Process and Systems Modeling of Hydrothermal Liquefaction for Fuels and Bioproducts from Wet Organic Wastes

Hydrothermal liquefaction (HTL) can convert wet organic waste into fuels and bioproducts, but variation in feedstocks and operating conditions can propagate from product yields to system performance and deployment. This study linked machine learning based HTL yield prediction with process simulation, techno-economic analysis (TEA), life cycle assessment (LCA), uncertainty analysis, and surrogate modeling. A literature data set of 402 experiments was used to predict biocrude, aqueous, gas, and solids yields, with 71.8% of predictions within ±5 wt% of experimental values. The predicted yields reproduced 77.1% of internal rate of return (IRR) and 83.2% of GWP 100 estimates within ±5% and ±10 kg CO 2 e/kg, respectively, and were integrated into two deployment configurations (centralized and decentralized HTL) and two product pathways (biobinder/biofuel and sustainable aviation fuel [SAF]). Process and system scale analyses showed that operating conditions strongly affecting HTL performance could have weaker or competing economic effects due to tradeoffs among product yields, water handling, energy requirements, and capital costs. Across representative feedstocks, decentralized biobinder systems achieved median IRRs up to 35.8%, compared with up to 8.0% for decentralized SAF systems. Feedstock price, followed by biocrude quality, and biocrude yield emerged as the leading economic drivers. Monte Carlo filtering identified waste tipping fee thresholds $25.1–196.7 per wet metric tonne for achieving the 10% IRR target. These results show that existing tipping fee structure can support HTL deployment in some cases, while higher tipping fees or additional financial incentives can expand viability and enable higher investment returns.