SPE 107701

Cement Design to Optimize Production in a Highly Active Waterdrive Reservoir

Carlos Capacho and Felipe Baquero, Perenco; Kris Ravi and Jose Vela, Halliburton

Copyright 2007, Society of Petroleum Engineers

expected, with low water cut. Yuralpa Field can become
This paper was prepared for presentation at the 2007 SPE Latin American and Caribbean uneconomical if cement sheath quality and reservoir standoff
Petroleum Engineering Conference held in Buenos Aires, Argentina, 15–18 April 2007.
to oil-water contact are not suitable.
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
presented, have not been reviewed by the Society of Petroleum Engineers and are subject to Introduction
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 The Hollin Sandstone formation is the primary objective in the
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
Yuralpa Field. The productive mechanism is bottom-water
for commercial purposes without the written consent of the Society of Petroleum Engineers is drive, which maintains almost a constant pressure during the
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 well’s productive life and actively influences early water
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.
breakthrough. The importance of a good cement job can help
maximize early oil production before water entrance.
Abstract However, if the cement sheath is not competent or debonds
This paper describes the improvement in well production from after cementing, water channeling behind the casing and next
using advanced well-construction techniques compared to to the formation during well production is almost immediate
conventional techniques. The advanced technique uses an once production begins, with an exponential growth in just a
engineering approach including finite element analysis (FEA), few days.
specialized erodability testing, rock mechanics lab testing, and To improve cement job results, multiple best cementing
conformance reservoir simulation predictions. practices have been identified and applied throughout the
In a highly permeable and very mobile water-drive decades. Currently, many best practices have been
environment where natural hydraulic seal is practically non- implemented at field operations using a score-card template,
existent, an effective cement sheath imposes a very good assigning different values to each particular practice
mechanical seal to shut the water influx. In this case, the depending on its importance. A final grade for each job’s local
cement sheath is not only a well-construction or pipe-stability score card enables comparison of different jobs under the
issue, it is a final or well-termination product for enhancing same criterion (Table 1). Considering that each field has
well life or last-recovery production. The FEA analysis studies different work methods and risks associated, a single score
the mechanical interaction among the formation, cement card cannot be used for all fields. Input from field personnel is
sheath, and casing while applying different stresses on the important in identification of key points and establish a special
system. Pressure and temperature changes inside the casing score card for each field.
during the life of the well affect the behavior from the cement Following best practices alone does not guarantee the best
sheath. To address the performance from the cement system results for a cement job. Rheology from fluids needs to be
design, the well history for water production requires review. designed to comply with high displacement efficiency. One
Production and simulated data from reservoir simulators are commonly used method to improve annulus mud displacement
compared to determine if water production is as expected or is to design the fluids’ yield point (YP) in a hierarchical
not. Conformance reservoir simulators give a good manner, beginning with the lowest YP on mud and increasing
comprehensive criterion to predict water behavior in the it on the spacer, lead, and tail cement, respectively. However,
reservoir and predict at earlier stages when water will affect one important question arises: what should be the adequate
well productivity. rheological properties for an optimum displacement?
In the study presented in this paper, three advanced Erodability gives a clue of how to determine optimum
solutions were applied to address an optimization in a highly rheological properties to erode the gelled mud and soft filter
active water-drive reservoir as: (1) spacer efficiency study, cake from the annulus. Using erodability theory, spacer can be
(2) cement-sheath characterization through measured designed to obtain the required shear stress to displace gelled
mechanical properties/FEA analysis, and (3) use of and dehydrated drilling mud and any soft filter cake from the
conformance simulators to determine the productive capacity annulus, using actual well conditions as hole/casing geometry
of the well. Results are compared with the real production data and flow rate.
to determine whether this engineering design improved well In addition to the slurry placement efficiency, it is
productivity. important to consider load on the casing, cement sheath, and
Results of using all new practices were evaluated by CBL formation during cement slurry hydration, well completion,
logs and production; production results were close to the and production stages. All loads have a direct influence over
2 SPE 107701

