 Research Objectives………………………………………....................... 11
1.3 Approach...……………………………………………………………… 11
II OILFIELD CEMENTING…………………………………………………………… 12
2.1 Well Cement…………………………………………………………….. 12

2.1.1 New Cements vs. Conventional Well Cements………………….. 17
2.2 Cement Additives……………………………………………………….. 21
2.3 Well Cementing Design Process………………………………………... 23
2.4 Cement Integrity Issues Due to Temperature and Pressure…………….. 25
2.4.1 Effect of High Temperature………………………………………. 25
2.4.2 Combined Effect of Temperature and Pressure…………………... 27
2.4.3 Casing-Cement –Formation Interactions…………………………. 29
ix
CHAPTER Page
III CEMENT FAILURE DUE TO STATIC LOADING……………………………… 33
3.1 Analytical Model of Wellbore Stresses……………………………….… 33

3.1.1 Background………………………………………………………. 33
3.1.2 Assumptions……………………………………………………... 35
3.1.3 Failure Criteria…………………………………………………… 35
3.1.4 Analytical Model………………………………………………… 38
3.2 Analytical and Finite Element Studies………………………….……… 45
3.3 Fatigue Studies……………………………………………………….… 49
IV RESULTS AND DISCUSSION…………………………………………………... 56
4.1 Static Studies…………………………………………………..……… 56

4.2 Fatigue Loading……………………………………………………….. 72
V CONCLUSIONS AND RECOMMENDATIONS…………………………….…... 88
5.1 Conclusions……………………………………………………..…....... 88
5.2 Recommendations for Future Work…………………………………… 90
REFERENCES…………………………………………………………………………. 91
VITA………………………………………………………………………..................... 94
x
LIST OF FIGURES
FIGURE Page
1.1 Fluctuating Tensile and Compressive Cyclic Stress ........................................... 6
1.2 Comparison of Fatigue Strain for High and Low Strength Cements ………….. 7
1.3 Comparison of Max. Stress Levels to Number of Cycles for Different Cement
Strengths………………………………………………………..………………. 8
1.4 Damage Development Depending on Different Cyclic Load Levels…………... 10
2.1 Stresses That Act on a Pre-Stressed Cement System…………………………… 18
2.2 Three Step Process for Cement Design …………..…………………………… 25
2.3 Tangential Stress for Hard and Soft Cement Systems……… ………………… 31
3.1 Radial and Hoop Stress Profile Due to Loading Conditions…………………... 34
3.2 Stresses Acting on a Cement Sheath ……………………….………………….. 34
3.3 Concrete Failure Criterion under Triaxial Compressive Stresses ……………… 37
3.4 Failure Envelope for Triaxial Compression and Tensile Stress State …………. 38
3.5 Contact Pressure on Casing-Cement Interface…………………………………. 39
3.6 Operative Stresses on Cement-Formation Interface ………………………….... 42
3.7 Contact Stresses on Cement Sheath..................................................................... 44
3.8 Equivalent Stress for Casing –Cement –Formation Model with Meshing ……. 48
3.9 Equivalent Stress Comparison for Analytical and FEA Models ………………. 48
3.10 S-N Curve for Fatigue Analysis in ANSYS…………………….……………… 50
3.11 Cyclic Loading Conditions for the Cement Sheath…….……….……………… 51
3.12 Goodman Diagram for Brittle and Ductile Materials….……….………………. 52

xi
FIGURE Page
3.13 Zero Based Loading, Goodman’s Diagram and Fatigue Options ……………... 53
3.14 Fully Reversed Loading, Goodman’s Diagram and Fatigue Options ………….. 54
6
4.1 Equivalent Stress with Formation E=3 x 10 psi and v=0.42 for Scenario 1………. 57
6
4.2 Equivalent Stress with Formation E=1 x 10 psi and v=0.3 for Scenario 1………... 58
4.3 Equivalent Stress in Three Cement Systems with pi=15000 psi and pf=1000 psi 59
4.4 Tangential Stress in Three Cement Systems with pi=15000 psi and pf=1000 psi 59
4.5 Radial Stress in Three Cement Systems with pi=15000 psi and pf=1000 psi....... 60
6
4.6 Von Mises Stress with Formation E=1 x 10 psi and v=0.3 for Loading case 2…… 61
4.7 Equivalent Stress in Three Cement Systems with pi=15000 psi and pf=0 psi...... 62
4.8 Tangential Stress in Three Cement Systems with pi=15000 psi and pf=0 psi..... 62
4.9 Radial Stress in Three Cement Systems with pi=15000 psi and pf=0 psi………. 63
4.10 Equivalent Stress in Three Cement Systems with pi=4000 psi and pf=10000 psi 64
4.11 Tangential Stress in Three Cement Systems with pi=4000 psi and pf=10000 psi 64
4.12 Radial Stress in Three Cement Systems with pi=4000 psi and pf=10000 psi…... 65
4.13 Equivalent Stress with pi=15000 psi, pf=1000 psi and ΔT =150oF ……………. 66
4.14 Tangential Stress with pi=15000 psi, pf=1000 psi and ΔT =150oF ……………. 66
4.15 Radial Stress with pi=15000 psi, pf=1000 psi and ΔT =150oF ………................ 67
4.16 Equivalent Stress with pi=15000 psi, pf= 0 psi and ΔT =150oF ……………….. 68
4.17 Tangential Stress with pi=15000 psi, pf= 0 psi and ΔT =150oF ………………. 68
4.18 Radial Stress with pi=15000 psi, pf= 0 psi and ΔT =150oF …………………… 69
4.19 Equivalent Stress with pi= 4000 psi, pf=10000 psi and ΔT =150oF.................... 70
xii
FIGURE Page
4.20 Tangential Stress with pi= 4000 psi, pf=10000 psi and ΔT =150oF ……….…. 71
4.21 Radial Stress with pi= 4000 psi, pf=10000 psi and ΔT =150oF ……..………... 71
6
4.22 Equivalent Stress for Static Loading with Formation E=1 x 10 psi and v=0.42..... 72
4.23 Equivalent Alternating Stress for Zero Based Loading………………………….. 73
4.24 Life Cycle for Zero Based Loading ……………………….……………………. 74
4.25 Fatigue Sensitivity to Life Plot for Zero Based Loading …………….................. 75
4.26 Fatigue Sensitivity to Safety Factor Plot for Zero Based Loading …………....... 75
4.27 Fatigue Sensitivity to Damage Plot for Zero Based Loading……………............ 76

6
4.28 Equivalent Stress for Static Loading with Formation Property: E=1 x 10 , v=0.3 … 77
4.29 Alternating Stress for Zero Based Loading with Formation Property:
6
E=1 x 10 , v=0.3………………………………………………………………... 77
4.30 Alternating Stress for Fully Reversed Loading with Formation Property:
6
E=1 x 10 , v=0.3………………………………………………………………… 78
6
4.31 Life Cycle for Fully Reversed Loading with Formation: E=1 x 10 , v=0.3……… 79
4.32 Equivalent Stress for Cement System 2 under Static Loading………………….. 80
4.33 Alternating Stress for Cement System 2 under a Zero Based Cyclic Loading … 80
4.34 Life Cycle for Cement System 2 under a Zero Based Cyclic Loading………… 81
4.35 Fatigue Sensitivity to Life for a Zero Based Cyclic Loading with Cement 2…. 82
4.36 Fatigue Sensitivity to Damage Plot for Zero Based Loading with Cement 2…. 82
4.37 Fatigue Sensitivity to Safety Factor Plot for a Zero Based Cyclic Loading with
Cement 2….......................................................................................................... 83
xiii
FIGURE Page
4.38 Equivalent Stress for Cement System 1 under Static Loading ……………....... 84
4.39 Alternating Stress for Cement System 1 under Zero Based Cyclic Loading....... 84
4.40 Life Cycle for Cement System 1 under Zero Based Cyclic Loading………….. 85
4.41 Fatigue Sensitivity to Life for Zero Based Cyclic Loading with Cement 1……. 86
4.42 Fatigue Sensitivity to Safety Factor for Zero Based Cyclic Loading with
Cement 1………………………………………………………………………… 86
4.43 Fatigue Sensitivity to Damage for Zero Based Cyclic Loading with Cement 1… 87
xiv
LIST OF TABLES
TABLE Page
2.1 ASTM Cement Classifications …………………………………………………. 15
2.2 API Cement Classifications…………………………………………………….. 16
3.1 Stress-Life Data for Cement System 1 (Strength; 26 MPa/ 3771 psi)………… 49
3.2 Stress-Life Data for Cement System 2 (Strength; 84 Mpa /12,183 psi)……….. 49
1
CHAPTER I
INTRODUCTION: RESEARCH MOTIVATION
For a well, whether oil or gas, to maintain its integrity and produce effectively and
economically, it is pertinent that a complete zonal isolation is achieved during well
completion. This complete zonal isolation, however, can be compromised due to factors
that come into play during the operative life of the completed well. Such factors may
come in the form of thermal or pressure loads generally regarded as HPHT (high
temperature-high pressure) loads which can manifest itself as a static/cyclic load or both,
depending on how it is exerted. Depending on the magnitude of loading (stress level),
the number of cycles and even the mechanical properties of the well cement, cyclic
loading could result in failure by extensive breakdown of the microstructure of the
cement.
There have been a lot of experimental investigations on the mechanism of fatigue
failure of structures like buildings and bridges but the fatigue behavior of well cement is
still relatively unknown to engineers. Research has led to improved cement designs and
cementing practices, yet many cement integrity problems persist and this further
strengthens the need to understand the mechanism of cement fatigue. Even though most
structural failures are as a result of fatigue rather than static loading, insights on the role
This thesis follows the style of SPE Journal.

2
of both static and fatigue loading conditions on the failure of cement sheath would
hopefully lead to improvements in well design.
1.1 Background
1.1.1 HPHT Well Integrity
A well can be said to have maintained its integrity if it effectively achieves zonal
isolation over its productive life. However, maintaining integrity is not always the case
in real life oilfield practice as case histories abound where the integrity of the well was
compromised due to failure of cement sheath and leading to loss of money and
production.
In order to keep up with the world’s energy demands, oil and gas producing
companies have taken the initiative to explore offshore reserves or drill deeper into
previously existing wells. The consequence of this, however, is that they have to deal
with the high temperatures and pressures encountered at increasing depths. The industry
acknowledges the threshold for high temperature and high pressure conditions as 3000F
and 10,000 psi respectively. For temperatures and pressures above these values, only a
rigorous design would ensure the integrity of the well. High pressure /high temperature
(HPHT) scenarios can be seen in the case of Tuscaloosa trend in Louisiana drilled to a
depth of 23,000 ft and with bottom hole temperature of 4000F and pressures between
17000 to 20000 psi or even in case of the Shearwater field in the East Central Graben
area of the North sea with a depth of 16000 ft, temperature of 3600F and pressure of
15,200 psi, to mention but a few.

