SPE 38978
Stress Analysis of Casing String Submitted to Cyclic Steam Injection
João C.R. Plácido, Ademar P. Jr., Luíz A.B.F. Paulo, PETROBRAS, and Ilson P. Pasqualino, Segen F. Estefen, COPPE
Copyright 1997, Society of Petroleum Engineers, Inc.
This paper was prepared for presentation at the Fifth Latin American and Caribbean
Petroleum Engineering Conference and Exhibition held in Rio de Janeiro, Brazil, 30 August–3
September 1997.
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. Electronic reproduction, distribution, or storage of any part of this paper
for commercial purposes without the written consent of 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 where and by whom the paper was presented. Write Librarian, SPE, P.O.
Box 833836, Richardson, TX 75083-3836, U.S.A., fax 01-972-952-9435.
Abstract
Oil well casing string may present intervals with cementing
failure. If the well is submitted to cyclic steam injection, some
problems may happen to this section. The heating and
cooling process may generate residual stresses (tension) in the
casing interval where cementing is not well bonded to the
formation or to previous casing. These residual stresses
decrease the casing collapse capacity. In this interval, may
exist fluid trapped behind the casing, which may be or
previous fluid that was not well displaced or free water from
cement. Therefore, in the next heating step, this fluid will
expand and may collapse the casing. A numerical model and
an experimental apparatus were developed to simulate this
situation. A finite element analysis incorporating geometric
and material non-linearities is carried out to evaluate the
residual stresses and their effects on the collapse pressure
capacity of the casing. The experimental work allowed to
understand this phenomenon. The numerical model is
validated using results from tests performed at PETROBRAS
and literature data.
Introduction
PETROBRAS used casing N-80 for all wells submitted to
cyclic steam injection. However, numerous problems have
happened. After a literature review, it was observed that
grade K-55 has succeed in cyclic steam injection wells1.
Therefore, PETROBRAS has decided to follow this
recommendation and, since then, none failures have been
registered. A project was initiated at PETROBRAS Research
Center (CENPES) to study the phenomenon that occurs in
casing sections submitted to cyclic steam injection where
cementing is not well bonded. This paper describes the main
failure mechanism. Mechanical properties and stress-strain
curves of casing grades H-40, K-55 and N-80 are presented.
An apparatus was developed and built at CENPES to simulate
physically a casing length with a window that represents a
free section where cement was not well bonded, which is
submitted to one or more thermal cycles. Three tests using
this apparatus are described. A numerical model was
developed and validated with results from theses tests and
with literature data2. Finally, it is concluded that this
numerical model is to be used as auxiliary casing design tool
for cyclic steam injection wells.
The Problem
According to Conceição1, failures that occur in casing
submitted to cyclic steam injection are the following:
collapse, rupture, connection leaking and corrosion.
When primary cementing is not perfect, fluid may exist
trapped behind the casing. It may be fluid of previous phase
that was not well displaced or free water from cement. Fig. 1
presents a steam injection well scheme, which shows a fluid
trapped in the free cement section.
When the first cycle occurs, casing temperature may reaches
350
C. The free casing section works as a shell with fixed
ends. When temperature increases, this section tries to
elongate. Since both ends are fixed, it will be submitted to a
compressive reaction. Fig. 2, extracted from Maruyama et
alli2, shows the stress cycle in function of temperature. When
yield stress is reached (B), the material is submitted to plastic
deformation that continues until the maximum temperature
(C). From C to D temperature is kept constant and stress
relaxation is observed (load decreases at constant
temperature). At point D, temperature starts cooling down
and the material deforms elastically until to reach point E,
which is the compressive yield stress. From E to F occurs
plastic deformation. When initial temperature is reached
again, the material is under a positive residual stress
(tension). Other cycles may occur and the tendency is to
follow the same hysteresis loop FCDEF.
Residual stress is considered the main cause of most casing
2 JOÃO CARLOS R. PLÁCIDO, ADEMAR POGGIO JR., LUÍZ A. B. F. PAULO, ILSON P. PASQUALINO, SEGEN F. ESTEFEN
failures in cyclic steam injection wells. Therefore, the study
developed at CENPES is concentrated in physical and
numerical simulation of this mechanism.
In this process is very important to know the mechanical
