HYDRO-THERMO-CHEMO-MECHANICAL MODELING OF CARBON DIOXIDE INJECTION IN FLUVIAL HETEROGENEOUS AQUIFERS

Understanding geological sequestration of carbon dioxide (CO₂) requires fully-coupled simulation of thermal-hydrological-mechanical-chemical (THMC) processes within well-characterized subsurface heterogeneity. A CO₂ injection pilot project was performed in the Cranfield site, Mississippi, USA, where subsurface heterogeneity reflects fluvial deposition. During the project, CO₂ was injected through an injection well and was monitored using two observation wells. We incorporate high-resolution three-dimensional heterogeneity model of the site into multiphase and multi-component flow and transport models. We validate our models and evaluate how bottom-hole pressure (BHP) in injection well, CO₂ breakthrough times at observation wells, and efficiency of trapping mechanisms (i.e., dissolution, snap-off trapping, mobile CO₂) are controlled by (1) non-isothermal CO₂ injection, (2) geochemical reactions, (3) capillary pressure heterogeneity, (4) geomechanical effects, and (5) permeability enhancement close to the injection well. Results suggest that neglecting thermal effects lowers BHP, shortens breakthrough times, overestimates dissolution, and underestimates snap-off trapping. The BHP remains unchanged, breakthrough times are overestimated, and solubility and snap-off trapping are underestimated when geochemical reactions are ignored. Ignoring capillary pressure heterogeneity results in underestimation of BHP and dissolution and overestimation of breakthrough times. Ignoring geomechanical process results in lower BHP, shorter breakthrough times, increased dissolution, and decreased snap-off trapping. Ignoring permeability enhancement results in underestimation of breakthrough times and solubility trapping, and overestimation of snap-off trapping. Our results have important practical implication for designing effective field scale experiments and numerical simulations of geological carbon sequestration. Other multiphase flow problems in which subsurface heterogeneities are the main controlling factor (e.g., hydrogen storage) may benefit from this work.