Simulation of Heterogeneously-Catalyzed Non-Isothermal Reactive Flow in Industrial Packed Beds

Abstract

Packed bed reactors are the most frequently used devices to perform heterogeneously catalyzed reactions on industrial scales. An industrial real-life heterogeneous catalysis is complex process that combines fully three-dimensional mass, momentum and energy transport on several scales. In the present work, we leverage our previously developed CFD solver for non-isothermal heterogeneously catalyzed reactive flow based on the finite volume method and couple it with our in-house DEM-based method for preparation of random packed beds. The resulting framework is verified in the simplified cases against available analytical solutions and correlations and is used to study an industrially-relevant case of ethylene oxychlorination performed in a tubular packed bed comprising CuCl₂-coated catalyst carrying particles. In particular, we compare properties of three different industrially used catalyst carrying particles: Raschig rings, Reformax, and Wagon wheels.

On Diffusion, Reaction, and Flow of High-Concentration Multicomponent Gas Mixtures

“…For the simplified case of a mass transfer in a Stefan tube, the new solver provides results in a perfect agreement with both the experiment and the available analytical solution. The proposed solver extension allows one to properly describe and optimize complex industrial processes, such as ethylene oxichlorination which is descibed in detail in (Hlavatý et al, 2023a) and illustrated in Fig. 2.…”

“…The presented model stems from our previously published OpenFOAM solver for the heterogeneously catalyzed reactive flow for the simulations of the catalytic filters in the automotive exhaust gas aftertreatment (Hlavatý et al, 2023b). Non-isothermal reactive flow of an ideal gas in the heterogeneous domain Ω is assumed, while the heterogeneity of Ω is ensured by dividing Ω in free space (Ω fs ) and porous media (Ω pm ), such that Ω = Ω fs ∪ Ω pm and Ω fs ∩ Ω pm = ∅, and by using (i) effective diffusion coefficients, and (ii) piece-wise continuous right hand side of governing equations, see (Hlavatý et al, 2022(Hlavatý et al, , 2023a.…”

“…where µ is the dynamic viscosity of the gas, λ eff is the effective heat conductivity, and c p is the heat capacity. For more information on parameters' computations see (Hlavatý et al, 2023a). The source term ρs f describes the resistivity of the flow through the porous medium (Ω pm ) and is defined based on the Darcy permeability model (Kočí et al, 2019).…”

“…In general, it is difficult to determine the Stefan-Maxwell binary diffusion coefficients D i,j . However, considering the ideal gas and heterogenous domain, Ω = Ω fs ∪ Ω pm , we can assume that the Stefan-Maxwell binary diffusion coefficients are equal to the effective Fick binary diffusion coefficients, D i,j = D eff i,j , whose calculation is taken from (Hlavatý et al, 2023a).…”