behavior of several casing grades. This is given by the
coefficient of linear expansion in different temperature levels
and also by stress-strain curves in controlled temperature.
Due to the lack of these data in the literature, several tests
were done to obtain these information for casing grades H-40,
K-55 and N-80. Table 1 presents coefficients of linear
expansion at several temperature levels. Tables 2, 3 and 4
present mechanical properties. Figs. 3, 4 and 5 present stress-
strain curves at 25
C.
Physical Simulation
To simulate physically this mechanism it was develop at
CENPES an experimental apparatus (Figs. 6 and 7). It
consists of two concentric pipes with cement in the annular
and fixed ends. The internal pipe simulates the production
casing and the external simulates either previous casing or
formation. The length of test sample is 2 m and a window is
cut at the mid span to simulate the free cement section (Fig.
8). The ends are welded in such way the external pipe
constrains internal pipe deformation. Pipes length and
diameters as well as window dimensions were chosen to
eliminate end effects. The internal casing is heat up with
electric resistors. To reach and hold the temperature at
325
C, it is necessary thermal insulation (Fig. 7).
Two strain gages and two thermocouples were used in each test
(Fig. 9) to measure and compensate the axial strain of the free
cement section during the thermal cycle.
To acquire data is necessary to get information from the
strain gages continuously (1 acquisition per minute), during
at least 48 hours. Therefore, it was set up a data acquisition
system (Fig. 10), which consists of microcomputer, analogic-
digital board and signal converter. Besides, temperature was
measured with thermocouples near to strain gages (Figs. 8
and 9). Temperature at the mid span of internal pipe is used
to test control.
Three tests were done. The external pipe diameter is always
9 5/8”. In the first test, internal pipe is grade H-40, nominal
diameter 5 1/2” and weight 14 lb/ft. Second test was
performed with internal pipe grade K-55, nominal diameter 5
1/2” and weight 15.5 lb/ft. In the third test, internal pipe is
grade N-80, nominal diameter 7” and weight 23 lb/ft. Figs.
11, 12 and 13 present results of strain in function of
temperature. The measured strain refers to axial strain gage
placed at mid span of internal pipe (Fig. 8 and 9).
Numerical Simulation
A numerical model was developed using a Finite Element
SPE 38978
Package, ANSYS3 . The analysis is non-linear with large
rotations and plasticity.
Finite element mesh consists of elements SOLID45, with
eight nodes and three degree of freedom per node (UX, UY
and UZ) to model internal pipe. Element CONTAC52 is used
to model contact elements simulating radial constraint due to
wellbore. The free cement section is modeled with fine mesh
and without radial constraint. This window can have any
length and any arc up to 360
(Fig. 14).
The system is loaded with thermal cycles (heating up and
cooling down) followed by external pressure.
This model is simplified because the system is symmetric.
The number of degree of freedom is reduced by modeling
only half pipe and half section.
The input data file consists of internal pipe geometry, type of
material and loads. Loading condition is given by initial and
final temperatures which is applied uniformly in the whole
mesh, external pressure applied in the free cement section,
boundary conditions of symmetry and radial constraints.
The model calculates residual stresses generated by the
unbalanced thermal expansion of the structure. Moreover, the
model gives the new collapse limit.
Model Validation
To validate the numerical model were used results from tests
done by Maruyama et alli2 and results of three tests performed
in the experimental apparatus built at CENPES.
Maruyama set up a equipment that consists of a casing
section with fixed ends. This apparatus considers this section
as the free length of casing where cement is not well bonded.
There are no radial constrains.
Fig. 15 shows results from numerical model that refers to
physical simulation presented in Fig. 16. It was used K-55
casing, nominal diameter 7”, internal diameter 6.366” and
weight 23 lb/ft. The free section was heated from room
temperature up to 340
C and cooled down to room
temperature. This cycle was repeated two times.
Fig. 17 shows results from numerical model that refers to
physical simulation presented in Fig. 18. In this simulation
was used K-55 casing with previous dimensions.
Fig. 19 shows results from numerical model that refers to
physical simulation presented in Fig. 20. In this simulation
was used N-80 casing with the same dimensions of previous
K-55. Figs. 17 to 20 present axial constraint force in function
of temperature when casing is submitted to three different
SPE 38978 STRESS ANALYSIS OF CASING STRING SUBMITTED TO CYCLIC STEAM INJECTION 3

