Catalyst deactivation, startup

An alternate startup procedure was also tested. Since CO2 at 80 bar and 95° C is less dense than the final reaction mixture, there was a concern of some initial catalyst deactivation during this startup period. To address this possibility, instead of initially charging the reactor with only carbon dioxide, a 70/30 mixture of C02/isobutane was added, followed by establishing the final temperature and pressure before adding the olefin. The results showed... 

For the most part we have labored over the analysis of steady-state problems, although there have been some important side trips into the unsteady state. Principal among these were the analysis of CSTR startup, visits to fixed-bed and CSTR dynamics arising from catalyst deactivation, and some discussion on adsorption variations. The purpose of this chapter is to pursue some of these topics in more detail the range of interests here is rather broad, but all can be linked through a common concern with fixed-bed dynamics. 

This is an example, taken from Dumez and Froment [1976], combining reactor simulation on the basis of a slightly simplified version of the model considered in this section but accounting for transients resulting from startup and from catalyst deactivation due to coke deposition. 

Ammonia feed is shut off by a low temperature switch when the operating flue gas temperature drops below the minimum recommended value. This prevents deactivation of the catalyst from ammonium bisulfate deposition. This control feature is also applied during system startup and shutdown. An economizer bypass is used to maintain the flue gas temperature above the minimum recommended SCR operating temperature and an SCR bypass is shown (though not often provided). The SCR bypass is used to protect the SCR catalyst during startup and shutdown when the flue gas temperature can be below its dew point. Economizer bypasses are used at the Chambers, Indiantown, and Keystone power plants (Franklin, 1993). 

The solution of the above system of partial differential equations (eqs 4-12) yields the concentration and temperature profiles inside the catalyst pellet, and if necessary across the external boundary layer, as a function of time. However, there are only few cases of practical importance where this complete solution is required, as for instance startup and shutdown periods, dynamic process control options such as the so-called Matros concept with flow reversals (for redox processes), or situations where the catalyst is rapidly deactivated. 

In this chapter, modeling of monolith reactors will be considered from a first-principles point of view, preceded by a discussion of the typical phenomena in monoliths that should be taken into account. General model equations will be presented and subsequently simplified, depending on the subprocesses that should be described by a model. A main lead will be the time scales at which these subprocesses occur. If they are all small, the process operates in the steady state, and all time-dependent behavior can be discarded. Unsteady-state behavior is to be considered if the model should include the time scale of reactor startup or if deactivation of the catalyst versus time-on-stream has to be addressed. A description of fully dynamic reactor operation, as met when cycling of the feed is applied, requires that all elementary steps of a kinetic model with their corresponding time scales are incorporated in the reactor model. 

Liu, X., Ruettinger, W., Xu, X., and Farrauto, R. Deactivation of Pt/Ce02 water-gas shift catalysts due to shutdown/startup modes for fuel cell applications. Applied Catalysis. B, Environmental,... 

Mechanical deactivation is due to strong stresses of packed catalyst beds during startups, shut-downs, and catalyst regeneration. 

---