1.

INTRODUCTION
In order for geological carbon sequestration to achieve
substantial reductions of greenhouse gas emissions,
many large injection projects will be required. Each
project is likely to require multiple wells, each injecting
millions of tons of CO
2
over many years. For storage in
saline formations, this is likely to create a large and
increasing pressure anomaly that will grow over the
duration of the injection project. The In Salah Project (a
joint venture of BP, Statoil and Sonatrach) includes a
CO
2
sequestration effort that has successfully injected
millions of tons of CO
2
into a deep saline formation
close to a producing gas field in Algeria [1, 2]. Since
2004, CO
2
has been separated from extracted natural gas
at the Krechba gas field at In Salah, Algeria, and
reinjected along the limbs of the trapping anticline as a
supercritical fluid. Three injectoion wells have been
used, targeting depths on the order of 1.8 km. We have
been jointly funded by the Joint Industry Project (A
consortium consisting of BP, Statoil and Sonatrach,
hereafter referred to as the JIP) and the U.S. Department
of Energy to investigate the role of injection-induced
mechanical deformation and geochemical alteration at
the In Salah CO
2
storage project. In this paper we will
focus upon the hydromechanical portion of this study.
Surface deformation has been observed associated with
the injection at the Krechba field at In Salah via
interferometric synthetic-aperture radar (InSAR). In
addition, CO
2
breakthrough has been observed at a
suspended appraisal well (see Ringrose et al. [3] for
details). The first work regarding surface deformation at
In Salah was to invert the observed surface deformation
to obtain permeability estimates within the Krechba
reservoir [4, 5]. More recently, Rutqvuist et al. [6] used a
sequentially coupled hydromechanical simulation to
model injection into the Krechba reservoir using a model

ARMA 10-307

A Study of Injection-Induced Mechanical Deformation at the In Salah CO
2

Storage Project

Morris, J. P.
*
, Hao Y., Foxall, W., and McNab, W.
Lawrence Livermore National Laboratory, P.O. Box 808, Livermore, CA 94551, U.S.A.
*
Now at Schlumberger Doll Research Center, One Hampshire St, Cambridge, MA, 02139U.S.A.


Copyright 2010 ARMA, American Rock Mechanics Association
This paper was prepared for presentation at the 44
th
US Rock Mechanics Symposium and 5
th
U.S.-Canada Rock Mechanics Symposium, held in
Salt Lake City, UT June 2730, 2010.
This paper was selected for presentation at the symposium by an ARMA Technical Program Committee based on a technical and critical review of
the paper by a minimum of two technical reviewers. The material, as presented, does not necessarily reflect any position of ARMA, its officers, or
members. Electronic reproduction, distribution, or storage of any part of this paper for commercial purposes without the written consent of ARMA
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 acknowledgement of where and by whom the paper was presented.


ABSTRACT: Large-scale carbon capture and storage projects involve injecting CO
2
into a porous, permeable formation that is
overlain by an impermeable caprock. The In Salah Project (a joint venture of BP, Statoil and Sonatrach) includes a CO
2

sequestration effort that has successfully injected millions of tons of CO
2
into a deep saline formation close to a producing gas
field in Algeria. We have performed detailed simulations of the hydromechanical response in the vicinity of the KB-502 CO
2

injector specifically because the morphology of the observed surface deformation differed from that above the other injectors at the
field. Associated with the injection, we have simulated the mm-scale uplift of the overburden and compared the results with
observed deformation using InSAR data. Our results indicate that the best fit is obtained through a combination of reservoir and
fault pressurization (rather than either alone). However, our analysis had to make assumptions regarding the mechanical properties
of the faults and the overburden. These results demonstrate that InSAR provides a powerful tool for gaining insight into fluid fate
in the subsurface, but also highlight the need for detailed, accurate static geomodels.