thermal cycles (250
C, 300
C and 354
C). Acknowledgments

Figs. 21, 22 and 23 show results from numerical simulation
that must be compared with results from three tests performed
in the experimental apparatus built at CENPES. Therefore,
they must be compared to Figs. 11, 12 and 13 respectively.
It can be seen that the numerical model shows good
agreement with physical simulation. Therefore, this model
can be considered validated.
Regarding to Maruyama experiments, there was a good
agreement quantitative and qualitative. However, in the
comparison with tests performed at CENPES, there was only
good qualitative agreement. The quantitative agreement was
not so good. This can be explained by following reasons :
a) It is necessary to obtain more data of casing mechanical
properties in function of temperature. In this numerical
simulation only the 25
C curve was used. The behavior
obtained with other temperatures was not reliable.
b) Maruyama used casing grade L-80. Since we did not have
data for this grade, it was used mechanical properties for
grade N-80.
We thank the following technicians from
PETROBRAS/CENPES/DIPLOT : Paulo Sérgio da Silva
Horta, Francisco de Paula dos Santos, Raimundo Nonato
Lopes Belfort and Carlos Eduardo L. Chiapim; and from
COPPE/UFRJ : Marcos P. Silva, Marcelo M. Pinheiro and
João Fabrício M. Castilho. They have done a very good
contribution in the development of the experimental part of
this work.
References
1- Conceição, A.C.F. “Revestimento para poços de vapor”,
IV Encontro Técnico de Revestimento e Cimentação, 1992.
2- Maruyama,IK.; Ogasawa,M. and Inoue,Y. “An
Experimental Study of Casing Performance Under Thermal
Recovery Conditions”, SPE 18776, 1989.
3- Pasqualino,I.P.; Estefen,S.P. and Plácido,J.C.R. “Análise
Numérica de Revestimento de Poço de Petróleo sob Cargas
Térmicas e Pressão Externa”, Seminário CAE’96, Rio de
Janeiro, 1996.

c) It must be consider that mechanical properties may vary

even for steel of same grade.

d) It was not possible to obtain a good thermal insulation

between internal and external pipe in the apparatus built at
CENPES. Therefore, the external pipe was also submitted to
some deformation, which was compensated with information
from strain gages placed at one end. However, it is difficult to
know the real effect of the deformation of external pipe in the
free cement section (internal pipe) in order to define correct
boundary conditions.

e) Relaxation is not include in the numerical model.

Conclusions
1) Based in the results obtained with physical simulation and
confirmed with numerical model, it was verified that the
residual tension level, for the same cyclic thermal loading,
increases according to the casing grade. Therefore, the
collapse strength of casing with high yield stress will be more
affected than casing with lower yield stress. For instance,
casing grade K-55 must be used instead of N-80 in cyclic
steam injection wells, however yield stress must be observed.

2) The numerical model obtained a good agreement with

physical simulation and it was considered validated.
Therefore, it can be used as auxiliary tool for casing design in
cyclic steam injection wells.
4 JOÃO CARLOS R. PLÁCIDO, ADEMAR POGGIO JR., LUÍZ A. B. F. PAULO, ILSON P. PASQUALINO, SEGEN F. ESTEFEN SPE 38978