3
These actual HPHT industrial experiences highlighted the inadequacy of
conventional cementing procedures to provide adequate zonal isolation. High
temperatures and pressures or even post-cementing stresses imposed on the cement
sheath as a result of casing pressure testing and formation integrity tests set in motion
events which could compromise the long term integrity of the cement sheath due to
fatigue. Knowledge of the mechanism of fatigue in cement and factors that affect it such
as the magnitude of the load, strength and composition of the cement, mechanical
properties of the cement and pattern of load cycles are important to achieve a realistic
design of a cement system that will be subjected to fatigue loading. Such a design will
go a long way to ensure the long term integrity of a well operating under HPHT
conditions.
1.1.2 Failure Criteria
Well cement is subjected to failure mainly by the process of
• debonding
• radial cracking
• cement plastic deformation
These are not new failure modes but just a petroleum engineer’s term for the usual failure
modes in mechanics of materials. Debonding can also be regarded as shear failure and can
exist in two forms - debonding from casing and debonding from formation. It is however
important to note that debonding can also occur as a result of cement shrinkage which in
this case cannot be regarded as a failure due to shear. Radial cracking is a failure mode by
4
fracture as a result of the gradual growth of cracks when the cement is subjected to fatigue
loading. Usually, the surface exhibits no sign of deformation and would finally fail under a
gradually increasing load perpendicular to the loading axis in tension and inclined to the
loading axis in compression. Plastic deformation is a result of yielding failure. It usually
results in the change of shape of the material involved.
Cook and Young1 discussed different classical theories of failure for brittle
materials including the maximum normal stress and Mohr-Coulomb’s criteria which may
partly define some of the failure modes described above. These failure criteria are used to
predict if a given material, in this case cement, will fail under a given stress condition.
Concrete under triaxial stresses fail in a unique manner and the Mohr-Coulomb’s criterion
can be used to approximately predict failure when concrete is under compressive and
tensile stresses. Neither this criterion nor the maximum normal stress criterion will suffice
in the case of triaxial compressive state. These failure criteria will be discussed in detail in
Chapter III. It is important to note that these failure criteria work by comparing a given test
condition such as a stress state with an already known parameter of the material like
ultimate strength.
1.1.3 Fatigue
Fatigue is a process of progressive and permanent internal damage in a material
subjected to repeated loading. Fatigue failure due to cyclic loading conditions occurs as
a result of frequently repeated stresses applied at stress levels lower than the ultimate
5
stress under static conditions. The magnitude of this stress decreases as the number of
cycles of stress increases.
The fatigue process occurs in three stages:
• Crack initiation which occurs as a result of cumulative damage in a localized
region under successive cycles of loading;
• Crack accumulation resulting in crack growth as a result of continued loading;
• Crack propagation where the specimen fractures and fails.
Since the third stage is very unstable, the fatigue life may be related to the transition
from the second stage to the third stage.
Results of static studies are used as the basis in defining a fatigue study. Fatigue
parameters include
• Fatigue life - the number of cycles to failure at a given stress level
• Fatigue Strength – the greatest amount of stress that can lead to the failure of a
material in a given number of cycles.
• Fatigue/Endurance limit- the maximum value of stress below which a material
can withstand an infinite number of cycles.
• S-N Diagram- a plot of fatigue strength against number of cycles to failure.
• Stress ratio -the ratio of the minimum to maximum stress
Fig. 1.1 shows the tensile and compressive stress fluctuations for a body under cyclic
loading.
6
Stress
Tension
σa
0
Cycles
σa
Compression
Fig 1.1: Fluctuating Tensile and Compressive Cyclic Stress
1.1.4 Fatigue in Cement/Concrete
The nature of fatigue in well cement is generally unknown and only a few studies exist on
the fatigue of construction cement. The differences between oil well and cement used in the
construction industry will be discussed in the next chapter. The fatigue strength of
cement/concrete can be affected by factors such as the composition and mechanical
properties of the cement, environmental and loading conditions and water-cement ratio of
concrete. A number of studies have been done on the fatigue of construction cement and it
was found that due to the heterogeneous nature of cement, experimental results show a
large scattering in the concrete behavior due to the cyclic loading and few data sets may not
be sufficient to give an adequate description of cement behavior under fatigue loading.
Studies were conducted by Kim and Kim2 on the fatigue behavior of high strength
concrete using a type I Portland cement to which Elkem micro silica (powder) was added.
A constant minimum stress level of 25 percent of the static uniaxial compressive strength
7
was maintained while the cyclic tests were conducted at maximum stress levels of 75, 80,
85 and 95% of the static strength. The first cycle of loading was loaded at a standard rate,
and the other cycles were loaded in the frequency of 1 Hz. The test results (Fig.1.3)
indicated that under the same stress levels, fatigue life decreases as the concrete strength
increases, and then the fatigue resistance of high strength concrete seems to be inferior to
that of low strength concrete. Figure 1.2 below shows the relationships between the number
of loading cycles and the fatigue strains of concrete low strength (LS) and high strength
(HS). Although the fatigue strain of HS concrete is smaller than that of LS concrete, the
slope of strain increment curve of HS concrete is steeper than that of LS concrete, i.e., the
rate of strain increment increases with the strength of concrete. Therefore high strength
concrete is more brittle than low strength concrete under fatigue loading.
Fig 1.2: Comparison of Fatigue Strain for High and Low Strength Cements [2]
8
Fig 1.3: Comparison of Max Stress Levels to Number of Cycles for Different Cement
Strengths [2]
Antrim3 conducted fatigue studies on hardened ordinary Portland (type I) cement
paste using 2 specimens; one with a high-water cement ratio of 0.7 and another with a
low water-cement ratio of 0.45. A high stress level of 80% (percentage of the ultimate
static compressive strength of the cement) was used in conducting this investigation. It
was observed that the degree to which shrinkage stresses were present in the specimens
were proportional to the water content. This lead to the 0.7 mixture undergoing more
shrinkage due to a more extensive capillary pore system. It was also observed that at
equivalent percentages of the compressive strength, the 0.7 water-cement mixture was
capable of withstanding more cycles to failure than the 0.4 water-cement mixture. From
these results, the author suggested that shrinkage stresses play a greater role in fatigue
strength because they serve to restrain crack propagation3. Crack propagation was slower
in the open capillary structure cement (0.7 mixture) than in the dense structure cement
9
(0.4 mixture) because the high water cement ratio paste is less brittle and can re-adjust
its structure thus delaying the build up of stress concentrations.
Breitenbucher et al. 4 noticed in their investigations that cyclic loading could lead
to the reduction of stiffness of concrete and that fatigue strain plays a role in the
degradation of the mechanical properties of concrete. The level of damage due to fatigue
loading can be ascertained from the degradation of stiffness at certain number of cycles.
It was also observed that as the longitudinal strain increases at 60% stress level, the
concrete properties (Young’s modulus, fracture energy) decreased faster up to the first
2.0 millions of cycles thereafter, whereas the compressive strength almost remained
constant. There was no observed failure due to fatigue for at least 25.5 million cycles.
Similar observations were made at 70% & 75% stress levels. This shows that the effects
of number of cycles appear to be negligible and therefore the damage is governed only
by the evolution of fatigue strain. These results were also corroborated by the findings of
Breitenbucher and Ibuk5, who in addition noticed that small differences in the upper load
can largely affect the formation of micro-cracks.
Hayeb et al.6 obtained results similar to those of Breitenbucher et al.4 The
inclusion of steel fibers in ordinary cement paste helped to improve its damage
resistance. From Fig. 1.4, it is observed that at stress levels of 80%, failure occurred at
2.7 x 104 cycles and at a reduced stress level of 72%, the specimens did not fail even
after 2 x 106 cycles. A sharp decline in Young’s modulus during the first 104 cycles was
observed with no appreciable decay in the composite strength. These were also in line
with the findings of Breitenbucher and Ibuk5.

10
Fig 1.4: Damage Development Depending on Different Cyclic Load Levels [5]
From the existing literatures reviewed, it can be observed that the fatigue of cement can
be affected by the following;
• Compressive strength of cement
• Water-cement ratio of cement mixture
• Stress levels at which cyclic loading occurs
• Elasticity of the cement
• Use of solid particles may increase fatigue resistance
It should be noted that these studies were conducted using construction cement/concrete
but they should be an indicative of the fatigue of well cement as they both contain the
same base material.

11
1.2 Research Objectives
The objective of this research is to develop a better understanding of the performance of the
casing – cement bond under HPHT well conditions, leading to a model to predict well life.
This would entail two major tasks:
• to identify the factors that affect the casing-cement integrity under HPHT
conditions
• to understand better the fatigue of well cement
Based on the knowledge acquired from completing these tasks, it will be possible to
analyze the mechanics of casing-cement systems under HPHT conditions for the long term
integrity of the system.
1.3 Approach
The approach of this research would involve a combination of analytical and finite element
studies to achieve the above mentioned objective.
The analytical model of well bore stresses will be developed to provide insights into
the casing –cement –formation interactions under static loading conditions and also
providing guidance for the finite element model.
The finite element model will seek to utilize the results from the analytical model to
develop an understanding of fatigue life for cement sheath under HPHT conditions.
12
CHAPTER II
OILFIELD CEMENTING
2.1 Well Cement
Well cements are specially formulated for the exploratory drilling of oil and gas wells.
Well cementing serves various purposes7 which includes
• to provide support and protection to the casing
• to enable zonal isolation by preventing the movement of fluids through the annular
space outside the casing, stopping the movement of fluids into fractured formation
or simply by plugging off an abandoned portion of the well. This is the primary
goal of well cementing.
The choice of a particular cement for zonal isolation centers largely on down hole and
formation conditions. But in general, oil well cements should have some features8 which
are necessary for a successful completion job and which would qualify them to be used
for well cementing purposes. Such features include the following:
• The cement should be able to maintain its integrity in terms of durability and
being free from strength retrogression during the operational life of the well at
the prevailing down hole conditions.
• There should be an optimal setting time for the cement .Too reactive a slurry will
result in a short setting time and an insufficiently reactive slurry may take too
long to set
• The cement slurry should have low viscosity to make it pumpable

13
• The cement should be high sulphate resistant.
• The cement should have a low permeability.
Well cement and construction cement have one thing in common- they are both
Portland cements. Their difference lies in the fact that well cement, in addition to its
Portland cement base, is also mixed up with additives in order to tailor it to a particular
application and is also manufactured to a higher level of consistency. Portland cement is
manufactured as a result of a chemical reaction between limestone and clay at temperatures
of about 2,600°F to 3,000°F. There are four principal compounds in Portland cement,
which are
• Tricalcium silicate C3S,
• Dicalcium silicate C2S,
• Tricalcium aluminate C3A,
• Tetracalcium aluminoferrite C4AF.
Portland cement, when set, develops compressive strength due to hydration as a
result of reaction between water and these constituting components of the cement. The
rate of hydration depends on temperature, size of cement particle and the percentage of
each component present, with C3A hydrating most rapidly followed by C3S and then by
C4AF and finally by C2S. This hydration reaction results in reduction of volume which
makes Portland cement to shrink when set. Expansive cement which is a modified
Portland cement is used to compensate for volume decrease due to shrinkage and to
induce a tensile stress in the reinforcement.

14
Pure Portland cement looses its compressive strength and increases its
permeability at temperatures above 230oF as a result of strength retrogression arising
from the breakdown of its crystalline structure at such temperatures. This would render
the Portland cement unusable for high temperature applications. The strength
retrogression could be explained as follows9:
When Portland cement is mixed with water, tricalcium silicate (C3S) and
dicalcium silicate (C2S) hydrate to form calcium silicate hydrate (C-S-H) gel and
hydrated lime (Ca(OH)2)2. At temperatures higher than 230°F, C-S-H gel converts to α-
dicalcium silicate hydrate (α-C2SH). Conversion to the α-C2SH phase results in the loss
of compressive strength and an increase in permeability. Conversion of C-S-H gel to α-
C2SH at 230°F and higher can be prevented by adding crystalline silica.
The American Society for Testing and Materials (ASTM) Specification C-150
provides for eight types of Portland cement I-VIII with type I cement being the normal,
general-purpose cement used for construction purposes10. More than 92% of Portland
cement produced in the United States are type I and II (or Type I/II). Type III accounts
for about 3.5% of cement production. Type IV cement is only available on special
request, and type V may also be difficult to obtain (less than 0.5% of production).
The American Petroleum Institute ,on the other hand, has defined Specifications
for materials and testing for well cements (API Specification 10A) , which includes
requirements for eight classes of oil-well cements (classes A through H) and three grades
(Grades O - ordinary, MSR - moderate sulphate resistant, and HSR - high sulphate
resistant). Each class is applicable for use at a certain range of well depths, temperatures,
15
pressures, and sulphate environments. Cement classes A, B, C, G, and H are primarily
used in the United States for well cementing. The petroleum industry also uses
conventional types of Portland cement with suitable cement-modifiers. Tables 2.1 and
2.2 shows the ASTM and API cement classifications and their uses.
Table 2.1: ASTM Cement Classifications [10]
ASTM Cement Use

Class
I General purpose cement, when there are no extenuating conditions.
Similar to API class A.
II Aids in providing moderate resistance to sulfate attack. Similar to
API class B.
III When a high-early strength is required. Similar to API class C.
IV When a low heat of hydration is desired (in massive structures).
V When high sulfate resistance is required.
IA4 A type I cement containing an integral air-entraining agent
IIA4 A type II cement containing an integral air-entraining agent
IIIA4 A type III cement containing an integral air-entraining agent

