Production optimisation of mixed oil (rubber seed oil–fish oil) feedstock using response surface methodology and artificial neural network

Abstract

This study is aimed at optimising the reaction parameters involved in transesterification of Mixed Non-edible Oil (rubber seed oil–fish oil) feedstock using response surface methodology (RSM) and artificial neural network (ANN). The reaction parameters such as the methanol-to-oil molar ratio (M:O), catalyst concentration (CC), reaction temperature (Rt) and the reaction time (RT) were investigated on biodiesel samples. The reaction process was optimised with a M:O of 7.5:1; a CC of 1.25 wt. %; Rt of 57.5°C and RT of 75 min with an optimum volumetric yield of 91.5 v/v%. The model attained from the results of analysis of variance indicates that the model developed was significant. A very small value of p (<0.0001) represents the strong correlation among biodiesel yield and model variables. Also, the R² values obtained from RSM and ANN were found to be 0.9918 and 0.8941 which indicates that model developed was fit and satisfactory with the actual experimental values. The fuel properties and FTIR analysis of biodiesel obtained with optimised conditions demonstrated its appropriateness as an alternative fuel for diesel engines.

KEYWORDS:

• Rubber seed oil
• fish oil
• transesterification
• optimisation
• Box–Behnken design (BBD)
• artificial neural network (ANN)

Acknowledgement

Authors are very grateful to Mr. Ramji Veerappan for his support throughout research.

Disclosure statement

No potential conflict of interest was reported by the author(s).