Predictive capability evaluation and optimization of sustainable biodiesel production from oleaginous biomass grown on pulp and paper industrial wastewater.

Biodiesel, as a green fuel, acts as a potential candidate to supplement conventional fossil fuels. This research study targets green environment (using biodiesel) and clean environment (reduce wastewater) by producing biodiesel through oleaginous biomasses (Yarrowia lipolytica , Metschnikowia pulcherrima and Lipomyces starkeyi) grown on pulp and paper industrial wastewater. Batch culture studies were explored for the potential feedstock of the oleaginous organism by the synthesis of single cell oil and fatty acid methyl ester (FAME) yield. Response surface methodology (RSM) was used to design the optimal experimental matrix and identify the optimal process conditions that enhance the FAME yield. To determine the inherent characteristics of the growth of oleaginous biomasses on the industrial wastewater, a data-driven adaptive neuro-fuzzy inference system (ANFIS) is implemented. Y. lipolytica strain cultured shown high biomass concentration of 32.36 g/l with biomass productivity of 5.39 g/l/d was considered for further scale-up for the transesterification process. Results indicated that the maximum yield of 0.48 (g-biodiesel/g-lipid) was obtained under the 2.5 g of lipid dosage with 0.02 g/ml of catalyst concentration by constant stirring at 70 °C. The optimum conditions to achieve maximum FAME yield of 1.154 g/g was obtained at 2.485 g, 70.87 °C and 0.021 g/ml for lipid dosage, temperature and catalyst concentrations, respectively. • Growth of Y. lipolytica, M. pulcherrima and L. starkeyi yeasts on paper and pulp wastewater explored • Bagasse seepage water, bagasse wash water and hardwood wash water were used • FTIR spectroscopical analysis confirmed the presence of prime biodiesel components • Enhanced FAME yield (1.154) was accomplished by response surface methodology optimization • ANFIS demonstrated the intrinsic characteristics of transesterification process for the better prediction of FAME yield [ABSTRACT FROM AUTHOR]