SPE 110470

Field Validation of Drillpipe Rotation Effects on Equivalent Circulating Density

Terry Hemphill, Halliburton; Peter Bern and Juan Carlos Rojas, BP Exploration; and Krishna Ravi, Halliburton

Copyright 2007, Society of Petroleum Engineers

these studies, the authors have looked at hydraulic effects in
This paper was prepared for presentation at the 2007 SPE Annual Technical Conference and terms of changes in annular pressure drop (ΔP). They
Exhibition held in Anaheim, California, U.S.A., 11–14 November 2007.
generally concluded that annular pressure initially drops with
This paper was selected for presentation by an SPE Program Committee following review of
information contained in an abstract submitted by the author(s). Contents of the paper, as
increasing drillpipe rotation speed, and later increases with
presented, have not been reviewed by the Society of Petroleum Engineers and are subject to increasing drillpipe rotation speed. In slim hole drilling
studies4,5,6, it was found that a rotating drillpipe can produce
correction by the author(s). The material, as presented, does not necessarily reflect any
position of the Society of Petroleum Engineers, its officers, or members. Papers presented at
SPE meetings are subject to publication review by Editorial Committees of the Society of
Petroleum Engineers. Electronic reproduction, distribution, or storage of any part of this paper
large increases in annular pressure drop when annular
for commercial purposes without the written consent of the Society of Petroleum Engineers is clearances are small. Later, with the advent of downhole
pressure drilling tools, researchers7,8,9 found that there was
prohibited. Permission to reproduce in print is restricted to an abstract of not more than
300 words; illustrations may not be copied. The abstract must contain conspicuous
acknowledgment of where and by whom the paper was presented. Write Librarian, SPE, P.O.
Box 833836, Richardson, Texas 75083-3836 U.S.A., fax 01-972-952-9435.
indeed a positive link between drillpipe rotation speed and
Equivalent Circulating Density (ECD), a term that
Abstract mathematically combines Equivalent Static Density (ESD)
It is important to predict the Equivalent Circulating Density and annular pressure drop converted to density by the
(ECD) of fluids as closely as possible for safe drilling and following equation:
completion of wells. This is especially true for those
challenging wells where the margin between hole collapse and ECD = ESD + ΔP/0.052/TVD……………………...………(1)
formation fracturing is very narrow. These characteristics of
narrow safe drilling windows are commonly seen in extended- Figure 1 and Figure 2 show results from published
reach drilling (ERD) and deepwater wells, and pose significant work8,9 that clearly demonstrate a relationship between
drilling and cementing challenges. drillpipe rotation speed and increasing ECD. Figure 1 shows
Over recent years, the changes in ECD caused by data for a 12.25-in section, while Figure 2 contains data for
fluid compressibility, downhole fluid rheology, drillpipe 8.75-in and 8.5-in sections. In presenting the data, the authors
eccentricity, and rate of penetration have been widely studied. did not publish a model to account for the measured changes
In contrast, work on the modeling of drillpipe rotation effects in ECD.
on pressure drop and ECD have received relatively little Today, researchers now generally agree that there is a
attention. Early studies have looked at the relationship positive relationship between drillpipe rotation speed and
between circulating pressure drop, especially in slimhole annular pressure drop / ECD. Until recently there has not
drilling configurations, and at very high drillpipe rotational been published a comprehensive mathematical model that
speeds. More recently the predicted effects of rotation on accurately predicts the effects of drillpipe rotation on annular
ECD in concentric and eccentric wellbore geometries have pressure drop / ECD. Results using a preliminary model10
been studied and model improvements have been made. were published for concentric wellbores, and subsequently
These technical approaches are summarized in this paper. results for a more comprehensive model11 were published for
The effect of drillpipe rotational speed on the eccentric wellbores.
predicted ECD for different drillstring combinations are Studies of drillpipe rotation effects in circulating
presented and discussed. The calculated effects of drillpipe invert emulsion drilling fluids were made, and the combined
rotation on pressure drop and ECD are compared to actual effects on ECD were measured. In these “fingerprinting”
field measurements using downhole annular pressure gauges. exercises performed on a North Sea deviated well, key
From the results discussed in this paper, the reader operational and drilling fluid parameters were recorded and
can gauge the accuracy of the current calculation methods. the subsequent ECD measurements were measured and
The changes in ECD with drillpipe rotation can now be better recorded. This body of work represents the most
predicted, something especially important for those wells comprehensive set of data from which to study the effects of
whose safe drilling window is narrow. Ultimately this reduces drillpipe rotation on annular pressure changes. In this paper,
drilling and completions risk and helps assure safe and the data set from the North Sea experiments is used to validate
efficient well construction. the mathematical model constructed for drillpipe rotational
effects in concentric and eccentric wellbores.
Introduction
Many studies on the effects of drillpipe rotation on annular
drilling hydraulics have been performed over the years1,2,3. In
2 SPE 110470

