Compositional numerical analysis of multiphase flow of crude oil in porous media under non-isothermal conditions

Abstract The present work details the development of a compositional model to replicate the heavy hydrocarbon flow in porous reservoir systems under non-isothermal conditions. The mathematical model considers mass and energy conservation equations describing the reactive of natural variables distributed in a multiphase hydrocarbon system. Such natural variable based compositional models better suit fully implicit numerical schemes with inexpensive Jacobian matrix computations. Further, the model accommodates a switch of primary variables for the disappearance and reappearance of a phase. The resulting nonlinear conservation equations are numerically discretized using a block-centered finite-difference scheme and solved with a quasi-Newton based implicit iterative solver. The present model is validated with the thermal profiles presented in the literature for the multiphase flow during the combustion of heavy crude oil in petroleum reservoir system with performance coefficient (R 2), mean absolute error (MBE), and maximum absolute percentage error (MAPE) of about 0.954, 0.37, and 0.01 respectively. The developed compositional model projected 26 and 72 % of light and heavy oil recoveries respectively in about 160 days with a maximum or peak temperature of about 798 K. Further, the thermal and production profiles projected by the sensitivity analysis on various operating parameters are presented. It is noteworthy that the present works aid in providing an economical numerical based tool in evaluating the flow and transport during underground or in-situ combustion process for efficient energy exploration.