Table 1 - Coefficient of linear expansion

T (
C) (E-6 /
C) (E-6 /
C) (E-6 /
C)
H-40 K-55 N-80 Table 3 - Material Properties (K-55)
20-30 3.9053 1.7292 3.1999 T Area Sy Su EL
30-40 4.5117 1.7292 3.1999 (
C) (mm2) (MPa) (MPa) (%)
40-50 4.8492 1.7292 4.9960 room 98,28 418 630 20
50-60 5.2017 4.5496 5.9639 75 88,75 432 655 21
60-70 5.1460 6.0686 8.2803 125 87,50 405 634 24
70-80 5.5824 6.7131 11.291 175 90,52 406 654 10
80-90 5.6886 8.3797 12.537 225 97,02 423 698 10
90-100 5.7823 10.538 14.457 275 95,00 390 747 22
100-110 6.0713 11.995 15.375 325 93,00 402 710 26
110-120 6.1121 15.774 16.836
120-130 6.4216 16.729 17.966 Table 4 - Material Properties (N-80)
130-140 6.4043 17.431 18.874 T Area Sy Su EL
140-150 6.4872 19.134 19.657 (
C) (mm2) (MPa) (MPa) (%)
150-160 7.1741 20.045 20.806 room 100,80 645 815 19
160-170 9.0541 21.617 21.699 75 102,09 559 766 16
170-180 7.4717 22.201 22.379 125 105,00 543 756 12
180-190 8.6549 22.870 23.410 175 103,32 546 777 14
190-200 7.5521 23.012 23.508 225 100,00 582 846 14
200-210 8.9612 23.677 24.625 275 103,75 550 883 23
210-220 10.119 24.521 24.639 325 101,25 551 860 22
220-230 10.505 24.604 25.204
230-240 10.503 24.617 25.158
240-250 12.457 24.517 25.078
250-260 14.931 25.952 25.882
260-270 18.088 25.463 25.219
270-280 20.040 25.133 25.242
280-290 20.773 24.553 25.263 trapped
fluid
290-300 20.583 24.226 24.923
300-310 21.131 23.185 25.106
310-320 20.968 23.185 24.618

Table 2 - Material Properties (H-40) cem ent

T Area Sy Su EL
(
C) (mm 2
) (MPa) (MPa) (%)
tubing
room 85,68 326 481 35
75 81,90 296 447 24 packer
125 76,25 289 450 14
175 76,25 291 479 22 reservoir
225 83,16 280 505 16
275 76,86 267 521 17
325 80,64 251 530 29
cem ent
where: Sy - Yield Stress
Su - Ultimate Stress
Fig. 1 - Scheme of a steam injection well with free-casing
EL - Elongation
section and fluid trapped
SPE 38978 STRESS ANALYSIS OF CASING STRING SUBMITTED TO CYCLIC STEAM INJECTION 5

800.00

600.00

Stress (MPa)
400.00

200.00

0.00
0.00 0.02 0.04 0.06 0.08 0.10
Stra in

Fig. 4 - Stress-Strain curve for grade K-55 at 25
C

8 0 0.0 0

Fig. 2 - Typical cyclic thermal loading history

(7”, 23 lb/ft, Grade K-55) - ref. 2 6 0 0.0 0
Stress (MPa)

500.00
4 0 0.0 0

400.00

2 0 0.0 0

300.00
Stress (MPa)

0 .00
200.00
0 .00 0 .02 0 .04 0 .06 0 .08
S train

100.00
Fig. 5 - Stress-Strain curve for grade N-80 at 25
C

0.00
0.00 0.02 0.04 0.06 0.08 0.10
S train

Fig. 3 - Stress-Strain curve for grade H-40 at 25
C

Fig. 6 - Experimental apparatus (w/o thermal insulation)

6 JOÃO CARLOS R. PLÁCIDO, ADEMAR POGGIO JR., LUÍZ A. B. F. PAULO, ILSON P. PASQUALINO, SEGEN F. ESTEFEN SPE 38978

4 00.0 0

0 .0 0

Strain ( µε )
-400 .0 0

-8 00 .0 0

Fig. 7 - Experimental apparatus (with thermal insulation) 0 .0 0 1 00.00 2 00.00 3 00.00 4 00.00

Tem perature (°C )

Fig. 11 - Strain vs temperature of one thermal cycle
with casing H-40 (Test # 1)

400.00

S train ( µε ) 0.00

-400.00
Fig. 8 - Experimental apparatus (free-cement section
with strain gage and thermocouple)
-800.00
external pipe
eletric resistor strain gages
internal pipe termocouples

-1200.00
0.00 100.00 200.00 300.00 400.00

Tem perature (°C )

Fig. 12 - Strain vs temperature of one thermal cycle
Fig. 9 - Position of strain gages and thermocouples with casing K-55 (Test # 2)

1 00 0.00
S train ( µε )

