Toward multiphase equilibrium prediction for ill-defined asymmetric hydrocarbon mixtures

Phase behaviour prediction for mixtures containing asymmetric components is of great importance in chemical and petroleum engineering applications and presents numerous challenges in general due to the number of phases encountered. Further, constituents such as bitumen or heavy oil are ill-defined on a molecular basis and their critical and other properties are not accessible experimentally. This imposes additional challenges for the development of reliable thermodynamic models. Group contribution methods that estimate thermodynamic properties based on measurable or known molecular structures present a computational option. In this contribution the appropriateness of group contribution methods proposed by Crampon et al. and Marrero and Gani in combination with the Peng–Robinson equation of state as a basis for such calculations is evaluated. Liquid–liquid–vapour three phase zones for ternary mixtures containing n-decylbenzene and n-eicosane + light hydrocarbons provide illustrative test cases. The results are compared to experimental data and to modeling results obtained using other thermodynamic models available in the literature. The performance of group contribution computations is further tested by predicting the phase behaviour trends that are observed experimentally for such mixtures. The results show that the group contribution method developed by Marrero and Gani out performs the Peng–Robinson equation of state by generating accurate phase behaviour and phase behaviour trend predictions for all cases evaluated. In combination with the group contribution based interaction parameter computation method (PPR78) developed by Jaubert et al., it has the potential to predict the phase behaviour of mixtures containing ill-defined hydrocarbons such as bitumen and heavy oil. Ill-defined hydrocarbon speciation continues to present significant experimental and computational challenges.