Optimization of Hydrogen Production from Pyrolysis of Biomass Waste

We developed an CFD aided ML-based tool for rapid assessment and optimization of different compositions of biomass in a fluidized bed reactor. First, we use CFD to simulate fluidized bed reactors with known inlet biomass mixtures and obtain corresponding syngas yields. A lumped kinetic mechanisms represents the conversion of cellulose, hemicellulose, and lignin, as well as subsequent cracking of tars into non-condensable gases (H2, CO, CO2, CH4). We use Bayesian analysis/optimization to obtain the ideal operational and mass flow conditions for hydrogen production.