Evaluation of drag models for CFD simulation of fluidized bed biomass gasification
Journal article, Peer reviewed
Published version
![Thumbnail](/usn-xmlui/bitstream/handle/11250/2459886/2017SolliEvaluation.pdf.jpg?sequence=7&isAllowed=y)
View/ Open
Date
2017Metadata
Show full item recordCollections
Abstract
Gasification of biomass into suitable feedstock has become a feasible alternative technology for reducing the use of energy feedstock from fossil sources. Usually, fluidized bed technology is used in the biomass gasification reactor. Optimization of a fluidized bed reactor needs to take into account the bed behavior in the presence of both biomass and bed material, as well as chemical conversion of particles and volatiles, among other process parameters. CFD simulation of the process is a valuable tool to go about the optimization. However, simulation result validation is limited by the accuracy of input parameters such as those characterizing several drag models given in the literature. This study is focusing on the drag model parameters. The simulation is aimed at validating some of the commonly used models for drag forces against the bed material(s) used in the fluidized bed gasification reactor. Drag models included in this study are those given by Syamlal and O’Brien, Gidaspow, and BVK. The MFiX CFD-software (version 2016.1) from The National Energy Technology Laboratory (NETL) is used. The Two-Fluid Model (TFM) are applied for comparison of the results. The key factors for validation of the drag models are based on the superficial gas velocity at the minimum fluidization condition and the degree of bed expansion. The simulation results show that the minimum fluidization velocity could be predicted using the Gidaspow and BVK drag models by adjusting the particle diameter used in the simulation. For the Syamlal & O’Brien drag model, two parameters are fitted to predict the minimum fluidization velocity. The bubbling bed behavior is not captured using the Syamlal & O’Brien drag model while Gidaspow and BVK drag models fairly captures this phenomenon. The bed expansion from the simulation is higher than that observed in the experiment, and the deviation is even higher with the Syamlal & O’Brien drag model.