Integrated kinetic modeling and Latin hypercube sampling-based sensitivity analysis with bio-oil characterization in the pyrolysis of Sargassum sp

This study investigates Sargassum sp. pyrolysis kinetics using two competitive lumped kinetic models: Di Blasi (1996) (Model 1) and Farag et al. (2014) (Model 2), across varying particle sizes. Sensitivity analysis was assessed using Latin Hypercube Sampling (LHS) and Spearman rank correlation. The results indicate that Model 1 provided a superior fit as reflected by higher R² and lower RMSE values. Sensitivity analysis revealed that the most influential parameters in Model 1 were Ea₃ (for 10–40 mesh) and Ea₁ (for 40–70 and 70–100 mesh), whereas in Model 2, Ea₄ (for 10–40 mesh) and Ea₃ (for 40–70 and 70–100 mesh) were most significant. Under ±30 % parameter variation, total error values predominantly remained below 10. A light-phase bio-oil (29.72 MJ/kg) was rich in alcohols, while a heavy-phase bio-oil (2.95 MJ/kg) dominated by carboxylic acids. These findings highlight the strong potential of Sargassum-derived light-phase bio-oil as an alternative to fossil fuels.