(286c) Cfd And Pseudo-Continuum Modeling Of A Low-N Packed Bed With Endothermic Reactions

Multitubular packed bed reactors with low tube-to-particle diameter ratios (N) are frequently selected for strongly endothermic reactions such as steam reforming and propane dehydrogenation. For low-N tubes, the presence of the wall causes changes in bed structure, flow patterns, transport rates and the amount of catalyst per unit volume. In particular, the particles close to the wall will behave differently to those inside the bed. The problem is that, due to the simplifying assumptions, such as uniform catalyst pellet surroundings, that are usual for the current heterogeneous reactor models, the effects of catalyst pellet design changes in the near-wall environment are lost. The challenge is to develop a better understanding of the interactions between flow patterns, species pellet diffusion, and the changes in catalyst activity due to the temperature fields in the near wall region for the modeling and design of these systems. To contribute to this improved understanding, a 3D Discrete Packing Model (DPM) was generated and analyzed with Computational Fluid Dynamics (CFD) to obtain detailed flow, temperature, and species fields. A 120º segment of an N=4 tube model with cylindrical particles, was studied under typical industrial conditions using the commercial CFD code FLUENT. In order to focus on the 3D intra-pellet distributions of temperature and species, diffusion and reaction were coupled to the external flow and temperature fields by user-defined code. The radial temperature and species distributions were obtained throughout the model. Additionally, a 3D pseudo-continuum model was considered with conventional simplifying assumptions, such as the utilization of correlation-based effective transport parameters, unidirectional velocity field, and reaction source terms using effectiveness factors. As typical examples of simplified unidirectional velocity fields, constant axial velocity and radial profiles of axial velocity were utilized. These profiles were obtained either from correlations or from the DPM. Effectiveness factors were also calculated from the DPM. The comparison of DPM and pseudo-continuum models was carried out for these endothermic reactions by considering the obtained particle heat uptakes, and temperature and species profiles. Different pseudo-continuum model settings with both correlation-based and DPM-based parameters were included. As a result of this comparison, the significance of utilizing the proper flow field and transport parameters was quantified for pseudo-continuum modeling approaches.