(430c) Incorporating Intra-Particle Temperature Gradient in Dpm Modeling of Biomass Fast Pyrolysis

Authors: 
Xiong, Q., Oak Ridge National Laboratory
Pannala, S., Oak Ridge National Laboratories
Daw, S. C., Oak Ridge National Laboratory

Incorporating intra-particle temperature gradient in DPM modeling of biomass fast pyrolysis

Qingang Xiong, Sreekanth Pannala, Stuart C. Daw

Oak Ridge National Laboratory, Oak Ridge, Tennessee 37831, USA

Emails: xiongq@ornl.gov (Q. Xiong), pannalas@gmail.com (S. Pannala), dawcs@ornl.gov (S. C. Daw)

Abstract Numerical simulation plays a vital role in the development of advanced technologies for biomass fast pyrolysis. Because of its relatively low thermal conductivity, biomass particles are always thermally thick and the effects of intra-particle temperature gradient on the overall reactor performance need to be considered. In this study, intra-particle temperature distribution was modeled and incorporated into the discrete particle simulation (DPM) for modeling biomass fast pyrolysis in fluidized-bed reactor. Each biomass particle was tracked individually and the inter-particle interactions were resolved directly via the soft-sphere collision model. A global multi-component multi-step kinetics was employed to model the biomass fast pyrolysis reactions. To consider the temperature gradient inside biomass, a mathematical distribution of intra-particle temperature which was developed from single-particle models was assigned to each biomass particle depending on its conversion history. With the assigned intra-particle temperature information, the pyrolysis reaction rates were obtained semi-analytically. The incorporation of the intra-particle temperature distribution into DPM was carried out in the open-source MFIX-DEM code. A small fluidized bed reactor consisting of several thousand biomass particles was simulated. The results were compared with those from the DPM simulation without consideration of the intra-particle temperature gradient. In the future work, sand particles will be included and parallel computing will be launched to simulate a laboratory-scale bubbling fluidized-bed reactor for experimental validation.