Statistical Modeling and Optimization of Biodiesel Conversion Produced From WCO Using Response Surface Methodology

Abstract

The growing demand for sustainable alternatives to fossil fuels has increased interest in biodiesel as an eco-friendly,
renewable energy source. However, maximizing biodiesel yield depends strongly on reaction conditions, including molar ratio, catalyst loading, and duration. This study applied Response Surface Methodology (RSM) coupled with a Central Composite Design (CCD) to statistically model and optimize biodiesel conversion. A quadratic regression model was developed and rigorously validated using ANOVA, residual diagnostics, and influence analyses. The model demonstrated high reliability (R² = 0.927, Adj. R² = 0.861), confirming its adequacy for prediction and optimization. Among the studied parameters, the molar ratio and catalyst weight were the most significant contributors, while reaction time played a minor role. Interaction and curvature effects were captured effectively, revealing nonlinear relationships between variables. Contour and 3D surface plots highlighted the optimal conditions for achieving conversions exceeding 90%, with maximum experimental yields reaching 97%. The findings demonstrate the effectiveness of RSM in reducing experimental effort while offering robust predictive power for process optimization. This work provides valuable insights into biodiesel production, highlighting the operational window for achieving high efficiency with minimal resource input.