Biodiesel production unpacked: Data-driven analysis of key operational factors

Biodiesel is a promising renewable alternative to depleting fossil fuels. However, variabilities in the operational factors that affect biodiesel yield make it difficult to ascertain which parameters mostly affect biodiesel production yield. The objective of the current study was to statistically evaluate the commonly reported major operational parameters that affect biodiesel yield using a wide dataset collected from the literature. CART@Regression results implied that feedstock type and temperature were the most and least important factors, respectively, for biodiesel production from both used and neat oils. Excluding temperature, all factors were above 73 % as important as feedstock type for used oils, suggesting that optimization of all five operational parameters (i.e., feedstock type, alcohol-to-oil-molar ratio, reaction time, catalyst type, and catalyst quantity) will increase biodiesel yield from used oils. In contrast, catalyst quantity, alcohol-to-oil molar ratio, and temperature were below 50 % as important as feedstock for neat oils, which implied that their control does not have much effect on biodiesel yield for the current dataset. Basic statistics (outlier test, normality check and correlation analysis), principal component analysis and two-dimensional surface contour plots were conducted. Multiple criteria analysis showed that potassium oxide is the best catalyst type for used oils, while muhua oil is the best feedstock for biodiesel production from neat oils. The present study offers fundamental knowledge regarding operational parameters which could support large-scale biodiesel production and hence boost the biorefinery sector.