Viscosity Estimation of Liquid Hydrocarbon Mixtures: An Application to Solvent-Based Enhanced Oil Recovery

Abstract

Abstract Various mixing rules exist to estimate the viscosity of liquid hydrocarbon mixtures given the viscosities of the individual components. These models generally work well for mixtures of components of similar viscosity but can produce quite inaccurate estimates for mixtures of light and heavy components. Enhanced oil recovery processes for heavy-oil reservoirs utilize light hydrocarbon solvents as an injectant to enhance the oil recovery. The mixture viscosity is a fundamental parameter for modelling solvent-based recovery techniques using reservoir simulation. In the present work, a large database of liquid mixture viscosity literature data has been constructed and used to test a diverse set of mixing rules. In addition to the general formulation, the impact of the component basis (i.e. the use of mole, mass, or volume fractions) was examined. The database includes a wide range of mixtures, which range in viscosity from 0.1 to about 100,000 cp. A power law mixing rule using pure component volume fractions calculated at standard conditions is shown to be, in general, very effective. Additionally, it was found that a modified Arrhenius model could provide similar accuracy as the proposed formulation but is somewhat more complicated apply.