(433b) Use of Computational Fluid Dynamics to Identify Transport-Derived Errors in Thermal Analysis Testing

The production of catalysts and other inorganic materials includes many energy-intensive thermal processing steps. To design efficient, optimised calcination and drying processes for new functional materials, a model-based approach could be used. This requires kinetic information, extracted from thermal analysis data. Although thermal analysis is a well-developed field, reaction engineering principles, such as reactor transport characterisation, are not widely applied.

It has been observed that the success of kinetic modelling with the modified Sestak-Berggren equation varies with different equipment configurations. Model criticism implies that source data could be affected by non-kinetic phenomena, such as transport limitations, or by a reverse reaction. To investigate possible transport phenomena occurring within thermal analysis reactors, a computational fluid dynamics (CFD) study has been carried out.

Transient tracer simulations (at a fixed temperature) were completed, along with a full temperature ramp simulation. The results from these were used to create residence time distribution curves for both the sample and the reactor (shown in Figure 1). Dimensionless analysis, in the form of Bodenstein number and Damköhler numbers (for the sample and reactor) have been calculated. These dimensionless numbers indicate the flow regime present, and if the system is mass transport limited.

This methodology has been applied to various configurations of thermal analysis reactor, which can be classified into pan-style or tubular reactors.

Dimensionless analysis indicates the pan style reactors are mass transport limited. The full temperature ramp CFD simulation indicated these pan-style reactors are also heat transport limited. Hence extracting kinetic data from these reactors is unlikely.

Although the tubular reactors are mass transport limited, the Bodenstein number indicates plug flow. This behaviour will likely manifest as a time delay, which could be accounted for during modelling. It is hoped that depending on experimental conditions, these tubular reactors could be capable of obtaining kinetically limited data.