IADC/SPE 115201

Achieving Long-Term Zonal Isolation in Heavy-Oil Steam Injection Wells:

Copyright 2008, IADC/SPE Asia Pacific Drilling Technology Conference and Exhibition

This paper was prepared for presentation at the IADC/SPE Asia Pacific Drilling Technology Conference and Exhibition held in Jakarta, Indonesia, 25–27 August 2008.

This paper was selected for presentation by an IADC/SPE program committee following review of information contained in an abstract submitted by the author(s). Contents of the paper have not
been reviewed by the International Association of Drilling Contractors or the Society of Petroleum Engineers and are subject to correction by the author(s). The material does not necessarily reflect
any position of the International Association of Drilling Contractors or the Society of Petroleum Engineers, its officers, or members. Electronic reproduction, distribution, or storage of any part of this
paper without the written consent of the International Association of Drilling Contractors or the Society of Petroleum Engineers is 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 IADC/SPE copyright.

Abstract
The world’s ever increasing demand for energy has forced operators to go after more difficult reservoirs. These challenging
reservoirs place a greater demand on the cement sheath’s ability to maintain a long term seal. To form an initially successful
annular cement seal, effective mud removal is required. If effective mud removal can be achieved, the resultant cement sheath
should be elastic and resilient enough to survive future downhole thermal stresses. Cyclic steam injection presents some of the
harshest conditions under which a cement sheath’s resilience and elasticity can be put to the test.
Derived from the industry’s latest innovative science in sheath durability, a third generation cement design has been
developed for a California heavy oil field. The paper will discuss the evolution from the initial wells to the improvements
made for the second generation and the extensive testing and qualification process that went into the third generation slurry
design. Best practices and lessons learned are covered.

Background
As in-place oil viscosity increases, its capability to flow to the well bore and up the production string decreases. The most
common method used to change the flow properties of heavy oil is to introduce steam directly into the formation. As the oil
absorbs heat from the steam, the viscosity decreases sufficiently to allow flow. In addition, the steam can increase or maintain
formation pressure further increasing productivity. Pumping super-heated steam down the casing places extreme stress loads
on the cement sheath. At the beginning of an injection phase the well is rapidly heated up, which causes the casing to expand.
During periods when the steam injection is halted the well can cool, which causes the pipe to contract. This contraction places
additional stresses the cement sheath. When injection begins again, the stress cycle is repeated. If the cement sheath is not
sufficiently elastic and resilient, the annular seal will fail during one of these stress cycles.
The Belridge fields are located 40 miles northwest of Bakersfield, California. The South Belridge field covers 2,400
productive acres. Two major hydrocarbon reservoirs are produced independently of one another. These are the Diatomite and
the overlying Tulare heavy oil sands. The Tulare has been developed with a mature steam drive. The Diatomite (specifically
the light oil Opal-A) is the reservoir of interest for this study.
The Diatomite reservoir produces 22—32 API gravity oil from depths of about 600 ft to 3,000 ft, depending on structural
position. The Diatomite is a biogenic silicious deposit consisting of the shells of diatoms with varying amounts of detrital
material (principally clay and sand) so that individual depositional cycles are identifiable. In the upper reservoir intervals,
diatoms are preserved as Opal-A, which is an amorphous, colloidal form of silica. Porosity ranges from 45—70% and
permeability is 0.1—4 md. The Opal-A is approximately 1,000 ft thick. With increasing depth and temperature, the dominant
mineralogic phase changes from Opal-A to porcelanite Opal-CT. The associated reduction in porosity (<#45%) makes the
Diatomite reservoir significantly less productive below the Opal-CT transition.
A Diatomite Opal-A waterflood has been in operation since the late 1980s to mitigate reservoir compaction and improve
hydrocarbon recovery. Vertical wells that which are hydraulically fracture stimulated in the stages over the gross interval are
used throughout most of the field.

History
Prior to the year 2000, wells were cemented with conventional cement in conjunction with a universal fluid (UF-Slag) mud
system that was pumped ahead of the cement. This was an effort to pump a heavier density conventional cement to combat the
2 SPE 115201