Fig. 1. InSAR surface relative displacement observed
above the KB-502 well (black line) after one year of
injection. Surface displacements exhibit a distinct two-
lobed structure.
that consisted of homogeneous layers of rock with and
without a vertically oriented fault zone and was able to
match the magnitude, of surface displacement observed
above the KB-501 CO
2
injector. However, in response to
the different morphology of uplift above the KB-502
injector, several teams have developed progressively
more complicated models for deformation associated
with the KB-502 injector [7, 8, 9].
In contrast with previous interpretations of the InSAR
data from Krechba, in this work we present our attempts
to perform forward models based upon the best available
data and compare against observed surface uplift. We
have assumed best estimate values for reservoir and
overburden mechanical properties and fault shear
properties in order to understand mechanical responses
to injection pressure. We have performed detailed
simulations of the hydromechanical response in the
vicinity of the KB-502 CO
2
injector specifically because
the morphology of the observed surface deformation
differed from that above the other injectors at the field
(see Fig. 1). Our results indicate that the best fit is
obtained through a combination of reservoir and fault
pressurization (rather than either alone). However, our
analysis had to make assumptions regarding the
mechanical properties of the faults and the overburden.
These results demonstrate that InSAR provides a
powerful tool for gaining insight into fluid fate in the
subsurface, but also highlight the need for detailed,
accurate static geomodels.
2. COMBINED MULTIPHASE FLOW AND
GEOMECHANICAL ANALYSIS
It is well established that fractures and faults that are
favorably oriented for slip (so-called critically stressed
fractures) tend to provide conduits for fluid flow [10].
Streit and Hillis [11] describe in detail how fault stability
and sustainable fluid pressures can be estimated for a
range of sequestration sites. Wiprut and Zoback [12]
discuss a specific example of fault activation in the
North Sea due in part to elevated pore-pressures. In
addition, many sequestration targets are effectively
closed on one or more sides by non-critically stressed
(impermeable or sealing) faults.
We performed critical stress analyses of the influence of
pore pressure on stability of fault stability within the
Krechba reservoir, using an approach similar to
Chiaramonte et al. [13]. These results used the estimated
in situ stresses corresponding to the KB-502 area. Fig. 2
shows our predictions for fault stability in terms of the
coefficient of friction required for stability and the
change in pore pressure anticipated to induce slip. It is
typically assumed that as the required coefficient of
friction approaches 0.6, the fault fails in shear and
becomes a conduit for flow. For example, it can be seen
that the F12 fault which cuts the KB-502 injector is
predicted to be a flow conduit at relatively low changes
in pore pressure. In contrast, the F9 fault is predicted to
be more stable and potentially act as a flow barrier.
A three-dimensional multiphase flow and transport
model, implemented using LLNLs NUFT simulator, has
been developed for the reservoir using porosity and
permeability data provided by the JIP. NUFT has been
previously demonstrated in the prediction of CO
2
storage
performance [14, 15]. In our reservoir scale modeling we
mainly focused on the KB-502/KB-5 area in order to
understand the early CO
2
breakthrough at the KB-5 and
the observed surface uplift. The preliminary fault map at
KB-502/KB-5 was incorporated into our model. Based
upon the fault stability analysis, several faults were
identified in the vicinity of KB-502 that could be fast
S
Hmax
0MPa
10MPa
0.6
0.0
N
F9 predicted to be
stable: Flow barrier
south of KB-502
F12 predicted to be
conducting

Fig. 2. Critical stress analysis provides overview of mechanical stability. At left: An estimate of coefficient of friction required
for stability Shear failure expected as value approaches 0.6. At right: The estimate of change in pore-pressure that will result in
shear failure of faults.
flow paths and flow barriers. Fig. 3 shows the how the
current model includes these features. This particular
model includes a hypothetical extension of the F12 fault
above and below the reservoir by 200m. All simulations
we have performed that include the conductive F12 fault
feature indicate that the fault leads to early arrival of
CO
2
at KB-5, consistent with observation (Figure 3).
We are interested in understanding the induced
displacement at the surface due to fluid displacement in
the subsurface in order to use the InSAR data to
constrain our model. This requires identifying
geomechanical treatments for the deformation due to
both the fluid in the reservoir and that within the fault. In
the analysis presented here, it is assumed that the
permeability field and mechanical response are not
tightly coupled. As discussed above, a geomechanical
analysis was performed to identify which faults are
permeable features and which are seals. The subsequent
NUFT simulation accommodates this information by
employing constant permeable or impermeable cells
along the fault traces. The NUFT model then predicts the
pore-pressure changes within the reservoir and faults
(see Figure 3).
Changes in pore-pressure induce local strains within the
rock that are transmitted through the overburden to the
surface. It is vitally important that we utilize appropriate
modes of induced deformation if we are to accurately
predict the surface displacements. In this study, we
employed the code SYNEF, which achieves a rapid
prediction of the surface displacement through
superposition of appropriate volume change and tensile
source terms. SYNEF (unpublished) is a general
purpose 3D elastic deformation code based on half-space
Green's functions for tensile, shear and dilatational
dislocation sources [16, 17].
It is clear from Figure 3 that the NUFT model predicts
pressurization of both the reservoir level and fault
portions of the storage domains. This begs the question:
Is the observed surface deformation due to:
1) Pressurization of the reservoir.
2) Pressurization of the fault portion of the storage
domain.
3) Combination from both fault and reservoir.
The geomechanical responses of reservoir and fault are
distinct and must be treated appropriately. Specifically,
the reservoir rock is approximately isotropic and
consequently, locally the reservoir rock will respond to
the increase in fluid pressure by attempting to expand
volumetrically in proportion to the fluid pressure change.
In reality, the induced strain field is more complicated,
but in this work we assume the volume change due to
poroelastic effects within the reservoir results in local
isotropic expansion. The contribution from reservoir
expansion leads only to uplift at the surface (Fig. 4) and
no relative subsidence or double-lobed deformation as
observed in the InSAR data (Fig. 1).
The second hypothesis assumes that pressurization of the
fault induces the observed surface deformation. In
contrast with the reservoir, a fault or other fracture-like
feature corresponds to a very weak plane and,
consequently, has a highly anisotropic response to
hydromechanical stress perturbations. The simplest

