Efficient Parallel Simulation of CO2 Geologic Sequestration in Saline Aquifers
Keni Zhang,Christine Doughty,Yu‐Shu Wu,Karsten Pruess
DOI: https://doi.org/10.2523/106026-ms
2007-01-01
Abstract:Efficient Parallel Simulation of CO2 Geologic Sequestration in Saline Aquifers Keni Zhang; Keni Zhang Lawrence Berkeley Natl. Laboratory Search for other works by this author on: This Site Google Scholar Christine Doughty; Christine Doughty Lawrence Berkeley Natl. Laboratory Search for other works by this author on: This Site Google Scholar Yu-Shu Wu; Yu-Shu Wu Lawrence Berkeley Natl. Laboratory Search for other works by this author on: This Site Google Scholar Karsten Pruess Karsten Pruess Lawrence Berkeley Natl. Laboratory Search for other works by this author on: This Site Google Scholar Paper presented at the SPE Reservoir Simulation Symposium, Houston, Texas, U.S.A., February 2007. Paper Number: SPE-106026-MS https://doi.org/10.2118/106026-MS Published: February 26 2007 Cite View This Citation Add to Citation Manager Share Icon Share Twitter LinkedIn Get Permissions Search Site Citation Zhang, Keni, Doughty, Christine, Wu, Yu-Shu, and Karsten Pruess. "Efficient Parallel Simulation of CO2 Geologic Sequestration in Saline Aquifers." Paper presented at the SPE Reservoir Simulation Symposium, Houston, Texas, U.S.A., February 2007. doi: https://doi.org/10.2118/106026-MS Download citation file: Ris (Zotero) Reference Manager EasyBib Bookends Mendeley Papers EndNote RefWorks BibTex Search Dropdown Menu toolbar search search input Search input auto suggest filter your search All ContentAll ProceedingsSociety of Petroleum Engineers (SPE)SPE Reservoir Simulation Conference Search Advanced Search AbstractAn efficient parallel simulator for large-scale, long-term CO2 geologic sequestration in saline aquifers has been developed. The parallel simulator is a three-dimensional, fully implicit model that solves large, sparse linear systems arising from discretization of the partial differential equations for mass and energy balance in porous and fractured media. The simulator is based on the ECO2N module of the TOUGH2 code and inherits all the process capabilities of the single-CPU TOUGH2 code, including a comprehensive description of the thermodynamics and thermophysical properties of H2O-NaCl- CO2 mixtures, modeling single and/or two-phase isothermal or non-isothermal flow processes, two-phase mixtures, fluid phases appearing or disappearing, as well as salt precipitation or dissolution. The new parallel simulator uses MPI for parallel implementation, the METIS software package for simulation domain partitioning, and the iterative parallel linear solver package Aztec for solving linear equations by multiple processors. In addition, the parallel simulator has been implemented with an efficient communication scheme. Test examples show that a linear or super-linear speedup can be obtained on Linux clusters as well as on supercomputers. Because of the significant improvement in both simulation time and memory requirement, the new simulator provides a powerful tool for tackling larger scale and more complex problems than can be solved by single-CPU codes. A high-resolution simulation example is presented that models buoyant convection, induced by a small increase in brine density caused by dissolution of CO2.IntroductionCO2 geologic sequestration in saline aquifers and in oil and gas reservoirs involves complex multiphase flow processes as well as geomechanical, and geochemical processes, such as advection and diffusion, convective mixing, phase appearance/disappearance, dissolution and precipitation of minerals, and other chemical reactions. Mathematical models are effective tools for understanding the behavior of CO2 in geological formations. Numerical modeling can in fact play an important role in evaluating the feasibility and reliability of CO2 disposal. However, modeling of CO2 geologic sequestration and migration processes in general requires fine spatial and temporal discretization, and represents a large computational challenge.One of the popular numerical simulators for CO2 geologic sequestration in saline aquifers is the ECO2N module of the general-purpose reservoir simulator TOUGH21–2. The code can be used to model non-isothermal multiphase flows of water, salt, and CO2 mixtures. TOUGH2/ECO2N represents fluids as consisting of two potentially mobile phases: a water-rich aqueous phase, a CO2-rich gaseous phase, and an immobile solid halite phase. Because of the complexity of subsurface flow processes, most simulations for these types of problems are limited to systems of up to several ten thousand gridblocks. For reliable field scale applications, however, hundred thousands and millions of gridblocks may be needed to represent both geologic heterogeneities and multiphase, multicomponent flow structures on different scales.In this study, a parallel simulator for large-scale, long-term CO2 geologic sequestration in saline aquifers has been developed that is based on the ECO2N module of the TOUGH2 code. The TOUGH2 code itself was originally parallelized on CRAY T3E and IBM SP supercomputers3–4, and later ported to Linux clusters and multi-core PCs5. The parallel version of TOUGH2, TOUGH_MP, has been successfully applied to solve multi-million gridblock multiphase fluid-flow problems and large-scale discrete fracture flow simulations6–7. The CO2 parallel simulator retains all the process-modeling capabilities of the original TOUGH2/ECO2N and parallel computation features of TOUGH_MP. The parallel simulator is a three-dimensional, fully implicit model that solves large, sparse linear systems arising from discretization of the mass and energy balance equations in porous and fractured media. The simulator provides a comprehensive description of the thermodynamics and thermophysical properties of H2O-NaCl-CO2 mixtures, and models single and/or two-phase isothermal or non-isothermal flow processes, two-phase mixtures, appearance or disappearance of fluid phases, as well as salt precipitation or dissolution. Keywords: Artificial Intelligence, flow in porous media, Upstream Oil & Gas, reservoir simulation, Sequestration, gridblock, Jacobian matrix, geologic sequestration, dissolution, Fluid Dynamics Subjects: Reservoir Fluid Dynamics, Reservoir Simulation, Flow in porous media This content is only available via PDF. 2007. Society of Petroleum Engineers You can access this article if you purchase or spend a download.