A modified NRTL model for accurate phases equilibria: Part I: application to vapor–liquid equilibrium

Abstract

• Modification of NRTL model. • Application in computing Vapor-Liquid Equilibrium (VLE). • VLE correlation for Binary and Ternary systems. • VLE Prediction for Quaternary systems. This study presents a modified Non-Random Two-Liquid (NRTL) model that enhances vapor–liquid equilibrium (VLE) correlations across binary and ternary systems and prediction of VLE in quaternary systems. The improvement is due to a structural modification that adds ternary interaction parameters. Then, the revised model requires four interaction parameters for binary systems: τ 12 , τ 21 , τ 121 , and τ 122 . To reduce this number that needed to be optimized from experimental data, values of binary parameters τ 12 and τ 21 were obtained from infinite-dilution activity coefficients using the UNIFAC group contribution approach. Thus, the modified NRTL model was rigorously validated using a dataset of 67 binary, 16 ternary, and 3 quaternary systems. A comparison between the results obtained using the present model and those from the classical NRTL model—with interaction parameters either estimated from infinite dilution activity coefficient data or taken from the literature—reveals that the modified version proposed in this work exhibits significantly smaller deviations: 0.005 for vapor-phase composition and 0.008 for pressure across all tested binary systems. Additionally, for ternary systems, the current model accurately reproduces experimental phase compositions and tie-line slopes, comparing other classical models. In quaternary systems, the modified NRTL model demonstrates strong predictive capabilities, highlighting its robustness and accuracy.