Maximizing biodiesel yield using mixture of waste cooking oil and mustard oil via RSM and ANN techniques

This research presents an exceptional study aimed at optimizing biodiesel production from an innovative blend of waste cooking oil (WCO) and Mustard oil (MO). Addressing the cold flow properties of WCO and the unique characteristics of MO, mixture of these two oils have been utilized and a combination of Response Surface Methodology (RSM) and Artificial Neural Network (ANN) techniques used to maximize biodiesel yield. The operating parameter, including reaction speed, RPM (100 – 500 rpm), reaction time (1-3 hours), catalyst concentration (0.2 – 2% w/w), methanol to oil ratio (8:1– 12:1 v/v), and reaction temperature (50 - 60 ℃ ). The maximum actual/experimental biodiesel yield of 95% was found at reaction speed of 500rpm, catalyst concentration of 1.1 w/w, reaction time of 2 hours, methanol to oil ratio of 10:1, and reaction temperature 55 ℃. By systematically comparing the efficiency of RSM and ANN, This extraordinary research not only advances the field of biodiesel production but also contributes significantly to addressing the pressing global challenges of energy sustainability and reduced reliance on fossil fuels. The interaction of the several operating parameters on biodiesel yield has been investigated using two methodologies: Response Surface Methodology (RSM) and Artificial Neural Network (ANN). RSM provides better modeling of parameter interaction, while ANN exhibits lower comparative error when predicting biodiesel yield based on the reaction parameters. The percentage improvement in prediction of biodiesel yield by ANN is found to be 12% as compared to RSM. For training, the R2 value is almost equal to 0.9991011, and for validation, it is 0.9540232. This study emphasizes the merits of both techniques’ response surface methodology and artificial neural network for the prediction of optimum biodiesel yield.