作者: Rebecca Allen , Tim Reis , Shuyu Sun
DOI: 10.2118/163658-MS
关键词:
摘要: The storage of CO2 in fluid-filled geological formations has been carried out for more than a decade locations around the world. After injected into aquifer and moved laterally under aquifer's cap-rock, density-driven convection becomes an important transport process to model. However, challenge lies simulating this accurately with high spatial resolution low CPU cost. This issue can be addressed by using lattice Boltzmann equation (LBE) formulate model similar scenario when solute diffuses fluid density differences lead convective mixing. LBE is promising alternative traditional methods computational dynamics. Rather discretizing system partial differential equations classical continuum mechanics directly, derived from velocity-space truncation kinetic theory. We propose extension LBE, which predict dissolved water, as step towards porous media simulations. achieved coupling two LBEs, one flow diffusion CO2. Unlike existing flow, our moment Crank-Nicolson discretization velocity-truncated equation. forcing terms are updated locally without need additional central difference approximation. Therefore preserves all advantages single-phase formally second-order accurate both space time. Our new also features novel implementation boundary conditions, simple implement does not suffer grid-dependent error that present standard "bounce-back" condition. The significance work ability efficiently simulate through water. From viewpoint, locality algorithm exploits massively parallel modern computer architectures, including graphics processing units (GPUs), would very fast computations scale linearly number processors.