various stresses encountered in the wells. The theory was, if cement to surface was not achieved, the universal fluid would
harden and provide adequate zonal isolation. Foam-Nitrified cement was used from 2001 to present.
However, communication between the injection intervals continued to occur. To address the ongoing communication
concerns, three types of quad steam-injection wells were designed. The first type of quad steam-injection well, a Type “E”
design, included the use of four strings of insulated 2 3/8-in. x 1 ¼-inch tubing (2 3/8-in. EUE x 1 ¼-in. coiled tubing),
cemented in a single wellbore with a foamed conventional cement. This design reduced the temperature on the cement sheath
to approximately 350F. The average injection temperatures ranged from 500–590°F. The reduction of temperature on the
cement was expected to increase the life of the cement sheath. The induced stresses continued to cause a 50—75% failure rate
in these wells after 1—3 years.
The second type of stem-injection well, a Type “J” design, is basically configured the same as the Type “E” design with
the exception of the 1¼-in. insulated tubing. To combat possible degradation of the cement sheath, special non-Portland high-
temperature cement, ThermaLock, was implemented. This special high-temperature cement allowed the steam to be injected
directly down the 2 3/8-in. tubing, saving the operator the cost of the insulated tubing. The cement sheath was to be exposed to
the high temperature ranges of 500—590º F without any adverse effect during testing.
The Type “J” designs wells worked well at first but then started failing after one year of service. Further evaluation of the
cement design and the well configurations were required. Analysis showed that a foamed version of conventional cement
added to silica could yield a superior solution. For this third version, the failure rate was reduced to 25% after 2 years. In a
quest to create an elastic, resilient cement sheath capable of maintaining the annular seal, further computer modeling and
laboratory testing were undertaken.
The third type, a dual-string injection well, was developed with two wellbores 20 ft apart from each other. The two dual-
string injection wells would be drilled and completed separately, therefore the injection zones were separated vertically by
~300 ft within a wellbore, instead of the ~120 ft used in the quad steam-injection wells. Additionally, 3 ½-in. × 2 3/8-in.
insulated tubing within this 300-ft blank interval was used to reduce cement temperatures and stresses. The increased elasticity
from foaming the cement+silica formulation helped the sheath better handle the steam-injection stress environment and
resulted in an improvement of 60% during the first year of the wells being put on line. For this third version, the failure rate
was reduced to 25% after 2 years. The quest to create an improved elastic, resilient cement sheath even better capable of
maintaining the annular seal is the basis for this paper and initiated further computer modeling and laboratory testing were
undertaken.

Computer Model
Finite element analysis software (FEA) can be used to model how future well events, may affect the stress imparted into the
cement. With an idea of the stress profile from this analysis, an effort to design slurry that could yield a set sheath capable of
surviving the calculated levels of stress can be undertaken. Figs. 1 and 2 show how the recommended system from the current
round of testing fairs compared to the best alternatives. In Fig. 1 the recommended slurry has over 40% remaining capacity
during injection, whereas the best alternatives have less than 20%. In Fig. 2 the slurry responses’ under an even higher stress
portion of the wellbore are shown. Here the recommended slurry still has greater than 30% remaining capacity while the other
slurries show a high likelihood of failure after steam injection commences.

Laboratory Testing
Before testing mechanical properties of any potential cement sheath can begin, base slurry testing must be completed. Initially
three things must occur: 1) the slurry must set up in a reasonable time, 2) the slurry must be stable, and 3) the slurry must be
mixable under typical field conditions.
High compressive-strength values are not normally a target when trying to design elastic stress-absorbing cement slurries.
This is because as Young’s Modulus (YM) decreases, the compressive strength typically decreases as well. However,
reasonable initial set times are still required. What is deemed a reasonable initial set time will vary from job to job and depend
primarily on how long the post-job rig operations are expected to take to prepare for the next scheduled activity. This time
period can be what dictates “reasonable” initial set.
Slurry stability is especially important when elastomers, fibers, or other additives that are substantially heavier or lighter
than the resulting cement slurry are included. A BP or modified BP settling test is one of the best ways to confirm whether or
not the slurry is stable. Ideally a column of cement is cured of sufficient length to be easily cut it into five samples. After
cutting the sample into sections, the density of each piece will be determined. If the density varies more than 1 or 2% from top
to bottom, the slurry is redesigned for increased stability.
Testing for mixability is more subjective. If all of the dry components do not easily slurry in a waring blend, chances are it
will be difficult to mix under field conditions. Determination of pore volume and yield point can also be useful to ensure field
mixability. On small number jobs batch mixing can be utilized to aid in the mixing process.
With confirmation that the base slurry properties are useable, mechanical testing can begin. Initially, large enough
quantities of samples are cured under downhole conditions to allow the collection of statically meaningful data. Curing time
will vary from a week or 10 days up to several months, depending on conditions being simulated. A simultaneous run in a
continuous compressive strength testing device can provide useful insight when steady state conditions are reached. Measuring
mechanical properties on a partially cured cement sample would be meaningless when looking for long term properties.
SPE 115201 3

