SPE

Society of Petroleum Engineers

SPE 18776

An Experimental Study of Casing Performance Under Thermal

Recovery Conditions
by K. Maruyama,* E. Tsuru, M. Ogasawara, and Y. Inoue, Nippon Steel Corp., and E.J. Peters,*
U. of Texas
*SPE Members

Copyright 1989- Society of Petroleum Engineers, Inc.

This paper was prepared for presentation at the SPE California Regional Meeting held in Bakersfield, California, April 5-7, 1989.

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 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. Permission to copy is restricted to an abstract of not more than 300 words. Illustrations may not be copied. The abstract should contain conspicuous acknowledgment
of where and by whom the paper is presented. Write Publications Manager, SPE, P.O. Box 833836, Richardson, TX 75083-3836. Telex, 730989 SPEDAL.

ABSTRACT stresses on well casing. In this process, steam is

injected into the formation through the tubing or
A study was conducted to investigate the be- casing from a period of a few days to several
havior of casing pipe body and connection under weeks. The steam is then allowed to soak for a
simulated thermal recovery conditions. Performed short period after which the well is placed on
in a new computer-controlled thermal well production. Thus, the casing is subjected to a
simulator, the study examined the thermal stress period of heating during the steam injection phase,
behavior and the leak resistance of the pipe and followed by a short period of nearly constant
connection at temperatures up to 354°C (670°F) under temperature during the soak period and then a
severe loading conditions similar to those en- period of cooling during the subsequent production
countered in thermal wells. Also studied was the of the well.
biaxial collapse resistance of the casing under
the large axial tension that would exist after the When casing is heated or cooled as in the
cooling period in a steam stimulation process. cyclic steam stimulation operation, one of two
things will happen. The casing will expand or
contract if allowed to do so. If the casing i_s
Results indicate that in a steam injection fixed at both ends as is often the case when the
process in which the casing is maintained at its casing is cemented to the surface, then compressive
maximum temperature for a period of time, stress and tensile stresses are generated in the casing.
relaxation can occur in a constrained production These stresses may be large enough to exceed the
casing, resulting in the development of excessive yield strength of the pipe or the joint strength of
tensile stress during the cooling of the casing. the coupling, resulting in casing failure [1-3].
Premium connections with metal-to-metal seal Injection temperatures up to 343°C (650°F) and
maintained gas leak resistance at temperatures up injection pressures up to 20.7 MPa (3000 psi) have
to 354°C (670°F) whereas the connections without been reported in steam injection wells [4,5].
the metal seal did not. The biaxial collapse
resistance of casing under large axial tension was Willhite and Dietrich [6) have presented a
dependent on the stress-strain characteristics of conceptual model of the thermal load history of the
the material. Hot-rolled K55 casing with work casing during cyclic steam stimulation. During
hardening characteristics showed a higher biaxial steam injection, the casing is heated and compres-
collapse resistance than expected from the von sive stress is created in the casing in proportion
Mises yield criterion. to the temperature change. If the temperature is
high enough, the yield strength of the casing
Based on the results of this study, it is material will be exceeded and the casing will
concluded that heavy weight K55 casing with premium become plastically deformed. Therefore, during the
connection is a good candidate for thermal well steam injection phase of the stimulation, the
completions. casing may fail due to plastic deformation and the
connection may fail due to excessive compressive
INTRODUCTION load. When the casing is cooled, tensile stress is
generated which may be high enough to cause tensile
Cyclic steam stimulation can impose severe failure of the pipe or the connection. Upon
cooling the casing to its original temperature
prior to steam injection, a permanent residual
Reference and figures at end of paper tensile stress will be left in the casing. In
305
 AN EXPERIMENTAL STUDY OF CASING PERFORMANCE
2 UNDER THERMAL RECOVERY CONDITIONS SPE 18776

