Experimental Studies of Heat and Mass Transfer in Fixed Beds - 1

ixed bed catalytic reactors are the most-used reactor type for large-scale heterogeneously catalyzed gas-phase reactions. In particular, multitubular fixed beds with low tube-to-particle diameter ratio (N) are used for strongly exothermic reactions such as partial oxidations and selective hydrogenations as well as strongly endothermic reactions such as steam reforming of methane. In these processes heat must be rapidly transferred into or out of a narrow reactor tube, while the need to reduce compressor costs dictates a low pressure drop along the tube, and thus the particle size cannot too small. These constraints combine to give tubes with low values of N, in which the presence of the tube wall has a strong influence on heat transfer which in turn affects reaction rates and selectivity. In such cases it is important to accurately describe the factors limiting the rate of heat transfer at the wall [Derkx, O. and Dixon, A.G., 1996; Dixon, A.G., 1996; Derkx, O. and Dixon, A.G., 1997].

The rates of heat and mass transfer from the wall of a fixed bed to a fluid flowing through it are necessary for equipment design. The wall viscous boundary layer and the change in packing structure next to the containing surface give rise to extra transport resistances, which are usually represented as being localized at the surface. The extra resistance to heat transfer at the wall of a catalytic fixed bed reactor, in particular, has received much attention. The extra resistance to mass transfer near the wall is also frequently studied, to give insight into the analogous heat transfer resistance in the fluid phase [Dixon, A.G., DiCostanzo, M.A. and Soucy, B.A., 1984; Dixon, A.G. and LaBua, L., 1985].

From our earlier work, the wall-to-fluid mass-transfer coefficient (dimensionless wall-to-fluid Sherwood number, Sh_wf) was relatively well-correlated for Re > 100 and for most N values greater than about 3. The two main methods of conducting the studies, dissolution of a coating and the limiting current technique, were thought to both give Sh_wf and were in good agreement. What was not known was the behavior of such low-N systems at low to moderate Re. We found [Dixon, A.G., Arias, J. and Willey, J., 2003] that the electrochemical method in fact gave an overall Sherwood number Sh_ov that included mass transfer resistance in the bed center. The dissolution method gave the near-wall coefficient Sh_wf. This caused the two dimensionless groups to behave differently as Re tended to zero. The main results are shown in the two figures on this page, click if you want more details.