Mixing issues in CO2 flooding: comparison of compositional and extended black-oil simulators

Abstract

Extended black oil simulators provide a convenient model for CO2-EOR flooding projects and are preferred because of less data requirements compared with compositional simulation models, therefore, underlying principles of extended black-oil simulation models deserve scrutiny. The main objective of this study is to investigate the impact of different numerical solution techniques on modeling of oil recovery by CO2 flooding and compare extended black-oil and compositional simulators in order to examine the capabilities of these simulators with an emphasis on mixing mechanism of oil and CO2. Thus, the 2D and 3D miscible and immiscible CO2 flooding cases will be used to compare the results from extended black-oil simulators (COZSim, Eclipse Solvent Model, Sensor-First Contact Miscibility Option) and compositional simulators (Eclipse and Sensor). Extended black oil simulators provide an alternative to compositional simulators for the prediction of CO2 flooding. An analysis provided for the capabilities, advantages, disadvantages and limitations of the simulators that were investigated. It is found that compositional simulators predict higher oil recoveries due to complete mixing and complete vaporization of oil assumptions. Moreover, it is shown that usage of a constant mixing parameter in extended black-oil simulators significantly affects oil and gas recovery predictions. CO2 solubility in water phase is investigated and results suggest that CO2 solubility is an important factor for the simulation of residual oil zones.