You are on page 1of 12

IADC/SPE-170480-MS

Challenges of Casing Design in Geothermal Wells


Angelito Torres, Chevron Geothermal Services Company

Copyright 2014, IADC/SPE Asia Pacific Drilling Technology Conference

This paper was prepared for presentation at the IADC/SPE Asia Pacific Drilling Technology Conference held in Bangkok, Thailand, 25–27 August 2014.

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
Casing and tubing in either oil and gas or geothermal wells are designed to withstand pressure loads (burst
and collapse) and axial tension - the typical parameters that should be considered in any casing design.
However, in geothermal applications, temperature is a major consideration in casing design because of
thermal stresses generated by the high temperature geothermal fluids. Elevated temperature reduces the
yield strength of the casing which essentially reduces the burst rating, collapse resistance and tensile
strength of the material.
High temperature will be experienced by the well once it is on production. The cemented casing strings
will be subjected to significant cyclical axial stress caused by changes in temperature due to production
of high temperature geothermal fluids that can then be followed by the injection of cold water for well kill
operations to perform wellbore maintenance. During production, the cemented casing will enter a
compressive state where plastic deformation occurs since the casing is not allowed to expand. When cold
water is injected to kill the well, axial tensile stresses are generated. This paper will explore this thermally
induced cyclical loading in geothermal wells, and discuss how to account for these loads in the design
process.
Another consideration in casing design is the selection of threaded connections that can maintain
strength and seals at elevated temperatures. It is also important to evaluate the potential of trapping water
in the annulus between two casing strings during cementing operations. When the well is put on
production the water expands due to the large increase in temperature, creating a significant amount of
pressure between the two casing strings, which can lead to a collapse failure of the inner string or a burst
failure of the outer string. Examples of these failure mechanisms will be discussed in detail.

Introduction
The major consideration that complicates well design in geothermal wells is temperature. Pressure and
axial stresses are the major concerns in casing design while thermal stresses are often neglected in oil and
gas wells. This is acceptable for most oil and gas wells since the wellbore temperature is not high enough
to cause significant yield strength reduction in the steel. Even though API does not publish data regarding
the reduction of yield strength at elevated temperatures, it is known that material properties of steel vary
with temperature. Aside from the yield strength reduction consideration, the two additional design criteria
2 IADC/SPE-170480-MS

that need to be included are Thermally Induced Axial Stress and Plastic Deformation Design in order to
fully account for temperature effects encountered in geothermal wells.
Selection of the casing connection thread is also critical in geothermal well design. The connection
should have the strength to withstand the compressive and tensile stresses generated when exposed to high
temperature. Use of special clearance connections, which usually reduce the tensile capacity of the
connection, should also be considered in the design.
Another area of concern is the tieback casing string which has the potential of trapping water in the
annulus due to poor cementing practices. Small amounts of water can be trapped in void spaces in casing
centralizers. When the well is put on production, the water expands due to the increase in temperature.
Water expansion can create a significant amount of pressure between the two casing strings. With the
outer casing being reinforced by cement and formation, the collapse resistance of the tieback string can
easily be exceeded causing collapse. It is important to provide sufficient annular gap between casings and
ensure good cementing practices, which include a thorough sweeping of trapped water especially across
the centralizer, to avoid failure.

Additional Well Design Considerations


There are different parameters in geothermal casing design that need special considerations due to
elevated temperature. Casing yield strength can be significantly reduced with high temperature and at the
same time an excessive axial stress is generated. Pipe body and connection should have the same
compression rating to avoid joint failure. Due to those conditions, geothermal casing design has additional
two design criteria, Thermally Induced Axial Stress and Plastic Deformation Design (Maximum Tensile
Load).
Design Safety Factors
Geothermal casing design follows safety factors suggested by API for burst, collapse and tensile loads.
Recommended safety factors for thermally induced axial stress and plastic deformation are shown below.