Model Description addition to ECD/rotation tests, the fingerprinting exercises

To accurately characterize the annular pressure changes with also included ESD measurement and flowback tests.
drillpipe rotation requires a complex set of equations, the Inside the 13.375-in casing, two sets of ECD tests
process of which is described below. These equations are were run with different surface fluid densities and rheological
valid for non-Newtonian drilling fluid moving in laminar flow properties. At each fluid density, three axial flow rates were
through a defined annular space. Key parameters used to used (900, 1000, and 1100 US gal/min). At each of these
characterize drillpipe rotation effects include: three flow rates, ECD was measured at four drillpipe rotation
• Axial velocities across the annular gap (no rotation) speeds (25, 80, 120, and 150 rev/min in FP1 and 20, 80, 120,
• Local fluid shear stresses and shear rates at the and 150 rev/min in FP2). Data collected using the fluid
conduit walls (no rotation) exhibiting the lower surface density and rheological properties
• Tangential velocities resulting from rotation across are labeled FP1 (Fingerprinting 1) while those for the fluid
the annular gap (no axial flow) having higher properties are labeled FP2. Inside the 9.625-in
• Local fluid shear stresses and shear rates near the casing, tests were run at a flow rate of 450 US gal/min, and
conduit walls caused by rotation (no axial flow) ECD measurements were made at four drillpipe rotation
• The coupling of axial flow and rotation effects on the speeds (30, 80, 120, and 140 rev/min). Data collected in the
resulting local shear stresses and shear rates near the smaller hole section are labeled FP3. Table 1 lists the key
conduit walls for fluids in helical flow. parameters used in the hydraulic modeling.
• The conversion of the local coupled fluid shear
stresses to pressure drop, with each conduit side Measured Results
(drillpipe wall and outer wall) calculated separately. Figure 4 shows the measured changes in ECD as a function of
drillpipe rotation speed for FP1and Figure 5 shows similar
Additionally, to accurately account for changes in measurements for FP2. Figure 6 shows similar data for the
annular pressure with drillstring rotation in eccentric FP3 series of experiments run at 450 gpm.
wellbores, adjustments must be made in calculated pressure In agreement with the results in Figure 1 and Figure
drops to account for angle arc distances at the drillpipe and 2, the data show that for all cases, the ECDs increased
outer walls. These annular cross sectional distances are shown compared to the original ECD values measured with no
in Figure 3. In concentric wellbores, for a given annular cross rotation. While there is some scatter in the measurements, the
section, the relation between the arc distances for the drillpipe fingerprinting data generally show a more rapid increase in
and the outer wall is constant. However, in eccentric ECD at the lower drillpipe rotation speeds, and a slower ascent
wellbores, this relation is variable with different annular cross- at the higher rotation speeds.
sectional areas, and adjustments to local pressure drops at the
outer wall must be made in order to determine the overall Hydraulic Modeling for the Fingerprinting Test
pressure drop. Conditions
With the available fingerprinting test data in hand, the annular
Fingerprinting Exercises in the Field pressures and ECD were first modeled to recreate original
In the North Sea series of experiments used to validate the testing conditions using the in-house hydraulic model. Several
drillpipe rotation model in this paper, a large set of data were steps were performed in this process:
collected to thoroughly study the effects of drillpipe rotation. • Each wellbore was divided into smaller hydraulic
Key operational and fluid parameters recorded before and sections to account for geometry and hole angle
after the study included: changes, a technique outlined in the current API
• Drilling fluid surface density and measured downhole RP13D12. A total of 5 hydraulic sections were created
ESD in the FP1 and FP2 cases, and 6 in the FP3 case.
• Defined downhole temperature profile • Using the defined wellbore static temperature profile,
• Defined hole/pipe geometry and wellbore deviation the downhole drilling fluid density (ESD) and fluid
profile rheological parameters were calculated.
• Drilling fluid rheological and other measured • Values for average drillpipe eccentricity (ε) for each
properties as reported on daily mud reports wellbore section were assigned. In deviated
• Controlled drillpipe rotation and pump rate speeds wellbores, the maximum drillpipe eccentricity is
• ECD measurements with axial flow only (no rotation calculated using the dimensions of the drillpipe tool
speed) at each flow rate joints.
• ECD measurements at each coupled axial flow rate • For each flow rate (with no drillpipe rotation), the
and drillpipe rotation speed predicted ECDs were compared to measured data and
adjustments were sometimes made to accurately
Since the fingerprinting exercises were performed reflect the ECD data measured during the
inside the 13.375-in and 9.625-in casings before drilling out, fingerprinting exercise.
the holes were considered clean and free from drilled cuttings, • The resulting pressure drops (with no drillpipe
so no adjustments were deemed necessary to compensate for rotation) for each wellbore hydraulic section were
additional ECD effects caused by annular cuttings loading. In calculated.
SPE 110470 3

