(419h) CFD Modeling of Binary Mixing of Particles in Fluidized Bubbling Bed Reactors

Biomass processing in fluidized bubbling bed is getting attraction for renewable liquid fuel productions in recent years [1,2]. However, there are still many difficulties and uncertainties regarding the process yield, stability and their scale-up. Instead of biomass pyrolysis or gasification, we took a different route by synthesizing liquid hydrocarbons directly. In this presentation, we will discuss how some of these issues were addressed using experimental data and computational fluid dynamics (CFD) modeling, especially regarding how the biomass and catalyst mixes in a bubbling bed with hydrogen as the fluidization gas.

The binary mixing process is designed to extract oxygenate vapor from biomass (non-food organic waste) and react it with hydrogen in a bubbling catalyst bed. The process requires proper dispersion of biomass with catalyst particles, in which it generates heat through deoxygenation reaction and produces hydrocarbon vapor under high temperature and high pressure. The yield depends critically on the degree of mixing, and the efficient generation of biomass oxygenate vapor at the early stage. If not properly designed, it can significantly impact reactor performance in many ways. The use of CFD enabled us to predict and identify possible issues early on. Interactions with lab experiments and the demonstration reactive unit further enhance our understanding of the interplays between experiments and CFD models. Additionally, the details in CFD allow the quantification of many scale-up rules, including solid flux, binary mixing and catalyst loss, etc.

However, here the discussion will be limited to the CFD modeling and quantification of the binary mixing process and how it impacts the overall reactor design. The CFD model used is based on multiphase-particle-in-cell (MP-PIC) approach [3] due to its efficiency in modeling particle-particle interactions and a commercial code was adopted [4]. The model was first tested against published work for binary mixing of particle and later was further fine-tuned to mimic the particle segregation time-scale measured in experiments. This establishes the fundamental mixing correlations between the particles of interest. Due to the computational costs of chemical reaction calculations, a simplified thermal mixing method was used to mimic the heat produced from the exothermic chemical reaction by assuming a high catalyst heat capacity that retains heat. The room temperature fed biomass particles provide the cold quenching experienced by the catalysts. The biomass particles can track the temperature values as they travel through the reactor and generate a temperature mixing history to validate required particle heating performance. When used with a demonstration unit data, it can show the necessary performance curve needed for the scale-up reactor to follow.

The experience we obtained from the interactions between CFD modeling and experiments are invaluable and provided many practical and useful lessons regarding how both can gain benefits from each other, especially clear and open communications and well defined processes are critical in generating positive feedbacks and creating productive results.

References:

[1] Troiano, et al., “Fluidized bed pyrolysis of biomass: A model-based assessment of the relevance of heterogeneous secondary reactions and char loading,” Energy Fuels, 36, 17, 9660-9671, 2022.

[2] Beheshti, et al., “Process simulation of biomass gasification in a bubbling fluidized bed reactor,” Energy Conversions and management, 94, 345-352, 2015.

[3] Peter J. O'Rourke and Dale M. Snider, "A newblended accelerationmodel for the particle contact forces induced by an interstitial fluid in dense particle/fluid flows," Power Technology, 256, 39-51, 2014.

[4] Barracuda Virtual Reactor 23.0.1, CPFD-Software Inc., 2023.