Simulation of surfactant/polymer floods with a predictive and robust microemulsion flash calculation

A compositional reservoir simulator that uses a predictive micro-emulsion phase-behavior model is essential for accurate estimation of oil recovery from surfactant/polymer (SP) floods. Current chemical-flooding simulators, however, use Hand's model (Hand 1939) for phase-behavior calculation. Hand's model can reasonably fit a limited set of experimental data, such as those of a salinity scan, but because it is empirical, it cannot predict phase behavior outside the matched data set. Hydrophyllic/lypophyllic difference (HLD) and net-average-curvature (NAC) equation of state (EOS) (Acosta et al. 2003) has shown great performance for tuning and prediction of experimental data. In this paper, the EOS model with the extension to two-phase regions has been incorporated for the first time into UTCHEM (2000) and our in-house general-purpose compositional simulator, PennSim (2013). All Winsor regions (Type II -, II+, III, and IV) are modeled by use of a consistent physics-based EOS model without the need for Hand's approach. The new simulator is therefore able to account correctly for gridblock properties, which can vary temporally and spatially, and significantly improve the modeling of phase behavior and oil recovery. The results show excellent agreement between UTCHEM and PennSim both in composition space and for composition/saturation profiles for the 1D simulation. The effects of varying pressure, temperature, equivalent alkane carbon number (EACN), and salinity on recoveries are demonstrated also in 1D simulations.