The conversion of biomass waste into bioenergy is one of the most important renewable energy production
strategies. However, the energy inputs and outputs for different conversion technologies have not been fully
comparatively evaluated. Herein, we developed a data-driven framework to optimize the process conditions of
conversion technologies, including hydrothermal carbonization, hydrothermal liquefaction, and hydrothermal gasification, anaerobic digestion (AD), pyrolysis, and gasification. Then the predictive properties of products from conversions based on optimal conditions were employed for following life-cycle energy profiles evaluation. The results showed that the developed machine learning models performed well with most of the R2 > 0.80 for all the targets from the six technologies. Energy profile evaluation indicated that the AD was the most potential one with respect to the energy return of investment by comparing with thermal conversions. The energy requirements from thermal conversions were mainly caused by the reactor heating
and feedstock drying for the hydrothermal and dry-thermal conversions, respectively.
Keywords waste to energy (WtE), sustainability, machine learning, optimization, biologic and thermal conversion