Pareto-hierarchical clustering framework for biodiesel transesterification

Abstract Biodiesel is commonly produced via transesterification process, usually through the use of batch type reactor. There are significant gaps in batch reactor technologies, where the importance of operating conditions are often concluded based on the steady-state conditions. In this study, a Pareto-Hierarchical framework were employed, based on results of a 4-factor 3-level full factorial Design of Experiments. Four parameters of interest such as agitation speed, catalyst loading, methanol-to-oil ratio, and temperature, were firsts analysed using transient Pareto approach followed by hierarchical clustering. The results indicate that low methanol: oil ratio favours transesterification at the beginning, with increasing importance as the process proceeds. From it, the magnitude of standardised effect increases from 0.38 to 14.21, representing an over 37-fold improvement. Conversely, agitation speed showed a reduction of 37-fold in standardised effect throughout the transesterification process, plateauing around mid-way of the reaction. Catalyst loading and temperature shared similar trends, as they influence the activation energy. Their standardised effects peak at 180 s which is in the middle of the physical-limiting regime. Therefore, this study provides a framework to develop an analyses to make dynamic optimisation strategy better in terms of economics and yield efficiency.