Numerical Simulation of CO2 Flood Performance

Abstract

AbstractA multicomponent, three-phase, three-dimensional (3D) reservoir simulator has been developed for predicting miscible flood performance with particular emphasis on CO2 flooding. Several features, not all normally found represented in either a black-oil or compositional simulator, are represented in the simulator described here. These features include (1) heavy-end dropout and concomitant reduction in phase mobility, (2) water blocking of oil from an invading solvent, (3) viscous instability (fingering) at the CO2 displacement front, (4) miscible/immiscible transition as a function of pressure and composition, and (5) loss of CO2 to the aqueous phase.The solution scheme employs an implicit in pressure, explicit in saturation and composition time integration scheme (IMPES). This scheme allows for ready use of a second-order correct spatial discretization. We have found that for field applications, a numerical grid fine enough to represent waterflood performance accurately is usually sufficient for representing CO2-miscible-type processes as well. The program automatically employs a stabilized Runge-Kutta time discretization to alleviate the usual stability limitations imposed by the IMPES procedure. Up to 49 times the usual stable timestep size can be employed for seven times as much computing work. Work required to simulate a given time period when stability is limiting was improved overall by a factor of seven. To date, we have not encountered a pattern flood displacement simulation where the increased timestep size provided by the Runge-Kutta scheme has not been adequate to maintain stability for timesteps chosen so as to limit truncation error.The CO2 flood simulator described has been in application for 4 years. The simulator has been used for design of CO2 EOR projects and interpretation of CO2 injectivity tests in several low-temperature west Texas San Andres dolomite reservoirs as well as high-temperature sandstone reservoirs. In addition to CO2 applications, the simulator has been used for the design of liquefied petroleum gas/dry gas flooding processes and for examining the effect of injection gas composition on oil recovery performance for immiscible gas drives.