Methanolysis of African pear seed oil catalyzed with acid activated empty palm fruit bunch ash: Optimization and sensitivity analysis

This study investigated the catalytic effect of acid activated empty palm fruit bunch ash (AAEPFBA) on the transesterification reaction of Africa pear seed oil (APSO), optimization of the process and sensitivity analysis. The AAEPFBA was synthesized from waste palm fruit bunch, and modified. The activation was best achieved by adding H₃PO₄ acid in a ratio of 1:2 (g/ml). The modification of the catalyst increased the alkaline properties and surface area and decreased the particle size and adsorption energy. The biodiesel produced was characterized and compared with standard properties for its application Comparing the three models prediction using response surface methodology (RSM), rsm-genetic algorithm (RSM-GA) and artificial neural network (ANN), the RSM-GA model values of the process variables at temperature of 61.21°C, reaction time of 3.3 h, 10.3:1 methanol/oil molar ratio, 3.16wt% catalyst concentration, and agitation speed of 320.51 rpm gave an experimental optimal yield of biodiesel of 90.1 %. In addition, the RSM, RSM-GA and ANN statistical percentage errors (SPE) are 0.64, 0.06, and 0.29, respectively with RSM-GA having the best prediction of biodiesel yield. The novel model analysis of the process variables connection weight values of MATLAB R2015a shows that the catalyst has the highest percentage (40.7 %) of relative impact on the yield of biodiesel while the methanol/oil ratio has the least percentage contribution less than 5 %. The overall result shows that the catalyst EPFBA activated with phosphoric acid presently has high potential over other catalysts in converting APSO to biodiesel, and the application of RSM-GA tool predict the require values of the variables for maximum yield of biodiesel.