Room: Poster Area
Date: Wednesday, 20 May 2026
Time: 15:00 - 16:00 CEST
Session code 4BV.5
Gasification, gas cleaning, upgrading, catalyst performance and modeling on component and system level
Physics-Informed Machine Learning Model For Optimization Of Syngas Production From Biomass And Waste Gasification
Short Introductive summary
Biomass and biogenic waste are attractive renewable energy sources, which can be used to produce syngas and green methane through thermochemical processes such as gasification. The produced gas can be used as energy source or can undergo further processing to produce renewable fuels such as methanol. In order to further develop and optimize the gasification process for different biomass and waste mixture streams and applications, modelling is required to support fast decisions on the operational parametric control during live production. The modelling of these processes is challenging due to the numerous complex reactions and extensive product slate. For this study, a novel physics-informed ML model was developed to predict the complete product slate of biomass and waste gasification under various conditions. The model was trained using an in-house generated dataset covering a wide range of biogenic and biomass waste streams under various conditions. In addition, to predicting the complete product slate, the model can be used to gain insight into the effect of different process conditions and feedstock characteristics.
Presenter
Surika VAN WYK
TNO Energy Transitions, THE NETHERLANDS
Biographies and Short introductive summaries are supplied directly by presenters and are published here unedited
Co-authors:
P.M.R Abelha, TNO Energy Transitions, Petten, THE NETHERLANDS
B.J. Vreugdenhil, TNO Energy Transitions, Petten, THE NETHERLANDS
Session reference: 4BV.5.17