The resulting information was then used as input into predicted, showing an overall better balance than that seen
the drillpipe rotation module to predict the effects on annular using the linear decay model. With use of the non-linear
pressure drop and ECD. Key considerations used in the decay model, absolute average errors were reduced
pressure drop modeling include: substantially, ranging between 0.0001 and 0.0581 lbm/gal.
• For each annular section, the drillpipe eccentricity (ε) The average absolute error for all 28 cases is calculated to be
was equivalent in value to that used in the hydraulic 0.0243 lbm/gal, a reduction in error of nearly 45%.
simulations for axial flow only. Possible changes in In Figure 9, the results for one case are shown in
eccentricity with drillpipe rotation were not more detail: FP1 at a pump rate of 1000 US gal/min. Here the
considered. calculated increases in pressure drop using the linear and non-
• The effects of drillpipe tool joints on annular pressure linear shear rate decay submodels are compared against the
drop and maximum possible drillpipe eccentricity measured pressure drop increases.
were taken into account. Pressure drops at the tool These results clearly show that, for non–Newtonian
joints were modeled as a short section with outer fluids in helical flow, to accurately model drillstring rotation
diameter equal to the tool joint diameter. These effects on ECD, a nonlinear approach must be taken for the
results were then integrated as a function of length coupling of axial and rotational shear rates. Such a nonlinear
with the drillpipe pressure drop results to get overall approach forces a sharper descent in shear rate as one moves
pressure drop. from the rotating drillpipe to the outer wall, and this sharper
• The local shear rates at the conduit walls were descent causes increased shear stress and pressure drop, and,
checked for possibilities of turbulence. No such cases hence, ECD at the walls.
were found, even at the highest levels of drillpipe As a general statement, the bulk of the increases in
eccentricity. Hence laminar flow modeling was used. pressure drop for the 3 fingerprinting studies is
• In the deviated sections, the rotating tool joints were underpredicted. Possible reasons for this include:
considered to be almost touching the wellbore wall on • Overestimation of effective drillpipe eccentricity
the low side of the hole and were thus assigned an • Assumption that effective drillpipe eccentricity
eccentricity value almost equal to 1 (0 = concentric remains constant while rotating
geometry, 1 = fully-eccentric geometry) . • No provision for resuspension of any residual barite
• For each flowrate / drillpipe rotation speed and solid particles picked up by the moving fluid
combination, the calculated increases in annular during the fingerprinting tests, which would increase
pressure drop caused by drill pipe rotation were then ESD
converted to ECD. • Redistribution of the axial flow due to imposed
• The predicted changes in ECD caused by drillpipe rotation (eg, drillpipe rotation forcoing fluid to flow in
rotation were then compared with the actual data the narrow part of the annulus).
collected in the fingerprinting exercise.
• The change in annular shear rate across the gap was It is also useful to study the distribution of pressure
modeled for linear and non-linear decay of helical drop increase in terms of location in the drill string: opposite
shear rate, with the necessary boundary condition of the rotating drillpipe, opposite the rotating tool joints, and
zero (0) velocity at the outer wall. opposite the rotating drill collars. In Figure 10 the data for
the FP1 case at 150 rev/min and 1000 US gal/min pump rate
Drillpipe Rotation Modeling Results are used, since in this case there was very little error between
Figure 7 shows the simulated ECD results plotted against the measured and calculated values of increased pressure. Here
measured ECDs for a total of 28 cases for the condition of the results show that the bulk of the pressure drop increase is
linear decay of helical shear rate across the annular gap. found opposite the rotating drill collars, which accounts for an
Absolute errors between the measured and calculated ECDs average of 71% of all pressure increase. The rotating main
ranged between 0.0043 and 0.077 lbm/gal. The average body of the drillpipe accounts for about 17% of the increase in
absolute error for all cases is calculated to be 0.0431 lbm/gal. pressure. The rotating tool joints, while only a small fraction
For FP1 and FP2, the modeled ECDs with drillpipe of the length of a joint of drillpipe, account for about 12% of
rotation are somewhat under-predicted, while those of the FP3 the annular pressure drop increase, a percentage level nearly
case at 450 US gal/min flow rate are both under-predicted and equal to pressure drop increase for the rest of the drillpipe.
over-predicted. The above results regarding the distribution of
Modeling efforts improved substantially when a non- pressure loss with fluids in helical flow are consistent with the
linear model for decay of shear rate across the annular gap was previously-published observations for slim-hole
used. The model used for non-linear shear rate decay is more geometries4,5,6. In cases where the annular clearances are
complex that for the coaxial viscometer12, in that it also takes small, the effects of drillpipe rotation can be quite significant.
into account the effect of axial shear rate. The end result is
that the composite (helical) shear rates are higher and their Summary and Conclusions
decay rates are faster. • The mathematical model that predicts the effect of
Figure 8 shows the results of measured vs. predicted drillpipe rotation on annular pressure drop can be
ECD for the same 28 cases. While most of the measured ECD used to predict the increases in wellbore ECD as a
changes were under-predicted by the model, some were over- function of drillstring rotation speed.
4 SPE 110470