addition to creating the potential for tensile (19.7 in) apart. The relative movement of the
failure, this residual tensile stress causes the rings is measured by magnetic scales which are
casing to be more susceptible to biaxial collapse located outside the heater and connected to the
failure [7,8]. rings by heat-resistant steel rods. If a thermal
elongation is about to occur, the computer permits
Operators of thermal wells are well aware of a servo valve to apply enough compressive load on
the dangers posed by high temperatures to the the test sample to keep its length constant. On
casing and often take measures in their well com- the other hand, if the sample is about to contract,
pletions to minimize the risk of casing failure the computer prevents this by applying sufficient
[5,6,9-13]. Among the steps usually taken are the tensile load to maintain its length constant.
use of high grade casing such as Pl05, PllO, S95
and Soo95 and the use of API buttress thread In order to test the sealing performance of the
coupling for thermal wells. The problem facing the connection, high pressure nitrogen gas was applied
casing designer is that there are very few experi- to the connection. Leaked gas was vented into a
mental data on the amount of thermal stress to be water vessel where it was detected by photoelectric
expected and the sealing performance of the various bubble detectors.
types of connections at steam injection conditions
upon which to base the casing design. Each test sample consisted of 4600 mm (15 ft)
of pipe with the connection at the center. Table 2
The objectives of this study were: (1) to shows the types of connections and the test
measure the thermal load histories of various conditions used in this study. API long round
casing grades under simulated cyclic steam stimula- thread coupling (LTC), API Buttress coupling (BTC)
tion conditions, (2) to test the leak resistance of and a premium connection with metal-to-metal seal
API connections and the new premium connections at ~ (Fig. 3) were tested. For each type of connection,
simulated steam stimulation conditions, (3) to K55 and L80 casing grades were tested. For the
measure the biaxial collapse resistance of the premium connection, a C95 casing grade was also
casing materials under tensile loads similar to tested. The physical properties of each grade at
those generated in the casing during cyclic steam room temperature and 354°C (670°F), its composition
stimulation and (4) to propose a new casing design and its heat treatment are shown in Table 3.
approach for ther-mal wells based on the results of
our experimental measurements. Since cemented casing is rigidly fixed, the
elongation or contraction of the casing is pre-
We measured the thermal load behavior of casing vented during cyclic steam injection operation.
pipes, high temperature leak resistance of various This situation was simulated in the laboratory by
connections and the biaxial collapse resistance of preventing the elongation or contraction of the
the casing at simulated cyclic steam stimulation test sample during the heating and cooling cycles
conditions using a computer-aided thermal well of the thermal well simulation test. To simulate
simulator and a biaxial collapse tester. This cyclic steam stimulation, the test sample was
paper.presents the results of our study which show heated to the required temperature in three to six
that thick-walled K55 casing equipped with a hours, maintained at the maximum temperature for
premium connection is a good candidate for high twenty four hours and then cooled by blowing air
temperature steam stimulation wells. over it. This heating cycle was repeated at least
twice at each test temperature. Before advancing
EXPERIMENTAL EQUIPMENT AND PROCEDURE the test to the next temperature, the sample was
relieved from the thermal constraint load. If the
Thermal Well Simulation Test connection leaked, the test was terminated at that
temperature. For each test, the internal pressure
The thermal well simulator was designed to was maintained at the steam saturation pressure at
subject a full size pipe and connection to cyclic the test temperature.
steam injection conditions of thermal load and gas
pressure. Fig. 1 shows a schematic diagram of the Biaxial Collapse Test
thermal well simulator. Fig. 2 shows a photograph
of the simulator with a test sample installed. The The residual tensile load developed in the
simulator consists of (1) a 1900 mm (6.2 ft) long casing during the cooling cycle in a steam stimu~
electric heater, (2) hydraulic power cylinders lation operation is so large that there is a
capable of applying 9807 KN (2,205,000 lbs) of load potential risk of biaxial collapse failure. To
and (3) an integrated measurement and control investigate this problem, collapse tests were
system. The specifications of the simulator are conducted on test samples subjected to tensile
summarized in Table 1. loads using a biaxial collapse tester. The tester
was designed for collapse testing under combined
A computer is used to control the operation of external pressure with axial load. Fig. 4 shows a
the thermal simulator. Feedback control systems schematic diagram of the tester with the test
are used to control the temperature, the thermal sample installed. The specifications of the
displacement and the thermal load. In order to collapse tester are given in Table 4.
control the temperature, six thermocouples are
attached to the test sample. The measured and Hot rolled K55 and quench and tempered N80
programmed temperatures are compared every second casing grades were tested at room temperature.
by the computer and appropriate signals are Fig. 5 shows the stress-strain curves for the two
transmitted to the temperature regulator to elimi- casing grades. It may be seen from the figure that
nate the difference between the two temperatures. the K55 grade is a work hardening material whereas
In order to control the thermal displacement, two the N80 grade is a perfect elastic-plastic materi-
clamp rings are mounted on the test sample, 500 mm al. The K55 samples were made uniform in shape and

306
SPE 1B776 K. Maruyama, E. Tsuru, M. Ogasawara, Y. Inoue, E.J. Peters 3

post annealed to remove the residual stresses after temperature over which plastic yielding occurs, the
they were received. The NBO samples were tested as larger will be the amount of residual tensile load.
received from the mill. Table 5 shows the mecha- Also, the larger the amount of stress relaxation,
nical properties of the test samples. the greater will be the residual tensile load.

