Oilwell Cement Shrinkage
Oilwell Cement Shrinkage
Oilwell Cement Shrinkage
September 1998, Volume 37, No. 9 The Journal of Canadian Petroleum Technology
sent the set-up cannot handle downhole pressures. The tempera-
ture influence on the measurements is large: both when heating up
the cell and when the set temperature is reached. When the cell is
heated, the water inside will expand and increase the pressure.
The shrinkage calculations are therefore taken from the point
when the test temperature or pressure levels off and stabilizes.
This has the advantage that it simulates field conditions; any
shrinkage occurring before placement of the cement in a well will
not be taken into account. However, some initial but minor shrink-
age may not be observed. A stable set temperature is important for
the same reason. At 180˚ C, a 1˚ C temperature fluctuation will
lead to an apparent change of 0.3 percentage points in the cement
shrinkage. This influence is reduced at lower temperatures.
FIGURE 2: Total shrinkage of slurry T25. FIGURE 4: Total shrinkage of slurry W90.
FIGURE 3: Total shrinkage of slurry T60. FIGURE 5: Total shrinkage of slurry W140.
the higher the shrinkage rate is. This correlation is to be expected Both peaks again coincide with the highest shrinkage rates, and
as the temperature evolution reflects the rate of hydration and the step-wise hydration rate may be influenced by a number of
hence the shrinkage. Note the two temperature peaks of slurry organic admixtures in the slurries.
W90 and W180 (with the exception of Figure 6), the first being at Some other slurries (not presented here) were tested at 90, 140,
five to ten hours and the secondary peak around 20 – 25 hours. and 180˚ C and they exhibited the same shrinkage and tempera-
T25 T60 W90 W140 W180 A140 B140 C140 D140 E140 F140
Test temperature,˚ C 25 60 90 140 180 140 140 140 140 140 140
Slurry density, g/cc 1.90 1.90 1.90 2.05 2.15 2.05 2.00 2.06 2.07 1.98 1.88
Cement, vol.% 39.48 40.62 39.84 27.37 27.47 26.63 29.73 33.12 33.22 37.67 39.02
Shrinkage at 20h, vol.% 3.44 3.29 2.08 1.50 2.65* 1.79 1.98 2.58 2.49 2.84 3.11
Additive g/cc
API class G cement 3.22 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00
Salt water, lhk** 1.03 45.61
Accelerator, lhk 1.36 2.00
Fresh water, lhk 1.00 45.09 35.70 51.81 43.59 45.52 39.96 44.28 44.28 44.28 44.28
Retarder, type A, lhk 1.20 0.3
Retarder, type B, lhk 1.08 1.20
Anti-gas migration, 1.10 3.50
type A, lhk
Dispersant, type A, lhk 1.18 3.00
Fluid loss control, 1.07 3.50
type A, lhk
Retarder, type C, 1.42 1.00
%bwoc***
Retarder, type D, 1.57 2.00
%bwoc
Fluid loss control, 1.48 0.75 1.00
type B, %bwoc
Weight material, 4.85 30.00 40.00 38.77 13.88 13.88 13.88 13.88
%bwoc
Anti-strength retrogr., 2.65 35.00 35.00 25.97 30.00 30.00
%bwoc
Anti-gas migration, 1.40 10.00 15.00 13.00 12.00
type B, lhk
Retarder, type E, lhk 1.18 1.25 1.25 1.25 1.25 1.25 1.25
Dispersant, type B, hk 1.21 3.00 3.00 3.00 3.00 3.00 3.00
Fluid loss control, 1.04 5.00 3.00
type C, lhk
Calcite flour, %bwoc 2.72 30.00
ture behaviour demonstrating that this is a general trend. The lished results.
behaviour is also confirmed by Lile et al.(9) using permeability, The results show that there is a marked contrast in shrinkage
tensile strength, hydrostatic pressure and temperature as indica- behaviour over the temperature range from ambient and up to
tors. Chenevert and Shrestha(4) tested fresh water slurries at 93, 180˚ C, where the tested slurries at 90, 160, and 180˚ C exhibited
121, and 177˚ C and their medium temperature slurry showed the two temperature peaks and S-shaped shrinkage curves. This dif-
fastest initial shrinkage. ference is most likely a temperature effect on the cement hydra-
The difference in shrinkage and temperature evolution that we tion chemistry.
see in the results may also be due to a temperature effect on the It was confirmed that there is a close correlation between total
cement hydration chemistry. It is well known that curing of chemical shrinkage and cement content. A low shrinkage and a
cement already at 50 – 70˚ C leads to an uneven distribution of short transition period (i.e., fast initial shrinkage) should reduce
hydration products with a densified calcium silicate hydrate the risk of gas migration.
