Residual trapping of carbon dioxide in aquifers during counter-current flow

Increasing concentrations of carbon dioxide (CO 2) in the earth's atmosphere has initiated a wide range of efforts to mitigate emissions and to capture and sequester CO2. Carbon sequestration in deep saline aquifers provides a potential site for long-term storing large amount of CO2 without significant leakage back to surface. However, the time-scales that characterize the immobilization of CO2 in an aquifer, by mechanisms such as capillary trapping, dissolution and chemical reaction, are not yet fully understood. The key contributions and findings from this research effort are: * A new dimensionless time-scale is proposed to describe residual entrapment of the non-wetting phase during counter-current flow. The new time-scale is demonstrated to outperform existing time-scales for a full range of relevant aquifer settings. * By combining experimental observations with numerical calculations, we demonstrate that the use of co-current relative permeability functions is insufficient to capture the migration dynamics of a non-wetting phase plume in counter-current flow settings.;In this research project, we focus on the time-scale that describes the rate of entrapment of a well-defined CO2 plume during counter-current flow. We use numerical calculations to study the interplay between capillary and gravity forces on migration dynamics. Based on numerical calculations, we propose a scaling analysis that allows us to estimate the rate and amount of entrapment over a range of relevant aquifer settings. In addition, current simulations models are commonly based on sequential displacements experiments (co-current flow) whereas any CO2 that is injected into an aquifer will migrate upwards due to buoyancy in a counter-current flow setting. We test the use of the co-current saturation functions to predict the migration of CO2 plume in an aquifer in the context of the counter-current flow settings. To test the accuracy of numerical calculations, a series of dynamic segregation experiments is performed with an analog brine/isooctane fluid system in a well-defined porous media to observe the changes in non-wetting phase saturation as a function of time. A comparison of the experimental observations with numerical calculations demonstrates that co-current relative permeability is inadequate to represent the migration dynamics and that counter-current relative permeability must be integrated in the simulation of aquifer storage processes.