the radial, axial, and tangential stresses imposed over the All new applied practices have provided a systematic
productive string and affecting the cement sheath behind the approach to obtain better results; however, spacer efficiency
casing. In recent years, the use from FEA analysis has allowed studies and FEA analysis have been the key to success on
operators to (1) simulate and predict the application from cement job design.
diverse loads on the well and (2) investigate the effect on
cement-sheath properties. Multiple cement designs can be Spacer Efficiency Study
evaluated in the rock mechanics laboratory to determine A weighted viscous spacer during production liner operations
specific properties and make FEA simulation more closely has been used ahead of cement slurry. Preflush use has been
evaluated. avoided considering the presence from unstable shale between
Using FEA results, operators can predict, from multiple Hollin superior and inferior formations and an inherent
cement designs, which one will be the optimum for a packing-off risk, especially when conventional hangers are
particular well. In some cases they can evaluate cement used. An extra cement volume is pumped behind the spacer to
sheaths that have shown failures through early water help increase mud removal efficiency as scavenger slurry.
production and understand the kind of failure present. Other Spacer rheology can be adjusted to perform its best mud-
important characteristics of FEA analysis are that it displacement, exerting the highest possible stress over gelled
(1) permits us to know the remaining capacity from a cement mud to remove it from the annulus. Erodability, reported by
sheath after it has passed all expected loads during its life and Ravi et al., defines a way to quantify dehydrated and gelled
(2) evaluates behavior from repeated or cyclic loads. mud resistance to be moved.1
Depending on the well position on the reservoir map and Erodability numbers can be found using large-scale
its distance to the water contact, water production does not erodability cells, where parameters influence as temperature,
follow the same trend. It is crucial to know the expected water pressure, fluid loss, shutdown period, etc., are taken into
production behavior to see whether a channel or debonding consideration. In simple cases, erodability can be determined
from cement sheath exist, to give a complete evaluation from using the mud static gel strength from gelled mud. In field use,
the cementing job and avoid costly workover remedial jobs erodability can be determined using Fann readings from 60-
when they are not necessary. Today’s conformance simulators minute mud gel strength to determine the minimum fluid shear
allow us to predict expected oil, water, and gas production, stress required to impose movement on the gelled mud from
and compare it with the production obtained from the well to the annulus side. However, this last method does not consider
determine the isolation from the cement sheath. Channels on the static time and mud-filtration characteristics, nor does it
the cement sheath can be simulated when production is not include the effects of temperature and pressure.
following the trend expected. Erodability (E) is defined in Equation 1, where τy equals
the minimum shear stress needed to impose movement over
Problem Analysis the dehydrated and gelled mud.
Main concerns in performing cement jobs over the Hollin
productive interval in Yuralpa’s field are the high permeability 600
from the zone, tectonics activity (maximum horizontal stress E= (Eq. 1)
τy
higher than overburden), and an active bottom-water drive
reservoir. According to the following criterion and depending on
Problems in the field can be reviewed on the log sections erodability numbers, mud systems can be classified as:
from Fig. 1, which is one of the first cement jobs performed in
the zone. Irregular bonding is observed along the permeable 5≤ E Mud systems hard to be removed
zones. The microseismograph plot indicates presence of some 5<E≤ 10 Mud systems moderately hard to remove
formation signals; however, pipe signals are influencing log 10<E≤ 20 Mud systems moderately easy to remove
amplitude, showing pipe debonding and channels present. 20<E≤ 30 Mud systems fairly easy to remove
It was thought that conventional slurries of 16 lb/gal could E>30 Mud systems easy to remove
affect log performance due to static slurry losses after
placement. We decided to decrease the slurry density to For the case in study, the 60-minute gel strength was
12 lb/gal, but cement results continued to be similar. We then
decided to continue using conventional cements from measured at 3 rpm to obtain mud shear stress (τy) which was
16 lb/gal, but added fibers to the slurry to help prevent losses. recorded as 14 lb/100 ft2. From the erodability relation, we
However, this change alone could not offer the required obtained E=42.9, which implies a mud system easy to remove
results. Cement practices and slurry design were reevaluated; and desirable (see Table 2).
some modifications implemented were: (1) increase slurry To find the required spacer rheology to erode mud in the
excess, (2) design centralization for maximum standoff, annulus, an erodability study was performed using a low-cost
(3) sand blast the pipe to increase cement adherence to the erodability cell which considers similitude from annular
tubular, (4) maximize conditioning and displacement rate, geometry, fluids velocity, and formation permeability (see
(5) use a thixotropic slurry, (6) use low Young’s modulus Fig. 2 for erodability cell description). Geometrical similitude
slurries, and (7) use expansion additives. Cement log between formation and casing for the experiment was made
evaluation has been closely followed, using impedance cement using a man-made core with a defined permeability and pore
values from ultrasonic lab devices.
SPE 107701 3

