Reservoir Oil Characterization for Compositional Simulation of Solvent Injection Processes

Abstract

Reliable design of solvent injection for oil recovery requires accurate representation of multiphase behavior in a wide range of pressure–temperature–composition conditions. Volume shift is often required to improve density predictions, separately from compositional behavior predictions, in conventional fluid characterization methods (CM). Thermodynamically, however, volumetric behavior predictions are consequences of compositional behavior predictions. We develop a new fluid characterization method (NM) that gives accurate multiphase behavior representation for oil/solvent mixtures without volume shift. The Peng–Robinson EOS is used with the van der Waals mixing rules. The NM is compared with the CM in terms of phase diagrams, minimum miscibility pressure calculations, and oil displacement simulations. It is shown that the CM with volume shift can give erroneous phase behavior and oil recovery predictions. The NM requires no volume shift to achieve accurate predictions of compositional and volumetric pha...