Ensuring That Correlation Parameters for Liquid–Liquid Equilibrium Produce the Right Results

Abstract

The correlation of experimental liquid–liquid equilibrium (LLE) data for ternary and quaternary systems using activity coefficient models such as nonrandom two liquid (NRTL) or universal quasichemical (UNIQUAC) is a very common practice to enable the design of equipment for separation processes. Consequently, many journals in this field, such as the Journal of Chemical & Engineering Data among others, frequently include several manuscripts in their issues dealing with this relevant aspect of phase equilibria. The thermodynamic background that must be used to conveniently proceed with this task is very well-known and established, though not always applied properly. Different tools are used to perform such correlations, either those included in process simulators or others designed by the corresponding authors. Regrettably, most of the commercial programs do not ensure the consistency of the solutions obtained. This fact, in addition to an extended lack of the custom of testing the validity of the solutions obtained, leads to a very undesirable situation where a surprisingly high number of papers presenting very serious mistakes and inconsistencies in the parameters published as a result of LLE correlations continuously appear in the literature.