Box-Behnken assisted RSM and ANN modelling for biodiesel production over titanium supported zinc-oxide catalyst

Global utilization of Green fuels mitigates the harmful effects of climate change caused by greenhouse gas emissions from fossil fuels. Biodiesel was identified as an alternative fuel to satisfy the growing need for energy. Waste cooking oil is applied as a feedstock due to legitimate worries about using food crops to produce fuel. In this study, optimization of biodiesel synthesized from waste cooking soybean oil in the presence of a novel enhanced titanium-supported zinc oxide (ZnO/TiO2) catalyst had been reported. The aim was to use DOE to correlate relationships between optimal biodiesel productivity and the operating parameters. The influence of catalyst loading, methanol-to-oil ratio, and reaction temperature were investigated using a time-efficient BoxBehnken design of response surface methodology and artificial neural network. The predicted and experimental yield was comparable with 94.04 % (BBD-RSM), 93.99 % (ANN), and 94.42 % respectively. A significant biodiesel yield of 94.93 % was obtained at optimal operating conditions of catalyst loading (21.9 wt%), reaction temperature of 55 OC, and methanol oil ratios of 8:1. Comparative analysis indicates higher prediction capabilities for RSM than the ANN model in terms of lowest error functionality and highest correlation coefficient. However, the obtained FAME has properties within the standard limits set for biodiesel.