(684d) Bridging the Gaps: From Particle-Resolved to Multi-Tubular Reactor to Process Simulation

Packed bed reactors are widely used, amongst others, for highly exothermic or endothermic catalytic reactions like reformation of hydrocarbons. Rigorous simulations can help to achieve improvements of the performance of these reactors with lower upfront investment. Recent advances in modeling particle-resolved packed beds allows detailed insight in the flow, species and temperature distribution in the beds. It also enables the calculation of reactions on the catalytic surface and pores, forming a basis to optimize the reactor on a micro-scale, e.g. develop new particle shapes. Although computational power has increased substantially in the last decades, a particle resolved multi-tubular reactor CFD simulation is too compute intensive for tuning of reactors on a daily basis. However, the output from these models can be used to create more accurate 1D or 2D models. These 1D models can then be embedded in multi-tubular reactor models. Then, in a last step these multi-tubular reactor models can be integrated into process models.

Methodology

In this contribution we will start with introducing a fully automated workflow to simulated particle-resolved packed beds in Simcenter STAR-CCM+ based on a coupled CFD-DEM approach. Based on these detailed CFD simulations effective transport parameters like dispersion and heat transfer coefficient are derived. Compared to experimental investigation this is a quicker, but still accurate and cost-effective way to determine such parameters. These derived parameters are used in a coupled simulation of the whole multi-tubular reactor by coupling CFD with the process simulation tool gPROMS. Finally this is then be used in a flow-sheet process simulation.

Results

We show based on several examples the potential of this method. Furthermore, we show based on a coupled optimization study of a multi-tubular reactor how the performance of the reactor can be improved and that the cost of such a steady is relatively low.