SPE 56535
Case Studies of Expanding Cement To Prevent Microannular Formation
Baumgarte, C., Thiercelin, M., SPE, & Klaus, D., Schlumberger Dowell
Copyright 1999, Society of Petroleum Engineers Inc.
This paper was prepared for presentation at the 1999 SPE Annual Technical Conference and
Exhibition held in Houston, Texas, 3–6 October 1999.
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
To improve our understanding of cement debonding linked to
microannulus formation and, in particular, to develop better
non-shrinking cements to ensure zonal isolation, a well
simulating annulus experiment has been developed and tests
with expanding cements have been performed. The expansion
experiments demonstrate that expanding cements can prevent
the formation of a microannulus if the cement properties are
designed with respect to the rock properties. A good
application for expanding cements is the isolation of gas
storage reservoirs. Two examples are shown to demonstrate
the efficiency of expanding cements.
Introduction
Cement in oil and gas wells is placed in the annular gap
between the drilled formation and the steel casing. The main
function of the cement is to prevent any fluid communication
between these drilled formations to provide zonal isolation.
Zonal isolation has to be achieved during the life of the well
and after its abandonment. Even when cement is correctly
placed in the well and provides initially a good hydraulic seal,
zonal isolation often disappears over time, which is in most
cases caused by microannular formation 1, 2.
Microannuli can either be formed between the casing and
the cement (inner microannulus) or between the cement and
the formation (outer microannulus). They can for example be
detected by the loss of cement bond log response (CBL) or by
observations of gas migration problems. An inner
microannulus can be created for instance by the radial
displacement of the casing resulting from wellbore
temperature or/and pressure changes; especially when the
wellbore pressure is decreased (change in mud weight for
example) once the cement has set.
It is also known 3, 4, 5 from the literature that an outer
microannulus can be generated by cement bulk shrinkage. For
example, in a previous work 5 we presented a new
experimental set up (“annulus experiment”) and demonstrated
the consequences of cement shrinkage. In these experiments
shrinkage had been forced by exposing the cement to water
desaturation. It was shown that bulk shrinkage can lead in the
worst case to the debonding of the cement from the bore hole
wall. An outer microannulus of 300µm ± 100µm had been
generated, and tensile cracks, which destroy the cement
integrity, had been observed through the whole cement sheath.
Expanding cements have been proposed to prevent the
formation of a microannulus 6. The concept is that the
expanding cement will fill any gap and ensure good bonding
either between the formation and the cement or between the
casing and the cement.
Some authors 7, however, have claimed that the expansion
of these cements always moves towards the formation and not
in the direction of the casing. Therefore expanding cements
could only be used to prevent outer microannular formation.
Also, the generation of tensile cracks due to tensions in the
cement sheath could be prevented.
In this study we will focus on the behavior of the cement
sheath with the recently developed annulus experiment to
determine whether expanding cements provide a viable
approach to microannulus prevention, and whether one could
define the optimum conditions for the successful use of such
cements.
We will first review the mechanisms associated with
shrinkage and expansion, present our experimental set-up and
experiments, which will lead to a better understanding of the
mechanisms of expansion. Recommendation on the use of
expanding cements will be provided. Finally we will present
some case studies which demonstrate the use of such cement
systems.
Shrinkage and Expansion Mechanisms
Shrinkage and Expansion in cement result from the formation
of hydration products having different volumes compared to
the hydrating components. The changes in external cement
sample dimensions are referred to as bulk shrinkage and bulk
expansion 8.
2 BAUMGARTE, C., THIERCELIN, M. & KLAUS, D.
Cement Chemical Shrinkage. Cement chemical shrinkage or
cement chemical contraction is a basic mechanism during the
hydration of Portland cement. The volume of the hydrating
components such as water and cement powder is larger than
the volume of the hydration products. This volume contraction
is referred to as the inner shrinkage or total chemical shrinkage
of the cement. Total chemical shrinkage can be measured by
placing cement slurry in a reservoir under free access to water.
The amount of water absorbed by the cement during the
hydration corresponds to the amount of the total chemical
shrinkage. These experiments often lead to an underestimation
of the total chemical shrinkage due to the continuously
decreasing permeability of the cement during hydration, which
hinders the water from penetrating completely through the
developing network of hydrates. The total chemical shrinkage