size and an inner cylindrical hollow metal device in order to For the purpose of this work, cement sheath was evaluated
be comparable with a 8.5-in. hole size and a 7-in. liner. to find its mechanical properties, using the following tests (see
First, fluid circulated through the erodability cell was mud Fig. 5):
under HTHP conditions and at a pump rate equivalent to field • Unconfined- and confined-compression testing
jobs. Enough time was allowed to construct a cake; then the • Brazilian test
core was removed, the remaining mud volume in the chamber
was measured, and the weight from core plus cake and filtrate Unconfined- and Confined-Compression Testing
was recorded. First approximation indicates that 9.5% of total A cement cylindrical core was axially loaded to failure with
mud volume was left as gelled-mud/cake and filtrate in the no confinement and with 994-psi confining pressure to find its
core and remaining mud was displaced by circulation (see compressive strength. Axial and lateral deformation was
Fig. 3). monitored to find the Young’s modulus and Poisson’s ratio as
After quantifying the amount of remaining gelled- shown in Figs. 6–7 and Table 3.
mud/cake in the core sample, the goal was to displace it from A plot from confining stress versus compressive strength
the core. For this purpose, a heavy-weight spacer was passed was drawn to find friction angle and cohesion (see Fig. 8). The
through the core annulus for 10 minutes, followed by a graph slope is defined as μ and the intersection with Y axis as
preflush train for 1 minute, and then by a heavy-weight spacer C0. Using μ and C0, combined with the following equations
for 10 minutes contact time. The test was continued until a (Eqs. 2 and 3) reported by Zoback et al.,2 the friction angle
spacer rheology for maximum displacement efficiency without (φ) and cohesion factor (S0) were found to be 20.9 degrees and
affecting the hierarchic modeling of fluids was obtained. The 879 psi.
final core weight showed displacement of 98.9% of total mud
in the annulus (see Fig. 4).
(μ 2
+1 + μ)
(μ +1 − μ)
An adjusted-rheology, heavy-weight spacer was used ϕ≈ 2
(Eq. 2)
ahead of the cement slurry. Results were good; however, for
outstanding results, especially in washed-out sections, the use
of a preflush could yield improvement in mud displacement S0 ≈
C0 (Eq. 3)
where a conventional liner could be replaced by expandable 2 * ( μ2 +1 + μ)
liner technology.
The erodability cell can give a guide on how to design Where:
spacer rheology for optimum mud displacement and give an μ = Slope from compressive strength vs. confining stress.
assessment of the optimum preflush/spacer train to obtain the C0 = Unconfined stress value, psi.
highest shear stress achievable for a particular job type. Φ = Friction angle, degrees.
S0 = Cohesion factor, psi.
FEA Analysis
Conventional cement-job designs normally do not include an
Brazilian Test
evaluation from drilling, completion, and production event The cylindrical core is cut perpendicular to its axis. Cut length
loads during the well active life. However, sometimes single is approximately 2 to 2.5 times the core diameter. The sample
or combined events have a dramatic effect on the performance is set on a rest, with its axis parallel to the rest surface and a
of a conventional cement sheath and can damage it. Damage compressional load is applied until failure as shown in Fig. 9.
may allow undesirable production fluids communication or Tensile strength is defined according Eq. 4, and results
early water breakthrough. obtained as shown in Table 4:
FEA software has been used to determine if actual cement
designs oriented to production strings are suitable for long-
term use. Mechanical interaction between casing, cement and σ ≈−
(2 P ) (Eq. 4)
rock are evaluated applying different loads during different ΠLD
stages from well life. FEA-predicted failure modes can be
Where:
evaluated as:
P = Compressional load, lbf.
• Debonding from casing. L = Core length, in.
• Debonding from formation. D = Core diameter, in.
• Radial cracks. σ = Tensile strength, psi.
• Cement plastic deformation.
Expansion properties of cement were adjusted to have no
Mechanical properties from the casing, rock, and cement shrinkage effect; care was taken to not exceed the required
sheath are the main input for FEA simulation. Normally the amount of expansive additives, which could impose a
casing provider has the mechanical properties for steel tensional failure over the cement sheath.
available. Rock properties come from field geomechanical FEA analysis for Well A (described in Table 5) was run
studies and sonic logs or from core testing. Cement sheath for two cases: conventional cement (Case A) and a lower
properties are obtained from cement sheath core testing in a Young’s modulus cement (Case B). Description of mechanical
rock mechanics lab. properties of the casing formation and both systems can be
4 SPE 107701

