Fluid Phase Equilibria, Vol.428, 4-17, 2016
Quantitative comparison between predicted and experimental binary n-alkane plus benzene phase behaviors using cubic and PC-SAFT EOS
The Perturbed-Chain Statistical Associating Fluid Theory (PC-SAFT) equation of state correctly predicts Type I phase behavior for binary n-alkane + aromatic and naphthenic mixtures, while cubic equations of state (Peng-Robinson (PR) and Soave-Redlich-Kwong (SRK)) predict a transition from Type I to Type II phase behavior within these families of mixtures, unless unconventional large and negative interaction parameters (k(ij)) are used. In this work, bubble pressures and L = V critical loci are evaluated for 13 binary n-alkane + benzene mixtures, including new data obtained for benzene + n-C-20, n-C-24, n-C-28 and n-C-36 binary mixtures. Computed bubble pressures for the Peng-Robinson (PR), Soave-Redlich-Kwong (SRK) and Perturbed-Chain Statistical Associating Fluid Theory (PC-SAFT) equations of state are compared with one another and with experimental measurements. The PC-SAFT EOS, with pure component parameters rescaled to conform with critical temperatures and pressures following the recommendation of Alfradique and Castier (2007), and interaction parameter values set to zero, yield accurate bubble pressures and critical loci for all benzene + n-alkane mixtures. By contrast, the PR and SRK EOS require mixture specific k(ij) values in order to provide quantitative bubble pressure and critical loci estimates, and best-fit let values exhibit significant temperature dependence. In the absence of experimental bubble pressures, options for estimating interaction parameters for cubic EOS for binary benzene + n-alkane mixtures, and for aromatic or naphthenic + alkane mixtures more broadly are discussed. While subject to further testing, selection of interaction parameter values for cubic EOS such that computed bubble pressures closely mimic bubble pressures predicted by the scaled PC-SAFT EOS is recommended. (C) 2016 Elsevier B.V. All rights reserved.