Axial Dispersion in Nonisothermal Packed Bed Chemical Reactors

Abstract

A criterion is developed to predict when axial dispersion is important in nonisothermal packed-bed reactors with cooling or heating at the walls. In contrast to the isothermal problem, the criterion does not depend on the length of the reactor, so that the importance of axial dispersion cannot be minimized by increasing the length of the reactor. An increase in flow rate does decrease the importance of axial dispersion. The criterion is applied to the experimental data presented by Schuler, et a/. (1 954)) for SO2 oxidation on an alumina catalyst impregnated with platinum, and the criterion suggests that axial dispersion is important. The experimental data apparently cannot be reconciled with a model excluding axial dispersion, but a model including both axial and radial dispersion correctly predicts the data. Mathematical models of chemical reactors are useful for predicting the conversion and temperature profiles in packed bed reactors. TT'hile very general models can be written donx, these are not often used either because of the computational complexity or because it, is difficult' or impossible to estimate the parameters in the model. We are concerned here with a cylindrical tube, which is cooled or heated at the walls, and which is packed with catalyst. We wish to determine when axial dispersion is important in such a reactor and relate the results to experimental data.