Optimization of PR Model and Compared with Fan and Yang Model to Predict the Viscosity of Heavy Liquid Mixture of Hydrocarbons with CO2

In this research, an integrated model to provide both pure and mixed viscosity hydrocarbons in areas; such as, supercritical gas and liquid using only one equation of state (Peng and Robinson) is provided. The main idea of this method is based on the similarity between data and theory TμP, PVT and balance. In this modeling, the constants of the viscosity state’s equation are calculated from a similarity between diagrams of PVT and TμP at the critical turning point. The equation of R constant viscosity (similar to universal gas constant parameter) is calculated from the dependence on the pressure in critical situations. In the model proposed by Fan Wang (the same model), critical compressibility factor values for each variable were reported. In this study, for the purpose of simplicity and better performance, a given amount (Zc = 0.3074) has been provided for this coefficient (index). Finally, a statistical objective function has been used for the purpose of making a comparison between the model’s performance and the experimental data. So that the mean absolute error for the mixture of heavy and light compounds by using modeling on 122 experimental data as well as for the purpose of extending the model to a mixture of two, three, four and five which for the thermal limitation based on Kelvin and the pressure’s limitation based on Bar for the compounds of three at the limitation of pressure (3.08 to 6.35), thermal (324.26 to 395.37); for the compounds of four at the limitation of pressure (2.73 to 4.12), thermal (359.82 to 395.37); for the compounds of four with Carbon dioxide at the limitation of pressure (25.14 to 48.28), thermal (324.26 to 395.37) and the incorporation of the new rules for more accurate modeling is used. Based on PRμ0 model, the calculated liquid viscosity has had less accuracy, so that by generating the optimization parameters in PRμ0 equation, deviation calculations will be reduced. Moreover, the average error rate for each of these mixtures is 5.04, 3.21, 5.16 and 2.9 percent recpectively, which is more accurate than the Wang technology for predicting the behavior of this mixture.