High-Order and Conservative Mathematical and Computational Modeling for Polymer Injection in Oil Reservoirs
Polymer Injection, Computational Model, High-Order, Conservative
Polymer injection in petroleum reservoirs is a highly promising alternative to optimize oil recovery in resevoirs. However, its performance can be compromised by phenomena such as adsorption and mechanical retention, which impact injectivity and, consequently, the effectiveness of the recovery process. Therefore, to fully understand the dynamics of polymer injection techniques, a robust, accurate, and comprehensive mathematical and computational model is required to precisely represent all the nonlinear couplings involved in the process dynamics. In this context, the computational model proposed in this study incorporates various phenomena associated with polymer injection, such as the influence of gravity, fluid pseudoplasticity, formation damage, and the dependence of relative permeability on concentration. The hydrodynamics are described by Darcy's law and the mass balance of the fluid phase, while polymer transport is governed by a convective-diffusive-reactive transport equation, with physicochemical adsorption modeled by the Langmuir isotherm and mechanical retention by a linear kinetic law. In the numerical model, based on the IMPESC algorithm, the set of equations is divided into two parts: the hydrodynamic system and transport. The discretization of the hydrodynamic problem uses a high-order conservative method, the hybrid dual method (HMFEM), in which we consider the Raviart-Thomas ($RT^1$) basis, allowing for greater accuracy in capturing velocity at the element centers. In turn, polymer transport is discretized using the Central-Upwind (CU) finite volume method, which accurately captures shocks and rarefactions, where we develop a new semi-discrete formulation with a variable porosity term. Furthermore, the use of the $RT^1$ basis improved the coupling between methods, as Gaussian integration is applied to the flux integrals on the dual volume. The proposed numerical methods and the coupled model are validated through accuracy and precision tests, proving effective in describing the polymer injection process in porous media.