A COMPARISON OF THE PREDICTIVE CAPABILITY OF DIFFERENT GROUP CONTRIBUTION METHODS

The knowledge of the real behavior of liquid mixtures plays an important role in different areas of application; this is especially true in the field of chemical engineering. Either gE-models or equations of state are used for the description of the real behavior. While 20 years ago a large number of measurements were required to describe the real behavior of multicomponent systems, great progress has been achieved in the calculation of the real behavior using binary information alone. However the required binary data are often missing and in such cases group contribution methods can be applied to complement the available experimental information. Different group contribution methods (ASOG, UNIFAC) have been developed for the prediction of vapor-liquid equilibria, so that VLE data have mainly been used for fitting the required group interaction parameters. The range of applicability was later extended to include other properties, whereby different weaknesses of the methods were observed. For example only qualitative agreement is obtained for hE- and γ∞, data and the method is not directly applicable to supercritical compounds. Modified forms of the UNIFAC method were developed to improve the situation, and not only VLE but also hE and γ∞ data were used for fitting the required parameters. At the same time the range of applicability has been extended to include supercritical compounds. This paper will discuss results for the different group contribution methods and their range of applicability. © 1990.