Predicting PVT data for CO2–brine mixtures for black-oil simulation of CO2 geological storage

Accurate modeling of the storage or sequestration of CO2 injected into subsurface formations requires an accurate fluid model. This can be achieved using compositional reservoir simulations. However, sophisticated equations of state (EOS) approaches used in current compositional simulators are computationally expensive. It is advantageous and possible to use a simple, but accurate fluid model for the very specific case of geological CO2 storage. Using a black-oil simulation approach, the computational burden of flow simulation can be reduced significantly. In this work, an efficient and simple algorithm is developed for converting compositional data from EOS into black-oil PVT data. Our algorithm is capable of predicting CO2–brine density, solubility, and formation volume factor, which are all necessary for black-oil flow simulations of CO2 storage in geological formations. Numerical simulations for a simple CO2 storage case demonstrate that the black-oil simulation runs are at least four times faster than the compositional ones without loss of accuracy. The accuracy in prediction of CO2–brine black-oil PVT properties and higher computational efficiency of the black-oil approach promote application of black-oil simulation for large-scale geological storage of CO2 in saline aquifers.