Modeling of a Micellar/Polymer Process

Abstract

Abstract A highly implicit, multidimensional, multicomponent, multiphase, unsteady-state flow model has been for-mulated to simulate a micellar/polymer process. Unlike most compositional approaches, the proposed model ac- counts for capillary pressure. In addition, the model describes the unsteady-state flow of fluids and accounts for additional pressure- and concentration-dependent variables such as average mass velocity, effective disper- sivity, and FVF that most compositional models do not. Numerical solutions to this model are obtained by a finite-difference method. For a one-dimensional (1D) case, the system is treated in terms of five pseudocomponents and two mobile phases. The proposed model is represented by a system of nonlinear partial differential equations in the dependent variables, component concentrations, and phase pressures. The model incorporates the process variables. These include those mentioned above plus in-terfacial tension (IFT), relative permeability, partition coefficient, adsorption concentration, and viscosity. The model was validated by history-matching with a laboratory core displacement test. The agreement of the numerical results and laboratory results shows the model's reliability and gives a realistic insight into its usefulness as a multidimensional, multicomponent, multiphase simulator. After testing, the model was used to investigate the effect of variations in the input parameters on the production history. Introduction The micellar/polymer process is designed to produce residual oil trapped by capillary forces. Papers pro-posing numerical models of the process have been documented. Most neglect the unsteady-state flow of fluids and the capillary pressure between phases by employing a fractional flow formula combined with Darcy's law to represent individual phase transport. They also approximate the physical dispersion by numerical dispersion with appropriate choice of timestep and spacestep. On the other hand, these models deal with the compositional effects and provide design criteria for chemical flooding. The purpose of this paper is to present a new set of micellar/polymer process model equations and methods to model process variables while requiring no restrictions for implicit formulation. This approach is presented as the first step for multidimensional simulation of reservoir flow systems where the proposed process variables govern. The proposed numerical model consists of a system of equations derived from the combination of mass balance, Darcy's law, Fick's first law, and the consideration of various forces. This system of equations was solved simultaneously for a 1D case by a finite-difference method. Included in this paper are solution methods, numerical techniques, verification of model by history-matching with a core displacement test, treatment of proc-ess variables, and effects of variations in the input parameters on the production performance. Model Equations General assumptions of a multidimensional, multicompo-nent, multiphase system are:the displacement process is carried out under isothermal conditions andthe total volume does not change with mixing of individual com-ponents. Based on these assumptions, the generalized model equations may be expressed for n components within in phases as ......................(1) SPEJ P. 617^