Exploring the Validity of TFM in Simulating Gas-Solid Heterogeneous Reactors for Slow Reactions

Hello everyone!

I have a query regarding the application of Two-Fluid Model (TFM) in simulating gas-solid heterogeneous reactors. Typically, in TFM simulations, the particle size is considered constant, which theoretically seems to undermine its suitability for modeling gas-solid reactions where particle size changes due to reaction should play a crucial role.

However, for slow reactions, it’s observed that TFM can approach a quasi-steady state over relatively short time scales. Within this timeframe, the consumption of particle size due to chemical reactions might be so minimal that the change in particle size becomes negligible. Does this imply that for academic research purposes, particularly when focusing on slow reactions, TFM could still be a valid approach for simulating gas-solid heterogeneous reactors?

Looking forward to hearing your thoughts and insights on this matter.

Best regards.

Hello,
It is a rather reasonable assumption when simulating a short period of flow or reactions. In fact, due to the limitation of computational cost, many literature perform simulations for only several to tens of seconds, under which the particle size could generally be seen unchanged. What’s more, in one investigation, the change of particle size with flow time is usually not the focus of study, and maybe it’s not necessary to consider the complex problem at all.

Hi, Zhiyong,

Thank you for your insightful reply. Indeed, I agree with your perspective. However, from the viewpoint of numerical computation, it does seem that only the gas-phase components are being consumed or produced in heterogeneous reactions, while the particulate phase remains unchanged, raising concerns about mass conservation in these simulations.

If this simulation approach is considered feasible, could it imply that any apparent lack of mass conservation due to the unchanged mass of the particulate phase is actually minor compared to the numerical errors inherent in the computational models?