Yield optimization and scaling of fluidized beds for tar production from biomass

A numerical study is performed in order to evaluate the performance and optimal operating conditions of fluidized bed pyrolysis reactors used for condensable tar production from biomass. For this purpose, a previously validated biomass particle pyrolysis model is coupled with a detailed hydrodynamic model for the binary gas particle mixture. The kinetics scheme is based on superimposed cellulose, hemicellulose, and lignin reactions. Any biomass feedstock can be simulated through knowledge of its initial mass composition with respect to these three primary components. The separately validated hydrodynamic model is based on a three-fluid model (gas, sand, and biomass) derived from the kinetic theory of granular flows. Separate transport equations are constructed for each particle class, allowing for the description of such phenomena as particle segregation and for separate temperatures for each particle class. The model is employed to investigate the effect of various operating conditions on the efficiency of tar collection in fluidized bed reactors. Results indicate that, at fixed particle size, the operating temperature is the foremost parameter influencing tar yield, The biomass feed temperature, the feedstock, and fluidization velocity magnitude, all have minor impact on the yield. The particle diameter has a considerable influence on the short-time tar yield, but it is inferred that it may have a more moderate influence on the steady-state tar yield. For the range of fluidizing gas temperatures investigated, optimum steady-state tar collection is obtained for 750 K under the assumption that the pyrolysis rate is faster than the feed rate; the predicted optimum temperature is only slightly higher if this assumption is not satisfied. Finally, scale-up of the reactor is addressed and is found to have a small negative effect on tar collection at the optimal operating temperature. It is also found that slightly better scaling is obtained by using shallow fluidized beds with higher fluidization velocity.