Thermally Induced Stress Safety Factor 1.0


Plastic Deformation (Maximum Tensile Load Safety Factor) 2.0

Casing Grade Selection


The weight and grade of casing to be used for each casing string will be dependent on the burst rating,
collapse resistance, tensile strength, thermal stress, and corrosion resistance requirements dictated by the
design criteria, geologic environment and operating conditions. Cemented casing should be designed to
handle stress due to thermal loads.
● K-55: This casing is a low strength carbon steel which is resistant to sulfide stress cracking (SSC)
at all operating temperatures and is more favorable for field welding than higher strength casing.
This is recommended for surface casing and perforated production liners.
● L-80: This casing is a restricted yield strength casing with a maximum requirement on the hardness
and heat treatment specification providing resistance to SSC allowing L-80 material to be used in
H2S environments at all operating temperatures.
● T-95. This type of casing is recommended for tieback casing in wells where the anticipated
flowing wellhead temperatures exceed 600F. When a T-95 casing string is to be attached to the
final wellhead, it is recommended to use a single joint of weldable alloy casing at the surface for
improved well integrity.
IADC/SPE-170480-MS 3

Table 1—Casing Yield Strength Degradation Factor (Grant PrideCo-TCA)


Temperature Standard API Casing Grades

Degrees F K-55 N-80 L-80 C-90 C-95 T-95

300 0.875 0.875 0.875 0.925 0.875 0.925


400 0.830 0.830 0.830 0.890 0.830 0.890
500 0.780 0.780 0.780 0.860 0.780 0.860
600 0.725 0.725 0.725 0.825 0.725 0.825

Yield Strength Reduction of Casing at Elevated Temperatures


Material properties of steel vary with increased temperature. Strength degradation factors for common
grades of casing are summarized in Table 1. This is based on test data presented in the technical
memorandum, “Casing Yield Strength Degradation”, provided by Grant PrideCo – TCA. Various casing
manufacturers have their own tables for yield strength degradation due to temperature which the designer
may opt to use.
In general, the burst rating, collapse resistance and tensile strength of casing are linearly proportional
to the yield strength of the material. Consequently, if there is 22% reduction of yield strength of K-55
tieback casing at 500°F, there will be a corresponding 22% reduction in the burst rating, collapse
resistance and tensile strength. Note that for certain conditions where the ratio of the pipe diameter to the
wall thickness is high, the collapse rating is no longer linearly proportional to the yield strength.
Thermally Induced Axial Stress
Once the casing is cemented, thermal expansion resulting from elevated temperature is restricted; thereby
creating axial compressive stress within the casing string. The maximum temperature change of the well
occurs during throttled/killing operations. In geothermal wells, wellbore temperatures can reach as high
as about 600°F when throttled and as low as 80°F when killed. With the high temperature change, material
properties of casing should be investigated for stress-temperature relationship to prevent casing failure.
Axial strain (␧z) is introduced when casing is subjected to a change in temperature, and follows this
relationship:

␧z ⫽ ␣T⌬T Where ␣T ⫽ coefficient of thermal expansion

When the casing is cemented in place, the induced strain is converted to stress, using this relationship:

␴z ⫽ ⫺E␣T⌬T Where E ⫽ Young’s modulus of elasticity

The product of Young’s modulus of elasticity (E) and coefficient of thermal expansion (␣T) gives the
relationship of stress with change in temperature. E and ␣T are inherent to the material, and according to
“Applied Drilling Engineering”, casing is low carbon steel with a 0.30% carbon content. Table 2 gives the
instantaneous Young’s modulus of elasticity and coefficient of thermal expansion from 70°F to 800°F.
Figure 1 shows the relationship of E and ␣T with temperature. Solving for the average value of the
function E ⫻ ␣T (T),
4 IADC/SPE-170480-MS

Table 2—Young’s Modulus of Elasticity and Coefficient of Thermal Expansion for Low Carbon Steels (C < 0.30%) at Different Temperatures
(Table E8.1 and E8.2, Piping Handbook)
Temperature (°F) Young’s modulus of elasticity (psi) Coefficient of thermal expansion (°Fⴚ1)

