Effective constitutive relations for simulating CO2 capillary trapping in heterogeneous reservoirs with fluvial sedimentary architecture

Carbon dioxide (CO2) storage reservoirs commonly exhibit sedimentary architecture that reflects fluvial deposition. The heterogeneity in petrophysical properties arising from this architecture influences the dynamics of injected CO2. We previously used a geocellular modeling approach to represent this heterogeneity, including heterogeneity in constitutive saturation relationships. The dynamics of CO2 plumes in fluvial reservoirs were investigated during and after injection. It was shown that small-scale (centimeter–meter) features play a critical role in capillary trapping processes and have a primary effect on physical- and dissolution-trapping of CO2, and on the ultimate distribution of CO2 in the reservoir. Heterogeneity in saturation functions at that small scale enhances capillary trapping (snap-off), creates capillary pinning, and increases the surface area of the plume. The understanding of these small-scale trapping processes from previous work is used to develop effective saturation relationships that represent, at a larger scale, the integral effect of these processes. Though it is generally not computationally feasible to represent small-scale heterogeneity directly in a typical reservoir simulation, the effective saturation relationships for capillary pressure and relative permeability presented here, along with an effective intrinsic permeability, allow better representation of the total physical trapping at the scale of larger model grid cells, as typically used in reservoir simulations. Thus, the approach diminishes limits on cell size and decreases simulation time in reservoir simulations.