Parallel Numerical Reservoir Simulations of Nonisothermal Compositional Flow and Chemistry

Abstract

Summary This paper describes an efficient numerical scheme for nonisothermal compositional flow coupled to chemistry. An iterative implicit-pressure/explicit-composition (IMPEC) method is applied to solve the flow problem using a volume-balance-convergence criterion. A backward-Euler mixed finite-element method (FEM) with lowest-order RT0 elements is applied to solve the pressure equation, and a component local mass-preserving explicit scheme is used to update concentrations. Chemical reactions are solved using explicit Runge-Kutta (RK) ordinary-differential-equation (ODE) integration schemes. A higher-order Godunov method and a backward-Euler mixed FEM are applied for thermal advection and conduction, respectively, in a time-split scheme. One of the major applications of the method is in the modeling of field-scale carbon dioxide (CO2) sequestration as an enhanced-oil-recovery (EOR) process or for containment in deep saline aquifers where chemical reactions and temperature variations may have an effect on the flow and transport of CO2. Leakage patterns when CO2 is injected near leaky abandoned wells, the displacement of methane from depleted gas reservoirs, and accurate modeling of geochemical reactions involving injected CO2 are other applications of interest. Results of a benchmark problem in multiphase flow with several hydrocarbon components in formations with highly heterogeneous permeability on very fine grids, as well as a large-scale parallel implementation of modeling CO2 sequestration, are presented to justify the practical use of the model. A parallel efficiency of approximately 80% was observed on up to 512 cores in the benchmark study. Results from a problem simulating injection of CO2 in deep aquifers including nonisothermal and chemical effects are also presented. The results indicate a good agreement of the solutions with published data, where available. Numerical modeling and simulation of CO2 sequestration plays a major role in future site selections and in designing storage facilities for effective CO2 containment. The main contribution of this paper lies in providing a parallel and efficient method of simulating challenging compositional flow problems, such as in the study of CO2 sequestration, as well as flow coupled to thermal and geochemical effects.