Process parameter assessment of biodiesel production from a Jatropha–algae oil blend by response surface methodology and artificial neural network

ABSTRACT

Biodiesel production from different feedstocks is one of the effective ways to anticipate the problems related with fuel crisis and environmental issues. In this study, the response surface methodology (RSM)-based Box–Behnken experimental design (BBD) is used to optimize the parameters of biodiesel production for the blend of Jatropha–algae oil such as molar ratio, temperature, reaction time, and catalyst concentration. A significant quadratic regression model (p < 0.0001) with R² of 0.9867 was achieved under the condition of molar ratio 6–12%, KOH 0–2%, reaction time 60–180 min, and temperature 35–55°C. The artificial neural network (ANN) with the Levenberg–Marquardt algorithm was also trained in this study with the topology 4-10-1 with a predicted correlation coefficient of 0.9976. From the results, it is also found that the predicted values of yield are in good agreement with the results of RSM correlations.

KEYWORDS:

• Biodiesel
• Response surface methodology
• artificial neural network
• transesterification
• Jatropha–algae oil