Effectiveness of RSM Based Box Behnken DOE over Conventional Method for Process Optimization of Biodiesel Production

Abstract

The conventional way of performing experiments has many drawbacks in terms of time and resources investments and lack of accuracy in analysis. Use of scientific tools for design of experiments (DoE) is a better approach for experimentation such that, minimum number of experiments can successfully give requisite aspects of analysis. Currently, many researchers still use the conventional way for performing experiments which results in higher utilization of resource and time investment, while the same objectives can be achieved by combined application of DoE with response surface methodology (RSM) which utilizes minimal efforts, time and resources. The experimental data from a published research paper have been considered as a case study in which conventional approach was used to decide to conduct experimental runs for transesterification process. The authors reported the optimized reaction conditions viz. methanol to oil molar ratio (6:1), catalyst amount (1 wt. %), reaction temperature (60 °C), and reaction time (15 min) for biodiesel production (98.1% yield) using hydrodynamic cavitation technique. They used a method of varying one factor at a time while the other three factors were constant. Twelve sets of experiments with total 176 number of sample observations were performed with varying factor conditions. Compared to this fifteen-run based on Box-Behnken technique with RSM has been used in the present study to predict the optimized reaction condition. The accuracy of the predicted yield and operating parameters were 99.99% and 99.8% respectively. Thus Box-Behnken DoE method with RSM is clearly a better approach as compared to the conventional one in terms of better resource utilization and time saving.