This paper demonstrates the use of coupled fluid flow and geomechanical fault slip (fault reactivation) analysis to estimate the maximum sustainable injection pressure during geological sequestration of CO2. Two numerical modeling approaches for analyzing faultslip are applied, one using continuum stress-strain analysis and the other using discrete fault analysis. The results of these two approaches to numerical fault-slip analyses are compared to the results of a more conventional analytical fault-slip analysis that assumes simplified reservoir geometry. It is shown that the simplified analytical fault-slip analysis may lead to either overestimation or underestimation of the maximum sustainable injection pressure because it cannot resolve important geometrical factors associated with the injection induced spatial evolution of fluid pressure and stress. We conclude that a fully coupled numerical analysis can more accurately account for the spatial evolution of both in situ stresses and fluid pressure, and therefore results in a more accurate estimation of the maximum sustainable CO2 injection pressure.

In situ fracture mechanical deformation and fluid flow interactions are investigated through a series of hydraulic pulse injection tests, using specialized borehole equipment that can simultaneously measure fluid pressure and fracture displacements. The tests were conducted in two horizontal boreholes spaced one meter apart vertically and intersecting a near-vertical highly permeable fault located within a shallow fractured carbonate rock. The field data were evaluated by conducting a series of coupled hydromechanical numerical analyses, using both distinct-element and finite-element modeling techniques and both two- and three-dimensional model representations that can incorporate various complexities in fracture network geometry. One unique feature of these pulse injection experiments is that the entire test cycle, both the initial pressure increase and subsequent pressure fall-off, is carefully monitored and used for the evaluation of the in situ hydromechanical behavior. Field test data are evaluated by plotting fracture normal displacement as a function of fluid pressure, measured at the same borehole. The resulting normal displacement-versus-pressure curves show a characteristic loop, in which the paths for loading (pressure increase) and unloading (pressure decrease) are different. By matching this characteristic loop behavior, the fracture normal stiffness and an equivalent stiffness (Young's modulus) of the surrounding rock mass can be back-calculated. Evaluation of the field tests by coupled numerical hydromechanical modeling shows that initial fracture hydraulic aperture and normal stiffness vary by a factor of 2 to 3 for the two monitoring points within the same fracture plane. Moreover, the analyses show that hydraulic aperture and the normal stiffness of the pulse-tested fracture, the stiffness of surrounding rock matrix, and the properties and geometry of the surrounding fracture network significantly affect coupled hydromechanical responses during the pulse injection test. More specifically, the pressure-increase path of the normal displacement-versus-pressure curve is highly dependent on the hydromechanical parameters of the tested fracture and the stiffness of the matrix near the injection point, whereas the pressure-decrease path is highly influenced by mechanical processes within a larger portion of the surrounding fractured rock.