Young’s modulus, Poisson’s ratio, tensile strength, and anelastic strain appear to be the most important measurements to help
determine a cement sheath’s ability to maintain an, effective annular in the presence of potentially damaging downhole
stresses.
YM and Poisson’s ratio can be calculated from carefully and closely monitored compressive strength tests. At the lower
stress levels the stress-strain curve is linear. YM is calculated from the slope of the linear portion of this curve. The bolded red
section of Fig. 3 represents the section of the curve used to determine the slope. YM can be used to represent the elasticity or
conversely the brittleness of the set cement sheath. As the YM increases, the slurry exhibits a more brittle behavior. Values of
2 to 6 MMpsi are common for cement. As the YM decreases, the set sheath properties become more elastic and less brittle.
With effort, values as low as 0.3 to 0.6 MMpsi can be achieved. A curve of very elastic cement can be observed in Fig. 4. In
this example the recommended system’s YM has been decreased all the way down to 0.1 MMpsi. Notice how the curve in
Fig. 4 was able to absorb over 1.2% of axial and 0.4% of radial strain without failing and in Fig. 3 which under most
comparisons would be considered a very elastic cementing material, absorbed only 0.35 % and 0.05% strain respectively.
Again, lower YM values are considered favorable with respect to a cement sheath’s capability to maintain an annular seal.
Tensile strength is most typically determined in one of two ways: 1) Dog-bone testing or 2) with a Brazilian test. In the
dog-bone test method, a dog-bone shaped mold is used to cure a sample, and then in the actual test, the ends are grabbed and
pulled apart (Fig. 5). The advantage of this method is that the measurement is a direct measure of tensile strength. The
disadvantage is that, if the sample does not fall perpendicular to the axis, the results will have to be mathematically corrected.
The Brazilian test is an indirect measure where a cylindrical wafer is crushed on its side (Fig. 6). Mathematics are required to
correlate the crush load to the tensile strength, but the measurement is more repeatable. Tensile strength for oilwell cements is
typically 10% of the compressive strength but can be increased to a higher percentage with fibers and other design changes.
Anelastic strain is a measure of the material’s likelihood to deform under repeat stress cycles. Fig. 7 illustrates elastic
cement’s capability to withstand cyclic load while exhibiting minimum deformation.
Possion’s ratio is the numerical value reflecting the relationship between longitudinal and lateral strain. To calculate this
value, a load is placed on a cylinder creating an axial compression and a radial expansion. The resulting lateral strain is
divided by the longitudinal strain.

Results
Normally, foaming cement slurry increases the elasticity of the set cement sheath and the higher strength microsphere designs
are found to be less elastic, but that was not the case with this particular study (Table 1). It is believed that addition of
elastomers and fibers to the base slurry is the reason for this behavior in spite of the fact that with conventional base slurries
the addition of nitrogen commonly results in a substantial increase in set sheath elasticity. However, the results of this study
seem to indicate that the addition of elastomers and fibers may be even more beneficial than nitrogen in decreasing YM.
Further research will be done in this area. Long-term results of the delivered zonal isolation with this new slurry design will
only be known after several years of use in the annular space of these steam-injection wells.

Conclusions

1. Cement slurries designed to survive 550°F can be made to have substantially more elasticity than conventional
cement systems.
2. Cement slurries designed to survive 550°F can be function at densities at least as low as 11 lb/gal.
3. 11-lb/gal cement slurries designed to survive 550°F can use either nitrogen or microsphere technology to reduce
density for use across weak or depleted zones.
4. At least under certain conditions, when used in combination with elastomers and fibers, microsphere designs can
equal or exceed the elasticity provided with foamed designs.

Table 1—Young’s Modulus of Recommend Design and Alternatives

YM
Slurry Type
(psi)
Elastomers, Fibers, and Microspheres 1.31E5
Elastomers, Fibers, and Nitrogen 3.49E5
Conventional Foamed Cement 4.15E5
4 SPE1115201

Fig. 1—Scenario 1 comparison of optimal design and alternatives.

Fig. 2—Second scenario comparison of optimal design and alternatives.

2-u

1000

900

800

700

600
Stress

500

400

300

200

100

0
-0.10% -0.05% 0.00% 0.05% 0.10% 0.15% 0.20% 0.25% 0.30% 0.35% 0.40%
Strain Comp Strength 850
Youngs Modulus 3.43E+5
Astrain in/in Rad Strain in/in Astrain Rad Strain Poisson's Ratio 0.20

Fig. 3—Slope of stress-strain curve highlighted for use in Young’s Modulus calculation.

1-u

450

400

350

300

250
Stress

200

150

100

50

0
-0.60% -0.40% -0.20% 0.00% 0.20% 0.40% 0.60% 0.80% 1.00% 1.20% 1.40% 1.60%
Strain Comp Strength 288 Youngs Modulus 1.17E+5
Poisson's Ratio 0.11
Astrain in/in Rad Strain in/in Astrain Rad Strain

Fig. 4—Stress-strain curve from recommended cementing system.

Fig. 5—Dogbone direct tensile strength test.

Fig. 6—Used sample from Brazilian indirect tensile strength test.

Slurry One Sample Eight

600

500

400
Stress

300

200

100

0
-0.04% -0.02% 0.00% 0.02% 0.04% 0.06% 0.08% 0.10% 0.12% 0.14% 0.16%
Strain

Astrain in/in Rad Strain in/in Astrain Rad Strain

Fig. 7—Cyclic loading and the minimal deformation from an elastic cement.3