Search
Close this search box.

Xu, Y.; Shahnam, M.; Fullmer, W. D.; Rogers, W. A., "CFD-Driven Optimization of a Bench-Scale Fluidized Bed Biomass Gasifier using MFiX-TFM and Nodeworks-OT", NETL-TRS-3-2019; NETL Technical Report Series; U.S. Department of Energy, National Energy Technology Laboratory: Morgantown, WV, 2019; p 28. DOI: 10.18141/1506664.

Abstract: Biomass is a widely available renewable energy source, which could be an alternative fuel source to fossil fuels to mitigate serious environmental problems. Gas-solid fluidized bed reactors offer excellent mixing, and heat and mass transfer between solid particles and a fluidizing gas. As such, a fluidized bed reactor is ideal for thermal/thermochemical conversion of biomass to valuable gas products. Despite many decades of research into improving the operation of biomass gasifiers, the process of biomass gasification is not yet well understood. Biomass gasification performance is strongly affected by operation conditions of the gasifier. Optimizing a reactor, such as a fluidized bed, experimentally is time consuming and expensive. Advancements in high performance computing, has made computer simulation a powerful tool for design engineers and decision makers. In this report, a bench-scale fluidized bed biomass gasifier is simulated using Multiphase Flow with Interphase eXchanges (MFiX) – two-fluid model (TFM). The Optimization Toolset (OT) in the National Energy Technology Laboratory’s (NETL) Nodeworks software is used to optimize the operation of the fluidized bed, such that the H2 to CO ratio in the syngas is 2.0. This optimization is achieved by optimizing the biomass feedstock flow rate, mass flow rate of fluidizing gas, and the amount of steam in the fluidizing gas. The performance of the optimal operating condition—evaluated via a surrogate model—is tested through a final computational fluid dynamics (CFD) simulation, resulting in a syngas composition within the expected accuracy of the model.
Keywords: Computational fluid dynamics, Design Of Experiments, Gaussian Process, Response Surface Model, Two-fluid model
Share
Share
Share
Share