Fig. 3.11. Experimental conversion as a function of feed position (symbols). The line shows the model prediction for a reaction rate constant of 12%

loading compared with the lab-scale column. The concentrations in the reactive sections were likewise similar in both cases. Nonetheless, the simulation showed that the model was able to describe observed concentration and temperature profiles of the laboratory column very well but failed in the simulation of the pilot column.

A similar observation was made by other authors in the investigation of the synthesis of methyl acetate. Whereas a 50 mm laboratory column could be described very well [16] with the equilibrium stage model, this held no longer in the case of a pilot column [17].

It is known that the geometry of the reactive packing types for lab- and pilot-scale applications differs. This is accounted for in the present model by using the appropriate values for NTSM (Number of Theoretical Stages per Meter) and catalyst mass per stage. A more detailed description of the hydrodynamics has hitherto not been included and remains necessary for future work. Possible explanations for the lower conversion compared to the lab column may be incomplete catalyst wetting due to maldistribution effects, mass-transfer phenomena, or the formation of temperature gradients in the reactive packing. As long as this puzzle remains unsolved, a reliable scale-up of RD processes is not possible without expensive pilot testing.

0 0

Post a comment