has been estimated on approximately 6.25ml/100g of cement
by assuming a 100% hydration 9.
The total chemical shrinkage resulting in external volume
changes of a cement specimen, is called bulk shrinkage. Bulk
shrinkage can be measured for example by placing cement
slurry in an impermeable rubber membrane, which is then
sealed. The volume variations of the sample during the
hydration correspond to the bulk shrinkage 4. Several
authors 3, 4, 10 using different methods (at different conditions)
have measured volumes varying between 0.5 and 5%. The
results show that the magnitude of bulk shrinkage depends
strongly on the environmental conditions 7. Free access to
additional water might for instance compensate for the bulk
shrinkage, so that no visible volume reduction of the cement
sample occurs. In contrast, no access to free water in addition
to pressure applied on the cement slurry from the surrounding
medium may lead to excessive bulk shrinkage.
Bulk shrinkage occurs mainly in the first period of the
hydration of Portland cement. In this period no rigid network
of hydrates has been built and the bulk shrinkage follows the
total chemical shrinkage 11. At the stage where the skeleton or
network of hydrates is strong enough to resist contracting
forces, the bulk shrinkage levels out and stops, whereas the
total chemical shrinkage still continues.
Chemical shrinkage is a linear function of the percentage
of the four major clinker minerals 12 and consequently
dependent on the cement class. A slight increase of the
chemical shrinkage with increasing water/cement-ratio has
been observed by Geiker & Knudsen (1982) 13. The
water/cement- ratio influences the magnitude and the rate of
the chemical shrinkage. The shrinkage rate rises also with an
increase of the curing temperature.
Cement Expansion. Cement expansion is an increase in bulk
volume of the initial cement volume, which can be achieved
by the addition of cement expanding agents to Portland
cement. Different types of expansion agents have existed for
several years. Some agents are based on calcium sulfate,
calcium sulfate hemihydrate (Plaster of Paris) or sodium
sulfate to form Ettringite 14. These expanding cements exhibit
more than ten times the expansion than that which a Portland
SPE 56535
cement slurry with salt would reach 6. Other expanding
cements are known which contain calcium and magnesium
oxides.
The expansion of the cements is related to chemical and
mineralogical changes, which results from the hydration and
crystallization of the agents. The magnitude of expansion
depends on the amount of expanding agent, cement powder,
slurry design and curing conditions (pressure, temperature).
Although the cement expands, it still sustains total chemical
shrinkage, but the bulk shrinkage will be superposed by bulk
expansion.
It was initially believed 6 that cement expansion would
always move towards the casing, but measurements of the
degree of acoustic coupling between the cement and the casing
indicated the fallacy of this idea. Beirute et. al. (1988)
assumed that expanding cement could improve the coupling of
the cement-formation interface by expanding outwards. As a
result the microannular formation could be prevented.
However, two drawbacks can be encountered using expanding
cement: one is the damage to the formation due to excessive
expansion, and the other is the risk of creating an inner
annulus in case of an unconsolidated formation 14.
Expanding Cement
Linear Expansion. The expanding cements used in this study
were generated by adding 10% BWOC of a calcium oxide-
based expanding agent to a conventional Class G Portland
cement.
Slurry Rheology. The slurry rheology is calculated from the
rheological parameters at test conditions measured with a Fann
35 viscometer applying the Bingham plastic model. The
parameters are shown in Table 1.
Table 1-Rheology Parameters
Plastic viscosity after mixing 69 [cp]
Yield point after mixing 7 [lb/100ft2]
Plastic viscosity after conditioning 79 [cp]
Yield point after conditioning 7 [lb/100ft2]
Mechanical Properties. Crush test measurements give values
of about 41 MPa for the compressive strength of the
expanding cements used in this study. The Young’s modulus,
determined from the best fitting curve of the linear part of the
stress strain plot, reaches values of about 6688 MPa.
Annular Expansion Mold
Experimental Set Up. The linear expansion of the cement
slurries has been measured with annular expansion molds. The
mold consists of a bottom plate, a split expandable ring with
two attached pins and a top plate. The expandable ring is
placed between the two plates, and a screw fixes the two plates
together.
SPE 56535 CASE STUDIES OF EXPANDING CEMENT TO PREVENT MICROANNULAR FORMATION
Carrying Out A Test. The molds were cured in a water bath
at a specific temperature before being filled with slurry. After
pouring the slurry into the mold, the initial distance (D1)
between the two pins on the split expandable ring is measured
with a micrometer as a zero-reading, then the molds are placed
in the preheated water bath at a specific temperature. The
slurry is in contact with water during the entire test. When the