found in Table 6. Applied production loads were 1,000 psi simulated data were completely different (see Fig. 17),
drawdown differential pressure, 3,500 psi to test casing demonstrating that a cement channel was not likely due to the
integrity, 0.9 lb/gal from differential pressure due to a fluids mismatch. These conformance simulations indicate that the
change while completing the well, and a shut-in period of 48 cement sheath was providing a complete seal around the
hours was taken for logging purposes. annulus.
As shown in Table 7 and Figs. 10–12, Case A cement
showed both formation debonding and radial cracks, while the Conclusions
Case B cement sheath remained intact for all loads. Case A • Score-card measurement is a good tool to compare
failures are attributed to lack of expansive additives and high performance of different jobs, and obtain
Young’s modulus cement. It was observed that use of low improvement from job to job. However, this method
Young’s modulus cement system can help improve the cement does not provide for optimization and maximization
sheath performance under production loads such as of reservoir production.
drawdown, increasing the remaining capacity of the cement • Spacer efficiency studies give the opportunity to
sheath. Expansion additives help prevent cement sheath obtain the required spacer rheology, spacer/preflush
debonding from the formation. When combined with low combination, volume, and pumping rate to obtain
Young’s modulus slurries, the additives can offer excellent efficient mud removal.
properties to help avoid tensile failures, thus helping prevent • An integrated tool such as FEA can help provide
radial-crack formation in the cement sheath.3 more accurate evaluation of diverse drilling,
After implementing all improvements from spacer completion, and production loads during the well life,
efficiency studies, FEA analysis, and water production and the influence on the cement sheath behavior.
followup, results have been clearly better than in the past (see When combined with a good mechanical study of
Fig. 13). cement properties, the FEA study gives more
accurate results.
Conformance Simulation • Use of a conformance reservoir simulator can help
In some cases where a good cement evaluation log is obtained, compare actual production with simulated data and
but the water production trend is more than expected and determine if a channel could be left behind in the
cement integrity is doubted, it is important to run a cement sheath, or if a complete isolation was
conformance simulator to see if water production is a obtained from the cementing job.
consequence of a cement sheath channel or an intrinsic • Well positioning in the reservoir and distance to oil-
characteristic from fluids movement into the reservoir. water contact (OWC) is critical when evaluating
To predict production for a well, a three-dimensional, production histories. Sometimes when the OWC zone
three-phase, and non-isothermal-coupled reservoir/wellbore is very close to the productive interval, the initial
simulator that numerically solves the partial equations for water-cut is higher than in wells that are more distant
multidimensional fluid and heat flow through porous medium from the OWC zone.
and the wellbore was used.4
Methodology was to construct a layered model that References
defined porosity, permeability, and water saturation by zone 1. Ravi, K.M., Beirute, R.M., Covington, R.L.: “Erodability of
(see Fig. 14), based on PVT parameters, relative permeability Partially Dehydrated Gelled Drilling Fluid and Filter Cake,”
curves, production data, and constraints. This information was paper SPE 24571 presented at the 1992 SPE Annual Technical
applied to the conformance simulator to obtain a prediction of Conference and Exhibition, Washington, D.C., 4–7 October.
oil, gas, and water production. Production was compared with 2. http://pangea.stanford.edu/~jack/GP170/MarkZoback.pdf.
simulated data to determine whether communication through 3. Bosma, M., Ravi, K., van Driel, W., Schreppers, G.J.: “Design
the cement sheath could be an issue. Using the 2-D viewer Approach to Sealant Selection for the Life of the Well,” paper
SPE 56536 presented at the 1999 SPE Annual Technical
from the model, water saturation vs. depth could be visualized Conference and Exhibition, Houston, Texas, 3–6 October.
(Fig. 15). As shown in Fig. 16, there was good correlation 4. Ansah, J., Soliman, M.Y., Ali, S., Moreno, C., Jorquera, R., and
between the simulated data and the well production trend, Warren, J., “Optimization of Conformance Decisions Using a
demonstrating that channeling through the cement sheath was New Well-Intervention Simulator,” paper SPE 99697 presented
not likely. To confirm this, the conformance simulation was at the 2006 SPE/DOE Symposium on Improved Oil Recovery,
run again, with an extra constraint that defined a small channel Tulsa, Oklahoma, 22–26 April.
in the cement using high porosity and permeability values in a
small grid near the wellbore to simulate a cement channel
behind the pipe. In this simulation, the production trend and
SPE 107701 5