16
Table 2.2: API Cement Classifications [8]
API class Operating Suitability

temperatures(0F)
A 80-170 Good for 0-6000ft depth. Used when special
properties are not required.
B 80-170 Good for 0-6000ft depth. Used for moderate to
high sulphate resistance
C 80-170 Good for 0-6000ft depth. Used for moderate to
high sulphate resistance and when high early
strength is required
D 170-230 Good for 6000-10,000ft depth. Used for
moderate to high sulphate resistance and
moderately high temperatures and pressures
E 170-290 Good for 10,000-14,000ft depth. Used for
moderate to high sulphate resistance and high
temperatures and pressures
F 230-320 Good for 10,000-16,000ft depth. Used for
moderate to high sulphate resistance and
extremely high temperatures and pressures
G 80-200 Good for 0-8,000ft depth. Used for moderate to
high sulphate resistance. Has improved slurry
acceleration and retardation.
H 80-200 Same as class G
17
2.1.1 New Cements vs. Conventional Well Cements
Special situations call for innovative actions and so, the need to drill deeper and produce
oil and gas under HPHT environments has motivated drilling engineers to come up with
what is regarded as “designer” or “supercement” systems which are actually
conventional cements that are modified so as to improve its long term sealing integrity in
HPHT wells. In some cases, these designer cements are non-Portland based. The non-
Portland based cements may include Pozzolanic cements, Epoxy Resin11, geo-polymers,
graphite and fibers (glass, steel) which polymerizes at suitable temperatures and/or
time to produce a flexible and mechanically improved cement system. Pozzolanic
cements are not actually cements but react at ordinary temperature with calcium
hydroxide in the presence of moisture to form compounds with cementitious properties.
The modified conventional cement systems include
• Expansive cements (e.g. super cement bond log ),
• Non-shrinking cement system ,
• Foamed cement,
• Thixotropic cement and
• POZMIX cement (Pozzolan and Portland cement mix).
Each of these “designer” cements is chosen based on the prevailing well
conditions as the cement for the design of one well may not be appropriate for the design
of another. Investigations11 have been conducted on the effectiveness of these new
cement systems in securing a long term integrity for HPHT wells. These have been
conducted using new cement systems which includes a Portland based expansive cement
18
system and a non-Portland based (Epoxy Resin) cement system. According to
experimental results, the expansive cement exhibited good qualities which make it a
good candidate for a HPHT scenario. When set, it generates an internal compressive
strength which enables it to counter tensile stresses as opposed to conventional cement
systems. The tensile stress generated by the pressure within the wellbore annulus serves
first to reduce the compressive pre-stress present in the cement before the material
realizes a net tensile stress. As a result, the effective compressive strength of the cement
is increased by the compressive preload applied. This is shown in Fig. 2.1 below:
Fig 2.1: Stresses That Act on a Pre-Stressed Cement System [11]
Some of the positive results with this kind of cement include;
• Improved Annular seal under HPHT conditions

19
Test results when compared to conventional Portland cement system (Portland
cement, retarded, with 35% silica) showed the expansive cement system providing a
very good seal in a hard formation. Based on the number of cycles at each pressure
applied (up to 10,000 psi with increments of 1000 psi and temperature of 2000F), the
expansive Cement absorbed 5 times the energy before failure than the conventional
cement system.
• Improved Mechanical Shear Bonding
From the mechanical shear bond testing, the conventional cement was about 140 psi
as opposed to 1,840 psi for the expansive cement, an increase of approximately 13
times.
• Improved Hydraulic Bonding
The resistance to various materials to allowing water to flow through or past a plug
was measured at ambient temperatures. The conventional system had a hydraulic
bond of 3800 psi as opposed to 6000 psi for the expansive cement system.
It should, however, be noted that expansive cement functions by expanding
against confinement and the higher the level of confinement, the better its performance
since expansive cements are generally strong in compression and weak in tension. As a
consequence, it would not be so good an idea to use expansive cement in soft/weak
formations as it tends to debond from the casing –cement interface.
The same investigators also conducted investigations on non Portland based
epoxy resin cement systems but with more inconclusive than positive results as
compared to the expansive cement system. They observed that for epoxy resin, the
20
pumping time is relatively insensitive to pressure and weighting materials, which makes
its design simpler than conventional cement systems. Different hardeners and diluents
can be added to push the temperature higher. The HPHT annular seal testing failed
laboratory investigations but was successful in field trials, and this led the investigators
to conclude that more tests and observation are necessary to understand the mechanism
of sealing using epoxy resin. The mechanical shear bond, however, was both high and
constant, and seems to be due to the material folding up and mechanically resisting the
imposed motion of the tubulars.
Foamed cement has also been employed to solve difficult HPHT well integrity
problems. This can be seen in the case of Shearwater field12,13. The presence of higher
than anticipated B annulus pressures in wells drilled at the Shearwater field in the East
central Graben area of the north sea resulted in serious concerns about the long term
integrity of the wells. The field is regarded as a HPHT well with initial reservoir
temperature and pressure of 360oF and 15200 psi respectively at 17900 MD. The
unusually high B annulus pressures were more likely as a result of
• The formation of micro annuli between the cement /formation interface
• Contamination of the cement by flowing hydrocarbons from tight crystalline
limestone within the surrounding formation.
The shearwater field team proposed the use of a foamed cement system citing the
following advantages:
• Improved Mud Displacement, Expansive Properties and Fluid Loss

21
Compared to conventional cement systems, foamed cement possesses superior mud
removal properties, has a lesser overall fluid loss and compensates for shrinkage
common with conventional cement systems.
• Improved Ductility
Compared to conventional systems, foamed cements are more flexible and possess the
ability to withstand both high temperature and high pressure cycling induced stresses.
• High Tensile Strength
Foamed cement high tensile strength would make it more resistant to tensile cracking
• Economy and Safety
Under a HPHT scenario, foamed cement provides a cost effective life cycle design even
though the initial cost may be higher than that of conventional cement, and it also
reduces health safety and environmental risks.
The use of foamed cement system proved very effective in dealing with the
shearwater field problems. However, it was pointed out that a comprehensive analysis is
required to assess the risk of damage to the cement sheath due to downhole well events.
It was also suggested that cement systems should be pre-tested in a laboratory to ensure
that it meets the requirements determined by the analysis.
2.2 Cement Additives
Depending on downhole conditions, certain qualities may be required of the cement used
in completing the well. Additives when added to the Portland cement base could be used
to achieve the desired qualities. They could also be used to extend the properties of the
22
base cement. For instance, with additives, Portland cement may be modified to sustain
very high temperatures up to 700oF and large pressures up to 30,000 psi.
Some of the most commonly used additives in oilfield cementing include
• Accelerators
These are cement additives that generally tend to reduce the thickening time of cement
slurry and increase the rate of development of compressive strength. Since the hydration
process, which results in the setting of cement, occurs at a faster rate at higher
temperatures, the setting of cement might be a problem while cementing wells drilled in
areas of low temperatures and also result in a lot of waiting time .To counter such,
accelerators like CaCl2, NaCl, sodium silicate, sea water etc are used to speed up the
thickening time.
• Retarders
Retarders are the opposite of accelerators as the name suggests. They act to increase the
thickening time of cement slurry. They are mostly lignosulfonate which are polymers
derived from wood pulp. Examples include calcium, sodium and chemically modified
lignosulfonates.
• Weighting Agents
These are added to cement to increase the density of the final cement mix. They are very
important when designing wells with high temperature and high pressure conditions in
order to give the base cement more strength to sustain high pressures and also to prevent
strength retrogression at high temperatures. Examples include ilmenite (FeTiO3),
Hematite (Fe2O3) and Barite (BaSO4).

23
• Fluid Loss Control Agents
These set of additives prevents from phase separation under downhole temperature and
pressure conditions. Such a separation would result in fluid being lost to the formation.
They are usually synthetic polymers.
• Extenders
These set of additives help to lower the density of the cement mix. Examples include
bentonite, pozzolans, microsphere, sodium silicates etc.
Other additives include dispersants and lost circulation control agents.
2.3 Well Cementing Design Process
The drilling and completion of a well is a capital project that runs into millions of dollars
and hence, it is necessary to have a comprehensive design of the cement used for
completion of a particular well and also to avoid remedial cement work which would
add extra cost to the project. Cement design is usually streamlined to a particular well
according to prevailing downhole conditions which is ensued by testing in the lab to
determine if the design would be satisfactory.
Ravi and Xenakis 14 discussed a three step approach to cement design. Step one
would involve a detailed engineering analysis. It would require identifying the nature of
the formation- is it a hard or a loose formation? It would require identifying all forces
that would come into play as the well is being produced- are there high temperatures,
high pressures or both? Is it normally or abnormally pressured? Step one also includes
static and fatigue loading analysis to determine if the cement sheath would sustain the
24
series of cyclic loads it would encounter during its lifetime. The answers to step one
questions would lead to step two which would involve designing the cement slurry based
on factors identified in step one. Here properties of the cement like tensile strength,
Young’s modulus, Poisson’s ratio, plasticity parameters, shrinkage/expansion during
hydration, and post-cement slurry hydration are chosen so as to effectively match the
effects of downhole conditions. Thereafter, laboratory investigations are conducted on
the designed slurry.
The data from the laboratory tests and the analysis of step one are then analyzed
together to evaluate performance. Step three would involve adhering to best drilling and
cementing practices such as centering of casing and effectively cleaning out hole of all
mud so as not to undermine the performance of the designed slurry. It also involves
monitoring during the life of the well. Fig.2.2 below summarizes the design process;
25
Step 1 -Effect of well operations on cement sheath

Engineering Analysis integrity
-Evaluate Properties of cement sheath to reduce
the risk of failure
Step 2 - Lab tests—Thickening time, mechanical

Cement Slurry Design properties
& Testing - Other tests—Wettability, hole cleaning and
slurry placement
Step 3 -People, equipment, quality process, HSE

Deployment and -Cement sheath evaluation, monitoring, learning,
Monitoring improvement
Fig 2.2: Three Step Process for Cement Design [14]
2.4 Cement Integrity Issues Due to Temperature and Pressure
The Knowledge of the stress and deformation fields in the vicinity of a HPHT well is very
critical in evaluating its structural capability. A combination of large temperature ranges
and high pressure variations exerts excessive load between the protection casing strings and
ultimately on the cement sheath. Studies have been conducted on the effects of high
temperature or high pressure or both on well integrity. Some studies describing such efforts
are presented below.
2.4.1 Effect of High Temperature
• Case Study 1
Very high temperatures change the crystalline structure of cement. Stiles15 conducted
studies to investigate the effect of ultrahigh temperatures on the mechanical properties of