cement sets and expands, the outside diameter of the
expandable ring will grow and the distance between the
attached pins will increase. After curing the mold the distance
(D2) between the two pins is measured in the same manner
again as the zero-reading. The linear expansion of the slurry is
calculated from the difference of the readings by multiplying
this value times a constant corresponding to the circumference
of the mold.
Slurry Preparation. The investigated cement slurries had a
density of 16.4 ppg and were generated with Class G cement,
mixed with fresh water at room temperature according to API
specifications. The expanding agent was dryblended with the
cement.
Annulus Experiment
Experimental Set Up. To measure the bulk expansion of
expanding cements and to investigate their cement sheath
behavior, annulus expansion experiments have been
performed. The experimental set up (Fig. 1) consists of two
concentric steel cylinders of different diameters (inner
cylinder: ∅i=111 mm, ∅o=113 mm; outer cylinder: ∅i=196
mm, ∅o=198 mm), 80 mm in height. The smaller cylinder
represents the casing (although very thin to allow for strain
measurement) and the larger one represents the formation.
Strain gages are mounted on the cylinders to monitor their
tangential deformation, and the consequent radial deformation
and displacement of the cylinders. Three gages are mounted
on the inner surface of the inner cylinder and three on the
outer surface of the outer cylinder, at 120° to each other.
The cylinders are fixed to a steel plate placed in an oven.
The cement slurry is then poured into the annular space. A
temperature sensor is also attached to the inside of the set-up
to measure the temperature in the cement and to detect when
the hydration process of the cement is completed. The strain
gages are connected via a Wheatstone bridge to a data
acquisition unit.
To simulate different geological formations, the thickness
of the outer cylinder has been varied during the tests between
1 mm and 10 mm (∅i=196 mm, ∅o=216 mm). The 1 mm
cylinder thickness corresponds to an unconsolidated sandstone
and the 10 mm to a harder Sandstone, as it will be
demonstrated in the discussion section.
Carrying Out A Test. The oven is preheated to the test
temperature until the signals of the gages are constant. At this
point the gages are reset to zero to avoid any influence of the
steel expansion on the gage readings. The prepared slurry is
poured into the annular gap between the two cylinders. The
3
slurries are covered with an approximately 10 mm thick water
layer to have free access to water. An additional oil layer was
added to avoid water evaporation (drying shrinkage). All the
tests have been performed at 170°F.
Slurry Preparation. 16.4 ppg cement slurries of Class G
cement were mixed with fresh water at room temperature and
conditioned at 170°F according to API specifications. The
expanding agent was dryblended with the cement.
Performed Tests. The expansion tests were mainly focused
on demonstrating the influence of the formation strength on
the expansion behavior of the cement sheath. Two different
test set-ups have been used:
Set up A: 1 mm outer cylinder thickness
Set up B: 10 mm outer cylinder thickness
Results: Annular Expansion Mold
The development of linear expansion of the investigated
expanding cement versus time is shown in Fig. 2. In this
example, all tests have been performed at 170°C with a
concentration of 10% BWOC of the expanding agent. Values
of linear expansion of almost 3% were reached at test
conditions after two weeks. The expansion starts after the
cement setting.
Results: Annulus Experiment
Set Up A. The results of the test with set up A are shown in
Fig. 3 and Fig. 4. In set up A the outer casing represents a soft
formation, for instance an unconsolidated sandstone. The
radial strain measured on the inner cylinder stayed constant at
about zero during the entire test. This means that the cement
sheath did not apply much stress on the inner cylinder, the
bonding between the cement and the casing was poor.
In contrary the gage signals of the outer cylinder increased
with time and reached maximum values of about 500 µstrain
after 7000 mn. The cement expanded and applied a pressure
on the outer cylinder, expressed in strain. The expansion of the
cement started after the setting of the cement.
The result of this experiment supports and proves the
theory of several authors 6, 13 that cement expands towards the
formation. The test results lead to the conclusion that the use
of expanding cement against soft formations, which cannot
resist surface-applied stress, bears the risk of creating an inner
microannulus.
Set Up B. The results of the experiment with Set up B, in
which the outer casing represents a hard formation, are
demonstrated in Fig. 5 and Fig. 6. Although the same
expanding cement has been used in the same test conditions as
in set up A, the results are significantly different. On the outer
cylinder a maximum strain of about 80 µstrain was measured,
whereas the signals of the inner gages increased to maximum
values of about 270 µstrain.
4 BAUMGARTE, C., THIERCELIN, M. & KLAUS, D. SPE 56535