The procedure for the collapse test was as Effect of Temperature, Heating Cycle and Casing
follows. Axial load was applied to the test sample Grade on Thermal Load Behavior
while the external pressure was raised until
failure. The maximum external pressure at which Fig. 7 shows the thermal load versus tempe-
failure occurred is the collapse resistance of the rature diagrams for K55 casing at three tempe-
sample. By varying the axial load, the biaxial ratures. At each temperature, the metallurgical
collapse resistance of the material was obtained. characteristics described previously were present.
However, there are three new findings due to the
RESULTS AND DISCUSSION effects of temperature and heating cycles. First,
the maximum tensile and compressive loads at 300°C
Thermal Stress Behavior Under Cyclic Steam (572°F) and 354°C (670°F) were essentially the same
Stimulation Conditions even though a larger plastic deformation occurred
at 354°C (670°F) than at 300°C (572°F). This
Fig. 6 shows the typical thermal load versus observation suggests that if there was significant
temperature diagram for 17B mm, 34.2 kg/m (7 inch, plastic deformation, the maximum compressive and
23 lbs/ft) K55 casing undergoing cyclic steam tensile loads would be the same regardless of the
stimulation operation. This diagram was obtained tempe.rature difference. Second, the amount of
experimentally using the thermal well simulator stress relaxation increased with temperature. It
under constrained thermal elongation and contrac- ranged from 200 KN (45000 lbs) at 250°C (4B2°F) to
tion at 354°C (670°F). Path A-B shows the elastic 400 KN (90000 lbs) at 354°C (670°F). Third, the
response of the material in which compressive load elastic limits in compression at 300°C (572°F) and
is developed in proportion to the temperature 354°C (670°F) were about 300 KN (67500 lbs) larger
change. The compressive load may be calculated than at 250°C (4B2°F). Thus, the pre-strain build-
from Eq.(l): up in the casing in the first beating cycle in-
creased the elastic limit of the casing during the
Lt = E(a~T)A ------------------------ (1) subsequent heating cycles due to strain aging.
Hence, the K55 casing will actually become stronger
At point B, the casing began to yield. Path B-C in the second and subsequent heating cycles in a
shows the region of plastic deformation in which cyclic steam stimulation operation.
the compressive load was observed to decrease
slightly. Path C-D shows stress relaxation in Fig. B shows the thermal load versus tempe-
which the compressive load continued to decrease rature diagrams for LBO casing grade at three
even though the temperature was constant. This temperatures. Like the K55 behavior shown in
stress relaxation phenomenon was not expected at Fig. 7, stress relaxation occurred. Unlike the
this low a temperature. Path D.-E shows the elastic K55 behavior, the thermal load changed elastically
response during cooling in which the tensile load throughout the cooling process without any evidence
increased along a path parallel to A-B. At E, a of yielding. This is in agreement with the load
yield behavior wa.s observed at which the rate of history presented by Willhite and Dietrich [6].
change of the tensile load decreased considerably. Neither strain aging nor work hardening of the
Because the thermal tension load of about 1000 KN material was observed. A similar thermal load
(225000 lbs) at point E is well below the actual behavior was obtained for C95 casing. However, as
pipe body yield strength of 1710 KN (3B5000 lbs), may be expected, the C95 casing had a smaller
the yield phenomenon from E to F is probably due to fraction of the residual load to pipe body yield
Bauchinger effect. If that is the case, the strength than the LBO casing. The residual tensile
compressive yielding encountered during heating stresses obtained from the thermal simulation tests
caused the early tensile yielding. After the for K55, LBO and C95 casing grades are summarized
completion of the first heating cycle at point F, a in Table 6.
large residual tensile load of 1400 KN (315000 lbs)
was left in the casing. High Temperature Leak Resistance of API and
Premium Connections
In the second heating cycle, the starting point
was offset by the residual tensile load from the Table 7 shows the high temperature leak resist-
first cycle. As the casing was heated again, the ance of API Buttress thread coupling (BTC), API
compressive thermal load progressed along path F-C long round thread coupling (LTC) and premium
elastically. From point C, the thermal load connection of 17B mm, 34.2 kg/m (7 inch, 23 lbs/ft)
followed essentially the same path as in the first K55 and LBO casing grades. The tests were
cycle. However, the final residual tensile load conducted with the test samples restrained from
was 140 KN (31500 lbs) higher than in the first expansion or contraction.
cycle. This may be caused by work hardening of the
material. In the third cycle, the load history was The results show that the BTC offered the worst
essentially the same as in the second cycle. leak resistance of the three connections tested.
The connection with manganese phosphate coating
From the thermal load response presented in leaked at a temperature as low as 200°C (392°F)
Fig. 6, it is apparent that the amount of residual while that with a copper plating leaked at 225°C
tensile load (F) is strongly dependent on the (437°F). A BTC connection requires the thread com-
degree of plastic yielding (B-C) and the amount of pound to plug the thread clearance of 0.03 mm
stress relaxation (C-D). The larger the range of (0.001 inch) in order to maintain leak resistance.

