Biodiesel production from waste cooking oil using palmyra fruit husk derived hydrochar catalyst: RSM-ANN optimization, simulation, and techno-economic analysis

Abstract

This study presents a novel integrated framework combining experimental optimization, machine learning, and techno-economic analysis for sustainable biodiesel production from waste cooking oil (WCO). The novelty of this work lies in the development of a low-temperature hydrochar catalyst derived from underutilized Palmyra fruit husk, specifically tailored to handle high-acid WCO through an integrated lab-to-economy assessment. The catalyst was synthesized from Palmyra fruit husk via hydrothermal carbonization at varying temperatures (180–220 °C). The catalytic basicity was directly quantified using the Hammett indicator method, revealing a substantial increase in active sites (3.33 mmol/g) upon functionalization. The catalyst produced by alkali treatment at 180 °C (HC-180B) exhibited a superior surface morphology, crystallinity, and functional properties, enabling a high catalytic efficiency. Process optimization using response surface methodology (RSM) and artificial neural network (ANN) models established a quantitative relationship between the reaction variables and biodiesel yield. Comparative analysis revealed complementary strengths between the models, where RSM provided higher accuracy in predicting the optimal conditions, while ANN demonstrated strong predictive capability across the entire dataset. The optimal conditions were identified at a 9.77:1 methanol: oil molar ratio, 150.6 min reaction time, and 7.73 wt.% catalyst loading, achieving a 92.50% biodiesel yield. The catalyst maintained stable performance (> 90% yield) for three reuse cycles, with deactivation primarily driven by active site leaching and surface fouling as confirmed by spent catalyst characterization. The physicochemical properties of the synthesized product complied with the ASTM D6751 and EN 14214 standard specifications for biodiesel fuel. Techno-economic evaluation using Aspen Plus simulation indicated commercial feasibility with a net present value of 19.71 M$, internal rate of return of 25.09%, and payback period of 4.70 years. This integrated experimental, modeling, and economic approach demonstrates the potential of agricultural waste valorization for cost-effective, scalable, and sustainable biodiesel production.