The expansion still applies a pressure on the outer cylinder, 1

but as the cylinder is too rigid to be displaced and resists the ur = − P1 r1 ……………………………………..….5
2Gr
expansion forces of the expanding cement, the expansion is
forced to move towards the inner cylinder. Under these Equating these two solutions gives:
conditions the expanding cement is pre-stressed and tightened
against the formation and the casing. Expanding cement is Gr =
(
G r22 − r12 ) ……………………………………6
therefore a good solution to get a good tight hydraulic seal, if (1 − 2ν )r12 + r22
used against hard rock formations.
Therefore, using G = 77 000 MPa and ν = 0.3 for steel, one
Discussion gets for the 1 mm thick cylinder a shear modulus value of Gr =
The elastic hollow cylinder solution (Fig. 7), which gives the 1112 MPa, the shear modulus of a very soft sandstone, and for
state of stress and displacement in the cylinder from the the 10 mm thick cylinder a shear modulus value of Gr= 10230
knowledge of the radial stresses at the inner and outer MPa. This is typical for the shear modulus of a medium to
boundaries, is used to interpret the results of the annulus hard rock.
experiment. As shown in Fig. 7, r1 and r2 are respectively the Note that the shear modulus of the expanding cement
inner and outer radius of the hollow cylinder, and P1 and P2 cured after three days is of the order of 2800 MPa, i.e.
are respectively the value of the radial stress at r1 and r2 . This between that of the equivalent rock for a 1mm thick cylinder
solution can be applied to both steel cylinders. The solution and that of the equivalent rock for a 10 mm cylinder.
cannot be applied to the cement during the hydration process, Qualitatively one could say that in order to have the expanding
when the cement behaves more like a compacting soil, with cement moving toward the inner casing, it needs to have a
changing properties over time, than an elastic material. The shear modulus value less than that of the surrounding rock.
inner cylinder and outer cylinder have respectively a zero The above equations allow us to estimate the radial stress
imposed radial stress condition on the inner surface and a zero at the cement steel interfaces from the strain measurements
imposed radial stress condition on the outer surface. obtained on the outer radius of the outer cylinder and on the
Continuity of radial stress and radial displacement between the inner radius of the inner cylinders. From the outer cylinder one
cylinders and the cement applied unless a microannulus is finds:
created.
εθ (r = r ) = σ r (r = r
(1 − ν ) (r22 − r12 ) .…..……...………….7
The two dimensional plane strain solution is used to 2 )
1 G r12
estimate the state of stress and displacement in each ring of
material. The radial stress as a function of radius in a hollow and for the inner cylinder:
cylinder is:
(1 − ν ) (r22 − r12 )
σr =
(P r 2
− Pr 2
) (P − P )r r 2 2
1 εθ ( r = r ) = σ r ( r = r ..………………..…….8
) − (r − r )
1 )
r22
(r
2 2 11 2 1 2 1 2 G
…….…………..…….1
2
2
−r 1
2
2
2
1
2
r2
which gives for the case where the outer cylinder thickness is
The tangential stress is: 1 mm, a radial stress on the outer boundary of the cement of

(P r )
about:
2
− P1r12 (P2 − P1 ) r22 r12 1
σθ = +
(r ) ( )
2 2
……..………….…..…2
2
2
− r12 r22 − r12 r2 σ r ≈ −1.1 MPa

and the radial displacement is: and no stress on the inner boundary. For the case where the

u=
(
1 − 2ν P r − P r r2
+
2
)
1 (P2 − P )r r 1 2 2 outer cylinder thickness is 10 mm, a radial stress on the outer
( ) ( )
2 2 11 1 2 1
.………….....3 boundary of the cement of about:
2G r −r 2
2
2G r22 − r
1
2
r 1
2

