Assessment of the Transesterification Stage of Biodiesel Production I: Application of a Plackett-Burman Design to Select the Process Variables

Although the independent effects of the predictors of the yield in transesterification are well documented, there are still gaps in our current understanding of how specific process variables cumulatively influence the yield of biodiesel. In this study, a two-level Plackett-Burman experimental design was used to assess the set of active process variables that most decisively influence the yield of biodiesel in transesterification. The results of the model in which the parameters were entered at once show in decreasing order, the magnitude of the effect of the process variables on the yield was as follows: amount of FFA > weight of catatlyst > rate of stirring > duration of reaction > temperature > methanol for the ranges of values that were studied. In the sequential/nested ordinary least squares regression model, the order of importance of the predictors was almost maintained however the coefficients varied somewhat. This indicates that the level and the order of entry of variables within each level makes a fundamental difference to the results. The methanol-oil ratio had no influence on biofuel yield within the range of values studied. Overall, the adjusted model in which most of the variables had been taken into consideration explained approximately 94 % of the total variance in the data.