Investigating the Predictive Capabilities of MARS for Biomass Pyrolysis Kinetics: A Case Study on an Oil Palm Empty Fruit Bunch

This study aimed to emphasize the potential of pyrolysis of oil palm empty fruit bunch (PEFB) biomass as a sustainable energy source and underscore the significance of considering multiple factors in implementing renewable energy. The kinetic analysis of PEFB pyrolysis was carried out using the Friedman, Kissinger-Akahira-Sunose (KAS), Ozawa-Flynn-Wall (OFW), and advanced Vyazovkin (VZ) methods, which yielded matching trends. The average activation energies (<i>E</i>) were found to be 187.40, 165.23, 166.24, and 183.23 kJ/mol for the Friedman, KAS, OFW, and VZ methods, respectively. Additionally, thermodynamic parameters, specifically the changes in enthalpy of activation (Δ<i>H</i>), entropy of activation (Δ<i>S</i> ^‡), and Gibbs free energy of activation (Δ<i>G</i> ^‡) were derived. Average values of Δ<i>H</i> ^‡, Δ<i>G</i> ^‡, and Δ<i>S</i> ^‡ were 161.44, 169.52, and -15.28 kJ/mol K, respectively. <i>E</i> was modeled using multivariate adaptive regression splines (MARS), considering temperature (<i>T</i>), heating rate (β), and conversion (α) as the contributing factors. The MARS model demonstrated a high coefficient of correlation (<i>R</i> ² = 0.987), low mean absolute error (MEA = 2.71), and root-mean-square error (RMSE = 4.94) to predict <i>E</i>. The results of this study highlight the potential of MARS to accurately represent complex relationships, predict biomass pyrolysis kinetics, and demonstrate the feasibility of PEFB pyrolysis as a sustainable energy source.