26
cement. Five different cement formulations were exposed to a high temperature of 645oF
and the variations of the Young’s modulus, tensile strength and Poisson ratio of these
cement systems at this temperature (and pressure of 2,133 psi) were observed over a 2
year period. The cement systems used are described below:
System 1- Conventional (class G) cement system with 40% BWOC silica flour, 4.3%
BWOC gypsum, fluid loss control additive and CaCl2 accelerator. It was mixed at 15.4
lb/gal.
System 2- A low density thixotropic cement system with 92% class A cement with 40%
BWOC silica flour, 8% gypsum, fluid loss control additive and CaCl2 accelerator and
mixed at 14.5 lb/gal.
System 3 – A foamed cement system of density of 11.65 lb/gal with class G cement (with
40% BWOC silica flour) as base slurry, a surfactant additive with 26% gas production
by volume. It was mixed at 15.8 lb/gal.
System 4 – High strength low density cement of density of 11.65 lb/gal, containing
hollow ceramic microspheres designed with maximized solid volume fraction technique
with silica content in excess of 40% BWOC and with added fluid loss control additive
and CaCl2 accelerator.
System 5- A flexible and expanding low density cement system of density of 12.5 lb/gal
containing solid flexible and an MgO base expanding agent designed with maximized
solid volume fraction technique similar to system 4.The flexible particles were added at
50% BVOB.
27
From the analysis of the experimental data, it was observed that the conventional
and foamed cement systems exhibited brittleness after curing at 6450F while other
systems possessed mechanical parameters (low Young’s modulus and high tensile
strength) that are as good as or better than the flexible blends. From this study it can also
be deduced that the ratio of tensile strength to young modulus gives an indication of
resistance to failure under tension. The thixotropic cement and the flexible cement had
the highest tensile strength to Young’s modulus ratio.
2.4.2 Combined Effect of Temperature and Pressure
The combined presence of high temperatures downhole with high pressure loads lead to
excessive pressure loading in annuli of the casing strings. The pressure load may be as a
result of an increase in pressure around the wellbore region due to pressure integrity
tests, increase of mud weight, casing perforation, stimulation, gas production etc and
these can result in considerable damage to the mechanical properties of cement sheath.
Godwin and Crook16 observed that failure in cement sheath due to excessive pressure
would normally occur in the bottom one half to three-quarters of the casing string and
due to excessive temperature in the upper one fourth to two-thirds.
• Case Study 2
Godwin and Crook16 investigated the effects of high temperatures and excessive annulus
pressure using laboratory experiments and field trials. They circulated hot oil at a
temperature of 350oF through the annulus of the test specimen while gradually
28
increasing the pressure up to 10,000 psi with 2,000 psi increments using the following
cement systems;
System 1- Cement/siliceous material mixture system with 30% BWOW latex, 1.25%
fluid loss control additive and 0.5% gelling agent. It was mixed at 12.1 lbm/gal with
10.81 gal mix water/sack. The yield was 2.49ft3/sack. Compressive strength, Young’s
modulus and Poisson ratio was 1000psi, 0.69 x 106 and 0.42 respectively.
System 2- Cement/Pozzolan mixture system with 30% BWOW latex, 2gal/sack fluid loss
control additive, 10 lbm silica additive and 0.25% gelling agent. It was mixed at 13.1
lbm/gal with 6.48 gal mix water/sack. The yield was 1.76 ft3/sack. Compressive strength,
Young’s modulus and Poisson ratio was 2500psi, 0.8 x 106 and 0.32 respectively.
System 3 is the same as system 2 but without the latex and was mixed at 13.1 lbm/gal.
Compressive strength, Young’s modulus and Poisson ratio was 2000psi, 0.9 x 106 and
0.3 respectively.
System 4- Class H cement system with 35% BWOC silica flour mixed at 18 lbm/gal.
Compressive strength, Young’s modulus and Poisson ratio was 9600psi, 2.4 x 106 and
0.11 respectively.
The results indicated a collapse in crystalline structure of system 1 at pressures of
8000 psi and no appreciable change in the permeability of systems 2 through 4 with
increasing casing pressures. All the cement systems withstood pressures of 2000 psi and
fractures where observed at pressures of 4000 psi with catastrophic effects after 6000 psi
stress cycles. A microanulus occurred in cement 2 with internal pressures up to 6000 psi.
System 1 exhibited elastic properties and provided full casing support until its failure at
29
6000 psi. Field trials were also conducted with systems 2, 3 & 4. After being subjected
to a 12,150 psi internal casing pressure (casing pressure plus casing fluid hydrostatic
pressure) system 4 failed while system 3 remained intact. No evaluation logs were run
for system 2.
2.4.3 Casing-Cement –Formation Interactions
As the awareness of the need to analyze the structural behavior of the cement is
currently increasing, so does the need for guidelines and quantitative results. This has led
to the use of simple models or the more complex finite element models to analyze the
casing-cement-formation system and to base the cement selection on the results of this
analysis. Fleckenstein et al.17 proposes finite element analysis as the best way to analyze
the casing-cement-formation properties during the design phase. This proposal has been
corroborated by others18,19,20,21.
Finite element methods offer a means of effectively modeling pressure and
temperature effects. Utilizing computer programs, analysis of the stress situation
downhole could be achieved in multiple dimensions by partially discretizing the system
and solving the problem using FEA. With FEA, stress variations at different points along
the cement sheath radius can be effectively modeled. Several studies 18,19,20,21 show how
finite element analysis was effectively used to model HPHT wells in order to understand
and overcome cement sheath failure problem. Ravi et al22 sounds a note that cement
which may be suitable under one set of conditions may not be suitable under a different
30
set of conditions. Thus, a rigorous design procedure using finite element analysis should
be used to select the cement system that will satisfy each specification.
• Case Study 3 - Casing-Cement Interactions
Fleckenstein et al.17 used finite element analysis to investigate the role which mechanical
properties of the cement plays in withstanding the internal casing pressure by modeling a
cemented wellbore with two cement systems- a soft and hard cement, the soft (ductile)
cement being one with a high Poisson’s ratio and low modulus of elasticity and the hard
(brittle) cement being one with a low Poisson’s ratio and relatively high modulus of
elasticity.
The hard cement was Class H cement mixed with 35% silica flour at 18 ppg,
with the following mechanical properties: compressive strength: 9,500 psi, Young’s
Modulus: 2,400,000 Psi, Poisson’s Ratio: 0.11 while the soft cement system was a
cement/siliceous material mixed with 30% latex at 12.1 ppg with Compressive strength:
1,000 Psi, Young’s Modulus: 690,000 psi, Poisson’s Ratio: 0.42. These cement systems
were also studied by Godwin and Crook16 .The results indicated that there is little
difference in the constraining effect of the different cement slurries. However, a
confining stress outside the cement sheath would increase the burst resistance of the
casing.
31
Fig 2.3: Tangential Stress for Hard and Soft Cement Systems [17]
• Case Study 4- Casing-Cement-Formation Interactions
The confining stress on the cement sheath may come from the formation and its effect on
the cement sheath was also modeled. It was run with a 10,000 psi internal burst pressure
acting upon the inner surface of the 5-½” casing and a 3000 psi confining stress from the
formation.
The results show reductions in von Mises stress from 6,099 psi to 4,292 psi and
in tangential stress from 894 Psi (tension) at the casing outer diameter to -283 Psi
(compression) at the borehole wall. The investigators noticed a difference in the von
Mises stress that is generated by the hard and soft cement systems. At 1000 psi, hard
cement generate twice the von Mises stress when compared with the soft cement but this
reduces as the confining stress increases with a reduction of less than 20% at 5000psi
confining stress.
32
From these investigations they deduced that radial cracking is less likely to occur
with soft cement systems because hard cement systems are likely to generate significant
tangential stresses (Fig. 2.3), which increases the likelihood of forming radial cracks in
the presence of high internal burst pressure pressures.
It should also be noted23 that with the use of flexible and expanding cement
systems, the stiffness of the formation plays an important role. If the stiffness of the
formation is low (low Young’s modulus) compared to the cement system, debonding and
formation of microannulus at the cement casing interface may occur.

33
CHAPTER III
CEMENT FAILURE DUE TO STATIC LOADING
3.1 Analytical Model of Wellbore Stresses
3.1.1 Background
As a result of production operations, a cemented casing is usually subjected to a variety
of stresses in the form of cyclic pressure and temperature variations. Considering an
infinitesimal element within the cement sheath, these stresses act in a three dimensional
fashion and the cement sheath can therefore be regarded as being under a triaxial stress
state, as shown in Figure 3.2. The third component (not shown in the diagram) is the
axial stress component which is perpendicular to the two stresses shown in Figure 3.2.
The radial stress is always compressive in nature while the tangential stress could be
tensile or compressive depending on the loading conditions. Figure 3.1 shows the nature
and profile of tangential and radial stresses under different loading conditions.
The casing-cement-formation set up can be analyzed as a pressurized composite
cylinder with three concentric cylinders and a perfect bonding is assumed to exist
between the cement and casing and the cement and formation. The pressure and/or
temperature changes induce stress concentrations near the casing – cement and the
cement-formation boundaries.
34
+
σr σr σθ
- σθ - -
(a)Internal pressure
(b)External pressure
Fig 3.1: Radial and Hoop Stress Profile Due to Loading Conditions [1]
Formation
Cement
Casing
Fig 3.2: Stresses Acting on a Cement Sheath [11]

35
3.1.2 Assumptions
The following assumptions were made to facilitate the analytical modeling of the
wellbore:
• Axisymmetric deformations exist for the composite cylinder
• The composite cylinder undergoes plane strain deformation. This implies that the
composite cylinder is under a triaxial stress state.
• Casing-cement and cement-formation interfaces are perfectly bonded with no
discontinuities. This implies that the radial displacements and radial stresses are
continuous across the boundary
• No initial stress exists in cement.
• The casing is regarded as a thin walled pressure vessel
• The cement sheath and formation is treated as a thick walled pressure vessel
3.1.3 Failure Criteria
The failure criteria employed in predicting the failure of cement sheath in this model
includes
• Maximum normal stress criterion
• Mohr-Coulomb’s criterion
• Experimental Investigation
The maximum normal stress criterion predicts that an isotropic material will fail when the
largest principal stress reaches a limiting value. This implies that failure would occur when
36
σ1
≥1 (3.1)
σf
where σ 1 is the maximum principal stress and σ f is the limiting stress. If σ 1 is tensile,
then σ f is the limiting tensile stress and the other two smaller principal stresses σ 2 & σ 3
play no role (with σ 1 〉σ 2 〉σ 3 ). If applied to a compressive stress state, this criterion
becomes
σ3
≥1 (3.2)
σf
where σ 3 is the magnitude of the minimum principal stress. It is should be noted that
this criterion would be inaccurate if all three principal stresses are compressive.
In the Mohr-Coulombs’ criterion, σ 2 does not to play any role and failure is predicted to
occur when
σ1 σ3
− ≥1 (3.3)
σ tensile σ compressive
where σ tensile and σ compressive are the tensile and compressive strengths respectively.
The classical theories of failure discussed above become insufficient even as an
approximation when the tangential, radial and axial stresses are all compressive. In such
cases, different failure theories based on experiments would be required to determine the
onset of failure. Avram et al.24 discussed concrete fracture under triaxial stresses and
proposed a new failure criterion given in Eq. (3.4) below which complies with Mohr-
Coulombs’ criterion.
37
0.86
σ1 ⎛σ ⎞
= 1 + 3.7⎜⎜ 3 ⎟⎟
fc ⎝ fc ⎠ (3.4)
where f c is the compressive strength of the cement, σ 1 is the major principal compressive
stress at failure and σ 3 is the minor principal compressive stress with σ 1 〉 σ 2 〉 σ 3 . Fig 3.3
and Fig 3.4 shows the experimental results and failure envelope respectively for concrete
under a triaxial compressive state.
Fig 3.3: Concrete Failure Criterion under Triaxial Compressive Stresses [24]
38
Fig 3.4: Failure Envelope for Triaxial Compression and Tensile Stress State [24]
It should be noted that this experimental failure criterion was proposed for plain concrete
and may only give an approximation of failure for well cement.
3.1.4 Analytical Model
In a composite cylinder model under consideration, the internal pressure pi acting on the
inner surface of the casing in conjunction with temperature increase will expand the
casing radially , while the cement sheath will resist the expansion. As a result, a contact
pressure (pc1) will develop at the interface between the casing and the cement.
39
Considering the casing-cement interface as shown in Fig. 3.5, pc1 is the contact
pressure formed at the cement-casing interface and pi is the internal pressure.
pc2
b pc1
a pi
pc1
(b) Casing (c) Cement