• The model accounts for all the major factors that can 10. Hemphill, T., and Ravi, K., “Calculation of Drill Pipe Rotation
affect pressure drop in an annulus while drilling. Of Effects on Fluids in Axial Flow: An Engineering Approach”,
all the major factors involved, drillpipe eccentricity paper SPE 97158, presented at the 2005 SPE Annual Technical
Conference and Exhibition in Dallas, October 9-12.
remains the unmeasured factor.
11. Hemphill, T. and Ravi, K.,”Pipe Rotation and Hole Cleaning in
• For fluids in helical flow, model predictions can be an Eccentric Annulus”, paper SPE 99150 presented at the 2006
significantly improved with use of a non-linear IADC/SPE Drilling Conference in Miami, February 21-23.
submodel to describe the decay of shear rate across 12. Bern, P., et al, “Modernization of the API Recommended
the annular gap. Practice on Rheology and Hydraulics: Creating Easy Access to
• The comparison of the modeled results with the Integrated Wellbore Fluids Engineering”, paper IADE/SPE
experimentally-measured results show that the model 98743 presented at the 2006 IADC/SPE Drillng Conference in
slightly underpredicts the annular pressure drop and Miami (February 21-23).
ECD increases with increasing drillpipe rotation.
However, the errors in the predictions are small and Nomenclature
give confidence to the user that the results can be ECD = Equivalent circulating density
used to better model hydraulics in challenging wells ERD = Extended reach drilling
where the safe drilling window is quite narrow. ESD = Equivalent static density
HB = Herschel-Bulkley rheological model
• The distribution of the increase in annular pressure
ID = Internal diameter
drop caused by drillpipe rotation has been studied,
MD = Measured depth
and the greatest percentage increase in annular
OD = Outer diameter
pressure drop occurs opposite the larger-sized OD
P = Pressure drop
sections of the drillstring.
TVD = True vertical depth
Acknowledgements
Δ = Change in
The authors would like to thank their respective companies BP
ε = Drillpipe eccentricity
and Halliburton for giving pemission to present this work to
the drilling industry.
0.3
M e a s u re d I n c r e a s e i n E C D