This set of equations can first be used to estimate what rock σ r ≈ −1.9 MPa
elastic properties the outer cylinder simulates. To carry out
this estimation, one has to compare the displacement of the and a radial stress on the inner boundary of about:
cylinder at its inner boundary with that of a semi-infinite rock,
under a pressure P1. For the steel one has: σ r ≈ −1.0 MPa

us =
1 − 2ν − P1 r1(
3
+
2
)
1 (− P1 ) r2 r1
( ) ( )
…………..…….4 Note that a compressive radial stress on the outer cylinder
2G r22 − r12 2G r22 − r12 will increase the tangential stress of about the same amount. In
For the rock, with r2 being infinite: this case the outer material is a semi-infinite material instead
of the outer steel cylinder. This increase of the tensile stress
creates concern of tensile failure of the rock while using
expanding cement. 1 MPa tensile stress on a 1000 MPa shear
SPE 56535 CASE STUDIES OF EXPANDING CEMENT TO PREVENT MICROANNULAR FORMATION
modulus rock is actually sufficient to cause concern, the rock
would have been under no confining pressure. However, in
downhole conditions, the compressive state of stress imposed
by the far-field condition and the near wellbore stress
concentration removes any concern of rock failure. Another
advantageous behavior is that, in the case of a hard rock, the
cement tends towards a compressive state of stress. Therefore,
since cements, like rocks, are pressure sensitive materials with
an increase of compressive strength with increase of confining
pressure, and a compressive strength about ten times higher
than the tensile strength, an expanding cement will also
become much more resistant than a non-expanding cement to
any pressure or temperature increase in the wellbore.
Finally it should be noticed that the amount of expansion is
a strong function of the boundary conditions. Linear expansion
of the expanding cements is one order of magnitude higher
than that of the cement when placed in the annulus
experiment. The chemical process of mineral growth is
strongly controlled by the state of stress, and mineral growth
will tend to occur where the stress value is the lowest, i.e. in
the pores or empty spaces. Expanding cements will therefore
be more efficient at filling gaps (dehydrated mud cakes, mud
layers, channels, already formed microannulus) than
compensating for a microannulus formed once the expansion
is completed. On the other hand, expansion lasts for days,
even weeks, which enables the cement to compensate for
change of mud weight during the drilling of the subsequent
section.
Case Studies
Application. One of the main applications of expanding
cements is their use in the cementation of gas wells. In
particular, various gas producers are now involved in the
storage of gas in downhole depleted reservoirs or in cavities.
Good zonal isolation of these reservoirs is required since the
gas flows easily towards the surface as soon as a microannulus
is present. Cement treatments which fail to provide adequate
zonal isolation may require expensive workover to stop this
unacceptable gas flow. The first main objective for gas storage
wells is to cement a production casing perfectly gastight.
Moreover, reservoirs for gas storage often have a salt
formation as a cap rock. An artificial reservoir could have
been made in the salt itself by creating cavities. Cementing in
salt zones often causes problems due to the fact that salt often
needs to be drilled with a high mud weight. The depleted
reservoir, however, needs to be drilled with a low mud weight.
If the casing has been set and cemented in the salt zone with a
high density fluid, then the subsequent drilling of the reservoir
will be made with a much lower mud weight. A decrease of
wellbore pressure can lead to the creation of a microannulus
between the casing and the cement.
Another risk of creating a microannulus in salt zones is
related to the dilution of the slurry by the salt. Experiments
demonstrated that cement slurries which were pumped across
a salt section and returned to the surface, gained more than 4%
salt 6.
5
This area provides an attractive application for expanding
cements, as they generate a good bonding to the casing and the
formation (salt is stiffer than cement), and salt can be used in
high concentrations in the cement slurry. The second main
objective is therefore to ensure that the cap rock is also
perfectly isolated.
In the following examples an expanding cement, which
was generated with the same expanding agent as in the
experiments described above, has been used to set production
casings across salt formations in order to obtain access to
underground gas storages.
Example 1: Wingas underground storage Rehden. The
underground gas storage Rehden operated by Wintershall for
Wingas is located in a Dolomite formation. On top of the
storage formation a massive Zechstein salt formation is found.
The drilled wells are extended reach wells with a long
horizontal section. The 12 ¼ ” open hole section of the wells
are highly deviated up to 85 °. They are completed with an
open perforated 7” liner. Attaining a gastight cementation for
the production casing was a key point for these wells.
The 13 3/8” casing was run as a liner and cemented with a
low density slurry, because of total losses occurring in a
Sandstone formation (Bundsandstein).
The 9 5/8” production casing was set at around 3000-
3400 m (TVD 2000 m) with a BHST of around 85-95°C. Two
different cement systems were used: a light weight slurry with
a density of 1.45 kg/l was pumped as a lead slurry and an
expanding cement as a tail slurry. The expanding cement
