Prediction of water solubility in ill-defined hydrocarbons at high temperatures: Modeling with the CPA-EoS

Abstract

Evaluation of phase behavior of mixtures containing reservoir fluids and associating (hydrogen-bonding) compounds like water is an important task in oil production and refining processes. In current work, a CPA computational approach was adopted to predict water solubility in pure hydrocarbons, petroleum fractions and heavy reservoir fluids. The reservoir fluid was considered as a single pseudo-component to simplify the model. The solvation (cross-association) between water and hydrocarbon molecules was taken into account while the self-association between hydrocarbon molecules was ignored. All the binary interaction coefficients were considered zero. Cross-association volume parameters of various mixtures were tuned to fit the experimental trend for each hydrocarbon and reservoir fluid as the only adjusting parameter of the model. The CPA model was applied to a number of pure hydrocarbons, petroleum fractions and heavy crudes with molecular weights more than 78 kg/kmol in a range of elevated temperatures suitable for thermal recovery processes. A generalized correlation is proposed to estimate the cross-association volume parameter of different hydrocarbon + water mixtures in terms of hydrocarbons molecular weight and Watson characterization factor. The suggested model is benchmarked against three complex water solubility data sets for Athabasca bitumen from the literature. The results are promising to extend the CPA model to hydrocarbons and reservoir fluids with a reasonable accuracy in the temperature range from 279 up to 644 K.