Reservoir Management

Recent advances in a multiscale flux-based method for simulating flow in fractured porous media.

Abstract

Computational simulation of reservoir flow is an important tool that provides valuable insight into the decision process in oil extraction. Several types of commercial software have been developed over the years for this application, the majority using low-order schemes, which can become prohibitive for very large models. This issue becomes more apparent since, nowadays, the accuracy of a simulator is dominated by the accurate simulation of the multiscale characteristics of a reservoir such as permeability heterogeneity. To capture these multiscale features in low-order schemes, very refined models are required. Therefore, developing a high-order scheme able to simulate fractured reservoir flow that is accurate and can efficiently capture the multiscale features of the reservoir is of great value for the field. With this motivation, this presentation reports on recent advances in a methodology to simulate flow in highly heterogeneous fractured porous media using the Multiscale Hybrid-Mixed (MHM) method with H(div)-confirming flux approximations. This method is particularly appealing because of its inherent properties such as local mass conservation, multiscale features, and strong divergence-free enforcement for incompressible flows. Flow in the porous media is modeled with traditional Darcy’s equations and the coupling between flow in the porous media and fractures is based on the conceptual Discrete-Fracture-Matrix representation, where the fractures are idealized as lower-dimensional elements at the interface of matrix elements. The methodology is compared with benchmark examples to demonstrate its robustness, accuracy, and efficiency.