(a)Casing–Cement Sheath
multi- cylinder
setup
Fig 3.5: Contact Pressure on Casing-Cement Interface
The hoop strain is given by
1
εθ = [σ θ − ν (σ z + σ r )] + αΔT (3.5)
E
and the axial strain reads
1
εz = [σ z −ν (σ θ + σ r )] + αΔT (3.6)
E
But since the axial strain is negligible considering the large depth, then εz≈0 (i.e. plane
strain assumption).It follows from Eq. (3.6) that
σ z = ν [σ r +σ θ ]− α E Δ T (3.7)
40
Substituting Eq. (3.7) into Eq. (3.5) yields
εθ =
1
E
[ ( ) ( )
σ θ 1 − ν 2 − ν + ν 2 σ r + (1 + ν )αEΔT ] (3.8)
The radial expansion is thus obtained as
δr =
r
E
[ ( ) ( )
σ θ 1 − ν 2 − ν + ν 2 σ r + (1 + ν )αEΔT ] (3.9)
Let the radii a, b and c be represented by ra , rb & rc respectively. Considering the casing
as a thin walled vessel we have at r = b,
prm
σ r = − p and σ θ = (3.10)
ts
where p = pi − pc1 , r m is the mean radius of the casing and ts is the thickness of the
casing. Substituting Eq. (3.10) into Eq. (3.9) gives
a (Pi − Pc1 ) ⎡ rm ⎤
δ r − ca sin g = { ⎢ ( 1 −ν s
2
) + (ν s + ν s2 )⎥ } + [ (1 + ν s )aα s ΔT ] (3.11)
Es ⎣ ts ⎦
∂ (ΔT )
Considering the cement sheath as a thick walled cylinder and assuming = 0 , the
∂r
tangential and radial stress is given by
pc1b 2 ⎛ c 2 ⎞ pc 2c 2 ⎛ b 2 ⎞
σr = ⎜1 − ⎟ − ⎜1 − ⎟ (3.12)
c 2 − b 2 ⎜⎝ r 2 ⎟⎠ c 2 − b 2 ⎜⎝ r 2 ⎟⎠
pc1b 2 ⎛ c 2 ⎞ pc 2c 2 ⎛ b 2 ⎞
σθ = ⎜1 + ⎟ − ⎜1 + ⎟ (3.13)
c 2 − b 2 ⎜⎝ r 2 ⎟⎠ c 2 − b 2 ⎜⎝ r 2 ⎟⎠
At r = b, Eq. (3.12) and Eq. (3.13) reduces to
σ r = − pc1 (3.14)
41
⎛ c2 + b2 ⎞ ⎛ 2c 2 ⎞
= ⎜ ⎟
σ θ pc1 ⎜ 2 2 ⎟ pc 2 ⎜⎜ 2 2 ⎟⎟
− (3.15)
⎝c −b ⎠ ⎝c −b ⎠
This gives the radial expansion in the cement sheath at r = b when Eq. (3.14) and Eq.
(3.15) are substituted into Eq. (3.9) as
b ⎡ ⎡ ⎛ b2 + c2 ⎞ ⎛ 2c 2 ⎞⎤ ⎤
δ r − cement = ⎢( 1 − ν c )⎢ pc1 ⎜ 2
2
⎜ ⎟
2 ⎟
− p ⎜
2⎜ 2
⎟
2 ⎟ ⎥ + p (
ν + ν 2
)⎥ + [ (1 + ν c )bα c ΔT ]
⎣ ⎝c −b ⎠ ⎝ c − b ⎠⎦
c c 1 c c
Ec ⎣⎢ ⎦⎥
(3.16)
Since both radial expansions are equal, it follows from Eq. (3.11) & Eq. (3.16) that
⎧⎪ b
( )
2 ⎡b + c ⎤
( a ⎡ rm
) ( 2 ⎤⎫
) (⎡b
) ⎛ 2c 2 ⎞
( ⎤
)
2 2
⎪
pc1 ⎨ [ 1 − vc ⎢ 2 ]
2
⎥ + v + v c + ⎢ 1 − v
2
+ v + v s ⎥ ⎬ − pc 2 ⎢ ⎜⎜ 2 ⎟ 1 − vc 2 ⎥ =
2 ⎟
⎣c − b ⎦ ⎝c −b ⎠
2
⎪⎩ Ec ⎦ ⎪⎭
c s s
Es ⎣ t s ⎣ Ec ⎦
pi a ⎡ rm
Es ⎣ t s
(
2
) (
2 ⎤
)
⎢ 1 − vs + vs + v s ⎥ + [ (1 + ν s )aα s ΔT ] − [ (1 + ν c )bα c ΔT ]
⎦
(3.17)
Eq. (3.17) can be put in the form
Apc1 + Bpc 2 = C (3.18)
where
⎧⎪ b
A = ⎨ [ 1 − vc (
2
)
⎡ b2 + c2 ⎤
⎢ 2 2⎥
+ vc + v ( 2
c ) ] + Ea ⎡⎢ rt (1 − v ) + (v + v )⎤⎥⎫⎪⎬⎪ (3.19)
m 2 2
s
⎪⎩ Ec ⎣c − b ⎦
s s
s ⎣ s ⎦⎭
⎡ b ⎛ 2c 2 ⎞ ⎤
B = - ⎢ ⎜⎜ 2 ⎟ 1 − vc 2 ⎥
2 ⎟
( ) (3.20)
⎣ Ec ⎝ c − b ⎠ ⎦
C= (
pi a ⎡ rm 2
) (
⎢ 1 − v s + vs + v
Es ⎣ t s
2
s )⎤⎥ + [ (1 +ν )aα ΔT
s s ] − [ (1 + ν c )bα c ΔT ] (3.21)
⎦
42
Similarly, considering the cement-formation interface as shown in Fig.3.6, pc2 is the
contact pressure formed at the cement-formation boundary as a result of the confining
pressure from the formation pressure, pf.
Pf
pc2
c
Pc2
b
pc1
(a) Cement-Formation (b) Cement (c) Formation

multi- cylinder setup
Fig 3.6: Operative Stresses on Cement-Formation Interface
∂ (ΔT )
Considering the cement sheath and assuming = 0 , at r = c,
∂r
σ r = − pc 2 (3.22)
⎛ 2b 2 ⎞ ⎛ c2 + b2 ⎞
σ θ = pc1 ⎜⎜ 2 2 ⎟⎟ − pc 2 ⎜⎜ 2 2 ⎟⎟ (3.23)
⎝c −b ⎠ ⎝c −b ⎠
43
This gives the radial expansion in the cement sheath at r = c when Eq. (3.22) and Eq.
(3.23) are substituted into Eq. (3.9) as
c ⎡ ⎡ ⎛ 2b 2 ⎞ ⎛ c 2 + b 2 ⎞⎤ ⎤
δ r − cement = ⎢( 1 − ν c )⎢ pc1 ⎜⎜ 2
2
⎟
2 ⎟
− p ⎜
2⎜ 2
⎟
2 ⎟ ⎥ + p 2
ν + ν 2
(
⎥ + [ (1 + ν c )cα c ΔT ) ]
⎣ ⎝c −b ⎠ ⎝ c − b ⎠⎦
c c c c
Ec ⎣⎢ ⎥⎦
(3.24)
Considering the formation as a thick walled pressure vessel with a finite radius d into the
∂ (ΔT )
formation and also assuming = 0 , at r = c
∂r
σ r = − pc 2 (3.25)
⎛ c2 + d 2 ⎞ ⎛ 2d 2 ⎞
σ θ = pc 2 ⎜⎜ 2 2 ⎟⎟ − p f ⎜⎜ 2 2 ⎟⎟ (3.26)
⎝d −c ⎠ ⎝d −c ⎠
when Eq. 3.24 and Eq. 3.25 are substituted into Eq. 3.9 , it follows that
c ⎡ ⎡ ⎛ c2 + d 2 ⎞ ⎛ 2d 2 ⎞⎤ ⎤
δ r − formation = ⎢( 1 − ν f )⎢ pc 2 ⎜⎜ 2
2
⎟
2 ⎟
− p ⎜
f⎜ 2 2 ⎟⎥
(
⎟ + pc 2 ν f + ν 2f ⎥ + [ (1 + ν f )cα f ΔT ) ]
E f ⎣⎢ ⎣ ⎝d −c ⎠ ⎝ d − c ⎠⎦ ⎥⎦
(3.27)
Since both radial expansions are equal, it follows from Eq. 3.24 & Eq. 3.27 that
[ (1 − v f 2 )⎢ d 2 + c 2 ⎥ + (v f ) ] + Ec [1 − v ]⎛⎜⎜ bc ( )⎫⎪⎬⎪ − p
⎧⎪ c ⎡c ⎤
⎡ ⎤ + c2 ⎞ ⎛ 2b 2 ⎞
( )
2 2 2
2 2
pc 2 ⎨ +v
2
⎟ − vc + v c1 ⎢
⎜⎜ 2 ⎟ 1 − vc 2 ⎥ =
− b 2 ⎟⎠ 2 ⎟
f c
⎣d − c ⎦ ⎝c −b ⎠
c 2
⎪⎩ E f c ⎝ ⎭ ⎣ Ec ⎦
⎡ c
pf ⎢
⎛ 2d 2 ⎞
⎜⎜ 2 (
⎟ 1 − vf 2
2 ⎟
)⎤⎥ − [ (1 + ν )cα ΔT ] + [ (1 + ν c )cα c ΔT ]
⎝d −c ⎠
f f
⎢⎣ E f ⎥⎦
(3.28)
Eq. (3.28) can be put in the form
Dpc1 + Kpc 2 = F (3.29)

44
⎡ c ⎛ 2b 2 ⎞ 2 ⎤
where D = - ⎢ ⎜⎜ 2 ⎟
2 ⎟
1 − v (⎥ ) (3.30)
⎣ Ec ⎝ c − b ⎠
c
⎦
[ (1 − v f 2 )⎢ d 2 + c2 ⎥ + (v f ) ] + Ec [1 − v ]⎛⎜⎜ bc −+ bc ⎞⎟⎟ − (v + v )⎫⎪⎬⎪ (3.31)

⎧⎪ c ⎡ 2 2
⎤ 2 2
2 2
2
K= ⎨ +v f c
⎣d − c ⎦
c 2 2 c
⎪⎩ E f c ⎝ ⎠ ⎭
⎡ p c ⎛ 2d 2 ⎞
F = ⎢ f ⎜⎜ 2 ⎟ 1− vf 2
2 ⎟
( )⎤⎥ − [ (1 + ν )cα ΔT ] + [ (1 + ν c )cα c ΔT ] (3.32)
⎢⎣ E f ⎝ d − c ⎠
f f
⎥⎦
From a simultaneous solution of Eq.(3.18) and Eq. (3.29), the contact pressures pc1 and
pc2 are given as
FB − KC
pc1 = (3.33)
DB − AK
⎡ FB − KC ⎤
C−⎢ A
⎣ DB − AK ⎥⎦
pc 2 = (3.34)
B
From the analysis presented above, the circumferential, radial and axial stresses present
in the cement sheath as shown in Fig. 3.7 can then be determined.
pc2
pc1
Fig 3.7: Contact Stresses on Cement Sheath

45
The radial, tangential and axial stresses in the cement sheath are then calculated using
the following formulas
b2 ⎡ c2 ⎤ c2 ⎡ b2 ⎤
σ r −cement = Pc1 ⎢1 − 2 ⎥
− Pc2 2 ⎢1 − 2 ⎥ (3.35)
c2 − b2 ⎣ r ⎦ c − b2 ⎣ r ⎦
b2 ⎡ c2 ⎤ c2 ⎡ b2 ⎤
σ θ −cement = Pc1 ⎢1 + 2 ⎥
− Pc2 ⎢1 + 2 ⎥ (3.36)
c2 − b2 ⎣ r ⎦ c2 − b2 ⎣ r ⎦
σ z − cement = ν [σ r + σ θ ] − αEΔT (3.37)
τ max , the maximum shear stress given by the expression;
(Pc1 − Pc 2 )b 2 c 2
τ max = (3.38)
(c 2
− b 2 )r 2
At r = b,
(Pc1 − Pc 2 ) c2
τ max = (3.39)
(c 2
− b2 )
At r = c
τ max =
(Pc1 − Pc 2 ) b 2 (3.40)
(c 2
− b2 )
3.2 Analytical and Finite Element Studies
The response of the cement sheath to different static and fatigue loading conditions will be
studied for three cement systems;
• Cement system 1 - Ductile cement system with compressive strength of 3000 psi,
tensile strength of 1000 psi, a young’s modulus of 0.69 x 106 psi and a Poisson ratio
of 0.4
46
• Cement system 2 - Brittle cement system with compressive strength of 9500 psi,
tensile strength of 3000 psi, a young’s modulus of 2.4 x 106 psi and a Poisson ratio
of 0.1.
• Cement system 3- A low young’s modulus and a low Poisson ratio cement system
with compressive strength of 2500 psi, tensile strength of 1000 psi a young’s
modulus of 1 x 106 psi and a Poisson ratio of 0.25
These cement systems will be studied for the following cases;
• Well pressure: 15,000 psi ; formation pressure; 1000 psi

• Well pressure: 15,000 psi ; formation pressure; 0 psi
• Well pressure: 4000 psi ; formation pressure; 10,000 psi
Other parameters used as inputs to the model include;

Temperature change ΔT: 150oF
Casing’s Young modulus and Poisson ratio: 2.9 x 107 psi and 0.3
Shear bond strength of cement: 1000 psi
Casing outer diameter, b: 9.625 in
Casing wall thickness, ts: 0.545 in
Casing inner diameter, a: 8.535 in
Cement wall thickness: 2.125 in
Formation outer diameter, d: 20 in
Formation Young’s modulus and Poisson ratio: 3 x 106 psi and 0.42
Density of cement mix: 14 lb/gal
Cement expansion coefficient of 0.000006 in/ oF
The finite element analysis was done with ANSYS workbench 11.0 and since the
casing –cement –formation is axisymetric, a quadrant was used for the 2D modeling.
47
To validate the casing-cement-formation model, the analytical result and finite
element simulations were compared. An internal pressure of 15,000 psi was applied inside
the casing with no formation pressure. The meshing was done with 6648 elements and
20,607 node density. Plane strain state was assumed. The boundary conditions applied
include
• U=0, τxy = 0 on X = 0
• V=0 , τyx= 0 on Y= 0
• σ rr = pi on r = a
• σ rr = p f on r = d
The casing –cement –formation model was also assumed fully bonded with no separation at
the interfaces.
Figure 3.8 shows the distribution of equivalent von Mises stress under the loading
condition described above and Figure 3.9 below compares the analytical values for von
Mises equivalent stress with those from the finite element analysis. The analytical and
finite element values were close with an error of about 1 %.