70 29.99 6.46
200 29.77 7.20
300 29.33 7.71
400 28.85 8.15
500 28.31 8.53
600 27.65 8.86
700 26.84 9.12
800 26.03 9.33

Figure 1—The product of Young’s modulus of elasticity and coefficient of thermal expansion Vs Temperature for low carbon steels (C < 0.30%)

Using the plot in Figure 1, an average value of 203.19psi/°F is obtained. This value is very close to 200
psi/°F, thus it is safe to use 200psi/°F as the relationship between thermal stress and temperature for
restrained casing strings. The final equation for the change in axial stress,
IADC/SPE-170480-MS 5

Figure 2—Heat Conduction of Steam from the Pipe to the Formation, with Thermal Conductivities and Wall Thickness of Cement and Casing Strings

For example, if a tieback casing was cemented in place at 100°F and the well flowed at a maximum
of 550°F, the casing string will theoretically experience a thermally induced compressive stress of 90,000
psi.
Figure 2 shows thermal conductivities across metal and cement used in calculating the most likely
temperature experienced by each casing string during production. If the maximum thermally induced
stress exceeds the yield strength of the casing at the maximum anticipated temperature where the casing
is exposed, plastic deformation design must be evaluated.
Heat Conduction Formula:

Plastic Deformation
Since temperature loading is a limited strain process, the casing can be designed such that yielding the
casing in compression can limit the applied stress. The concept of plastic deformation design for casing
has many applications in geothermal well design.
For example, if tieback casing cemented at 100°F was subjected to flowing temperature of 550°F, the
theoretical thermally induced compressive stress on the casing would exceed the yield strength if L-80
casing was used. However, once the casing reaches the yield point, the compressive stress remains
constant while the temperature continues to increase effectively shortening the length of the casing. When
the casing is cooled to 80°F (quenching) after yielding, the casing will be shorter that its original length
and tensile stress will be generated as the temperature decreases.
Sample Calculation: Thermal cycling of the well from cementing temperature of 100°F to maximum
flowing temperature of 550°F (throttling) to a kill temperature of 80°F.
Temperature Cycle: 100°F - 550°F - 80°F
Casing Material: L-80 (80,000 psi yield strength)
Calculating the temperature at which L-80 casing starts to yield:
Tyield ⫽ 425.8°F
Yield strength at 425.8°F ⫽ 65,166 psi
Residual stress generated in the casing from cooling casing from 550°F to 80°F during well kill.
Temperature when axial stress is zero:

Residual Tensile Stress when temperature reduced to 80°F


6 IADC/SPE-170480-MS

Figure 3—Thermal Stress vs YS Deration

Figure 4 —Thermal Stress vs Temperature Plot for Plastic Deformation

In Figure 4, the casing temperature increases to 425.8°F at which point the L-80 material begins
yielding in compression, maintaining constant stress of 65,166 psi. As the casing is cooled from 550°F,
the axial compressive stress decreases linearly back to 0 psi at temperature of 224°F. As temperature
decreases to 80°F, a tensile stress of 28,800 psi was generated. Thermal induced axial stress is limited by
casing grade after yielding in compression and generates 28,800 psi tensile stress upon cooling. If the
casing connection has the capability to handle the maximum compressive stress, this design is acceptable.
Although the casing connections may resist a catastrophic failure, they are unlikely to remain gas tight
once the yield strength of the casing is exceeded. The casing connection should be tested according to the
full thermal cycle environment to which it will be exposed.
However, if the sample casing was subjected to a higher production temperature of 620°F with the
same assumptions as the example above, tensile failure of the casing may occur. Casing will yield in
compression at 65,166 psi upon reaching 425.8°F. As the temperature increases to 600°F, stress is
IADC/SPE-170480-MS 7

constant at 65,166 psi. As temperature decreases, stress will decrease to 0 psi at 294°F. Upon reaching
quenching temperature of 80°F, the tensile stress generated in the casing is 42,800 psi. This value is
approximately 54% of the yield strength of L-80 casing at ambient condition. This design would not meet
the required safety factor of 2.0 for plastic deformation design. Increasing the casing thickness does not
help in reducing thermal stress, however, increasing the casing steel grade does. It is generally acceptable
to use a simple consideration of temperature in calculating compressive yielding, but if the stress value
is close to yielding, a more accurate approach using triaxial stress is recommended.