References
1. Luo, Y. and Peden, J.M., “Flow of Drilling Fluids Through 0.25
Eccentric Annuli,” paper SPE 16692 presented at the 1987 SPE 0.2
(l b m / g a l )

Annual Technical Conference and Exhibition in Dallas,

September 27-30. 0.15
2. Walker, R.E. and Al-Rawi, R., “Helical Flow of Bentonite 0.1
Slurries”, paper SPE 3108 presented at the 45th Annual Fall
Meeting of the SPE of AIME in Houston (October 4-7, 1970). 0.05
3. Bailey, W.J. and Peden, J.M., “A Generalised and Consistent 0
Pressure Drop and Flow Regime Transition Model for Drilling 0 50 100 150 200 250
Hydraulics Suitable for Slimhole, Underbalanced and Horizontal
Drillpipe Rotation Speed (rev/min)
Wells”, paper SPE/IADC 39281 presented at the 1997 SPE /
IADC Middle East Drilling Conference in Bahrain, 23-25
November.
4. McCann, R.C., Quigley, M.S., Zamora, M., and Slater, K., Figure 1: Measured changes in ECD with drillpipe rotation speed
in 12.25-in interval. Data read from Reference 9.
“Effects of High-Speeed Pipe Rotation on Pressures in Narrow
Annuli”, paper SPE 26343 presented at the 1993 SPE Annual
Technical Conference and Exhibition in Houston, October 3-6.
5. Delwiche, R.A., Lejeune, M., Mawet, P., and Vigheto, R.,
M e a s u r e d C h a n g e in E C D (lb m /g a l)

0.3
“Slimhole Drilling Hydraulics”, paper SPE 24596 presented at
the 1992 SPE Annual Technical Conference and Exhibition in 0.25 Elf 8.75-in A
Washington, DC, October 4-7. Elf 8.75-in B
0.2
6. Bode, D.J., Noffske, R.B., and Nickens, H.V., “Well-Control
Total 8.5-in A
Methods and Practices in Small-Diameter Wellbores”, JPT 0.15
(November 1991) 1380-1386. Total 8.5-in B
0.1
7. Ward, C. and Adreassen, E., “Pressure While Drilling Data Total 8.5-in C
Improve Reservoir Drilling Performance”, SPE Drilling & 0.05 Total 8.5-in D
Completion (March 1998) 19-24.
0
8. Isambourg, P., Bertin, D., and Branghetto, M., “Field Hydraulic
0 50 100 150 200 250
Tests Improve HPHT Drilling Safety and Performance”, SPE
Drilling &Completion (December 1999) 219-227. Drillpipe Rotation Speed (rev/min)

9. Charlez, P., Easton, M., and Morrice, G., “Validation of

Advanced Hydraulic Modeling Using PWD Data”, paper OTC
8804 presented at the 1998 Offshore Technology Conference in Figure 2: Measured changes in ECD with drillpipe rotation speed
Houston, May 4-7. in 8.75-in and 8.5-in intervals. Data read from References 8 and 9.
SPE 110470 5