Table 1—Scorecard for the Yuralpa Field

Ideal Actual Actual

Parameters Ideal Measurement Score Measurement Score Comments
Well Parameters and Design
Optimum caliper condition to obtain an efficient Yes 10 10 The caliper showed hole size average 8.91 in.
mud displacement from annular (review zones as
washouts, breakouts and undergage zones).

Optimum clearance annular hole. 1 to 1.5 in. 10 10

The standoff of the centralization program is Yes 10 10 Used 27 centralizers, 3 per joint below WOC.
superior to 70% in production zones and overlap.

Use of centralizers with high deformation and Yes 10 0 Rigid Centralizers

impact resistance.
Gas flow potential (GFP) from the production Yes 10 10 GFP = 1.41
zone known and the slurry designed according to
this value.
Production slurry designed using well life. Yes 10 10
Using expansive additives to compensate the Yes 10 10
cement shrinking in order to avoid microannulus.
There are no injection wells near the actual well. No 10 10

A PRIZM analysis was done to evaluate the Yes 5 5 Calculed average hole size 8.911 in.
openhole logs.
Use a temperature software to evaluate a well Yes 5 0
temperature profile.
ECD gap between fracture and pore gradient with Yes 5 5
a 0.5 lb/gal tolerance.
Drilling Fluid Parameters
Mud type Polymer of low content 10 10 MBT 5 lbs/bbl
of solids
YP o Yz at BHCT less than 15 lb/100 sqft. Yes 5 3 Yp = 22, PV = 20; MWde 8.9 lb/gal.
Flat gels profile 10s/10m and 30 m at the Yes 5 5 Geles: 5/7/9
cementing moment (e.g. 7/9/12)
Mud filtrate API<12 and HPHT<15 Yes 5 5 API Filtrate 5.6
Cementing Fluids Parameters
Preflush Length 1000 ft minimum or 5 10 5 Any preflush was used (small volume).
* Any preflush should be avoid when the well has min contact time
risk of influence or instability, is better to replace
this volume with spacer.
Spacer height Minimum 1000 ft or 5 10 10 50 bbls spacer equivalent to 1200 ft.
*The spacer density should be 1 lb/gal over mud min contact time
weight
YP at BHCT from displacing fluid from 20% to Hierarchic rheologies 10 10 YP Mud 22, Yp Spacer 24, YP Slurry 37
30% higher than displaced fluid
YP Mud<YP Spacer<YP Lead<YP<Tail
Surfactant additives used in preflush and spacer. Yes 10 10 Used 0.46 gal/bbl Clay Seal, 0.4 gal/bbls de Sem-8 and Clean
Bore A
Chemical reaction, breaking mud cake using the Yes 10 10
preflush.
Pipe Running and Drilling Fluid Conditioning
Circulate the well in front of permeable zones Yes 5 5
while running the casing in the well.
