2025 AIChE Annual Meeting

(586e) Predictive Models for the Hydrothermal Liquefaction Products of Brown Macroalgae

Authors

Fernando Resende, University of Texas at Tyler
This study presents novel predictive models for hydrothermal liquefaction (HTL) of brown macroalgae describing the formation of biocrude, gas, biochar, and water-soluble compounds as products. The models account for the chemical composition of the macroalgae and explain the effects of time, temperature, pressure, and water-to-biomass ratio as input variables to estimate product yields from HTL. To achieve this goal, we used experimental kinetic data to develop a process simulation for the batch HTL of macroalgae. We then applied the design of the experiment to generate simulation runs at different combinations of process variables. The results were used to develop predictive models describing the effects of such process conditions on product yields from the HTL of macroalgae. Next, the predictive models generated were used to optimize the yield of biocrude produced. Also, we used the response surface methodology to visualize the effect of process variables on product yields. Additionally, the models were validated against experimental data from the literature, with 91% agreement within the 95% prediction interval for the biocrude yield model. Analysis of variance showed that the selection of operational parameters significantly affects the biocrude yield. The optimal biocrude yield was 23% at 283 °C, 200 bar, 54 min, and a water-to-biomass ratio of 10:1, with temperature and residence time as the significant variables that affect biocrude yield. Sensitivity analysis of the reaction rate constants allowed for the identification of significant paths that affect biocrude yield. The workflow presented in the study and the predictive models provide an accurate path for modeling various products from the HTL of kelp.