48
Fig 3.8: Equivalent Stress for Casing –Cement –Formation Model with Meshing
9000
8500
8000
Equivalent stress
7500
FEA
7000
Analytical
6500
6000
5500
5000
4.5 4.7 4.9 5.1 5.3 5.5 5.7 5.9
Radius(inches)
Fig 3.9: Equivalent Stress Comparison For Analytical and FEA Models
49
3.3 Fatigue Studies
To study the fatigue of the cement sheath, finite element analysis was utilized. The fatigue
properties of concrete were used for this analysis. Such properties include S-N curves,
strain-life curves and stress-strain curve. Tables 3.1 and 3.2 below gives the fatigue
material data used in the finite element modeling for cement systems 1 and 2 based on data
extracted from Fig.1.3 using a special software. The fatigue life for the cement systems will
be predicted from these curves using ANSYS. The S-N curve is shown in Fig.3.10.
Table 3.1: Stress-Life Data for Cement System 1 (Strength; 26 MPa/ 3771 psi)
No of Cycles Stress(psi)
0.99692 4332.322
4.65426 4031.543
57.8064 3590.166
683.674 3203.529
8084.64 2852.871
100398 2535.519
1.00E+06 2276.105
Table 3.2: Stress-Life Data for Cement System 2 (Strength; 84 Mpa /12,183 psi)
No of Cycles Stress(psi)
1.01527 13376.842
1.87358 13009.757
19.6824 11594.051
165.699 10353.456
1740.34 9208.586
16164.6 8206.784
1.46E+05 7314.037
50
Newman and Choo25 gave a relationship between the number of cycles to failure
and load ratio derived through experimental studies for concrete with densities greater
than 12.5 lb/gal as
f c. max
= 1 − 0.0685(1 − R) log10 N 3.26 (3.41)
fc
Where f c. max is the maximum compressive stress of the cyclic loading,
f c is the compressive strength of the concrete,
f c. max
R is the stress ratio and
f c. min
N is the number of cycles to failure.
Fig 3.10: S-N Curve for Fatigue Analysis in ANSYS

51
The fatigue properties of the cement sheath were also studied under two loading conditions;
• Cyclic loading at the casing coupling and
• Cyclic loading within the casing annulus.
These loading conditions shown in Fig. 3.11 below, presents a cyclic loading condition due
to shear at the coupling and that due to the internal and formation pressure on areas where
there are no couplings.
Casing
τ
Pi
Casing Po
Coupling
Cement
a) Cement sheath under b) Cement sheath under Pi

shear at coupling and Po
Fig 3.11: Cyclic Loading Conditions for the Cement Sheath

52
For both cases, a constant amplitude loading condition is assumed .A fully reversed
loading is also assumed while modeling the fatigue behavior of the cement under shear
cyclic loading while a zero based (compression) loading is assumed for the internal and
external pressures. The Goodman diagram for fatigue analysis was chosen as it gives a
good description of the fatigue behavior of brittle materials. Figure 3.12 shows the
Goodman diagram. Figs. 3.13 and 3.14 present the different loading conditions and options
available with ANSYS.
Fig 3.12: Goodman Diagram for Brittle and Ductile Materials [1]
According to the Goodman diagram, for brittle materials;
σ σm 1
a
+ = (3.42)
σ fs σu SF
53
where σm =
(σ max + σ min ) (3.43)
2
σa =
(σ max − σ min ) (3.44)
2
Fig 3.13: Zero Based Loading, Goodman’s Diagram and Fatigue Options
54
Fig 3.14: Fully Reversed Loading, Goodman’s Diagram and Fatigue Options
Some of the result outputs for fatigue evaluation include:
• Fatigue life - The fatigue life plot gives an indication of number of cycles to failure
of a particular material;
• Damage - The fatigue damage plot gives the fatigue damage at a given design life;
55
• Factor of safety – This gives the factor of safety with respect to fatigue failure at a
given design life;
• Fatigue sensitivity – This plot gives an indication of how loading conditions affect
the fatigue performance of the material;

56
CHAPTER IV
RESULTS AND DISCUSSION
4.1 Static Studies
• Case Scenarios without Temperature
To understand the effect of static loading on the integrity of cement sheath, the analytical
model and finite element analysis were used to examine the responses of different cement
systems subjected to different magnitudes of internal pressure and formation pressure as
described in the previous chapter. The temperature change was neglected and the analytical
model was used to show the trends in the von Mises equivalent, tangential and axial
stresses in the cement sheath for the three cement systems.
1. High Inner Pressure and Low Formation Pressure
With an inner pressure of 15,000 psi and a formation pressure of 1000 psi, the finite
element model gives us an idea of the amount of stress generated within the steel casing
and the actual amount transferred to the cement sheath through the casing-cement interface.
The same can also be said of the transfer of stresses from the formation to the cement
sheath through the cement-formation interface. Figures 4.1 and 4.2 shows the trend for von
Mises stress using cement system 2, which has the same mechanical properties for steel
casing but with different formation properties. From the diagrams, it is evident the amount
of support the steel casing provides the cement sheath. In Fig. 4.1, the equivalent stress
within the casing decreases non-linearly from 84,178 psi to 66,784 psi. At the casing-
57
cement interface, there is a large decrease from 66,784 to 8,384 psi. Due to the fact that the
properties of the cement (E=2.4 x 106 psi, v=0.4) and the formation (E=3 x 106 psi, v=0.42)
are similar, there is a little change in the von Mises equivalent stress across the cement-
formation interface (5,806 to 5736 psi). Fig. 4.2 shows somewhat a similar trend but with
higher stresses imposed on the steel casing (93,575 to 74,325 psi) and a lower stress on the
cement sheath (8,160 to 5,560 psi).As opposed to Fig. 4.1, there a significant difference in
the equivalent stress at the cement-formation interface (5,560 to 2,776 psi).
Fig 4.1: Equivalent Stress with Formation E=3 x 106 psi and v=0.42 for Scenario 1
58
Fig 4.2: Equivalent Stress with Formation E=1 x 106 psi and v=0.3 for Scenario 1
Keeping the formation and casing properties unchanged, the analytical model was
used to study the response to a high internal pressure and low formation pressures on the
casing-cement-formation system as shown in Figs. 4.3 to 4.5. The results show that cement
system 1 generates compressive (negative) radial and tangential stresses. The tangential
stress fluctuates between a maximum value of -1,161 psi to a minimum of -590 psi. In
contrast, cement systems 2 and 3 generate tensile tangential stresses. Cement system 3
posses more of a ductile quality and generates less von Mises stress than system 2 but a
little more than system 1.

59
10000
9000
8000
Equivalent Stress(psi)
7000
6000
cement system 1
5000 cement system 2
3000
2000
1000
0
4.5000 5.0000 5.5000 6.0000
Radius(inches)
Fig 4.3: Equivalent Stress in Three Cement Systems with pi=15000 psi and pf=1000 psi
6000
5000
4000
Tangential stress(psi)
3000
cement system 1
cement system 3
1000
0
4.5000 5.0000 5.5000 6.0000
-1000
-2000
Radius(inches)
Fig 4.4: Tangential Stress in Three Cement Systems with pi=15000 psi and pf=1000 psi
60
600
-400
4.5000 5.0000 5.5000 6.0000
-1400
Radial stress(psi)
-2400 cement system 1

cement system 2
-4400
-5400
-6400
Radius(inches)
Fig 4.5: Radial Stress in Three Cement Systems with pi =15000 psi and pf =1000 psi
2. High Inner Pressure and Zero Formation Pressure
The trend in the radial and tangential stresses in the three cement systems subjected to a
high inner pressure of 15,000 psi and a zero formation pressure is strikingly similar to the
trend in scenario 1 with values for radial and tangential stresses being a little higher as
shown in Figs. 4.7 to 4.9.The maximum equivalent stress in this case is 8,692 psi for
cement system 2 as opposed to 9,000 psi for the same cement system under scenario 1.The
same can be said about the values for the radial and tangential stresses for these two
loading cases.
61
Fig 4.6: Von Mises Stress with Formation E=1 x 106 psi and v=0.3 for Loading Case 2
The finite element modeling of case 2 with cement system 2 (Fig. 4.6) shows a
higher amount of equivalent stress on the casing (91,151 to 73,398 psi) in comparison with
case 1 with the same formation properties (84,178 to 66,784 psi).The transitions at the
boundary show a different trend from that of scenario 1.At the casing-cement interface, the
pressure reduces from 73,398 psi in the casing to 8,210 psi in the cement .The trend is the
opposite at the cement-formation boundary with 5,586 psi at the cement side of the
boundary and 6,150 psi on the formation side.

62
10000
9000
Equivalent Stress(psi) 8000
7000
6000
cement system 1
3000
2000
1000
0
4.5000 5.0000 5.5000 6.0000
Radius(inches)
6000
5000
4000

cement system 2
1000
0
4.5000 5.0000 5.5000 6.0000
-1000
Radius(inches)
63
600
4.5000
-400 5.0000 5.5000 6.0000
-1400
Radial stress(psi)
cement system 1
cement system 3
-3400
-4400
-5400
Radius(inches)
Fig 4.9: Radial Stress in Three Cement Systems with pi=15000 psi and pf=0 psi
3. Low Inner Pressure and High Formation Pressure
Fig 4.10-4.12 below presents the case when a casing pressure of 4000 psi and a high
formation pressure of 10,000 psi is applied to the casing –cement-formation model. The
trend here is significantly different t from that seen in the first two cases. Here, both the
tangential and radial stresses are compressive in nature. Cement systems 2 and 3 generates
almost equal amounts of tangential stresses and are much lower than those of cement
system 1.However, cement system 1 has the lowest value of radial stress (-5297 psi) of the
three cement systems.

64
7000
6000
5000
4000
cement system 1
cement system 2
2000
1000
0
4.5000 5.0000 5.5000 6.0000
Radius(inches)
0
4.5000 5.0000 5.5000 6.0000
-1000
-2000

cement system 2
-5000
-6000
-7000
Radius(inches)
65
500
300
100
-100
4.5000 5.0000 5.5000 6.0000
Radial stress(psi)

cement system 2
-700
-900
-1100
-1300
Radius(inches)
Fig 4.12: Radial Stress in Three Cement Systems with Pi=4000 psi and Pf=10000 psi
• Case Scenarios Considering the Effect of Temperature
1. High Inner Pressure, Low Formation Pressure with Temperature Change
It is no doubt that temperature change plays an important role in adding to the stresses that
would ultimately lead to the failure of the cement sheath. As expected, the combined effect
of temperature and pressure shows a trend that is a bit different, in terms of the magnitude
of the stress, from that when the effects of pressure are considered alone. Figs. 4.13 to 4.15
give us a view of that trend when an inner pressure of 15,000 psi, formation pressure of
1000 psi and temperature change (ΔT) of 150oF respectively are applied to the casing-
cement-formation model.
66
12000
Equivalent Stress(psi) 10000
8000
cement system 1
cement system 3
4000
2000
0
4.5000 5.0000 5.5000 6.0000
Radius(inches)
Fig 4.13: Equivalent Stress with pi=15000 psi, pf=1000 psi and ΔT =150oF
6000
5000
4000
3000
cement system 1
cement system 3
1000
0
4.5000 5.0000 5.5000 6.0000
-1000
-2000
Radius(inches)
Fig 4.14: Tangential Stress with pi=15000 psi, pf=1000 psi and ΔT =150oF
67
600
-400
4.5000 5.0000 5.5000 6.0000
-1400
Radial stress(psi)

cement system 2
-4400
-5400
-6400
Radius(inches)
Fig 4.15: Radial Stress with pi=15000 psi, pf=1000 psi and ΔT =150oF
With the effect of temperature change, the trend in the tangential stress distribution
across the cement sheath remains pretty much the same although the stress value is higher.
The highest von Mises stress with ΔT=0 is about 9,000 Psi (cement system 2) compared
with 10,000 psi with the same cement system with ΔT=150. The tangential stress
distribution, however, differs as it fluctuates between a negative (compressive) value for
systems 2 to a positive (tensile) value for cement systems 1 and 3.

