Simulation of Channelized Gas Flow Pattern in Heterogeneous Porous Media: A Feasibility Study of Continuum Simulation at Bench Scale

Abstract

In remediation technologies, air or O gas is injected into a polluted aquifer to remove organic contaminants. A gas injection process is also applied in enhanced oil recovery and CO sequestration. The wide range of gas injection processes into porous media necessitates accurate gas flow modeling. The focus of this study was to determine whether a continuum model with stochastic parameters can represent the properties of gas flow patterns. We applied stochastic heterogeneity in TOUGH2 simulation and used the experimental observations to validate simulation results. We created two types of stochastic heterogeneity: correlated heterogeneity, generated by an assigned covariance function, and uncorrelated heterogeneity, which distributes uniformly. We used Euler characteristics to quantify the gas channel properties and to determine whether the heterogeneity applied in a continuum model can represent gas channel properties. The results of this investigation confirm that the continuum model, which includes correlated heterogeneity, can predict the quantitative properties of gas flow, such as gas volume. Also, it can create gas channels that have Euler characteristics close to those obtained by experimental observation. In conclusion, the continuum model can predict the properties of gas flow if the spatial variability and correlation structure of the characteristic properties applied to the model are representative of the medium.