Catalyst poisoning and fixed bed reactor dynamics

Abstract

Abstract-The poisoning kinetics of thiophene on Ni-kieselguhr catalysts and the deactivation behavior of nonisothermal fixed bed reactors have been studied experimentally using benzene hydrogenation as a model exothermic reaction. The time dependent axial temperature profiles in the reactors were measured and compared with values evaluated from a dispersion model , the parameters of which have been determined in separate experimentation.Poisoning kinetics were measured in a series of differential reactor experiments at atmospheric total pressure, thiophene partial pressures of 0·037-0·19 torr, hydrogen to benzene molar ratios >8/1 and temperatures from 6O--180·C. Excellent agreement was found with a power law equation for the rate of change of activity with time , fir st order in catalyst acti vity and in thiophene concentration , with an experimental activation energy of 1080 kcal/kmole. This correlation of poisoning kinetics, however, was not able to predict the propagation of the zone of activity (hot-spot) on poi soning of an integral fixed bed reactor. Initial (steady state) temperature profiles were modeled satisfactorally , but the rate of migration of the hot spot was found experimentally to be more rapid than that predicted from the correlation of poisoning kinetics. A ·semi-empirical two site deactivation model is shown to resolve the discrepancy.

Kinetic, transport, and deactivation rate interactions on steady state and transient responses in heterogeneous catalysis

“…(1975) This experiment and Price and Butt ( 1977) 1 tions, it is assumed that hydrogen is present in essentially constant concentration, that the contribution of the poisoning to the energy balance is negligible in comparison to the heat of reaction, and that the poisoning kinetics can be represented as separable. By the latter it is meant that the reaction rate at any time is given by the product of the reaction rate on undeactivated catalyst times the activity variable s determined by the poisoning kinetics, The forms of the rate equations for hydrogenation and poisoning together with the associated kinetic and capacity parameters have been determined in separate experimentation and are discussed elsewhere by Kehoe and Butt (1972b) and Weng et al ( 1975). The kinetic correlations given here have been developed for a rather narrow range of variables characteristic of the low benzene and high hydrogen concentrations employed in the experiments and are not intended to represent a general correlation.…”

Progress toward the a priori determination of catalytic properties

“…In fact, this type of nonselective poisoning is probably fairly rare, as the data of Figures 3 and 4 would indicate and as we have had occasion to find out in this laboratory for sulfur poisoning of nickel catalysts (Weng et al, 1975). For sintering of supported metals, the rate of loss of active surface appears in most cases to be well correlated by dA dt…”

“…with a and n constants and C, the weight of coke on the catalyst, This has been shown to work on anything from A (alumina) to Z (zeolites); two examples are shown in Figure 20 (Voorhies, 1945;Butt et al, 1975). Note, however, that the correlation is for weight of coke on catalyst, not catalyst activity, so the question of the nature of the relationship between the two again arises.…”

Improvements in acetone adsorption efficiency

“…As pointed out b y Finlayson, two major drawbacks to FDM solutions of typicel convection-diffusion equations are 1) the numerically generated oscillations that can occur for large Peclet numbers, and 2) the inaccuracies that can occur at sharp gradients over inadequate meshes. Two alternative ways to avoid these problems without an objectionably large number of spatial grid points are 1) to use higher-order differencing approximations[35], or 2) to use an adaptive (variable) mesh [7,15,25,36, 371.Thomson [38] recommends (in agreement with others[39,40]) that usin adative meshes with standard solution methods is pre 8' erah P e to higher-order-accurate schemes with uniform meshes.…”