A robust statistical model for optimising biodiesel production from waste cooking oil using non-synthetic caustic potash

Abstract

A Response Surface Methodology was applied as predictive tool for biodiesel optimisation from remediated Waste Cooking Oil (WCO). Caustic potash was synthesised from plantain peels and used for WCO conversion into biodiesel. Four process variables, temperature, time, alcohol to oil molar ratio and catalyst concentration, were considered in thirty experimental runs using a Central Composite design. Optimal conditions for biodiesel yield from WCO transesterification were determined using numerical optimisation. Highest biodiesel yield of 93.6% was produced at 65°C temperature, 90 min' time, alcohol to oil molar ratio of 9:1 and catalyst concentration of 2%. At probability level of 5%, the RSM model was established to be significant with R-squared value of 0.9966 and Adequate Precision of 52.802. From analysis, optimum biodiesel condition of 93.06% was predicted at reaction conditions of 62.07°C temperature, reaction time of 92.82 min, alcohol to oil ratio of 8.31, and catalyst concentration of 1.85%.