Production of Purified H2, Heat, and Biochar from Wood: Comparison between Gasification and Autothermal Pyrolysis Based on Advanced Process Modeling

Biomass gasification is an interesting route for renewable hydrogen production, but it is still hampered by technical, environmental, and economic issues. A first key step toward its development is the quantification of mass and energy balances of the integrated process. This work compares different processes to produce purified H2 from wood and also other products (heat, biochar) at medium-scale power (20 MW of biomass power inlet that corresponds to 3.7 tdry/h). Three complementary processes were modeled under Aspen Plus, including biomass drying, gasification-pyrolysis reactors, and advanced syngas upgrading units. The first two cases are based on oxygen/steam gasification (1) with or (2) without catalytic reactors (steam reforming and water–gas shift). The third case is an autothermal oxidative pyrolysis resulting in biochar and syngas. The syngas cleaning process was detailed with a special focus on a partial oxidation (POX) unit to reduce the tar content. This unit was modeled by coupling Aspen Plus with Chemkin to predict tar and syngas composition by detailed elementary mechanisms. A hybrid hydrogen separation process is proposed combining membrane and pressure swing adsorption. A cape-open module for membrane modeling (called Memsic) was included in the whole process model. The global energetic efficiency is 75.4, 77.8, and 80.4% net for scenarios 1, 2, and 3, respectively. The hydrogen yields are 79, 26, and 18 gH2/kgbiomass,dry after separation, and heat efficienciescorresponding to hot water productionare 23.4, 60.0, and 49.0% net, respectively. Option 3 produces 110 gbiochar/kgbiomass,dry, which is a carbon sink. All utilities and consumables were also determined. This model can be used for technoeconomic and life cycle assessment studies. This methodology is also of interest to model all other thermochemical processes with detailed kinetics embedded in process models.