0.25

M e a su re d In c re as e in E C D
0.2

(l b m / g a l )
0.15

0.1

0.05

0
0 20 40 60 80 100 120 140 160
Drillpipe Rotation Speed (rev/min)
concentric eccentric
Figure 3: Schematic showing dimensions of sides of individual
sectors changing with inner pipe eccentricity. Figure 6: Measured increase in ECD at 450 US gal/min flow
rate, Fingerprinting 3 (FP3).

0.25

C a lc u la te d D e lta E C D (lb m /g a l)
0.12
0.2
M e a s u re d I n c re a s e in E C D

0.15 FP 1
0.08 FP 2
(lb m / g a l)

FP1 at 900 gpm

0.1 FP 3
FP1 at 1000 gpm
FP1 at 1100 gpm
0.04 0.05

0
0 0 0.05 0.1 0.15 0.2 0.25
0 50 100 150 200 Measured Delta ECD (lbm/gal)
Drillpipe Rotation Speed (rev/min)

Figure 7: Measured changes in ECD vs calculated changes in

Figure 4: Measured increase in ECD for 3 flow rates, ECD for all fingerprinting exercises using a linear decay
Fingerprinting 1 (FP1). submodel for annular helical shear rate.

0.3
C a lc u la te d D e lta E C D (lb m /g a l)

0.16
M ea su red In cre as e in E C D

0.25

0.12 0.2
FP 1
FP2 at 900 gpm
(l b m / g a l )

0.15 FP 2
0.08 FP2 at 1000 gpm FP 3
0.1
FP2 at 1100 gpm
0.04 0.05

0
0 0 0.05 0.1 0.15 0.2 0.25
0 50 100 150 200 Measured Delta ECD (lbm/gal)
Drillpipe Rotation Speed (rev/min)

Figure 8: Measured changes in ECD vs calculated changes in

Figure 5: Measured increase in ECD for 3 flow rates, ECD for all fingerprinting exercises using a non-linear decay
Fingerprinting 2 (FP2). submodel for annular helical shear rate.
6 SPE 110470

In c re a s e in A n n u la r P re s s u re
Table 1: Parameters Used in Hydraulic Simulations
20

16 Parameter FP1 FP2 FP3

Drillpipe OD (in) 12.347 12.347 8.861
12
Drillpipe ID (in) 5.5 5.5 5.5
(p s i )

8 Tool joint OD (in) 7 7 7

4
Flow rate (US gal/min) 900, 900,
1000, 1000, 450
0 1100 1100
25 80 120 150 Drillpipe rotation speed 25, 80, 20, 80, 30, 80,
Drillpipe Rotation Speed (rev/min) (rev/min) 120, 150 120, 150 120, 140
Measured Linear shear rate decay Non-linear shear rate decay Surface fluid density 11.2 12.6 11.5
(lbm/gal)
Surface fluid HB ‘n’ 0.785 0.77 0.725
Figure 9: Increases in annular pressure drop: measured data
vs. predictions using linear and nonlinear shear rate decay Surface fluid HB K 0.253 0.466 0.461
models. Data from FP1 at pump rate of 1000 US gal/min. (lbf/100 sq ft s^n)
Surface fluid HB τ0 8.5 10.1 15.7
(lbf/100 sq ft)
17% ECD at 0 rev/min 11.36- 12.79 13.15
(lbm/gal) 11.4 12.85
Max. deviation (°) 51 51 57
HWDP OD (in) 5.5 5.5 5.5
HWDP Tool joint OD (in) 7 7 7
12%
Drill collars OD (in) 8 8 6.75

71%

% opp DP % opp TJ % opp DC

Figure 10: Distribution of calculated annular pressure drop

increase caused by drillpipe rotation. Data from FP1 at 150
rev/min and flow rate of 1000 US gal/min.