Stability Analysis of CO2-Brine Immiscible Flow in Homogeneous Core Samples
Abstract
Carbon Capture and Storage (CCS) in saline aquifers is drawing attention as a potential method to reduce the CO2-level in the atmosphere and hence mitigate the effect of global warming. In order to understand the complicated physics between CO2 and brine at large depths, laboratory core-scale experiments with reservoir conditions can be performed, and the knowledge gained can help us maximize the amount of CO2 stored. This thesis uses results from a previous study by the author (Westergaard, 2012), where capillary pressure and relative permeability saturation relationships were derived for a set of homogenous core samples. The results were used in the present study, for numerical simulations of the multi-phase immiscible flow of CO2 displacing brine. The stability of the flow was investigated through a series of numerical simulation scenarios, where the interplay of viscous, capillary and gravitational forces was evaluated. Initially, an attempt was made to predict the stability of the flow using Buckley-Leverett Frontal Theory without considering capillary pressure and gravity. Predictions of the stability were made using the shock-front mobility ratio for a set of relative permeability curves related to the pore structure of the core samples. The predictions were then evaluated with numerical simulations. Although some of the predictions corresponded well to the simulations, a great number of them did not. Viscous fingering or channeling was seen in most of the simulated cases, and it was difficult to detect a clear separation of stable and unstable flow. Furthermore, new scenarios were made to investigate the stabilizing effect of capillary pressure and the segregation of gravity. Capillarity stabilizes the flow at great flow lengths that are considered realistic aquifer dimensions. The flow is stabilized since capillary pressure suppresses the formation of fingers. Gravitational effects cause a gas override in the displacement for low flow rates, but this effect is diminished as the gravity number decreases, and the viscous regime starts to dominate. The stabilizing effect of capillary pressure is not as powerful at increasing length scales, when gravitational forces are considered. Gas override quickly dominates the CO2-plume migration as the length of the domain increases, and no signs of projected streaks of gas can be seen. This conclusion of this thesis is that the density difference between CO2 and brine should be the main concern for a CO2-sequestration project, and not the viscosity ratio.