Recent Progress in Predicting Permeability Distributions for History Matching Core Flooding Experiments

Abstract

Abstract Laboratory core flooding experiments coupled with CT scanning has been shown to be very useful for examining CO 2 -brine displacement processes. These experiments can be used to measure core average properties such as absolute and relative permeability, and also to examine sub core-scale saturation and porosity distributions. By examining the sub core scale fluid distributions during the displacement process, it is possible to study the displacement efficiency of CO 2 -brine drainage processes, residual trapping and fluid saturation at the millimeter to sub-millimeter scale. One potentially useful tool for studying CO 2 -brine systems is using numerical simulation to replicate and study these core flooding systems. This could be used to study the interactions and relative impact of different parameters such as capillary pressure, relative permeability and heterogeneity on brine displacement by CO 2 under various flow conditions. One challenge to successfully conducting such numerical experiments has been accurate representation of the permeability distribution inside the core at the millimeter and sub-millimeter scale. Other simulation parameters can all be measured using laboratory experiments, but permeability must be derived from other properties at the core and sub core-scale. Previous work has shown that predicting sub core-scale permeability distributions based on porosity does not result in accurate representation of permeability at such a small scale. To improve these predictions, a new method based on capillary pressure and was developed and used to accurately predict sub core-scale permeability distributions in a relatively homogeneous Berea sandstone. The work presented in this paper uses the same method to calculate permeability in a strongly heterogeneous sandstone core from the Otway Basin Pilot Project in Australia. Simulations show that the results are consistent with previous results in the homogeneous cores, with statistically significant capability to predict sub core-scale CO 2 distributions in the core. Due to the extreme heterogeneity of the core used in this study, the average match is not as good as for a relatively homogeneous rock core, however, a visual comparison shows that the results are still very good, and that the new method used to calculate permeability may still be valid even in the presence of strong heterogeneity.