CFD simulations of biofuel bed conversion: A submodel for the drying and devolatilization of thermally thick wood particles

Artikel i vetenskaplig tidskrift, 2013

Biomass can be converted into heat and power in fixed or fluidized fuel beds. Even though the use of these combustion techniques is widespread, detailed knowledge of the combustion behaviors in such beds, which could be used to optimize performance (especially that of fixed-bed boilers), is lacking. Computational fluid dynamics (CFD) is useful in the optimization and in obtaining detailed knowledge of the fuel conversion process, regarding parameters such as burnout, emissions, fuel flexibility, and material wear. However, for such simulations to be viable, computationally efficient models that can handle the most important features of the fuel conversion processes are needed. In the current work, a model is derived for the drying and devolatilization of moist wood particles in an inert atmosphere. The model is compared to three sets of experimental data from two independent research groups. The proposed model predicts drying and devolatilization behavior in agreement with the experimental data using only a few variables. By describing the heat of devolatilization as a linear function of temperature, excellent predictions of the particle core temperature (and thereby, of the rate of devolatilization) are obtained after the drying is complete. For beech and poplar, the average heat of devolatilization is approximately 100 kJ/kg. The implementation of the particle submodel into a CFD code is described in detail. The submodel can predict 2–10 seconds of real time per second of computation, which is shown to be sufficient to treat a bed of particles in 2D or 3D. The discussion addresses how effects on the multi-particle scale, such as local collapses of a fixed fuel bed, can be dealt with in the most efficient manner.