307.
 AN EXPERIMENTAL STUDY OF CASING PERFORMANCE
4 UNDER THERMAL RECOVERY CONDITIONS SPE 18776

The poor leak resistance observed is due to the K55, 140 mm, 29.7 kg/m (5-1/2 inch, 20 lbs/ft)
fact that the dropping point of the thread compound premium connection subjected to a severe thermal
of 165°C (329°F) was exceeded at the test tempe- tensile load.
rature of 200°C (392°F). Therefore, the thread
compound readily flowed out of the threads, result- The results show a greater deformation of the
ing in the low leak resistance of the connection. regular coupling compared to the strong couplings.
The poor sealing performance of the BTC at high As a result, the regular coupling has less contact
temperatures is significant although BTC has tradi- pressure than the strong couplings. The calculated
tionally been recommended for thermal well comple- maximum contact pressures were 127 MPa (18000 psi)
tions. for the regular K55 coupling, 255 MPa (37000 psi)
for the heavier K55 coupling and 353 MPa (51000
The results indicate that the leak resistance psi) for the higher grade C95 coupling. The
of the LTC connection depends on the yield relative magnitudes of the contact pressures are
strength of the material. The K55 LTC connection shown by the stress contours in the seal portion of
leaked at 225°C (437°F); however, the L80 LTC the connection. Therefore, the use of heavier or
connection maintained leak resistance up to 275°C higher grade coupling is recommended for premium
(527°F). It leaked at 300°C (572°F). Since 275°C connections for high temperature service.
(527°F) is higher than the dropping point of the
thread compound, the flow of the thread compound is The second method to maintain the contact
not the main cause of the leakage of the LTC pressure of the metal-to-metal seal is to reduce
connection. The main cause is the yielding of the the adverse effect of thermal stress relaxation of
test material. It can be seen from Fig. 9 that the the material. Stress relaxation reduces the hoop
K55 and L80 LTC connections leaked at temperatures and radial stresses in the seal portion of the
slightly greater than the yield temperatures of the connection thereby reducing the contact pressure.
materials. Therefore, it is concluded that plastic The reduction in the contact pressure can be esti-
deformation of the wedge type thread is the main mated by measuring the change in the seal diameter
cause of leakage of the LTC connection. after the thermal well simulation test.

The results show that the premium connection Fig. 12 shows representative reductions of the
offered the best leak resistance of the three seal diameters for three different grades of
connections tested. The connection maintained leak connections maintained at the prescribed tempe-
tightness up to 325°C (617°F) even for the K55 ratures for 48 hours. Seal interference reductions
grade but leaked at 350°C (662°F). The good seal- of 37% and 27% were observed for the K55 and C95
ing performance of the premium connection is due to connections, respectively. However, C95 containing
its metal-to-metal contact seal [14-16]. 0.45% molybdenum showed very little reduction in
the seal diameter. Therefore, this material has
Fig. 10 summarizes the sealing performance of good resistance to stress relaxation and would be a
the three types of connections at high tempe- good material for premium connections.
ratures. The results of the tests on additional
grades and sizes of premium connections are Collapse Pressure with Axial Tensile Load
included in the figure. In general, the sealing
ability of the BTC connection is limited to about The collapse resistance of quench and tempered
200°C (392°F). Plastic deformation induced by casing and as rolled casing showed significantly
thermal stress limits the sealing ability of the different characteristics under tensile load.
LTC to 300°C (572°F) for high grade casing such as Fig. 13 shows the measur~d collapse pressure of N80
L80 and to a much lower temperature for K55 casing. casing which is a quench and tempered material.
The ~etal-to-metal seal of the premium connection The solid line is the collapse resistance predicted
provides good sealing performance up to 350°C by Eq. (2) which is a modified version of the API
(662°F) for all the casing grades tested. collapse formula based on the von Mises yield
criterion.
Premium Connections for High Temperature Service

Since the high temperature leak resistance of

the premium connection is due to the contact where
pressure in the metal-to-metal seal portion of the
connection, the performance of the connection can P0 y = 2oy(D/t - l)/(D/t) 2 -------------------- (3)
be enhanced by maintaining or increasing the
contact pressure. There are two methods for It should observed that the measured collapse
maintaining or ~ncreasing the contact pressure. pressures are in good agreement with those pre-
The first method is to use a strong coupling. The dicted by Eq. (2).
second method is to modify the coupling material to
minimize the adverse effect of stress relaxation. Fig. 14 shows the measured collapse pressure
for the K55 casing which is an as rolled material.
Fig. 11 presents the results of stress analysis The solid line is the collapse pressure predicted
of a premium connection equipped with three types by Eq. (2). It can be observed that at low axial
of coupling; a regular K55 coupling, a strong stresses, the measured collapse pressures are in
coupling made of heavier K55 material than the good agreement with those predicted by ·the API
regular coupling and a strong coupling made of the formula. However, at high axial stresses, the
higher grade C95 material than K55 material. The measured collapse pressures are significantly
figures show the deformation and the contact higher than those predicted by the API formula.
pressure of the connection obtained by a finite This deviation is caused by the work hardening
element analysis. The analysis was performed for a nature of the K55 casing.