Circulation lost / influx while running casing, mud No 10 10
conditioning.
Packing while running casing / circulation No 10 10
Circulation pressure vs Real P. are similar. Yes 5 5 WELLPLAN pressure 580 psi; real pressure 650 psi.
Was the well circulated at 100% of annular Yes 5 5 Circulated 120 minutes before setting the liner.
capacity before setting the liner hanger?
Was the well circulated at 100% of annular Yes 5 5 Circulated 60 minutes.
capacity after setting the liner hanger?
Maximum circulation rate before the cementing 8 to 10 bbl/min 5 5 Circulating at 10 bbl/min.
job?
Use of liner with expandable technology. Yes 5 0 Used conventional liner.
Job Execution
Job execution following the program. Yes 10 10 Bump the plug with 1400 psi.
Static time <5 min. Yes 5 5
Circulation lost / influx during the cementing No 10 10
and/or after doing the cementing job.
Packing while the cementing job? No 10 10
Pipe Reciprocation/Rotation while cementing No 10 0 Liner does not permit reciprocating the pipe before setting the
job? liner.
Float Equipment and Alternative Equipment Function
Use of floating shoe with eccentric nose. Yes 5 0
Bump plug Yes 5 5 Bump plug with 1400 psi
Shoe track length > 80 ft Yes 5 5 83 ft
Proper function of float equipment, retain Yes 5 5 Flowback 0.6 bbls
differential pressure.
Liner setting ok. Yes 10 10
Liner release without any problem. Yes 10 10
Cement Log Evaluation
The logging company used the same acoustic Yes 5 5 Z= 4.86
impedancy as the value obtained from the UCA
test for the cement evaluation.
Optimum results from cement log evaluation. Yes 20 20
Remedial cementing jobs needed? No 20 20
Total for the ideal score 350 Job Total 308
Yz = (2 x Fann 35 3 RPM Dial Reading) - Fann 35 6 RPM Dial Reading

Efficiency to apply the Best Practices: 88 %

6 SPE 107701

Table 2—Rheological Properties of Yuralpa Field Mud

Angular Velocity, rpm Fann Readings Fann Readings
at 20°C at 53°C
600 78 63
300 60 49
200 50 43
100 48 33
60 32 28
30 25 22
6 15 13
3 12 10
Gel at 3 rpm (10 seconds) 12 10
Gel at 3 rpm (10 minutes) 15 12
Gel at 3 rpm (60 minutes) - 14
Plastic Viscosity (cP) 35.2 29.1
2
Yield Point (lbf/100ft ) 28.5 23

Gel at 3 rpm and 60 minutes Erodability Observation

2
(lb/100 ft ) Factor
14 at 53°C 42.9 Mud system easy to remove

Table 3—Mechanical Properties for Different Samples on Yuralpa Slurry

Sample Confining Compressive Young’s
Pressure Strength Modulus
psi psi
1 0 2,819 1.13E+06
2 0 2,204 1.05E+06
3 994 3,812 8.08E+05
4 998 3,424 8.01E+05

