화학공학소재연구정보센터
Industrial & Engineering Chemistry Research, Vol.53, No.30, 12129-12142, 2014
Simulation of Bed Dynamics and Primary Products from Fast Pyrolysis of Biomass: Steam Compared to Nitrogen as a Fluidizing Agent
Fast pyrolysis of biomass, using steam as a fluidizing agent, provides several benefits. In this paper, an unsteady multiphase computational fluid dynamics (CFD) model coupled with a comprehensive kinetic scheme for primary pyrolysis is used to obtain the formation rates of primary products and compare the profiles when operating with steam and nitrogen. The model only considers the physical effects of the fluidizing gas at the moment, although a literature review indicates the existence of various chemical and surface-interacting effects. At stabilized pyrolysis reaction rates, the product yields were compared to data found in the literature, which indicated similar yields; this supports the correct implementation of the kinetic model. However, the difference in overall rate and composition is very small when steam is compared to nitrogen. The simultaneous simulation of bed dynamics indicate a shifted formation rate of primary products toward the lower part of the fluidized bed, with an increase in solid vapor contact time and better temperature distribution as a result. More specifically, total heat flux to the biomass increased by 1396 in the lowest part of the reactor. In addition, more heat from the sand is carried through the gas phase when using steam: an increase by 9% in the overall reactor (25% in the lowest part), as indicated by the results. Finally, since no substantial differences in overall product formation rate and composition were found, the considerable effect of steam found in experiments and the literature is mainly (not exclusively) attributed to the chemical and surface-interacting mechanisms. Because of the complex nature of secondary pyrolysis in this process, a comprehensive gas-phase kinetic model is needed to investigate the effects of steam further. Coupling of both is difficult, because of computational constraints, as the present model already is very demanding. The obtained profiles of formation rate of primary products can however be used as an input to another model specifically made for studying homogeneous secondary pyrolysis reactions.