308
SPE 18776 K. Maruyama, E. Tsuru, M. Ogasawara, Y. Inoue, E.J. Peters 5

It may be concluded that the biaxial collapse thermal well designs of Table 8 is unsatisfactory.
pressure depends on the stress-strain characteris- Further, the use of the BTC connection is unsatis-
tics of the material. For the quench and tempered factory because this connection does not have
material which is a perfect elastic-plastic materi- adequate leak resistance at this temperature.
al, the biaxial collapse pressure is predicted However, if the temperature was 300°C (572°F) or
adequately by the API formula. Since L80, C95 and less as normally encountered in many steam injec-
PllO casing grades are all quench and tempered tion projects, then light weight HC95 casing would
materials, it is expected that their biaxial have sufficient collapse resistance to withstand
collapse resistance can be predicted by the API the maximum formation pressure, but the BTC may
formula. For the as rolled material which is a still be unsatisfactory.
work hardening material, the biaxial collapse
resistance is much higher than predicted by the API Fig. 16 shows the biaxial collapse resistance
formula. It seems reasonable to take advantage of of heavy weight 178 mm, 43.1 kg/m (7 inch, 29 lbs/
this work hardening characteristic of the as rolled ft) K55, N80 and C95 grades. At the extreme tempe-
material in designing thermal well casing. rature of 354°C (670°F) at the maximum formation
pressure of 17.2 MPa (2500 psi), only the K55
NEW CASING DESIGN APPROACH FOR.THERMAL WELLS casing would have enough collapse resistance to
withstand the effect of the residual thermal
In this section, we present a new casing design stress. Therefore, under high temperature condi-
approach in which the biaxial coalapse resistance tions, a heavy weight K55 casing would be superior
of the casing and the leak resistance of the to a light weight higher grade casing for thermal
connection are .taken into account. It will be well completion. Also, only the premium connection
shown that because of its work hardening character- would provide adequate leak resistance at this
istic, heavy weight K55 casing may be superior to temperature. Therefore, our recommended casing
higher grade casing for thermal well service. design for the thermal wells shown in Table 8 would
be 178 mm, 43.1 kg/m (7 inch, 29 lbs/ft) K55 casing
Table 8 shows the design for the production with premium connection.
casing for two thermal wells reported in the
literature [4]. The temperature for the Cat Canyon CONCLUSIONS
well was reported to be as high as 354°C (670°F)
and that at the Cold Lake well was 336°C (636°F). A thermal well simulation study was conducted
In both designs, 178 mm, 34.2 kg/m (7 inch, 23 lbs/ on full size casing pipe and connection to measure
ft) high grade casing Soo95 with buttress connec- the temperature-generated stresses in the casing,
tions were selected. It is assumed that the light the high temperature leak resistance of the connec-
weight casing was selected to minimize the casing tions and the biaxial collapse resistance of the
cost. Our objective is to evaluate the casing casing. Based on the results of the study, the
designs for the two wells by taking into account following conclusions may be drawn.
the biaxial collapse resistance of the casing and
the high temperature leak resistance of the connec- 1. The complete thermal stress behavior of casing
tions obtained in this study. including stress relaxation and Bauchinger
effect has been successfully measured.
Fig. 15 shows the biaxial collapse resistance
of 178 mm, 34.2 kg/m (7 inch, 23 lbs/ft) K55, N80,
C95 and HC95 [8] casing under axial stress. In 2. Premium connections with metal-to-metal seal
this figure, the residual stresses obtained from will provide excellent seal tightness in thermal
the thermal simulation tests are superimposed on wells at temperatures up to 354°C (670°F).
the biaxial collapse behavior of the various casing
materials.· The symbols in the figure represent 3. The seal integrity of premium connections can be
the residual axial stresses for each material at enhanced by use of thick and/or higher grade
the temperatures indicated in the legend. It coupling than the pin and by use of coupling
should be noted that the Soo95 casing used in the made of quench and tempered steel containing
thermal well designs of Table 8 has biaxial molybdenum.
collapse behavior similar to HC95 casing. The
horizontal dashed line is the formation pressure of 4. The seal limits of API buttress thread coupling
17.2 MPa (2500 psi) which was reported for the Cat is 200°C (392°F) and that of API long round
Canyon well. It is assumed that this pressure thread coupling is 300°C (572°F).
constitutes the collapse load on the casing.
Therefore, to prevent collapse, the biaxial 5. Hot rolled K55 casing has higher biaxial
collapse resistance of the casing must be greater collapse resistance under large axial tension
than 17.2 MPa at the residual tensile stress at than may be expected from the von Mises yield
the temperature of interest. criterion.