68
2. High Inner Pressure, Zero Formation Pressure with Temperature Change
12000
10000
8000
cement system 1
cement system 3
4000
2000
0
4.5000 5.0000 5.5000 6.0000
Radius(inches)
7000
6000
5000
4000
cement system 1
cement system 3
2000
1000
0
4.5000 5.0000 5.5000 6.0000
-1000
Radius(inches)
Fig 4.17: Tangential Stress with pi=15000 psi, pf= 0 psi and ΔT =150oF
69
600
4.5000
-400 5.0000 5.5000 6.0000
-1400
Radial stress(psi)
cement system 1
cement system 3
-3400
-4400
-5400
Radius(inches)
Fig 4.18: Radial Stress with pi=15000 psi, pf= 0 psi and ΔT =150oF
In this case, the radial stress distribution profile is similar to that of Fig. 4.9 with
almost the same magnitude of stress but the tangential stress profile is different. While the
tangential stress profile in Fig. 4.8 was all negative (compressive) stress for cement system
1, it fluctuates from 85 psi (tensile) to -468 psi (compressive) when the effect of
temperature change is considered.. The equivalent stresses decreases non -linearly from the
inner surface of the cement ( 2928 psi for cement 1, 9,652 psi for cement 2, 4,687 psi for
cement 3) to 1,977 psi for cement 1, 6,802 psi for cement 2, 3,164 psi for cement 3 at the
outer surface. These are shown by figures 4.16 to 4.18 above.

70
3. Low Inner Pressure, High Formation Pressure with Temperature Change
This case generates the higher values for tangential and radial stresses in the three cement
systems as compared to the case without temperature change. The radial stress in the three
cement systems also increases non-linearly from the inner surface of the cement to the
outer surface where the opposite is the case without temperature change. The results are as
shown in Figs. 4.19 to 4.21.
6000
5000
4000
cement system 1
cement system 3
2000
1000
0
4.5000 5.0000 5.5000 6.0000
Radius(inches)
71
0
4.5000 5.0000 5.5000 6.0000
-1000
-2000

cement system 2
-5000
-6000
-7000
Radius(inches)
Fig 4.20: Tangential Stress with pi=4000 psi, pf=10000 psi and ΔT =150oF
600
-400
4.5000 5.0000 5.5000 6.0000
-1400
-2400
Radial stress(psi)
-3400
cement system 1
cement system 3
-5400
-6400
-7400
-8400
-9400
Radius(inches)
Fig 4.21: Radial Stress with pi=4000 psi, pf=10000 psi and ΔT =150oF
72
4.2 Fatigue Loading
The aim of the fatigue study was to determine if a cement system capable of sustaining a
static load would be able to sustain a similar load under cyclic loading conditions. A
fatigue analysis together with a related static analysis was done on a casing – cement -
formation setup with ANSYS using cement systems 2 with an inner pressure of 7,000 psi
and formation pressure of 2,000 psi. As was done in the static study, the properties of the
casing remained the same while the fatigue behavior of cement was studied with two
different formation properties; E=3 x 106 psi, v=0.42 and E=1 x 106 psi, v=0.3. Fully
reversed and zero based loading conditions were used in the fatigue analysis.
Fig 4.22: Equivalent Stress for Static Loading with Formation E=1 x 106 psi and v=0.42
73
With 7,000 psi internal pressure and 2000 psi formation pressure, the finite element
model in Fig. 4.22 gives an indication that cement system 2 would withstand such pressure
loads under static loading conditions. The effective von Mises stress changes from 22,760
psi at the cement-casing boundary on the casing side to 4330 psi on the cement sheath side
and from 2994 psi on the cement side of cement-formation boundary to 1,842 psi on the
formation side of the same boundary. The equivalent alternating stresses for a zero based
cyclic loading condition with the formation properties E=1 x 106 psi, v=0.42, however, is a
bit different as shown in Fig. 4.23.
Fig 4.23: Equivalent Alternating Stress for Zero Based Loading

74
The zero based fatigue loading results in a lower equivalent alternating stress in the
casing but higher values in the cement and the formation, as shown in Fig. 4.23.The
equivalent alternating stress decreases non-linearly from 18,199 psi to 13,719 psi at the
casing-cement boundary with a life of 5.86 x 105 cycles (Fig. 4.24). The cement part of the
casing –cement boundary has a life of 5.6 x 104 cycles (Fig.4.24) and an alternating stress
of 7,783 psi (Fig 4.23) psi as opposed to 4,330 psi (fig 4.22) under static loading
conditions. The progression from the cement to the formation at the cement –formation
boundary however sees a large increase in the equivalent alternating stress from 2,992 psi
to 1 x 1032 psi.
Fig 4.24: Life Cycle for Zero Based Loading

75
Figures 4.25 to 4.27 show the fatigue sensitivity plot for life, damage and safety
factor but does not explicitly show the fatigue performance of the different regions of the
casing-cement –formation model.
Fig 4.25: Fatigue Sensitivity to Life Plot for Zero Based Loading
Fig 4.26: Fatigue Sensitivity to Safety Factor Plot for Zero Based Loading
76
Fig 4.27: Fatigue Sensitivity to Damage Plot for Zero Based Loading
With the formation properties changed to E=1 x 106 psi and v=0.3, while the casing
and cement properties remain the same, an insight is gained on how mechanical properties
like Young’s modulus and Poisson ratio affects stress distribution and ability to withstand
fatigue loading. Figure 4.27 shows the response to fatigue with zero based load and
formation properties E=1 x 106 psi and v=0.3.Under the different formation properties, both
show similar trends in the static equivalent stresses and alternating stresses but as shown in
Figs. 4.28 and 4.29, the formation with more ductile properties (E=1 x 106 psi and v=0.3)
has slightly higher stress values for the static equivalent stresses and almost the same value
alternating stress.
77
Fig 4.28: Equivalent Stress for Static Loading with Formation Property: E=1 x 106, v=0.3
Fig 4.29: Alternating Stress for Zero Based Loading with Formation Property: E=1 x 106,
v=0.3
78
FEA modeling of the fully reversed cyclic shear loading of cement sheath at the
couplings (Figs. 4.30 and 4.31) also showed that the casing-cement boundary, which is the
point of maximum shear, appears more vulnerable to fatigue failure.
Fig 4.30: Alternating Stress for Fully Reversed Loading with Formation Property:
E=3 x 106, v=0.3

79
Fig 4.31: Life Cycle for Fully Reversed Loading with Formation Property: E=3 x 106,
v=0.3
The fatigue sensitivity plots for life, damage and safety factor provides us with the
number of life cycles to failure, the amount of damage done and the factor of safety for the
casing-cement-formation system, but they do not clearly express the effects of fatigue
loading on the cement sheath. In order to achieve this, the cement sheath was isolated and
the contact pressures obtained from the analytical model were applied. A temperature
change of 1500F was also considered. The response of cement systems 1 and 2 were
modeled. The results are as shown in Figs. 4.32 to 4.34.

80
Fig 4.32: Equivalent Stress for Cement System 2 under Static loading
Fig 4.33: Alternating Stress for Cement System 2 under a Zero Based Cyclic Loading
81
Fig 4.34: Life Cycle for Cement System 2 under a Zero Based Cyclic Loading
The fatigue sensitivity plots are as shown in Figs. 4.35 to 4.37 below.
82
Fig 4.35: Fatigue Sensitivity to Life for a Zero Based Cyclic Loading with Cement 2
Fig 4.36: Fatigue Sensitivity to Damage plot for a Zero Based Cyclic Loading with Cement
2
83
Fig 4.37: Fatigue Sensitivity to Safety Factor plot for a Zero Based Cyclic Loading with
Cement 2
From the fatigue sensitivity plots, it can be seen that the maximum load ratio that
can be supported by cement system 2 before failure is about 0.75. The same analysis for the
ductile cement (cement system 1), and the results are shown in Figs.4.38 to 4.40 below.
84
Fig 4.38: Equivalent Stress for Cement System 1 under Static Loading
Fig 4.39: Alternating Stress for Cement System 1 under Zero Based Cyclic Loading
85
Fig 4.40: Life Cycle for Cement System 1 under Zero Based Cyclic Loading
For cement system 1, the magnitude of the alternating stress appears to be smaller
than the equivalent stress under static loading (Figs. 4.38 & 4.39).This trend is quite
different from what was obtained with cement system 2 where the minimum alternating
stress under a zero based cyclic loading is greater than the maximum equivalent stress
under static loading (Figs. 4.32 and 4.33).The fatigue sensitivity plots (Figs.4.41 to 4.43)
show that the maximum load ratio that can be supported by cement system 1 before failure
is about 1.1.
86
Fig 4.41: Fatigue Sensitivity to Life for Zero Based Cyclic Loading with Cement 1
Fig 4.42: Fatigue Sensitivity to Safety Factor for Zero Based Cyclic Loading with Cement
1
87
Fig 4.43: Fatigue Sensitivity to Damage for Zero Based Cyclic Loading with Cement 1
88
CHAPTER V
CONCLUSIONS AND RECOMMENDATIONS
5.1 Conclusions
The drilling and completion of a well is a capital project that needs to be executed
properly. As a consequence, a detailed design is required, putting into consideration all
forces that may affect the integrity of a well throughout its life span.
The findings of this investigation highlight the importance of cement fatigue in
cement design. Designs based solely on static loading conditions may or may not be
enough to ensure long term integrity depending on prevailing downhole conditions thus
the need to take the analysis further by also examining the effect of fatigue.
Life prediction is possible through models that define the operating stress
principles in a well. Such are incorporated into the analytical model presented in this
thesis, which can be used to evaluate stresses in the cement sheath based on actual
wellbore parameters and in synergy with finite element models, can be combined
effectively to evaluate the fatigue and static loading behavior of the well thereby helping
to predict the life of the well. These tools will also help in optimizing the design with
regards to the material properties of both the cement and casing which would also help
save cost in addition to a good well design.
Insights were also thrown into some fatigue and static behavior of well cements.
Fatigue failure in cement occurs when microscopic damage within the microstructure of
the cement caused by initial cyclic loading turns into macroscopic cracks under
89
gradually increasing loads. Cyclic loading impacts initial damage and if loading is
continued at load ratios above the critical ratio for a particular cement mix, failure is
imminent but may undergo many cycles when loaded below this ratio. Loading
conditions may affect the fatigue property of cement only when the mechanical
properties are such as to withstand static loading.
The mechanical properties of cement play a very important role in the static and
fatigue performance of cement. Ductile cement s systems – cements with low Young’s
modulus and a high Poisson ratio generally perform better under static and cyclic
loading conditions as compared to brittle cement systems i.e. cement systems with a high
Young’s modulus and low Poisson ratio. Ductile cement systems generates a
significantly lower value of tangential and radial stresses while brittle cements are more
likely to generate higher tensile and radial stresses within its microstructure under a
particular loading condition.
The magnitude of confining stress and the mechanical properties of the formation
also play an important role in the static and fatigue behavior of both the cement and
casing. A large far field stress (formation pressure) act to increase the performance of the
casing and counteracts high internal pressures ensuring a minimal transfer to the cement
sheath. Also the more brittle the formation (in terms of Young’s modulus and Poisson
ratio), the more stresses that will be transmitted to the casing and cement sheath.
90
5.2 Recommendations for Future Work
This work includes report on the effect of both static and fatigue behavior of well cement
based on analytical and finite element models. A lot of experimental work is required in
the following areas;
Developing equations specific to well cement. Equations 3.3, 3.4 and 3.26 can be
derived specifically for well cement from experimental data and linear regression
analysis. New failure mechanism, crack initiation and propagation and failure theories
can also be developed from these data. This would help expand the analytical model to
include fatigue life prediction
• The findings reported in this work are centered mainly on the mechanical
properties of the cement and on loading conditions. The effect of other factors
like cement-water ratio etc should be investigated through experimental studies.
• The effect of additives on the static and fatigue properties of well cement.
• Performance of new cement system with special properties like foam and
expansive cements should also be studied and data generated for them
91
REFERENCES
1. Cook, R.D. and Young, W.C.: Advanced Mechanics of Materials, 2nd ed. Prentice Hall,
Upper Saddle River, NJ. 1999.
2. Kim, J. and Kim, Y.: “Experimental Study of the Fatigue Behavior of High Strength
Concrete,” Cement and Concrete Research, Vol. 26, No. 10, 1996. 1513-1523.
3. Antrim, J.D.: “The Mechanism of Fatigue in Cement Paste and Plain Concrete”,
Highway Research Record, 210 no 6, 1965.
4. Breitenbucher, R., Alawieh, H., and Ibuk, H.: “Influence of Cyclic Loading on the
Degradation of Mechanical Concrete Properties,” Advances in Construction Materials,
(August 2007) 317–324.
5. Breitenbucher, R. and Ibuk, H.: “Experimentally Based Investigations on the

