Stochastic reactor modeling of biomass pyrolysis and gasification

Abstract

In this paper, a partially stirred stochastic reactor model is presented as an alternative for the modeling of biomass pyrolysis and gasification. Instead of solving transport equations in all spatial dimensions as in CFD simulations, the description of state variables and mixing processes is based on a probability density function, making this approach computationally efficient. The virtual stochastic particles, an ensemble of flow elements consisting of porous solid biomass particles and surrounding gas, mimic the turbulent exchange of heat and mass in practical systems without the computationally expensive resolution of spatial dimensions. Each stochastic particle includes solid phase, pore gas and bulk gas interaction. The reactor model is coupled with a chemical mechanism for both surface and gas phase reactions. A Monte Carlo algorithm with operator splitting is employed to obtain the numerical solution. Modeling an entrained flow gasification reactor demonstrates the applicability of the model for biomass fast pyrolysis and gasification. The results are compared with published experiments and detailed CFD simulations. The stochastic reactor model is able to predict all major species in the product gas composition very well for only a fraction of the computational time as needed for comprehensive CFD.