A Mixed Natural and Concentration Variable Formulation for Chemical Flood Simulation

Abstract

Abstract Correlations for PVT properties used in chemical flood simulation are frequently written in concentration variables. Therefore, a mixed variable formulation has been proposed to implement chemical flooding option in an existing natural variable simulator. The new formulation results in a straightforward and efficient implementation of published PVT correlations as well as fast convergence. A new three-layer pseudo component framework (i.e., phase – pseudo component – pure component) has been adopted to effectively model a group of pure components that can have the same composition and properties in each phase where it exists while allowing the flexibility of using different variable types on the pure component level, such as phase mole fraction and molar concentration. This ensures the same composition for pseudo component in different phases without introducing extra constraint equations. In our implementation, we use a symmetric binodal curve and tie lines to describe the liquid phase behavior based on the pseudo-component concept. Phase densities and viscosities are computed using correlations with corresponding pseudo-component values and volumetric concentrations. A Corey-type relative permeability model and shear thinning effect on viscosity are also implemented. Furthermore, both cation exchange on surfactant/clay and adsorption of polymer/surfactant on rock are efficiently implemented. Our algorithms have been validated against other state-of-the-art chemical flood simulators with both small conceptual test cases and large scale models. A very good agreement in results and a high nonlinear performance have been observed. Mixed natural and concentration variable formulation proposed in this paper is a novel method to implement chemical flood capability in an existing natural variable simulator. The combination of mass and natural variables for surfactant flood simulation into a single formulation is an industry first. Its extensibility and robustness give us confidence for further extensions along these lines.