Ethanol and methanol Unifac subgroup parameter estimation in the prediction of the liquid-liquid equilibrium of biodiesel systems

The biodiesel conventional production method is based on the homogeneous transesterification reaction, in which the quality and quantity of the raw materials used have considerable effect on the separation of phases of the generated products. The presence of glycerol in biodiesel, for example, has a maximum permissible value of only 0.02% (EM 14214). On the other hand, high purity glycerol needs to be marketed separately. In this way, a better understanding of the phase equilibrium becomes fundamental for the proper design, operation and optimization of the process. In the literature, the thermodynamic modeling focuses on the UNIQUAC and NRTL models. Such models can predict phase equilibrium for each system separately, and the interaction parameters are distinct and unique. However, the use of several types of vegetable or residual oils has been reported for biodiesel production. On the other hand, group contribution thermodynamic models, such as UNIFAC, could be used to predict the phase equilibrium behavior of reactional biodiesel systems formed from these different types of oils. This work contributes to this subject, providing generalized UNIFAC interaction parameters for the already specified subgroup methanol (CH3OH) and a new specific subgroup, ethanol (CH3CH2OH). The UNIFAC-LL database parameters were used, except for methanol and ethanol subgroups proposed here, for calculations with ten vegetable oils, for which experimental data are available. The mean square error (0.92%, methanol; 1.20%, ethanol) was considerably lower when compared to the available UNIFAC-LL data (10.58%, methanol; 8.89%, ethanol) and work of the literature (2.21%, methanol; 3.92%, ethanol) that uses a greater number of adjustable parameters. The results show that the methodology employed is consistent and can be useful in predicting liquid-liquid phase equilibrium for both reactive methyl and ethyl biodiesel systems considered. (C) 2019 Elsevier B.V. All rights reserved.