Fig. 3. Result of NUFT model, including hypothetical extension of fault F12 into the overburden after 1 year of injection into
the KB-502 injector. The saturation field (top right) indicates that the high permeability of the fault combined with buoyancy
effects allows CO2 to migrate to the top of the fault.
appropriate representation for the mechanical response
of the fault to pressurization is to assume that it
undergoes mode I (tensile) opening. The consequential
surface displacements are shown in Fig. 5. The surface
deformation induced by the fault is both qualitatively
and quantitatively different from that due to
pressurization of the reservoir (Fig 4.). Although the
induced deformation due to the fault results in a dual-
lobed structure at the surface, the spacing of the lobes
and the region of subsidence between them is not
consistent with observation (Fig. 1).
These results indicate that the details of the surface uplift
are captured by neither the fault induced or reservoir
induced displacement alone. The surface deformation
due to the reservoir alone (Fig. 4) lacks the morphology
observed in the InSAR data, although the magnitude is
reasonably correct. Additionally, these results indicate
that the uplift lobes associated with the fault alone (Fig.
5) exhibit greater separation than those in the data.
Additionally, the pressurization of the fault alone leads
to a surface depression which is not observed in the data.
The third hypothesis is that the observations are due to
the combined effect of pressurizing the reservoir and the
fault portion of the storage domain. This solution is
shown in Fig. 6, alongside the corresponding InSAR
data. Combining the influence of the fault with that of
the reservoir has two effects:
a. The peaks of the lobes are brought closer together, to
be more consistent with the data
b. The depression is cancelled by uplift due to the
reservoir pressurization.
Consequently, with this model, it is only when the
hydromechanical effect of the hypothetical vertical
extension of F12 is added to the deformation due to the
reservoir that the surface deformation adopts the shape
and magnitude observed via satellite.
CONCLUSIONS
We have performed detailed simulations of the
hydromechanical response in the vicinity of the KB-502
CO
2
injector in an attempt to explain the morphology of
the observed surface deformation differed from that
above the other injectors at the field. Our analysis took
the best available data for the permeability within the
reservoir and included forward models of CO
2
injection
and hydromechanical response for comparison against
InSAR data. Associated with the injection, we have
simulated the mm-scale uplift of the overburden and
compared the results with observed deformation using
InSAR data. By including conducting and bounding
faults into the model we achieve better agreement with
the observed net uplift at the ground surface, but not the
shape of the observed uplift. However, by including flow
into a hypothetical fault, our simulations better match
the morphology of the surface deformation observed via
InSAR. Our results indicate that the best fit is obtained
through a combination of reservoir and fault
pressurization (rather than either alone).
These results demonstrate that InSAR provides a
powerful tool for gaining insight into fluid fate in the
subsurface. However, we have also identified some of
the limitation of such a methodology. Firstly, our work
indicates that at the depth in question, it is difficult to
determine the precise vertical depth of the fault. In
addition, our analysis had to make assumptions

Fig. 4. Predicted surface deformation above KB-502 due to
reservoir pressurization alone

Fig. 5. Predicted surface deformation above KB-502 due to
fault pressurization alone


Fig. 6. Predicted surface deformation above KB-502 due to
combination of reservoir and fault pressurization (left)
compared with the InSAR observation.