Casing Coupling Connection


The standard API buttress coupling has been the most commonly used casing connection in geothermal
wells throughout the world. API buttress casing connections are inexpensive and readily available in all
casing sizes, weights and grades and are recommended for use where a gas tight seal is not required and
axial stresses are not excessive.
The primary exception to using buttress connections is the tieback production casing where a gas tight
seal is desired and high axial loads are likely to be present due to thermally induced stress. Gas tight
connections with a metal to metal seal and a positive torque shoulder are recommended for the tieback
casing. When selecting a connection for the production tieback string with a plastic deformation design,
it is important to check the compression rating of the coupling to determine if the coupling will maintain
a gas tight seal throughout the thermal cycles of the well.
Casing dope that is applied prior to making up a casing joint has a friction factor that can impact the
torque make-up of the joint. If the friction factor is known it should be conveyed to the casing running
crew so that appropriate corrections are made prior to commencing the operation. Dope should be
designed as well with the temperature exposure. Dope trapped within the threads can cause casing collapse
as it expands due to high temperature. Torque-turn equipment should always be employed when premium
connections are run to ensure proper make-up.

Common Casing Failures


Common casing problems observed in geothermal wells are connection failures and casing collapse. It is
very critical that casing grade and coupling are properly evaluated for the maximum anticipated
temperature and load scenarios throughout the life of the well. Also, slurry design suited for high
temperature and good cementing practices play significant roles in preventing casing integrity issues.
Other considerations that need special attention are casing sizing (annulus gap) and centralizer design.

Connection Failure
Connection failures are most likely due to connection yielding when subjected to high compressive stress
as the wellbore temperature increases to flowing condition. If the casing connection has a lower yield
strength than the body, use the connection compressive stress as the design limit. There are three common
casing connection failures observed in geothermal wells – pin compression deformation (compression of
the pin against coupling shoulder leading to deformation of the pin end, pin jump-in (compression of the
pin against the coupling shoulder leading the pin end to “jump-in” into the casing and with some of the
threads disengaged), and pin pull-out (occurs when a “jump-in” pin gets to be subjected to tension again,
the threads are no longer engaged, thus, a “pull out” effect happens exposing some of the threads). Shown
below are some downhole camera photos showing connection failure.
A. Pin Compression Deformation
8 IADC/SPE-170480-MS

B. Pin Jump-in (After Milling)

C. Pin Pull-out
IADC/SPE-170480-MS 9

Casing Collapse
The casing body collapse is most likely attributable to trapped free water in the annulus between two
casing strings. Casing collapse due to the expansion of trapped water in between casing strings has been
evaluated and verified through finite element analysis, computational fluid dynamics and actual specimen
sample testing. Trapped water between casing strings can be attributed to the development of viscous
fingering. Fingering develops when the bottom liquid has significantly higher viscosity than the top fluid.
Propagation of the finger that traps water is affected by pump rate, annular gap, exposure time to the
cement slurry during the pump job, and centralizer configuration. The first picture shows a collapsed
casing section while the second picture was after milling operations to reestablish full ID, exposing the
bow spring centralizer.

Studies Conducted to Understand Common Casing Failures


In order to understand the dynamics that contribute to casing failures, various studies and tests were
performed. Finite Element Analysis (FEA) was conducted to evaluate the reasons behind failures that
occurred in premium casing couplings. To simulate casing collapse caused by trapped water, an actual
specimen testing was conducted. A Computational Fluid Dynamics study was also conducted in order to
check the cement placement efficiency between casing strings with centralizer.

