An Application of Multivariate Regression Analysis in Yield Optimisation of Biodiesel Synthesis from Used Sunflower Oil
摘要
The upsurging price of fuel along with the rising scarcity of it across the world has put countries in a fix. They are unable to meet the rising demand due to depletion in supply. The continuous burning of the fuels has also put a negative impact on the environment. To save the environment and to provide a solution for this drastic change that the fuel industry has been going through in the recent years, we need to shift to renewable energy. This can be made possible by generating fuel which is not only eco-friendly but can also be renewed easily and biodiesel is the most suitable alternative for it. In this paper, an effort has been made to obtain biodiesel from used sunflower oil. A yield optimisation-based Design of Experiment (DOE) for biodiesel made from used sunflower oil has been conducted in this work. The factors chosen were methanol to oil molar ratio (6:1, 9:1, and 12:1), catalyst amount (0.5%, 1%, and 1.5%), duration (45 min, 75 min, and 105 min), and temperature (50℃, 55℃, and 60℃) as the operational conditions. For determining yield, these parameters were run at different treatment conditions, which were conducted using Taguchi's orthogonal array L9. To find the optimal solution, Multivariate Regression Analysis with the data has been carried out using MS excel data analysis tool pack. Comparative study between yield obtained from experimental transesterification result and regression data analysis has been done in the aforementioned studies. In the study, R2 value of 0.8098 showed a better data fitting. It was also experienced that methanol and oil molar ratio with p-value of 0.0411 from ANOVA study was discovered to have the most substantial outcome on the yield followed by catalyst amount.