regarding the mechanical properties of the faults and the
overburden, highlighting the need for detailed, accurate
static geomodels. Our future work will focus upon
reducing the uncertainty by including new detailed
overburden data as it becomes available. In addition, tilt
meter and GPS measurements will be included as they
are made available by the In Salah JIP.
ACKNOWLEDGEMENTS
This research was supported with cofounding and data
provided by the In Salah Project JIP (a consortium of
BP, Statoil and Sonatrach). This work was performed
under the auspices of the U.S. Department of Energy by
Lawrence Livermore National Laboratory under
Contract DE-AC52-07NA27344.
REFERENCES
1. Riddiford, F., Akretche, S., and Tourqui, A., 2003,
Storage and sequestration of CO
2
in the In Salah Gas
Project: 22nd World Gas Conference, Tokyo, 7p.
2. Riddiford, F., Wright, I., Bishop, C., Espie, T., and
Tourqui, A., 2004, Monitoring geological storage: The
In Salah gas CO
2
storage project: GHGT-7, Vancouver,
6 p.
3. Ringrose, P., Atbi, M., Mason, D., Espinassous, M.,
Myhrer, O., Iding, M., Mathieson, A. & Wright I.
(2009), Plume development around well 307 KB-502 at
the In Salah CO2 storage site, First Break, 27, 8589.
4. Vasco DW, Ferretti A, Novali F, 2008a, Reservoir
monitoring and characterization using satellite geodetic
data: Interferometric synthetic aperture radar
observations from the Krechba field, Algeria,
GEOPHYSICS 73(6), WA113-WA122.
5. Vasco DW, Ferretti A, Novali F, 2008b, Estimating
permeability from quasi-static deformation: Temporal
variations and arrival-time inversion, GEOPHYSICS,
73 (6), O37-O52.
6. Rutqvist, J., Vasco, D.W. and Myer, L., 2009, Coupled
reservoir-geomechanical analysis of CO
2
injection at In
Salah, Algeria, Energy Procedia, 1, 1847-1854.
7. Morris, J.P., 2009, Simulations of Injection-Induced
Mechanical Deformation: A Study of the In Salah CO
2

Storage Project, Presented at the Society of Exploration
Geophysicists 2009 Summer Research Workshop,
Banff, Canada, August, 2009.
8. Morris, J.P., McNab, W., Carroll, S., Yue, H., Foxall,
W, 2009, Status Report September 2009 Injection and
Reservoir Hazard Management: The Role of Injection-
Induced Mechanical Deformation and Geochemical
Alteration at In Salah CO
2
Storage Project, Lawrence
Livermore National Laboratory Technical Report,
LLNL-TR-418848
9. Vasco, D., Rucci, A., Ferretti, A., Novali, F., Bissell,
R,. Ringrose, P., Mathieson, A., Wright, I., 2009,
Satellite-based measurements of surface deformation
reveal fluid flow associated with the geological storage
of carbon dioxide, Presented at the American
Geophysical Union Fall Meeting, San Francisco
10. Barton, C. A., Zoback, M. D., and D. Moos, Fluid flow
along potentially active faults in crystalline rock,
Geology, Vol. 23, no. 8, 683-686, 1995.
11. Streit, J. E., Hillis, R. R., Estimating fault stability and
sustainable fluid pressures for underground storage of
CO
2
in porous rock, Energy, 29, 1445-1456, 2004.
12. Wiprut D.J. and M.D. Zoback, Fault reactivation,
leakage potential, and hydrocarbon column heights in
the northern North Sea, Hydrocarbon Seal
Quantification (NPF Special Publication) 11, 203-219,
2002.
13. Chiaramonte, L., M.D. Zoback, S.J. Friedmann, and V.
Stamp, CO
2
Sequestration, Fault Stability and Seal
Integrity at Teapot Dome, Wyoming, NETL 5th Annual
Conference on Carbon Sequestration, Alexandria, VA,
Exchange Monitor Publications, 2006.
14. Johnson, J.W., Nitao, J.J., and Morris, J.P., 2005,
Reactive transport modeling of cap rock integrity
during natural and engineered CO
2
storage: in Carbon
Dioxide Capture for Storage in Deep Geologic
Formations: Vol. 2, Thomas, D.C., and Benson, S.M.,
eds., p. 787-813.
15. Johnson, J.W., Nitao, J.J., Morris, J.P., 2004, Reactive
Transport Modeling of Cap Rock Integrity During
Natural and Engineered CO
2
Storage, in Carbon
Dioxide Capture for Storage in Deep Geologic
Formations (Elsevier: ISBN0080445705)
16. Okada, Y., 1985, Surface deformation due to shear and
tensile faults in a half-space: Bull. Seismol. Soc. Am.,
75, 1135-1154.
17. Okada, Y., 1992, Internal deformation due to shear and
tensile faults in a half-space: Bull. Seismol. Soc. Am.,
82, 1018-1040.