New Sub-grid Model for Convective Mixing in Field-Scale \({\textrm{CO}}_2\) Storage Simulation

Solubility trapping involves dissolution of supercritical carbon dioxide (CO\(_\text {2}\)) into the resident brine and is considered an important trapping mechanism for any Carbon Capture and Storage (CCS) project. Previous experimental and numerical studies indicate that density-driven convective mixing can greatly enhance solubility trapping, and is thus a key mechanism to capture when assessing the capacity of industry-scale CCS projects. However, convective mixing is a centimeter-scale phenomenon that is computationally challenging to resolve in standard reservoir simulation that uses coarse grid blocks. Therefore, the goal of this work is to incorporate convective mixing as a sub-grid effect within the traditional formulation of slightly miscible two-phase flow. We do this by adapting the classical model applied in individual grid cells to account for the transient behavior of the underlying convection and the downward transport of dissolved CO\(_\text {2}\)through the cell. The new sub-grid model for convective mixing is a mechanistic formulation based on observations from high-resolution simulations. The model employs a set of non-dimensional parameters, calibrated against 2D simulations that allow it to be applied generally to any reservoir properties. We show that the calibrated sub-grid model is easily implemented in a 3D reservoir simulator and benchmark it against a fully resolved field-scale simulation of CO\(_\text {2}\)injection in a sloping aquifer. The sub-grid model shows a marked improvement in computing the total amount of CO\(_\text {2}\)dissolved over time compared with the classical model for the tested cases. The new implementation is further applied to the openly available model for the Smeaheia storage site in the Norwegian North Sea, to demonstrate the utility of the new model to improve estimates of CO\(_\text {2}\)dissolution and explore parameter sensitivities for realistic storage projects.