Pressure build-up during CO2 storage in partially confined aquifers

Abstract Geologic CO 2 sequestration in saline aquifers is an attractive option for GHG mitigation, but an associated risk is pressure buildup in the aquifer, which could increase the probability of fracturing the seal or activating a fault. We report on a series of simulations to quantify this risk factor. This work also evaluates injectivity limitations from the “backpressure” at sealing faults, a potentially important factor in assessing the number of wells (and hence the cost) needed for storage. We simulate injection of CO 2 at rates and durations appropriate for capture from coal-fired power plants. The target formations are deep saline aquifers partially confined by faults. We inject at fixed rate, subject to an upper bound on the injection pressure; none of the simulations shown here reach this threshold. Compositional (Peng–Robinson equation of state) simulations are carried out with CMG’s GEM reservoir simulator with different locations and geometries of sealing faults in aquifers, with several values of rock compressibility. We evaluate two parameters: CO 2 injectivity vs. time, and pressure profile in the aquifer. From the latter we obtain a risk parameter defined as the location of a contour of 50 psi above hydrostatic. Such a parameter is suitable for inclusion in a Certification Framework for geologic storage ([C.M. Oldenburg, S.L. Bryant, J.-P. Nicot, N. Kumar, Z. Yingqi, Certification framework based on effective trapping for geologic carbon sequestration (abs.), in: Seventh Annual Conference on Carbon Capture & Sequestration: Addressing the knowledge, policy, regulatory and technology gaps to expedite CCS deployment, Pittsburgh, Pennsylvania, May 5–8, Abstract #817, 2008.)]. A single sealing fault has little influence on injectivity as long as it is beyond the radial extent of the CO 2 plume. Rock compressibility has negligible influence on injectivity and pressure contours. As the number or proximity of faults increases, the injectivity decreases slightly. In contrast to injectivity, contours of elevated pressures are sensitive to faults. They extend farther as the number or proximity of faults increases, increasing the area of influence and thus the risk of failure (seal fracture, fault activation) significantly. Thus well placement relative to known faults is an important design consideration. The effect of aquifer depth on pressure build-up due to injection is also investigated. The variation of fluid viscosity with pressure and temperature (brine viscosity is much more sensitive than CO 2 viscosity) is the dominant effect on injectivity and pressure build-up. An important overall message is that contours of elevated pressure extend much farther into the aquifer than the CO 2 plume itself. Thus risk assessment that focuses exclusively on CO 2 may underestimate actual project risk.