Predicting biocrude oil yields from hydrothermal liquefaction of an energy grass: Machine learning approach and experimental validation

In recent years, our world has been grappling with the challenges of global warming, prompting a search for alternative energy sources that can provide clean and sustainable power without adversely impacting the environment. Biomass has emerged as a viable option due to its potential as a renewable and environmentally friendly energy source. Among various processes available, hydrothermal liquefaction (HTL) has gained significant attention for its capability of converting wet biomass into biocrude oil in a relatively short timeframe. Previous studies have explored integration of machine learning with HTL to forecast the yields of biocrude oil. This work aims to develop a machine learning model specifically tailored to predicting the biocrude oil yields from the HTL process. Furthermore, experiments were done using Napier 4190 (Giant Juncao Grass) as a biomass source to verify the accuracy of this machine learning model. The findings revealed that the machine learning model achieved an R² value of approximately 0.8. Model predictions and experimental results show a discrepancy of approximately 6-14%.