Table 4—Brazilian Tensile Strength for Different Samples on Yuralpa Slurry

Sample Tensile
Strength
Psi
A 317
B 284
SPE 107701 7

Table 5—Well A: FEA Input Parameters

Parameter Units Value
Surface Temperature °F 80
Temperature Gradient °F/ft 1.4
Vertical Depth ft 7887
Hole Size in. 8.5
Casing OD in. 7
Casing ID in. 6.276
Drilling Fluid Density lb/gal 9.3
Cement Slurry Density lb/gal 16
Compl. Fluid Density lb/gal 8.4
Vertical Stress Gradient psi/ft 1
Maximum Horizontal Stress Ratio — 1.35
Minimum Horizontal Stress Ratio — 0.95
Pore Pressure psi/ft 0.43

Table 6—Well B: Casing, Rock, and Cement FEA Input Parameters

Property Units Casing Case A Case B Rock
Young’s Modulus Mpsi 29.02 1.2043 0.843 0.8706
Poisson Ratio — 0.3 0.1 0.108 0.2
Volumetric Specific Heat btu/ft3·°F 59.04 59.04 57.56 30.996
Cohesion psi NA 3132.8 879 —
Friction Angle degrees NA 17.1 20.9 —
Cement Tensile Strength psi NA 217.557 261 NA
Cement Hydration Volume Change % NA -4 0.5 NA
Casing Weight/Unit Length lb/ft 26 NA NA NA
Casing Yield Strength Mpsi 6.04E+05 NA NA NA

Table 7—Well A: FEA Results

Case A—Conventional Cement, Case B—Low Young’s Modulus Cement
8 SPE 107701

Fig. 1—Well B CBL and openhole log, conventional neat cement

at 15.8 lb/gal.
SPE 107701 9

Remaining
Remaining Heavy w eigth
Cake spacer
Fluid
Output
Fluid
input
A B

Fig. 4— Mud displacement stages: (A) Using 10 min. contact time

from a heavy weight spacer; (B) using 1 min. contact time from a
preflush and 10 min. contact time from a heavy weight spacer.

Man
made
core

Fig. 2—Erodability cell diagram.

Unconfined Confined 994 psi

Mud

Brazilian

Fig. 5—Cement core samples for unconfined, confined, and

Cake Brazilian stress tests.
thickness
Mud cake over reference
manmade core

9392-51-1-1-u

3000

2500
Liner
7/8-in.
2000
Stress

1500

Mud cake formed at 500 psi and 128°F over manmade core.
1000

Fig. 3—Mud cake deposition over the filtrate medium.

500

0
-0.20% -0.10% 0.00% 0.10% 0.20% 0.30% 0.40%
Strain Comp Strength 2819
Youngs Modulus 1.13E+6
Astrain in/in Rad Strain in/in Astrain Rad Strain Poisson's Ratio 0.20

Fig. 6—Strain vs. stress plot for Sample No. 1 (unconfined).

10 SPE 107701

9392-51-1-6-1k

4000

3500

3000

2500
Stress

2000

1500

1000

500

0
-1.00% -0.50% 0.00% 0.50% 1.00% 1.50% 2.00% 2.50% 3.00% 3.50%
Strain Comp Strength 3424
Youngs Modulus 8.01E+5
Poisson's Ratio 0.09
Astrain in/in Rad Strain in/in Astrain Rad Strain

Fig. 7—Strain vs. stress plot for Sample No. 4 (confined 994 psi).

S1 vs S3

10,000
8,000
Sigma 1

6,000
4,000
2,000
0
0 200 400 600 800 1000 1200
Sigma 3

Fig. 8—Mohr’s plot for friction angle estimation.

Fig. 10—Cement remaining capacity for different well life events.

Fig. 9—Application of Brazilian test.

SPE 107701 11

Fig. 11—Case A, cement debonding at formation.

Fig. 12—Case A, cement radial cracks.

12 SPE 107701

Fig. 13—Well A CBL log; low Young’s modulus cement at

16 lb/gal.
SPE 107701 13

Fig. 14—Petrophysical model from Well C.

Fig. 15—Water saturation output from conformance simulator for Well C (incipient coning).
14 SPE 107701

Fig. 16—Strong comparison between original production data and simulated data for coning.

Fig. 17—Comparison of original production and simulated data for channeling through cement sheath. Lack of correlation demonstrates that
channeling is not likely.