Casing Connection Failure Analysis


A Finite Element Analysis (FEA) for a premium connection with 100% compression rating relative to
pipe body was conducted. The simulation assumed that the connection was properly made-up and there
is perfect cement around the casing. For thermal cycle assumption, the casing was cemented in place at
80°F. The connection was then heated up to 500°F and cooled down to 80°F.
Figure 5 shows the stress developed in the casing connection after being exposed to 500°F. The
connections yielded when heated and the thermally induced compressive stress was transferred to the pipe
body around the coupling. The eventual cool down left residual stress on some threads near the pin area
but it did not show pull-out of the pin.

Casing Collapse Specimen Testing


An actual free water collapse test was conducted. This actual test entailed cementing a 10-3/4⬙ 45.5 L-80
casing inside 13-3/8⬙ 68ppf L-80 casing with intentionally trapped water in the annulus. The test specimen
was gradually heated up using ceramic heating elements. The casing collapsed in precisely the location
10 IADC/SPE-170480-MS

Figure 5—Stress Developed in casing Connection After Heat-up to 500°F

of water in the annulus. The full size of the collapse was attained when the bulge developed a slit allowing
the internal pressure to bleed off (see pictures below).

PICTURES OF 10-3/4” CASING PIPE (COLLAPSE AREA EXPOSED)

Study on Cement Placement Efficiency in the Annulus and Around Casing Centralizer
The displacement of water by viscous cement flowing in a concentric annular region containing a bow
type centralizer has been studied numerically. The study primarily examined the effects of annular gap
size and exposure time on the efficiency of the displacement process. Narrow annular gap widths have
been shown to be problematic in field examples where ultimate casing collapse has been attributed to poor
cement bonding adjacent to centralizers. These observations all occurred in wells having narrow annular
gaps and near the surface of the well where cement contact times were limited.
The result shows that the displacement of water by viscous cement is over two times more effective
in the larger annulus than in the smaller one. With the narrow annular gap, a significant amount of cement
remained behind the centralizer blades, slowing the propagation of the displacement interface. Whereas
with the wider annulus, the displacement interface remained relatively flat and little water was trapped by
IADC/SPE-170480-MS 11

the centralizer. In addition, a test was performed to determine the effect of a very slow flow rate - 2 bpm
on the displacement process. This simulation was performed using the large gap geometry only. The
results show an additional improvement in the overall displacement efficiency is achieved by lowering the
flow rate from 6 bpm to 2 bpm.
Viscous fingering would likely be present in the annular space of both geometries. Experiments
suggest that the fingering would likely be more pronounced in the narrow annulus than in the wider
annulus. This fingering would greatly affect the efficiency by which the cement displaced the water phase.

Results and Discussion


From the results of actual test and theoretical calculations, it is clear that the major issue which needs more
careful consideration in geothermal well casing design is attributed to elevated temperature that the casing
is being exposed. Casing collapse is most common in a cemented tieback string with a tight annular
clearance due to trapped water. Trapped water cannot be mitigated by increasing the casing grade as it can
generate pressure way more than the collapse resistance of usual API casing. A rule of thumb is that for
every 1°F temperature increase, there is 50 psi pressure increase. To avoid casing collapse, the potential
for trapping water between casing strings should be carefully evaluated and mitigated. The potential for
trapping water can be reduced, if not completely eliminated, by improving the cement placement
efficiency, choosing the proper centralizer, and selecting the appropriate casing grade.
Cement placement is usually provided by the cementing company and the procedure is basically
acceptable worldwide. In order to attain efficient cement placement, there are number of items that needs
to be considered. It is important to understand the effect of centralizer configuration to cement flow
behavior and the potential for viscous fingering. Bow spring centralizers have angles that can trap water
and start developing viscous fingers during cementing operations. Viscous fingering can be reduced and
potentially eliminated by pumping more excess cement, lowering the pump rates, and using rigid spiral
centralizers.
Another issue of equal importance that needs to be considered is the proper selection of casing
connection. Although the FEA did not show any signs of connection failure, the fact is that this type of
failure occurred many times. This condition should be discussed with the casing and connection supplier
in order to make the required testing and qualification needed in order to have further assurance of the
failure not happening again. Connection failure might be often neglected in casing design since it is not
a major design item and it is often relied on as a supplier recommendation.