Degradation-Process of Concrete under Cyclic Load,” Materials and Structures, (June
2006) 39,717–724.
6. Nayeb, H., Hashemi I., Cohen, M. D., and Erturk, T.: “Evaluation of Fatigue Damage
on the Mechanical Properties of Fiber Reinforced Cement Pastes,” Cement and Concrete
Research,(May 1985)15, 879-888.
7. Bourgoyne, A.T., Millheim, K.K., Chenevert, M.E., and Young, F.S.: Applied Drilling
Engineering, SPE Textbook Series, Vol. 2, 1986.
8. Nelson, E.B.: “Well Cementing,” Development in Petroleum Science, Vol. 28, 1990.
9. Halliburton Energy Services: “Halliburton Cementing Technology Manual”, 2007
10. US Department of Transport: “Portland Cement”,

www.fhwa.dot.gov/infrastructure/materialsgrp/cement.html , 2007.
11. Edgley, K., D., Sabins, F., L. and Watters, L., T.: “Supercement for Annular Seal and
Long Term Integrity in Deep, Hot Wells “ ‘Deep Trek’,” Phase II Annual Report, CSI
Technologies, Houston. August 2005.
12. Pine, M., Hunter, L., Mutch, J., Adam, J., and Griffith, E.J.: “Selection of Foamed
Cement for HPHT Gas Wells Proves Effective for Zonal Isolation-Case History,” paper
IADC/SPE 79909, IADC/SPE Drilling Conference, Amsterdam, Netherlands, February
2003.
92
13. Ravi, K., Bosma, M., and Hunter, L.: “Optimizing the Cement Sheath Design in
HPHT Shearwater Field,” paper IADC/SPE 79905, IADC/SPE Drilling Conference,
Amsterdam, Netherlands, February 2003.
14. Ravi, K and Xenakis, H.: “Cementing Process Optimized to Achieve Zonal
Isolation”. Halliburton, January 2007.
15. Stiles, D.: “Effects of Long-Term Exposure to Ultrahigh Temperatures on

Mechanical Parameters of Cement,” paper SPE 98896, IADC/SPE Drilling Conference,
Miami, Florida, February 2006.
16. Godwin, K.J., and Crook R.J.: “Cement Sheath Stress Failure,” SPE Drilling
Engineering, pp. 291-296, December 1992.
17. Fleckenstein, W.W., Eustes, A.W., and Miller, M. G.: “Burst Induced Stresses in
Cemented Well Bores,” paper SPE 62596, SPE/AAPG Western Regional Meeting, Long
Beach, California, June 2000.
18. Gray, K.E., Podnos, E., and Becker, E.: “Finite Element Studies of Near-Wellbore
Region during Cementing Operations: Part 1,” Paper SPE 106998, SPE Production and
Operations Symposium, Oklahoma City, Oklahoma, April 2007.
19. Heathman, J., and Beck, F.E.: “Finite Element Analysis Couples Casing and Cement
Design for HP/HT Wells in East Texas,” paper SPE 98896, IADC/SPE Drilling
Conference, Miami, Florida, February 2006.
20. Rodriguez, W. J., Fleckenstein W.W., and Eustes A.W.: “Simulation of Collapsed
Load on Cemented Casing Using Finite Element Analysis,” paper SPE 84566, SPE
Annual Technical Conference, Denver, Colorado, October 2003.
21. Schubert, J.J., Shahri, M.A., and Amani, M.: “Detecting and Modeling Cement
Failure in High Temperature/High Pressure (HP/HT) Wells Using Finite Element
Methods,” paper IPTC 10961, IPTC Conference, Doha, Qatar, November 2005.
22. Ravi, K., Bosma, M., and Gastebled, O.: “Improve the Economics of Oil and Gas
Wells by Reducing the Risk of Cement Failure,” paper SPE 74497, IADC/SPE Drilling
Conference, Dallas, Texas, February 2002.
23. Krusche, K., Johnson, C.R., Braud, N.Y., and Ghazi, H.B.: “Application of Cement
Engineered Solutions to Solve Long Term Cement Integrity Issues in Tunisia,” paper
SPE 100390, SPE Annual Technical Conference, San Antonio, Texas, September 2006.
93
24. Avram, C., Facaoaru, I., Mirsu, O., Filimon, I., and Tertea, I.: “Concrete Strength and
Strains,” Developments in Civil Engineering, Vol. 3, Elsevier Scientific Publishing
Company, New York, 1981.
25. Newman, J., and Choo, B.S.: “Concrete Properties”, Advanced Concrete Technology,
Elsevier Butterworth-Heinemann, Boston, MA. 2003.
Supplemental Sources Consulted
1. Magalhaes, A.G. et al. : “Mechanical Behavior of Cementitious Matrix Composite”,

Cement and Concrete Composites, (1996)18 no 1, 9-22.
2. Bocca, P. and Crotti, M.: “Variations in the Mechanical Properties and Temperature
of Concrete Subjected to Cyclic Loads, Including High Loads,” Materials and
Structures, (February 2003)36, 40-45.
3. Morris, A. D., and Garrett, G. G.: “A Comparative Study of the Static and Fatigue
Behavior of Plain and Steel Fiber Reinforced Mortar in Compression and Direct
Tension,” The International Journal of Cement Composites and Lightweight Concrete,
(May 1981) 3, no 2, 73-91.
4. Khandka R.K.: “Leakage behind Casing” M.S. Thesis, Norwegian University of

Science and Technology, Trondheim, 2007.
5. Alliche, A., and Francois, D.: “Fatigue Behavior of Hardened Cement Paste,” Cement
and Concrete Research, (January 1986)16, 199-206.
6. LeRoy-Delage, S., Baumgarte, C., Thiercelin M., and Vidick B.: “New Cement
Systems for Durable Zonal Isolation,” paper SPE 59132, IADC/SPE Drilling
Conference, New Orleans, Louisiana, February 2000.
7. O’Brien, B.T.: “Case against Cementing Casing - Casing Annuli,” paper SPE 35106,
IADC/SPE Drilling Conference, New Orleans, Louisiana, March 1996.
8. Hancq, D.A.: “Fatigue Analysis using ANSYS”, ANSYS Inc.

http://www.caeai.com/papers/Fatigue_ANSYS.pdf, 2007.
9. Hahn, D.: “Why Do HP and HT Completions Require Rigorous Engineering Design?”,

paper SPE 101509, SPE Distinguished Lecture Series.
10. Thiercelin, M.J., Dargaud, B., Baret, J.F., Rodriguez, W. J.: “Cement Design Based on
Cement Mechanical Response,” SPE Annual Technical Conference, San Antonio, Texas,
337–348, 1997.
94
VITA
Name: Ignatius Obinna Ugwu
Address: Department of Mechanical Engineering
c/o Dr Xin-Lin Gao
College Station, TX 77843-3123
Email Address: [email protected]
Education: B.Eng., Mechanical Engineering, Federal University of Technology,

Owerri, Nigeria, 2005
M.S., Mechanical Engineering, Texas A&M University,

College Station, 2008

Ugwu Thesis

Uploaded by

Copyright:

Available Formats

Ugwu Thesis

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Ugwu Thesis

Uploaded by

Copyright:

Available Formats

CEMENT FATIGUE AND HPHT WELL INTEGRITY WITH

APPLICATION TO LIFE OF WELL PREDICTION

IGNATIUS OBINNA UGWU

Submitted to the Office of Graduate Studies of

Major Subject: Mechanical Engineering

APPLICATION TO LIFE OF WELL PREDICTION

IGNATIUS OBINNA UGWU

Submitted to the Office of Graduate Studies of

Co-Chairs of Committee, Xin-Lin Gao

Major Subject: Mechanical Engineering

Ignatius Obinna Ugwu,

B.Eng., Federal University of Technology, Owerri, Nigeria

Co-Chairs of Advisory Committee, Dr. Xin-Lin Gao

and pressures encountered at increasing depths.

economically, it is pertinent that a complete zonal isolation is achieved during well

fatigue failure as a consequence of extensive degradation of the microstructure of the

of research and experimental investigations on the mechanism of fatigue failure of

further strengthens the need to understand the mechanism of cement fatigue.

This research seeks to develop a better understanding of the performance of the

Based on the findings of this investigation, the mechanical properties of the

and fatigue failure of well cement.

its progress is dedicated to Him.

To my parents who have been shinning beacons and continuous source of

continued guidance, understanding and support is something I highly appreciate.

to serve on my thesis committee.

for funding this project.

HPHT High Pressure -High Temperature

S-N Diagram Stress-Cycle Diagram

BWOC By Weight of Cement

BVOB By Volume of Blend

BWOW By Weight of Water

FEA Finite Element Analysis

TABLE OF CONTENTS………………………………………………………………. viii

LIST OF TABLES……………………………………………………………………… xiv

I INTRODUCTION: RESEARCH MOTIVATION……………………...................... 1

2.1 Well Cement…………………………………………………………….. 12

III CEMENT FAILURE DUE TO STATIC LOADING……………………………… 33

3.1 Analytical Model of Wellbore Stresses……………………………….… 33

IV RESULTS AND DISCUSSION…………………………………………………... 56

4.1 Static Studies…………………………………………………..……… 56

V CONCLUSIONS AND RECOMMENDATIONS…………………………….…... 88

1.1 Fluctuating Tensile and Compressive Cyclic Stress ........................................... 6

1.4 Damage Development Depending on Different Cyclic Load Levels…………... 10

2.1 Stresses That Act on a Pre-Stressed Cement System…………………………… 18

2.2 Three Step Process for Cement Design …………..…………………………… 25

3.1 Radial and Hoop Stress Profile Due to Loading Conditions…………………... 34

3.2 Stresses Acting on a Cement Sheath ……………………….………………….. 34

3.3 Concrete Failure Criterion under Triaxial Compressive Stresses ……………… 37

3.5 Contact Pressure on Casing-Cement Interface…………………………………. 39

3.6 Operative Stresses on Cement-Formation Interface ………………………….... 42

3.7 Contact Stresses on Cement Sheath..................................................................... 44

3.10 S-N Curve for Fatigue Analysis in ANSYS…………………….……………… 50

3.11 Cyclic Loading Conditions for the Cement Sheath…….……….……………… 51

3.12 Goodman Diagram for Brittle and Ductile Materials….……….………………. 52

4.23 Equivalent Alternating Stress for Zero Based Loading………………………….. 73

4.24 Life Cycle for Zero Based Loading ……………………….……………………. 74

4.27 Fatigue Sensitivity to Damage Plot for Zero Based Loading……………............ 76

4.32 Equivalent Stress for Cement System 2 under Static Loading………………….. 80