Modeling and optimization of Fischer–Tropsch synthesis in the presence of Co (III)/Al2O3 catalyst using artificial neural networks and genetic algorithm

Fischer–Tropsch synthesis is a collection of chemical reactions that converts a mixture of carbon monoxide and hydrogen into hydrocarbons. In this study, application of FTS is studied in a wide range of synthesis gas conversions. Artificial neural networks (ANN) were used to predict the molar percentage of CH4, CO2 and CO in the Fischer–Tropsch process of natural gas and also genetic algorithm (GA) was applied to obtain the optimum values of operational parameters. The input parameters consist of a 3-dimensions vector which includes the reaction time, operating pressure and temperature and also the output was molar percentage of CH4, CO2 and CO. Topology and decision parameters have been calculated by trial and error and acceptable correlation coefficients (R2 = 0.94 for CH4, R2 = 0.93 for CO2 and R2 = 0.96 for CO) were obtained. Also the results obtained by sensitivity analysis represent that operation time has significant influence on molar percentage of CH4 as desired product with respect to other operational parameters. Finally the results justify that GA-ANN could be effectively used for FTS as a powerful estimation technique.