This method was the first model to incorporate the concept of "local composition". The basic idea is that the composition in the vicinity of a specific molecule in solution will differ from that of the bulk liquid because of intermolecular forces. Two parameters are used for each binary to describe the degree to which each molecule can produce a change in the composition of its local environment.
Two parameters are used to compute the liquid activity coefficient in the Wilson equation. An additional two constants may be added to introduce a temperature term in the parameters.
The Wilson method has been used successfully to correlate mixtures containing polar components. However, it cannot describe local maxima and minima in the activity coefficient. Its largest limitation, however, is that it is mathematically unable to predict the splitting of a liquid into two partially miscible phases. Therefore, it is completely unsuitable for systems which exhibit liquid-liquid phase splitting.
See "Wilson Equation" under "Liquid Activity Coefficient Methods" in the PRO/II Reference Manual, Volume I, Chapter 2, for more information on this method.