Optimization of Vegetable Oil-Based Biodiesels by Multi-Response Surface Methodology (MRS) using Desirability Functions

Abstract

Environmental concerns associated with petroleum resources have propelled the development of sustainable and renewable alternatives to petroleum based products. Vegetable oil is one amongst the foremost abundant bio-based feedstocks. The interest in using vegetable oils and low molecular weight alcohols by direct transesterification have shown great potential as alternatives to petroleum-based diesel, and the production of bio-based diesel continues to increase. Utilization of multi-response surface methodology (MRS) for the most effective combination effect or response from the uses of input to output variables to optimize the yield and higher heating values (HHV) of biodiesels was investigated. In this work, utilizing variety of non-edible vegetable oils like castor (Ricinus communis L), jatropha (Jatropha curcas), and neem (Azadirachta indica) seeds and several process variables or inputs, including mixing time, mixing speed, process temperature and catalyst dosage to formulate high quality renewable fuels were further explored. The outputs were yield, viscosity, higher heating value, density and turbidity. The proposed optimization scenarios for biodiesel using the statistical (MRS) models was aimed to optimize the processes to achieved high conversion and higher heating values, while reducing the reaction time, turbidity, density, and viscosity in the samples. The results showed catalyst dosage as the most important variable for all the three samples. For maximum yield of 100%, the molar ratio of 6.25, catalyst of 0.75 wt.%, reaction speed of 499.99 rpm, reaction time of 19.88 min and temperature of 24.50 oC were found as optimal conditions; while the molar ratio of 5.60, catalyst of 1.01 wt.%, the reaction speed of 499.5 rpm, reaction time of 20.00 min and temperature of 35.50 oC were optimal conditions for maximum biodiesel yield.