Publication | Peer reviewed papers | Modellierung und Simulation
Multi-scale modelling of a fluidized bed biomass gasifier of industrial size (1 MW) using a detailed particle model coupled to CFD: Proof of feasibility and advantages over simplified approaches
Published 15 June 2023
Citation: von Berg L, Anca-Couce A, Hochenauer C, Scharler R. Multi-scale modelling of a fluidized bed biomass gasifier of industrial size (1 MW) using a detailed particle model coupled to CFD: Proof of feasibility and advantages over simplified approaches. Energy Conversion and Management. 15 June 2023.286:117070
Abstract
Fluidized bed biomass gasification is a complex process whereby gas source terms are released by reactions at the particle level during the movement of fuel particles throughout the reactor. The current study presents for the first time the application of a multi-scale modelling approach for a fluidized bed biomass gasifier of industrial size, coupling a detailed one-dimensional particle model based on the progressive conversion model (PCM) with a commercial CFD software. Results of particle movement and gas source terms are compared with results of an additional simulation employing the simplified uniform conversion model (UCM) which is commonly used in literature. Validation at the particle level showed that the UCM leads to a massive underprediction of the time needed for pyrolysis whereas the PCM is in good agreement with experimental data. This heavily influences the gas sources released during pyrolysis of the biomass particles in the coupled reactor simulations. Volatiles are much more concentrated to the close proximity of the fuel feed when using the UCM whereas the PCM leads to a more homogeneous distribution over the reactor cross-section. The calculation time analysis of the coupled simulations showed that despite the increased complexity, the PCM shows only an increase of 20% in calculation time when compared to the UCM, whereas it is much better suited for these conditions. The coupled multi-scale simulations using the PCM showed the numerical feasibility of the modelling approach for 1,200,000 bed parcels and about 80,000 reacting fuel parcels and furthermore highlighted the importance of a comprehensive description of the particle level.