Investigation of different activity coefficient models in thermodynamic modeling of wax precipitation

The deposition of wax in oil and the gas industry is a serious problem causing difficulties in the production, transportation and processing of petroleum fluids. In order to prevent this phenomenon, it is essential to be able to predict the cloud point temperature and wax content at below cloud point temperature by using a reliable thermodynamic model. The description of the solid phase non-ideality is the main obstacle in the modeling of the solid-liquid equilibrium of hydrocarbons. In this work, 25 combinations of five models (regular solution theory, UNIFAC, predictive UNIQUAC, predictive Wilson and ideal solution model) for the description of solid and liquid phases have been surveyed to get the best results near the experimental data without using any adjustable parameters. For validating the models, some binary, ternary, quaternary and multi-component systems as the experimental data have been used. which are for 121 equilibrium data points and a total of 76 various mixtures. The results prove that predictive Wilson, together with one of the models, regular solution, UNIFAC, ideal or predictive UNIQUAC is more appropriate than the others except in two cases that predictive UNIQUAC-UNIFAC and predictive UNIQUAC-regular solution models give better results for the description of the non-ideality solid and liquid phase behavior, respectively. These models, in comparison with multi-solid models, show better agreement with the experimental data. (c) 2006 Elsevier B.V. All rights reserved.