Mass balances, catalyst deactivation

The approach of this work is to measure product compositions and mass balances in much detail in a time resolved manner and to relate this to the controlling kinetic principles and elemental reactions of product formation and catalyst deactivation. Additionally the organic matter, which is entrapped in the zeolite or deposited on it, is determined. The investigation covers a wide temperature range (250 - 500 °C). Four kinetic regimes are discriminated autocatalysis, retardation, reanimation and deactivation. A comprehensive picture of methanol conversion on HZSM5 as a time on stream and temperature function is developed. This also explains consistently individual findings reported in literature [1 4]. 

The packed bed reactors section of this volume presents topics of catalyst deactivation and radial flow reactors, along with numerical techniques for solving the differential mass and energy balances in packed bed reactors. The advantages and limitations of various models (e.g., pseudo-homogeneous vs. heterogeneous) used to describe packed bed reactors are also presented in this section. 

The general approach for modelling catalyst deactivation is schematically organised in Figure 2. The central part are the mass balances of reactants, intermediates, and metal deposits. In these mass balances, coefficients are present to describe reaction kinetics (reaction rate constant), mass transfer (diffusion coefficient), and catalyst porous texture (accessible porosity and effective transport properties). The mass balances together with the initial and boundary conditions define the catalyst deactivation model. The boundary conditions are determined by the axial position in the reactor. Simulations result in metal deposition profiles in catalyst pellets and catalyst life-time predictions. 

This type of solution method is possible for reactions where deactivation is slow, and a pseudo steady-state assumption can be made when solving the mass balance equations. Thus, these equations are applicable to reactions where the activity loss is first-order in both the poison and the active sites, and where deactivation is slow compared to the main reaction. A similar type of approach was taken by Johnson et al. (5), for oxygen consumption and carbon content during catalyst regeneration and by Bohart and Adams (6), for chlorine consumption and absorbence capacity of charcoal. 

Since the n-heptane reforming rate is much faster than catalyst deactivation rate, the reaction system considered, with appropriate assumptions, is represented by the following quasi-steady state mass balance equations ... 

We now use the mass balance discrepancy to subtract the contribution of any "accumulation-reaction" processes present from the measured responses. For example, the corrected response curve for CO, shown in Fig. 12, was obtained at each instant in time by adding to the measured outlet CO concentration the amount of CO that was converted to CO2 by reaction with oxygen atoms stored in the catalyst, as determined from the oxygen balance. The fact that the corrected CO response and the corrected CO2 response, shown in Fig. 13, do not match the instantaneous response demonstrates the action of an "activity change" type of transient chemical process. The process resulted in lower-than-expected CO conversion, since the activity change process included in the simulation was partial deactivation of the catalyst in lean exhaust (e.g., by oxidation of the precious metal). 

The data in Fig. 5 also show that coke formation occurs rapidly at first and becomes increasingly slow as coke piles up. This behavior reflects the deactivating effect of coke on the coking reaction. Since the TEOM microbalance maintains temperature and the pressure time invariant, the mass balance equation for coke-on-catalyst can be described as [17] dC. 

The concentration of ArOQ depends on the amounts of ArOK, KX, and KI, and the initial usage of the catalyst QX. Combining the mass balance equation for initial RX in the organic phase with Eq. (134), a deactivation function cj) can be introduced in the situation under decline of catalytic efficiency, leading to the following equations ... 

For a plug-flow reactor in which the catalyst undergoes either uniform or shell-progressive deactivation, an additional mass balance for the species causing deactivation is required along with Eqs. 10.16 through 10.18 ... 

Process simulations with time-varying catalyst activity were performed based on a quasi-steady-state approximation (Lababidi et al., 1998). The underlying principle is that because catalyst aging is a relatively slow process in the operating cycle timescale, it can be assumed that the process is stable during short periods of time. In this case, it is considered that this time period is equal to the duration of the mass-balance runs during the catalyst stability tests (12 h). The simulation runs start at t=0 with the catalyst in its fresh state = 1.0 for the entire catalyst length). The concentration and temperature profiles are established from the steady-state solution of the heat and mass balances, as described previously. The next step is to estimate the local amount of MOC from the axial metal profiles in this period and after that to evaluate the deactivation functions for each reaction. The time step is increased and all the calculations are repeated. 

Chapter 8 is dedicated to the modeling of heavy oil upgrading via hydroprocessing. Experimental studies for generation of kinetic data, catalyst deactivation, and long-term stability test are explained. Mass and heat balance equations are provided for the reactor modeling for steady-state and dynamic behavior. Simulations of bench-scale reactor and commercial reactor for different situations are also reported. 

A particular issue is the deactivation of methanation catalysts by carbon formation. Kuijpers et al. [345] observed significant carbon formation over a nickel/kieselgur catalyst containing 54wt.% nickel when exposed to a mixture of 10 vol.% carbon monoxide, 15 wt.% hydrogen, with a balance of nitrogen at 0.6 bar pressure and a 250 °C reaction temperature. Carbon filaments were found, which contributed to 10 wt.% of the catalyst mass at the inlet of the fixed bed. A nickel/silica catalyst showed practically no coke formation for 1000 h duration under the same conditions. 

---