Modeling Effect of Geochemical Reactions on Real-Reservoir-Fluid Mixture During Carbon Dioxide Enhanced Oil Recovery

Abstract

Summary Carbon dioxide (CO2) injection in oil reservoirs has the dual benefit of enhancing oil recovery from declining reservoirs and sequestering a greenhouse gas to combat climate change. CO2 injected in carbonate reservoirs, such as those found in the Middle East, can react with ions present in the brine and the solid calcite in the carbonate rocks. These geochemical reactions affect the overall mole numbers and, in some extreme cases, even the number of phases at equilibrium, affecting oil-recovery predictions obtained from compositional simulations. Hence, it is important to model the effect of geochemical reactions on a real-reservoir-fluid mixture during CO2 injection. In this study, the Gibbs free-energy function is used to integrate phase-behavior computations and geochemical reactions to find equilibrium composition. The Gibbs free-energy minimization method by use of elemental-balance constraint is used to obtain equilibrium composition arising out of phase and chemical equilibrium. The solid phase is assumed to be calcite, the hydrocarbon phases are characterized by use of the Peng-Robinson (PR) equation of state (EOS) (Robinson et al. 1985), and the aqueous-phase components are described by use of the Pitzer activity-coefficient model (Pitzer 1973). The binary-interaction parameters for the EOS and the activity-coefficient model are obtained by use of experimental data. The effect of the changes in phase behavior of a real-reservoir fluid with 22 components is presented in this paper. We observe that the changes in phase behavior of the resulting reservoir-fluid mixture in the presence of geochemical reactions depend on two factors: the volume ratio (and hence molar ratio) of the aqueous phase to the hydrocarbon phase and the salinity of the brine. These changes represent a maximum effect of geochemical reactions because all reactions are assumed to be at equilibrium. This approach can be adapted to any reservoir brine and hydrocarbon as long as the initial formation-water composition and their Gibbs free energy at standard states are known. The resultant model can be integrated in any reservoir simulator because any algorithm can be used for minimizing the Gibbs free-energy function of the entire system.