system was salt saturated with a density of 1.94 kg/l. The
slurries were batchmixed prior to the job. The top of the
cement was 100 m above the 13 3/8” liner. A saltwater spacer
was pumped ahead. The cement job ran as planned. The
displacement rate was 3000 l/min. USIT and CBL logs
showed excellent bonds with lead slurries for the wells.
Example 2: EWE Underground Gasstorage Huntorf. EWE
has a underground storage Huntorf in the area of Oldenburg in
north-western Germany. These storage wells were built as
caverns which were created by dissolving salt pumping
freshwater through a massive salt formation. EWE
recommended a gastight cementation of the 13 3/8”
production casing.
For this section the cementation had to be cemented up to
surface. Therefore the job was done as a stab-in cementation.
The casing was set at 1050 m, around 500 m deep into the salt
formation (BHST 55°C). Expanding cement was used as a tail
slurry for the 13 3/8” production casing and a Portland cement
as a lead slurry. Both slurries were salt-saturated. The
displacement rate was 1500 l/min. CBL log have not been
done yet, but no gas migration problems in the annulus have
been observed after one year to date.
6 BAUMGARTE, C., THIERCELIN, M. & KLAUS, D.
Conclusion
It was shown by annulus expansion experiments that the use of
expanding cement systems in soft formations (e. g.
unconsolidated sandstone) may run the risk of creating an
inner annulus between the cement and the casing string, as the
expansion moves radial outwards in the direction of the least
resistance, in this case towards the simulated unconsolidated
sandstone.
Expanding cement systems will therefore work best in hard
rock formations, which are able to resist the expansion forces
of the expanding cement. As a result the cement is pre-stressed
and a good hydraulic seal is achieved.
Expanding cement only begins to expand after the cement
has set, which can take days. The quality of the bonding
between the cement/formation and cement/casing will
therefore increase with time.
The shear moduli of the rock and the cement are important
parameters determining whether the cement will expand only
towards the formation or towards both the formation and the
casing.
Case studies demonstrated the efficiency of expanding
cements used in salt formations
Acknowledgments
The authors wish to thank Schlumberger Dowell, Wingas and
EWE for permission to publish this paper.
Nomenclature
r1 =inner radius of hollow cylinder
r2 =outer radius of hollow cylinder
P1 =radial stress at r1
P2 =radial stress at r2
σr =radial stress in a hollow cylinder
σθ =tangential stress in a hollow cylinder
Gr =shear modulus of the rock
G =shear modulus
us =radial displacement of steel
ur =radial displacement of rock
v =Poisson’s ratio
BHST =Bottomhole Static Temperature
BWOC =By Weight Of Cement
References
1. Goodwin, K.J. and Crook, R.J.: “Cement Sheath Stress
Failure,” SPEDE (Dec. 1992) 291-296.
2. Jackson, P.B. and Murphey, C.E.: “Effect of Casing Pressure
on Gas Flow Through a Sheath of Set Cement,” SPE 25698
presented at the 1993 SPE/IADC Drilling Conference,
Amsterdam, Feb. 23-25.
3. Justnes, H., Reynier, B., van Loo, D. and Sellevold, E.J.: “An
Evaluation Of Methods for Measuring Chemical Shrinkage of
Cementitious Pastes,” Nordic Concrete Research, Publication
No. 14, 1994, 45-61.
SPE 56535
4. Justnes, H., van Loo, D., Reynier, B., Skalle, P., Sveen, J. and
Sellevold, E.J.: “Chemical Shrinkage of Oil Well Cement
Slurries,” Advances in Cement Research (Apr. 1995), 7, No. 26,
85-90.
5. Thiercelin, M., Baumgarte, C. and Guillot, D.: “A soil
mechanics approach to predict cement sheath behavior,” SPE
47375 presented at the SPE/ISRM Eurock held in Trondheim,
Norway, Jul. 8-10, 1998.
6. Seidel, F. A. and Greene, T. G.: “Use of Expanding Cement
improves Bonding and Aids in Eliminating Annular Gas
Migration in Hobbs Grayburg-San Andres Wells,” SPE 14434
presented at the 60th Annual Technical Conference and
Exhibition held in Las Vegas, NV, Sept. 22-25, 1985.
7. Beirute, R. M., Wilson, M. A. and Sabins, F. L.: “Attenuation
of Casing Cemented with Conventional and Expanding
Cements across Heavy-Oil and Sandstone Formations,” SPE
18027 presented at the SPE Annual Technical Conference and
Exhibition held in Houston, Oct. 2-5, 1988.
8. De Rozieres, J.: “Shrinkage and Expansion of Oil Well
Cements,” Report of the API work group on shrinkage,
February 1995.
9. Powers, T. C.: “Structure and Physical Properties of Hardened
Cement Paste, “ J. Am. Ceram. Soc. (1958), 41, No 1, 1-6.
10. Parcevaux, P. A. and Sault, P. H.: “Cement Shrinkage and
Elasticity: A New Approach for a Good Zonal Isolation,” SPE
13176 presented at the 59th Annual Technical Conference and
Exhibition held in Houston, Texas, Sept. 16-19, 1984.
11. Setter, N. & Roy, D. M.: Mechanical Features of Chemical
Shrinkage of Cement Paste. Cement and Concrete Research,
(1978), 8, No. 5, 603-610.
12. Powers, T. C.: Ind & Eng. Chem. (1935), 790.
13. Geiker, M. and Knudsen, T.: “Chemical Shrinkage of Portland
Cement,” Cement and Concrete Research (1982), 12, No 5,
603-610.
14. Moran, L. K., Murray, T. R. and Moyer, W. R.: “Cement
Expansion: A Laboratory Investigation,” SPE 21685 presented
at the Production Operations Symposium held in Oklahoma
City, Oklahoma, April 7-9, 1991.
SI Metric Conversion Factors
in. × 2.54* E−02 = m
(oF-32) / 1.8* E+00 = oC
psi × 6.894 757 E+00 = kPa
ft ×3.048* E-01 = m
ppg 1.198264* E+02 = kg/m3
*Conversion factor is exact.
SPE 56535 CASE STUDIES OF EXPANDING CEMENT TO PREVENT MICROANNULAR FORMATION 7

