Optimization of biodiesel yield from waste cooking oil and sesame oil using RSM and ANN techniques

Abstract

In the era of global energy crises and the pressing concern of fossil fuel depletion, the quest for sustainable alternatives has become paramount. The current study aims to optimize biodiesel extraction from a combination of waste cooking oil (WCO) and sesame seed oil (SSO) through response surface methodology (RSM) and artificial neural network (ANN). The cold flow properties of biodiesel produced from WCO are a major obstacle to the commercial use of biodiesel. On the other hand, SSO possesses better oxidation stability and cold flow properties. A mixture of waste cooking oil (i.e. 70 % by volume) and sesame seed oil (i.e. 30 % by volume) has been prepared for biodiesel production via a microwave-assisted transesterification process. For biodiesel yield optimization, the interaction among the operating parameters is developed by RSM, whereas biodiesel yield is predicted by ANN. The operating parameters include reaction speed, RPM (100–600 rpm), reaction time (1–5 min), methanol to oil ratio (8:1–12:1 v/v), and catalyst concentration (0.1–2 % w/w). The highest biodiesel yield of 94 % is found at a reaction speed of 350 rpm, reaction time of 3 min, catalyst concentration of 1.05 w/w, and methanol to oil ratio of 10:1. Furthermore, it is discovered that when estimating biodiesel production rate depending on reaction constraints, ANN shows lower comparative error compared to RSM. The results show that ANN outperforms RSM in terms of percentage improvement when it comes to biodiesel production prediction.

Keywords

• Biodiesel yield
• Waste cooking oil
• Sesame seed oil
• Response surface methodology
• Artificial neural network
• Optimization