Prediction of Fuel Properties of Biodiesel Produced by Sequential Esterification and Transesterification of Used Frying Soybean Oil Using Statistical Analysis

Purpose The present work envisages developing an efficient method for maximization of biodiesel yield from used frying soybean oil (UFSO) containing high free fatty acid (FFA). Multivariable statistical correlations capable of predicting fuel properties as a function of operating variables have been established. Inter-relations among important fuel properties of the combined-biodiesel have also been presented. Method A sequential biodiesel synthesis process comprising of H2SO4 catalyzed esterification of physically pretreated UFSO followed by NaOH-catalyzed transesterification of unreacted triglyceride was conducted separately to maximize the overall yield of fuel-grade biodiesel. HPLC analysis was conducted for quantification of FAME. Linear least square techniques have been applied to correlate fuel properties as a function of major operating variables and to develop inter-relationship among the fuel properties. Results The yield of used frying soybean oil methyl ester (UFSOME) was 94.7 and 97% at the best combination of operating parameters governing the esterification and transesterification steps, i. e. 1, 4 h reaction time; 0.30, 0.20v/v alcohol to oil ratio; 0.75, 1.0 wt% catalyst concentration; 50, 60 degrees C reaction temperature; 600, 400 rpm stirrer speed, respectively. Conclusion The esterification step could significantly improve the overall yield of biodiesel from UFSO ensuring effective waste valorization. The developed multivariable regression equations can predict the important fuel properties of the combined-biodiesel as a function of operating variables. Correlations between fuel properties are useful in assessing product quality.