Numerical investigation and comparison of coarse grain CFD - DEM and TFM in the case of a 1 MWth fluidized bed carbonator simulation

- CFD-DEM and Euler-Euler comparison for CFB simulation.
- EMMS drag model increases the accuracy of both models.
- Cell length to parcel diameter ratio of less than 3 reproduced results of adequate accuracy.
- Parcel diameter to particle diameter of below 60 reproduced good results for the pressure profile.

This work focuses on a comparison between the Euler-Euler Two Fluid Model (TFM) approach and the coupled coarse grain discrete element CFD-DEM numerical model for the simulation of a 1 MWth CFB carbonator reactor located at TU Darmstadt (TUD). The effect of the drag force formulation and its associated application in the numerical model for both approaches in terms of their numerical accuracy, compared to experimental data is investigated, by implementing either the conventional Gidaspow model or the advanced EMMS one. Moreover, for the coarse grain CFD - DEM model, the range of values for important numerical parameters as the particle per parcel and cell to parcel size ratios are investigated to shed light on the necessary resolution such a model should have in order to reproduce valid and not parameter dependent numerical results. An adequate cell length to parcel diameter ratio is found to be around 2.6 while as concerns the parcel to particle diameter ratio a value around 58.5 proved to be sufficient, at least for the range of parameters investigated in this paper (size of riser, flow rates and particles average diameter). The EMMS model improved the accuracy of results derived by the coarse grain CFD - DEM model, while further research on the appropriate drag models for the coarse grain CFD - DEM is a sine qua non for its successful implementation in similar studies. For instance it is of interest to answer whether the individual particles slip velocity instead of the particles cell averaged slip, should be used for the calculation of the momentum interexchange coefficient (β) as well as the treatment of different particle diameters in the EMMS equation scheme.

112