It is clear from Fig. 15 that if the casing 6. Thick-walled K55 casing with premium connection
were subjected to a temperature of 354°C (670°F) at is a good candidate for high temperature steam
the maximum formation pressure of 17.2 MPa (2500 injection wells.
psi), none of the casing grades shown in the figure
would have enough collapse resistance under the NOMENCLATURE
t~ermal load generated by cyclic steam stimulation.
rhus, under this extreme temperature condition as A cross-sectional area of pipe, mm 2 (in 2 )
might occur in steam injection through the tubing/ D outside diameter of pipe, mm (in)
casing annulus or ~s a result of a packer leak, the
use of the light weight 34.2 kg/m casing in the E Young's elastic modulus, MPa (psi)

309
 AN EXPERIMENTAL STUDY OF CASING PERFORMANCE
6 UNDER THERMAL RECOVERY CONDITIONS SPE 18776

temperature-generated load in pipe, KN (lb) 12. Greer, F.C. and Shyrock, S.H.: "New Technique
Improves Steam Stimulation Completions, Jour.
biaxial collapse pressure, MPa (psi) Pet. Tech. (July 1968) 691-699.
P0 y internal yield pressure of pipe, MPa (psi)
13. Burkill, G.C.C.: "Thermal Well Completion
t wall thickness of pipe, mm (in) Design with Openhole Gravel Packed Liners and
a coefficient of linear expansion, l/°C Methods for Selective Steam Injection," The
Oil Sands of Canada-Venezuala, CIM Special
liT temperature change, °C (°F)
Publication, 1977, 595-608.
Oy yield strength of pipe, MPa (psi)
14. Mimaki, T., Ogasawara, M. and Yazaki, Y.:
o2 axial stress applied to pipe, MPa (psi) "Analytical Investigation for The Performance
Properties of Oil and Gas Well Pipe Connec-
ACKNOWLEDGEMENT tions," ASME ETCE Drilling and· Production
Symposium (Feb. 1987).
The authors wish to thank Shunji Nishi for his
great work of designing the thermal well simulator 15. Mimaki, T., Ogasawara, M. and Yazaki, Y.:
and supervising its construction. "A Tentative Criterion for Gas Tightness of
Premium Connection under Thermal Mechanical
REFERENCES Cycle," ASME ETCE Drilling and Production
Symposium (Feb. 1985).
1. Bissoondatt, J.C.: "Casing Failure in Steam
Stimulation Wells," SPE 5951. 16. Payne, M.L.: "Qualification and Use of a Mill-
premium Connection in a Geopressured Gulf of
2. Dietrich, W.K. and Willhite, G.P.: "The Casing Mexico Application," Jour. Pet. Tech. (August
Failure Problem in Steam Injection Wells, Cat
1988) 1033-1040.
Canyon Oil Field, Santa Barbara County, Cali-
fornia," presented at ASME 21st Annual
Petroleum Mechanical Engineering Conference,
New Orleans Sept. 18-21, 1966.

3. Farouq Ali, S.M.: "Current Status of Steam

InJection as a Heavy Oil Recovery Method,"
Jour. Cdn. Pet. Tech. (Jan.-Mar. 1974) 1-15.

4. Farouq Ali, S.M. and Meldau, R.F.: "Current

Steamflood Technology," Jour. Pet. Tech.
(October 1979) 1332-1342.

5. Gates, C.F. and Holmes, B.G.: "Thermal Well

Completions and Operation," Paper PDll,
presented at the Seventh World Pet. Congress,
Mexico City, 1967, vol.3, 419-429.

6. Willhite, G.P. and Dietrich, W.K.: "Design

Criteria for Completion of Injection Wells,"
Jour. Pet. Tech. (Jan. 1967) 16-21.

7. Tamano, T., Inoue, Y. and Mimura, H.: "Exami-

nation of Commercial Casing Collapse Strength
Under Axial Loading," Jour. of Energy
Resources Tech. (Dec. 1982), vol.l04.

8. Marlow, R.S.: "Collapse Resistance of Mill

Selected High Collapse Grade 95 Casing," SWRI
Project 06-8685, Southwest Research Institute
(Aug. 1987).

9. Bleakley, W.B.: "Steamed Wells Need Good

Completions," Oil and Gas Jour. (April 4,
1966) 136-139.

