Avaliação de modelos preditivos para a estimativa de propriedades físicas de ésteres de ácidos graxos

Abstract

To design and optimize processes involving biodiesel, knowledge of a variety of physicochemical properties of fatty acid alkyl esters is crucial. Among these properties, the following ones are of great importance: normal boiling temperature, critical temperature, critical pressure, acentric factor, vapor pressure, and enthalpy of vaporization. Although reliable data are already available for some esters, experimental determination of these properties for all the existing biodiesel components is an expensive and hard task. The reasons include the difficulty in obtaining the esters in pure forms and in carrying out the experiments, because these substances degrade at severe temperatures. Hence, the application of thermodynamic models becomes essential. In this work, predictive models capable of estimating the previous properties for fatty acid methyl esters and fatty acid ethyl esters have been meticulously compared. All the models studied are based on the Group Contribution concept or on the Corresponding States Principle. The accuracy of each model was tested by comparing its output values to experimental data obtained in a literature survey. Specifically for normal boiling temperature, critical temperature, critical pressure, and acentric factor, the estimation models were also tested in terms of the plausibility of combining experimental and calculated values without violating the expected physical behavior for corresponding pairs of esters and for esters belonging to the same homologous series. The results for vapor pressures indicate that the models may generate estimates outside the desirable accuracy for engineering calculations. For this reason, component-specific parameters for Antoine and Wagner vapor pressure correlations have been proposed. The Wagner constants were generated by using a special methodology, which allows their extrapolations beyond the limits used in the data correlation. A comparative analysis of the results produced by the proposed parameters and by others available in the literature encourages the use of this work’s values in future applications. For all properties, it is expected that the reported results become a guidance for selecting thermodynamic models in process simulation involving biodiesel.