Local composition modelling of the thermodynamic properties of refrigerant and oil mixtures

Six local composition models of the thermodynamic behaviour of mixtures are described. Using data from the literature and a non-linear regression analysis, a comparison of the predictive abilities of the models is undertaken for R12, R22, R134a and R125 with various oils. The Wilson and Heil equations provide the most consistent results, with the Heil equation providing a modest improvement over the Wilson model. Using a 95% confidence interval, the Heil equation predicted the behaviour of R12 with a paraffinic mineral oil to within 3.1%; its worst-case 2-σ error was 10.4% (R22 with a polyol ester oil), and its average 2-σ error for all of the mixtures was 6.2%. Using model parameters and error estimates from the regression analyses, pressure-temperature-concentration behaviour for these mixtures can be predicted for system design and simulation.