Computational Fluid Dynamics Simulation and Reaction Network Modelling for the Hydrothermal Liquefaction of Shorea Wood Sawdust

This study investigates the hydrothermal liquefaction (HTL) aqueous phase (AP) of Shorea sawdust in a semi-flow batch reactor, focusing on the reaction network and computational fluid dynamics (CFD) simulation. High-performance liquid chromatography (HPLC) was used to detect lignocellulosic decomposition compounds, revealing the presence of glucose, galactose, xylose, furfural, ethanol, and other undefined compounds due to lignocellulosic decomposition. Reaction ordinate (R0) indicates that the reaction progresses steadily as time increases, and higher temperature leads to a greater reaction ordinate, agreeing with Arrhenius’ assumption that gained energy enables molecules to overcome the activation energy barrier. However, saccharide C6 and C5 yield at 220 °C fluctuates as the reaction increases, suggesting secondary reactions. A kinetic model was built based on a reaction network, which was developed based on HPLC results. Arrhenius parameters revealed that reaction yield is influenced by temperature and time, whereas galactose, xylose, and ethanol production are time dependent. In contrast, glucose formation is influenced by both time and temperature. The prediction of saccharide yields by the model confirmed that 220 °C is the optimal temperature for glucose and ethanol production, balancing slow reactions and rapid degradation. CFD simulations show a uniform pressure distribution inside the reaction chamber with high localised pressure at the input (1570 Pa). In addition, feedstock particles tend to distribute along the chamber wall because of the laminar flow, which is consistent with the observation of the experiment. The findings highlight the intricate relationship between reaction conditions and the composition of the HTL product, contributing to a more comprehensive understanding of the process.