Simulation of the impact of faults on CO2 injection into sandstone reservoirs
Article first published online: 8 APR 2013
© 2013 John Wiley & Sons Ltd
Volume 13, Issue 3, pages 344–358, August 2013
How to Cite
Geofluids (2013) 13, 344–358
We examine the impact of faults on the flow of fluid-phase CO2 injected into formation water in sandstone reservoirs. We estimate the permeability for a range of faults and show that low-permeability deformation-band faults may impede CO2 flow and increase the volume ratio of CO2 stored in a compartmentalized reservoir. Conversely, fracture-dominated faults act as bypass conduits and allow CO2 to escape out of the target reservoir, and into overlying rocks or to the surface.
- Issue published online: 23 JUL 2013
- Article first published online: 8 APR 2013
- Manuscript Accepted: 22 FEB 2013
- Manuscript Received: 14 SEP 2012
- U.S. Department of Energy. Grant Numbers: DE-FG03-00ER15043, DE-FG03-00ER15042, DE-FG03-95ER14526
- Carbon Capture Project
- CO2 injection;
Numerical simulations of multiphase CO2 behavior within faulted sandstone reservoirs examine the impact of fractures and faults on CO2 migration in potential subsurface injection systems. In southeastern Utah, some natural CO2 reservoirs are breached and CO2-charged water flows to the surface along permeable damage zones adjacent to faults; in other sites, faulted sandstones form barriers to flow and large CO2-filled reservoirs result. These end-members serve as the guides for our modeling, both at sites where nature offers ‘successful’ storage and at sites where leakage has occurred. We consider two end-member fault types: low-permeability faults dominated by deformation-band networks and high-permeability faults dominated by fracture networks in damage zones adjacent to clay-rich gouge. Equivalent permeability (k) values for the fault zones can range from <10−14 m2 for deformation-band-dominated faults to >10−12 m2 for fracture-dominated faults regardless of the permeability of unfaulted sandstone. Water–CO2 fluid-flow simulations model the injection of CO2 into high-k sandstone (5 × 10−13 m2) with low-k (5 × 10−17 m2) or high-k (5 × 10−12 m2) fault zones that correspond to deformation-band- or fracture-dominated faults, respectively. After 500 days, CO2 rises to produce an inverted cone of free and dissolved CO2 that spreads laterally away from the injection well. Free CO2 fills no more than 41% of the pore space behind the advancing CO2 front, where dissolved CO2 is at or near geochemical saturation. The low-k fault zone exerts the greatest impact on the shape of the advancing CO2 front and restricts the bulk of the dissolved and free CO2 to the region upstream of the fault barrier. In the high-k aquifer, the high-k fault zone exerts a small influence on the shape of the advancing CO2 front. We also model stacked reservoir seal pairs, and the fracture-dominated fault acts as a vertical bypass, allowing upward movement of CO2 into overlying strata. High-permeability fault zones are important pathways for CO2 to bypass unfaulted sandstone, which leads to reduce sequestration efficiency. Aquifer compartmentalization by low-permeability fault barriers leads to improved storativity because the barriers restrict lateral CO2 migration and maximize the volume and pressure of CO2 that might be emplaced in each fault-bound compartment. As much as a 3.5-MPa pressure increase may develop in the injected reservoir in this model domain, which under certain conditions may lead to pressures close to the fracture pressure of the top seal.