0 .0 0

-1 00 0.0 0

Fig.10 - Data acquisition system

-2 00 0.0 0
0 .0 0 1 00 .0 0 2 00 .0 0 3 00 .0 0 4 00 .0 0

T em peratu re (°C )
Fig. 13 - Strain vs temperature of one thermal cycle
with casing N-80 (Test # 3)
SPE 38978 STRESS ANALYSIS OF CASING STRING SUBMITTED TO CYCLIC STEAM INJECTION 7

Fig. 14 - Finite element mesh used in numerical model

3 00 0 .0 0

First C ycle
Secon d C ycle
2 00 0 .0 0
Fig. 16 - Experimental curve (ref. 2): K-55, fixed ends, two
thermal cycles
T h er m a l C o n s tra in t L o a d (K N )

1 00 0 .0 0

0 .0 0

-1 0 0 0 .0 0
3 00 0 .0 0
Grade K -55 T emp. (C) R esidual S tress
(N /mm 2)

-2 0 0 0 .0 0 250 481

2 00 0 .0 0 300 490

354 504
T h er m a l C o n s tra in t L o a d (K N )

-3 0 0 0 .0 0
1 00 0 .0 0
0 .0 0 1 00 .00 2 00 .00 3 00 .00 4 00 .00
T em p e ra tu re ( o C )

0 .0 0
Fig. 15 - Numerical model simulation : K-55, fixed ends, two
thermal cycles (compare to Fig. 16)
-1 0 0 0.0 0

-2 0 0 0.0 0

-3 0 0 0.0 0

0 .0 0 1 00 .0 0 2 00 .0 0 3 00 .0 0 4 00 .0 0
T e m p e ra tu re ( o C )

Fig. 17 - Numerical model simulation : K-55, fixed ends,

three different thermal cycles (compare to Fig. 18)
8 JOÃO CARLOS R. PLÁCIDO, ADEMAR POGGIO JR., LUÍZ A. B. F. PAULO, ILSON P. PASQUALINO, SEGEN F. ESTEFEN SPE 38978

Fig. 20 - Experimental curve (ref. 2): L-80, fixed ends, three

different thermal cycles
Fig. 18 - Experimental curve (ref. 2) : K-55, fixed ends, three
different thermal cycles

4 0 0 0.0 0
G r ad e N -8 0 Te m p . (C ) R e sid u al S tr es s
( N /m m 2 )
3 0 0 0.0 0 25 0 563 0.00
30 0 591

35 4 609
2 0 0 0.0 0
Th erm al C o n strain t L oa d (K N )

G rade H -40 (Fixed E nd)

1 0 0 0.0 0 A xial Strain

6
S train x 10

0 .0 0 -2000 .00

-1 00 0 .0 0

-2 00 0 .0 0

-4000 .00
-3 00 0 .0 0
0 .00 1 00 .00 2 00 .00 3 00 .00 4 00 .00
T em p eratu re (o C )
-4 00 0 .0 0

0 .0 0 1 0 0 .0 0 2 0 0 .0 0 3 0 0 .0 0 4 0 0 .0 0
Fig. 21 - Numerical model simulation : axial strain, H-40,
T em p eratu re (oC ) fixed ends, one thermal cycle (compare to Fig. 11)

Fig. 19 - Numerical model simulation : N-80, fixed ends,

three different thermal cycles (compare to Fig. 20)
SPE 38978 STRESS ANALYSIS OF CASING STRING SUBMITTED TO CYCLIC STEAM INJECTION 9

2 00 0.00

G rad e K -55 (F ixed E nd)

A x ial S train
0 .00
6
S train x 10

-200 0.0 0

-400 0.0 0

0 .0 0 1 0 0 .0 0 2 0 0 .0 0 3 0 0 .0 0 4 0 0 .0 0
T e m p e ratu re (o C )

Fig. 22 - Numerical model simulation : axial strain, K-55,

fixed ends, one thermal cycle (compare to Fig. 12)

6000.0 0

G rade N -80 (Fixed E nd)

4000.0 0
A x ial Strain
6
S train x 10

2000.0 0

0.00

-2000.0 0

-4000.0 0

0 .00 1 00 .00 2 00 .00 3 00 .00 4 00 .00

T em peratu re (o C )

Fig. 23 - Numerical model simulation : axial strain, N-80,

fixed ends, one thermal cycle (compare to Fig. 13)