Comparison of MOSCED (NRTL) model results with regular correlative and predictive models based on vapor-liquid equilibrium calculations for azeotropic systems

Abstract

The Modified Separation of Cohesive Energy Density (MOSCED) model is an attractive and accurate engineering tool for process design and modeling. MOSCED is able to make qualitative as well as precise quantitative predictions of principal molecular-level interactions. However, the capability of this model in predicting Vapor-Liquid Equilibrium (VLE) of different azeotropic systems has not yet been fully assessed, when compared with the conventional methods. In this study, the ability of MOSCED model to predict the binary interaction parameters of NRTL model has been evaluated by comparing the obtained isobaric phase diagrams of the experimental sets of data for 11 binary isobaric systems. In order to better evaluate the performance of MOSCED model, the same calculations were performed via correlative (Wilson and NRTL) and predictive (UNIFAC) models. The values of statistical parameters have demonstrated acceptable predictions for the studied systems. Overall, the MOSCED model has provided an Average Absolute Relative Deviation(s) (AARD %) of 0.1978% and 4.3239% for temperature and K-value, respectively; which illustrate excellent performance of this model in spite of no dependence on experimental data, which is a big advantage over other methods.