10. Leutuyler, K. and Bigelow, H.L.: "Temperature

Effects on Subsurface Equipment in Steam
Injection System," Jour. Pet. Tech. (Jan.
1965) 93-101.

11. Wen-Zhang, L.: "Pilot Steam Soak Operation in

Deep Wells in China," Jour. Pet. Tech. (Nov.
1987) 1441-1448.

310
 JJ!E l-877 6
Table 1 - Specifications of the thermal well Table 4 - Specifications of the biaxial collapse
simulator tester

Loading · Heater max 400°C (752°F), Loading · External pressure max 260 MPa (37740 lbs/in2)
· Axial load max ±9B07 KN (2205X103 lbs), · Axial tension or
max 7.BMN (175X104 lbs)
· I nterna I pressure max 196 MPa (2B450 lbs/in 2 ) compression load
·Torsion max 0.2MN·m (137X103 lbs·ft)
Test Pipe · Outside diameter max 245 mm (9·5/B inch),
· Length exposed to 1900 mm (6.23 ft) Test pipe · Outside diameter max 219 mm (B.6 inch)
heating · Length exposed to
max 1270 mm (50 inch)
external pressure
Test mode · Displacement constraint test
· Temperature-load control test
· Pretension test

Table 2 - Thermal simulation test samples and

test conditions Table 5 - Mechanical properties of casing used in
the collapse tests
Size Internal
Types of
Coupling
O.D.
(mm)
Weight
(kg/m)
Grade
Temp.
(oC)
Pressure
(MPa) Grade O.D.
Size
Thickness D/T
Yield
Strength
(MPa)
Tensile
Strength
(MPa)
0 va ,.,I Y
(%)
I'"'"";.
city
(%)
Residual
Stress
(MPa)
(mm) (mm)
250 3.9B
5.0 27.9 stress
275 5.9B machined to
K55 140 422 784 uniform pipes relief
7.6 18.4 annealing
K55 300 B.62
NSO 178 9.6 18.5 674 745 a.o9 1 3.9 19
327 12.4
354 17.2
17B 34.2
250 3.9B
275 5.9B
Premium LBO 300 B.62
Connection 327 12.4
354 17.2
K55
17B 43.2 LBO 354 17.2 Table 6 - Residual tensile stresses for different
C95 casing grades
250 3.9B (MPa)

~·
245 69.9 C95 300 B.62
G 250°C 300°C 354°C
354 17.2
K55 270 330 330
K55 225 2.55
200 1.57 LBO 190 350 500
API
250 3.9B C95 BB 275 549
LTC 17B 34.2 LBO
275 5.9B
300 B.62
API K55 200 1.57
17B 34.2
BTC LBO 225 2.55

Table 3 - Chemical compositions and tensile proper-

ties of casing used in the thermal Table 7 - High temperature leak resistance of API~_
simulation tests and premium connections

Chemical Composition (wt%) Tensile Yield Tensile

Heat Test Strength Strength 400 500 600 700 °F
Grade Treatment Types of
Temp. Grade Coating I
c Si Mn p s Cr Others (oC) (MPa) (MPa) Coupling oc
200 250 300 350
Room I I
Temp. 392 775
K55
Hot
Rolled .47 .24 1.30 .018 .005 -·
AI
N 354 363 735
API
Round
K55 Cu plate •
Quenched Ti Room 588 696
Long LBO Mn-phos. 0 0 0
•
LBO and
Tempered
.17 .24 1.42 .015 .013 - AI
B
Temp. API K55 Mn-phos. •
N
Room
354 497 642 Buttress LBO Cu-plate •
Ti
C95
Quenched
and .19 .27 1.42 .015 .006 .2 AI
B
Temp. 706 784
Premium K55 Cu plate 0 0 0 0
•
Tempered 354 569 710 Connection LBO Mn-phos. 0 0 0
N

Note: •; Leak, o; No Leak

311
 Table 8 - Example casing design for thermal wells

Steam Steam Production Casing

Depth Injection Injection
Field Operator Temperature Size Weight Types of
Pressure Grade Prestress
(m) (MPa) (oC) (mm) (kg/m) Coupling

Cold Lake Esso 457 13.8 335 178 34.2 Soo-95 BUT Yes
Cat Canyon Conoco 1067 17.2 354 178 34.2 Soo-95 BUT Yes
Recommendatory casing Premium
Formation pressure 17.2 178 43.2 K55 -
based on the study Connection

OISPI.ACEI'£NT I
SENSOR
(MAGNETIC' SCALE) ~~
ILil

'"--r~ :
LEAK DETECTOR
1
• RECORDER

Fig. 1 - Schematic diagrqm of the thermal well

simulator.

Fig. 2 - Photograph of the thermal well simulator.

