Date of Award
Master of Science in Chemical Engineering - (M.S.)
Chemical Engineering and Chemistry
Dimitrios P. Tassios
Richard Clyde Parker
David S. Kristol
The Wilson equation is chosen as typical of the new correlations for activity coefficients gaining wide distribution for use in the correlation and prediction of vapor-liquid equilibria today. Vapor liquid equilibrium data for 247 binary systems have been regressed against each of ten (10) objective functions and values of the Wilson energy parameters obtained. For the typical binary system, a wide variation of the Wilson energy parameters are observed when different objective functions are used with the given set of binary data. These are each examined with respect to their capability to correlate multicomponent vapor-liquid equilibrium data for seventy-three (73) systems. In general the best results are obtained from the objective function that minimizes the relative error between the calculated and experimentally derived values of the binary activity coefficients.
The number of parameter sets for the Wilson equation that can be obtained from a given set of binary vapor-liquid equilibrium data is also examined. It is proven mathematically that only one set of Wilson energy parameter values exist for the case of positive deviations from Raoult's Law; however, as many as three parameter sets can be derived from experimental data for the case of negative deviations from ideal behavior. The criteria with which to determine the number of such sets is developed. Best correlation of both binary and multicomponent vapor-liquid equilibrium data appear to result from the set of Wilson energy parameters that are smallest in absolute value.
Silverman, Norman I., "Vapor-liquid equilibrium at low pressures : correlation of binary and prediction of multicomponent data" (1977). Theses. 2001.