Maximizing Local Capillary Trapping During CO2 Injection

Abstract

Local capillary trapping (LCT) occurs during buoyancy-driven migration of bulk phase CO 2 within a saline aquifer exhibiting spatially varying properties (permeability and capillary entry pressure). When the rising CO 2 plume encounters a region where capillary entry pressure is locally larger than average, CO 2 accumulates beneath the region. The benefit of LCT, applied specially to CO 2 sequestration, is that saturation of stored CO 2 is larger than the saturation predicted for other trapping mechanisms, yet the CO 2 is not susceptible to leakage through failed seals. In this paper, the effects of the injection rate, anisotropy, formation dip, aquifer types and residual gas saturation on the LCT saturation/mass were systematically investigated including the injection period. The domain with heterogeneous capillary pressure was generated by using geostatistical permeability realization and Leverett-J function to assign an individual drainage curve to each grid block. After injection, leakage was simulated that represents localized caprock breach induced by geochemical process or by stress/strain variation within the reservoir/caprock. The trend of the fraction of CO 2 above residual that remains trapped after seal rupture was evaluated. Residual and dissolution trapping were assessed and compared to LCT over time during and after injection. Results indicate that local capillary traps in the near-well region can be fully filled during injection in the domain. Moreover, they remain filled after post-injection buoyancy-driven flow ends. The flow regimes are dependent on the injection rate, but LCT mass fraction is almost insensitive to the CO 2 injection rate (gravity number). This implies that local capillary trapping is primarily determined by the geological model. Anisotropy does not change the LCT structures although it has some impact on the LCT CO 2 saturation and plume dynamics. Formation dip angle is demonstrated to be the most influential parameters on the structures and mass fraction of injected CO 2 held in LCT, as it changes the effective horizontal permeability correlation length. The key finding from this research is that LCT could be as significant as residual trapping in immobilizing CO 2 during the intermediate period in the open aquifer.