Modified Guo viscosity model for heavier hydrocarbon components and their mixtures

Abstract

Guo et al. (2001) equation of state based viscosity model is simple and is able to predict viscosity for lower carbon n-alkane components reliably. However, it loses its reliability when extended to higher carbon number hydrocarbon components. In this work, Guo et al. viscosity model has been modified to improve its predictive capability and its application to heavier components. First modification has been done to improve component level viscosity prediction and then binary interaction parameters have been developed to improve viscosity prediction for hydrocarbon mixtures.New set of correlations for parameters in Guo et al. (2001) model for each of three main hydrocarbon groups paraffins, naphthenes, and aromatics, have been developed as function of carbon number of the components. Correlation for binary interaction parameters has also been developed for paraffinic hydrocarbon components.Four thousand viscosity data points for n-alkanes, 850 points for naphthenes, and 608 points for aromatics have been used to modify correlations for improving the reliability of Guo et al. model. The modified Guo et al. viscosity model has AARDs of 8.3%, 6.89%, and 9.09% in viscosity prediction as compared to 33.95%, 19.6%, and 48.9% from Guo et al. model for paraffinic, naphthenic, and aromatic groups of hydrocarbons respectively. The AARD in viscosity prediction for liquid hydrocarbon mixtures from modified Guo et al. with optimized binary interaction parameter is 3.95% whereas that from Guo et al. model with default binary interaction parameter is 24.6%. The modified Guo et al. model has significantly better predictive capability for pure components and their mixtures. It has potential for use in viscosity model for crude oils in wider specific gravity range.