312
 RE 18776

1 Vessel 2 Test Specimen 3 Cylinder for Axial Loeding

Fig. 4 - Schematic diagram of the biaxial collapse

Fig. 3 - Premium connection with metal-to-metal
seal. tester.

3000.-------------------,

e: - - - 1 st cycle
o: -----2nd cycle
2000 f-
F

1000

1400 ...c
·~
1200 If)
§
(.)

10
~ NBO
E -1ooo
BOO ~
2§ 1-
If)

~
IJl
600 K 55
-2000
400

200 I
I I
100 200 300 400
Temperature ('C)
10 15 20
Strain e (%)

Fig. 5 - Typical tensile stress-strain curves for Fig. 6 - Typical cyclic thermal load history of
test pipes. work hardening material.

313
 ..
,_
~.
;::-~

~
.jl-
-~.
.~_,

~';'.
~
~- ::is-! SPE 18776

3000 3000

\
Temp.
("C)
Residual
stress
(MPa) \ Temp.
("C)
Residual
stress
(Mf'a)

' .
0 250 270 0 250 190
2000 r- 6 300 330 2000 r- .,, 6 300 350
• 354 330
Ll'\::1
.""- 354 500

""""---~.,.._ ............. '.c., "".

,LI,LI,, ""'
'~"'"'·~
2 1000 2 1000 r
~ ~
-g -o
'0
0
~ Ll' •

r·~·~~
...J
...<: c ...J ~ b. .c.,
0 <: 0 ~,~0~~. I
-~ -~
0
(.)

\ ~o~~ 0
~, \ "LI300"C ~t
\::":\\
(.) \
(ij

~~
E
Ql
.s::
-1000 r- ~ -1000-
1-- 1!

\~J._J
1--

I
' 0 \

-2000 r- ~ 1-·-
--6-.-~-
-2000-

I I I I I I
- 3000 o~------1~00--------20L0-------3~0-0------~400 100 200 400
300
Temperature Cc) Temperature ("C)

Fig. 7 - Effect of temperature on the cyclic Fig. 8 - Effect of temperature on the cyclic
thermal load history of work hardening thermal load history of quench and
material. tempered material.

Connection No I eak I eak

Premium
API BTC
0
6.
•
.A

BOO
API LTC 0
•
400r---------------------------------~
Region for application of
premium connection

eo (() <0
600
~ 8 0
6 ~ 0
·-E LBO e
~. j Seal lim1t
...J
400
• • 2i
of LTC

~~.
0
~
~
r-
w
200
Thermal stress

100 200 300 400 K55 LBO C95

Temperature ('C) Grade

Fig. 9 - Elastic limit of casing versus temperature. Fig. 10 - Sealing performance of connections at high
high temperature.

314
 JPE 18776
IDeformation j I Contact Pressure I
I>
2)
3>
-4.000
-3.000
-2.000
MPa
4l -1.000
5) o.o

(a) Regular K55 coupling

(b) Heavy K55 coupling

(c) Regular C95 coupling

Fig. 11 - Deformation of couplings

due to axial stress.

t:. [ C95 (0.45%Mo) J

0 I C95 J
D I K55 I
0.7 !:============------.
80
0.6 -6
'E
sQ)
0.5-
D
0
1::
~ 0.4-
~
e
E
0.3-

(ij
Q) 0.2 r-'
(/)

0.1 ,....

j_ I I
Room 300 354
Temp.
Temperature (t) 600 BOO
Axial Stress (MPa)

Fig. 12 - Seal interference reductions at high Fig. 13 - Biaxial collapse resistance of 178 mm,
temperatures. 38.7 kg/m, N80 (7 inch, 26lbs/ft) casing.

315
 ~ ....
.·-~::.

I JeE 18776
h\ t
·.

50r-----------------------------------------,

0 thick wall . 24.1 kg/m

e thin wall. 162 kg/m

10

100 200 400 500 600 700

Axial Stress (MPa)

Fig. 14 - Biaxial collapse resistances of 140 mm,

16.2 kg/m and 140 mm, 24.1 kg/m (5. 5
inch, 11 lbs/ft and 5.5 inch, 16 lbs/ft)
K55 casings.

60
Residual axial stresses obtained
250t: 300t: 354 t;
K55
N80
0 ()
••
40
;f
~
•
.,
300t: 364t: :I
()
• ].,
"' ••
30
[)

il' li !
8 20

10

100 200 400 500 600 700 800

Axial Tensile Stress (MPa)
Axial Tensile Stress (MPa)

Fig. 15 - Biaxial collapse resistance of 178 mm, Fig. 16 - Biaxial collapse resistance of 178 mm,
34.2 kg/m (7 inch, 23 lbs/ft) casing 43.1 kg/m (7 inch, 29 lbs/ft) casing with
with residual axial tensile stress. residual axial tensile stress.

316