Recommendations
Given the extreme temperature to which casing is being exposed in geothermal operations, failures cannot
be totally eliminated. But there are mitigations to reduce the severity and rate of occurrence. Failures
encountered should be used as lessons learned so as not to repeat the same mistakes. Further analysis
should be made to understand the cause of failure and apply the improvement. Below is a list of
recommendations to further understand and address casing failures in geothermal well design.
1. Since the FEA for casing connection did not show any failure even though there was actual failure,
it is an additional safeguard to ensure the connection used satisfies the requirement in Thermal
Well Casing Connection Evaluation Protocol (TWCCEP). This document provides procedures for
assessing suitability of a threaded casing connection for intermediate or production casing exposed
to temperature ranges of 180°C to 350°C. It would be helpful also to evaluate the connection
tensile reduction if special clearance connections were used.
2. Most of the casing failures observed happened to 10-3/4⬙ tieback cemented inside 13-3/8⬙ casing.
The annular clearance between these two casings is tight and thus prone to development of viscous
fingering. Viscous fingers arebelieved to be the pockets that trap water and was the major root
12 IADC/SPE-170480-MS

cause suspected in all of the casing collapse events. It is recommended that an engineering study
be conducted to further evaluate the technical soundness and feasibility of cementing 10-3/4⬙
tieback inside 13-3/8⬙ casing where the annular space is only 0.8325⬙ on the pipe body, about half
an inch narrower than the 9-5/8⬙ ⫻ 13-3/8⬙ completion configuration.
3. Development of viscous fingers can be related not only in tight clearances but also in cement
quality and placement practices. It is good to note that viscous fingers are developed when the fluid
above has significantly lower density than the fluid below. A more comprehensive cement slurry
formulation study should be conducted where the focus is not only on the attainment of zero free
water. The objective of the study is to evaluate what overall system properties are needed to
address the challenge of cementing long strings with narrow annular space. Slurry density as well
as the properties of the final product like compressive strength, pump rate and stiffness should also
be evaluated.
4. Since the major cause of casing collapse is trapped water in the pockets of viscous fingers
developed across centralizers, it is recommended to use spiral rigid centralizers instead of bow
spring centralizers in the tieback casing string. The displacement efficiency of the water by the
cement through the spiraled centralizer significantly better than that computed using the bow type
centralizer. The spiral blade centralizer did not produce significant quantities of discrete water
pockets, left in the cement stream after displacement of the water through the centralizer.
5. Most of the casing collapse and connection failure occurred in the shallow section (not below 750
ft.). This may be associated with inefficient displacement of water trapped across centralizer and
viscous fingers. In order to improve water displacement, it is good to consider pumping more
excess cement to increase the volume fraction of cement contact time. It will significantly increase
water sweeping action in the annulus and in breaking viscous fingers.

References
1. Swanson, R.J., “Unocal Geothermal Well Design Guidelines”. Internal CGPHI Document, 2001
2. Nayyar, M.L., 2000. “Piping Handbook”. 7th Ed. McGraw Hill. pp. E101–E107
3. Wells, M.R., “The Process of Cement Displacement of Water in an Annular Region while Flowing
Through a Topco Centralizer”. Report for Unocal.
4. Naïve, P.T., 2005. “Update Report on the 10-3/4⬙ TC- II Casing Collapse and Connection Failure
Analyses”. Internal CGSC Document.
5. Thermal Well Casing Connection Evaluation Protocol (TWCCEP), Domain Edition 1.2, 2005

You might also like