Modeling, Analysis, and Computation for Water-Drive Oil Recovery

University Of Oklahoma Norman Campus, Norman OK

Investigators

Abstract

Accurate mathematical analysis and efficient numerical methods play an increasingly important role in studying partial differential equation models in the petroleum industry. The goal of this project is to carry out research in mathematical analysis and design of high-order-accuracy numerical methods for the modified Buckley-Leverett (MBL) model, which is a partial differential equation model for water-drive secondary underground oil recovery. When an underground source of oil is tapped, a certain amount of oil flows out on its own due to pressure difference. After the flow stops, there is typically a significant amount of oil still left in the ground. One standard method of "secondary recovery" is to pump water into the oil field through an injection well, forcing oil out through a production well. In this process there will be a water and oil mixture created. The MBL equation models the evolution of the oil saturation in the entire oil reservoir, in particular, at the production well. This research project combines mathematical analysis, design of numerical schemes, and computational techniques. The mathematical analysis is based on partial differential equation theory, and the numerical methods under development are based on state-of-the-art discontinuous Galerkin (DG) schemes. The results will be cross-validated using data from laboratory experiments. The investigator plans to carry out the following specific research tasks: (1) extend the well-developed 1D MBL model to 2D and 3D fully nonlinear and linearized MBL models; (2) determine the approximation error induced by employing the MBL model on a spatial domain smaller than an entire reservoir; (3) design high-order-accuracy discontinuous Galerkin (DG) methods to numerically solve the MBL model; (4) study the nonlinear asymptotic stability of the traveling-wave solutions of the MBL model; (5) employ experimental data to cross-validate both the analytical and computational conclusions. The project involves a postdoctoral associated and a graduate student in the research. The investigator also plans to develop a new graduate-level course on numerical solutions to water-drive oil recovery models.