Identification of UNIQUAC binary interaction parameters in liquid-liquid equilibrium

Abstract

An algorithm is presented for the estimation of the UNIQUAC interaction parameters for liquid-liquid equilibrium of ternary systems. The algorithm is based on two optimization levels. In the inner level the algorithm performs the minimization of an objective function based on the isoactivity conditions. The outer level aims to minimize the error between calculated and experimental compositions. The Common Tangent Plane condition is checked at the end to guarantee a thermodynamically consistent representation of the phase behavior of ternary liquid systems.The algorithm is challenged with a historical Type 1 ternary liquid-liquid equilibrium system from the seminal study of Anderson and Prausnitz in which the authors showed the limitations of the original UNIQUAC model and justified its amendment in the modified UNIQUAC model. The present algorithm makes available single temperature and temperature-dependent interaction parameters enabling accurate and thermodynamically correct description of the experimental data with the original UNIQUAC model, therefore without the need of any model modification. This outcome does not change when the interaction parameters from the binary partially miscible constituent pair are first regressed and kept constant during the estimation of the remaining parameters on ternary equilibrium data. This investigation confirms that a model cannot be judged if the correctness of the model parameters has not been proved first.