Monolithic catalysts heat transfer coefficient

The relative methane conversions for the entire series of tests are shown in Fig. 12. From these data the heat transfer coefficients that were calculated (Table 3) show a significant improvement of the monolith bed over the packed bed. It is apparent that a combination of (1) the endothermic reaction rate and (2) heat transfer via conduction and convection are balanced over these flow rates in this monolith bed. However, based on the unfavorable shift in the extent of methane conversion from the monolith to the packed bed as flows increase, shown in Fig. 12, it appears that the mass transfer of reactants to the catalyst surface in this monolith bed may be rate-limiting at flow rates at and above 0.50 L/sec, probably a function of the high void fraction of this bed design. 

Several runs were also made using monoliths that had reaction pass walls coated with copper chrome catalyst. The catalyst coated monoliths have much higher overall heat transfer coefficients, so that there is equalization of the coolant and reactant temperatures in three out of four monoliths. 

Figure 4.9 Overall heat transfer coefficient (a) and heat transfer parameters (b) of a highly conductive structured catalyst in methanol synthesis as a function of the syngas stoichiometric number in the fresh feed stream (Mp). (Squares) Commercial Lurgi multitubular packed-bed reactor (PB) (circles) copper honeycomb monoliths (HM) (triangles) open-cell foams (OF). In Figure 4.9b, the radial effective thermal conductivity is plotted with solid symbols and the wall heat transfer coefficient, h, with empty ones. Reprinted from Montebelli etal. [162], with permission from Elsevier.

Transfer coefficients in catalytic monolith for automotive applications typically exhibit a maximum at the channel inlet and then decrease relatively fast (within the length of several millimeters) to the limit values for fully developed concentration and temperature profiles in laminar flow. Proper heat and mass transfer coefficients are important for correct prediction of cold-start behavior and catalyst light-off. The basic issue is to obtain accurate asymptotic Nu and Sh numbers for particular shape of the channel and washcoat layer (Hayes et al., 2004 Ramanathan et al., 2003). Even if different correlations provide different kc and profiles at the inlet region of the monolith, these differences usually have minor influence on the computed outlet values of concentrations and temperature under typical operating conditions. 

It should be emphasized that Oh and Cavendish assumed that the reactions only occur on the surface of the channel wall. This assumption is less realistic for a layer of washcoat (typically y-alumina) dispersed with catalyst applied on to the wall surface. Ramanathan, Balakotaiah, and West showed that the diffusion in the washcoat has a profound influence on the light-off behavior of a monolith converter. They derived an analytical light-off criterion based on a onedimensional two-phase model with position-dependent heat and mass transfer coefficients. The derivation of this criterion is based on the two key assumptions a positive exponential approximation (i.e., the Frank-Kameneskii approximation) and negligible reactant consumption in the fluid phase. The light-off is defined as the occurrence of multiple steady states with the attainment of the ignited steady state. Here, we discuss only the results of their analysis, without going into the details of their derivation. 

In a conventional monolith the contact between the coated walls and the reactants is carried out by means of diffusion. Therefore, in order to achieve a high conversion during the catalytic reaction, high geometrical area of the coated catalysts is required. If the flow of reactants presents a radial component, mass transport to the walls may be enhanced. Pressure drop compared with conventional systems at the same density will increase, being compensated by the increasing of heat and mass transfer coefficients. 

---