0.6

0.5

0.4

Strain [mstrain]
0.3

0.2

0.1

strain gage 0

−0.1
0 1000 2000 3000 4000 5000 6000 7000
Time [mn]

113 mm Fig. 3.—Results of test with Set up A, 1mm outer casing.

198 mm Strain as a function of time. () : outer gages, (---) : inner
gages.
Fig. 1.—Set up A of the annulus experiment. The outer cylinder
has a thickness of 1mm.
85

84

83

3.5 Temperature [C] 82

81
3
80
2.5 79
Linear Expansion [%]

78
2
77
1.5
76

1 75
0 1000 2000 3000 4000 5000 6000 7000
Time [mn]
0.5

0 Fig. 4.—Results of test with Set up A, 1mm outer casing.

0 5 10 15 20
Slurry and oven temperature as a function of time. () :
Time [day]
oven temperature, (---) : slurry temperature.
Fig. 2.—Development of linear expansion versus time
measured with annular expansion mold.
8 BAUMGARTE, C., THIERCELIN, M. & KLAUS, D. SPE 56535

0.05

0
Strain [mstrain]

−0.05
P2
−0.1 P1 r1 r2 r
−0.15

−0.2

−0.25

0 1000 2000 3000 4000 5000 6000 7000

Time [mn]

Fig. 5.—Results of test with Set up B, 10mm outer casing.

Fig. 7.—The hollow cylinder geometry.
Strain as a function of time. () : outer gages, (---) : inner
gages.

85

84

83

82
Temperature [C]

81

80

79

78

77

76

75
0 1000 2000 3000 4000 5000 6000 7000
Time [mn]

Fig. 6.—Results of test with Set up B, 10mm outer casing.

Slurry and oven temperature as a function of time. () :
oven temperature, (---) : slurry temperature.