(CSH) layer around the remaining unhydrated cement grains
which eventually will slow down the following hydration rate. It
is possible that, at even higher curing temperatures, such a densi- Acknowledgement
fied layer is formed initially and broken off after some further
moderate hydration (rate increasing with increasing temperature), The authors would like to thank The Research Council of
leading to a renewed acceleration period of the cement hydration Norway, Norsk Hydro, Saga Petroleum and Statoil for supporting
explaining the step-wise hydration rate shown as the S-shaped this work. We also thank Vincent H.J. Bosch for carrying out
shrinkage curves in Figures 4, 6, and 7. some of the experiments.
Given that the shrinkage rate reflects the hydration rate, an ini-
tial fast shrinkage covering the transition period will limit the time REFERENCES
in which gas can enter the cement and thus reducing the likeli-
1. LEVINE, D.C, TINSLEY, J.M., and SUTTON, D.L., Annular Gas
hood of gas migration. This is the case for slurry W140. Flow After Cementing: A Look at Practical Solutions; Paper SPE
8255 presented at the 1979 SPE Annual Fall Technical Conference
Shrinkage versus Cement Content and Exhibition, Las Vegas, NV, September 23 – 26.
2. SABINS, F.L., THOMAS, E.W., BEZNER, H.P., and TOLLE, G.C.,
Elvebakk(10) used the methods of Lile et al. (9) to study the influ-
Transition Time of Cement Slurries Between the Fluid and Set State;
ence of additives on gas migration. He found that the hazard of SPE Journal, Vol. 22, No. 6, pp. 875-882, December 1982.
gas migration generally increased when the recipe was simplified.
3. CHEUNG, P.R., and BEIRUTE, R.M., Gas Flow in Cements;
The same slurries were also tested for shrinkage. The shrinkage
Journal of Petroleum Technology, Vol. 37, No. 7, pp. 1041-1048,
curves of the six slurries A140 through F140 are presented in June 1985.
Figure 8 and the slurry recipes are shown in Table 1. The differ-
ence in shrinkage level was found to be correlated with the 4. CHENEVERT, M.E., and SHRESTHA, B.K., Chemical Shrinkage
Properties of Oilfield Cements; SPE Drilling Engineering, Vol. 6,
cement content as shown in Figure 9 where the straight line fitted No. 1, pp. 37-43, March 1991.
to the data points is forced through zero, as zero cement content
should yield no shrinkage. 5. SABINS, F.L., and SUTTON, D.L., Interrelationship Between
Critical Cement Properties and Volume Changes During Cement
This correlation is self-evident and known(4,5) , but has not Setting; SPE Drilling Engineering, Vol. 6, No. 2, pp. 88-94, June
before been linked to the gas migration problem. As long as the 1991.
strength development of the cement slurry is satisfactory, more
6. JUSTNES, H., VAN LOO, D., REYNIERS, B., SKALLE, P.,
use of an extender will reduce the shrinkage and the risk of gas SVEEN, J., and SELLEVOLD, E.J., Chemical Shrinkage of Oil
migration. And if the extender particles are small enough, they Well Cement Slurries; Advances in Cement Research, Vol. 7, No. 26,
will fill the space between the cement particles, lowering the pp. 85-90, April 1995.
cement permeability and thus reducing the likelihood for gas 7. PARCEVAUX, P.A., and SAULT, P.H., Cement Shrinkage and
intrusion even further. Elasticity: A New Approach for a Good Zonal Isolation; Paper SPE
13176 presented at the 1984 SPE Annual Technical Conference and
Exhibition, Houston, TX, September 16 – 19.
Conclusions 8. STEWART, R.B., and SCHOUTEN, F.C., Gas Invasion and
Migration in Cemented Annuli: Causes and Cures; SPE Drilling
A new and simple method for measuring total shrinkage of cur- Engineering, Vol. 3, No. 1, pp. 77-82, March 1988.
ing oil well cements at elevated temperatures has been developed. 9. LILE, O.B., ELVEBAKK, H., BACKE, K.R., SKALLE, P., and
It has proved to be reliable and sensitive for monitoring the LYOMOV, S., A New Technique for Measuring Permeability and
cement hydration process. The recorded shrinkage at 20 hours Tensile Strength of a Curing Oil Well Cement; Advances in Cement
varied between 1.4 and 3.5 vol.% which is in agreement with pub- Research, Vol. 9, No. 34, pp. 47-54, April 1997.
Authors’ Biographies
Knut Backe is currently a research scien-
tist at the Norwegian University of Science
and Technology (NTNU), Department of
Petroleum Engineering and Applied
Geophysics in Trondheim, Norway. His
main areas of interest are cement, drilling
fluids and electrical geophysical methods.
He received a M.Sc. in petroleum engineer-
ing from NTNU in 1987.