Combustors catalytic reactor

A derivative of the Claus process is the Recycle Selectox process, developed by Parsons and Unocal and Hcensed through UOP. Once-Thm Selectox is suitable for very lean acid gas streams (1—5 mol % hydrogen sulfide), which cannot be effectively processed in a Claus unit. As shown in Figure 9, the process is similar to a standard Claus plant, except that the thermal combustor and waste heat boiler have been replaced with a catalytic reactor. The Selectox catalyst promotes the selective oxidation of hydrogen sulfide to sulfur dioxide, ie, hydrocarbons in the feed are not oxidized. These plants typically employ two Claus catalytic stages downstream of the Selectox reactor, to achieve an overall sulfur recovery of 90—95%. 

The performance of a fluidized bed combustor is strongly influenced by the fluid mechanics and heat transfer in the bed, consideration of which must be part of any attempt to realistically model bed performance. The fluid mechanics and heat transfer in an AFBC must, however, be distinguished from those in fluidized catalytic reactors such as fluidized catalytic crackers (FCCs) because the particle size in an AFBC, typically about 1 mm in diameter, is more than an order of magnitude larger than that utilized in FCC s, typically about 50 ym. The consequences of this difference in particle size is illustrated in Table 1. Particle Reynolds number in an FCC is much smaller than unity so that viscous forces dominate whereas for an AFBC the particle Reynolds number is of order unity and the effect of inertial forces become noticeable. Minimum velocity of fluidization (u ) in an FCC is so low that the bubble-rise velocity exceeds the gas velocity in the dense phase (umf/cmf) over a bed s depth the FCC s operate in the so-called fast bubble regime to be elaborated on later. By contrast- the bubble-rise velocity in an AFBC may be slower or faster than the gas-phase velocity in the emulsion... 

Modeling of catalytic combustors has been the subject of a number of studies. The models used varied in degree of complexity and could therefore answer various types of questions. General issues of modeling monolith catalytic reactors are discussed in Chapter 8 of this book and in the reviews of Irandoust and Andersson [57] and Cybulski and Moulijn [58]. Hence, only topics that are specific to the modeling of catalytic combustion in monolith catalysts are considered here. A description of some important aspects of different types of models are as follows. 

A radial concept is used for the integration of the combustor and catalytic reactor where concentric functional chambers build outward from a central core (Figure 4). The innermost sections comprise the combustor unit, where air is carried to the interior and raffinate gas from the membrane penetrates through the wall to form a combustion flame along the itmer wall of the reaction chamber. Preheated and mixed steam and fuel enter the reaction chamber and are catalytically converted to reformate. Reformate exits the reactor and is then sent to the... 

Ideally. CFB risers operate at a relatively uniform temperature, which is achieved by a high solids recycle rate renewing the inventory of the riser. Catalytic reactions are generally carried out at relatively low temperatures (250-650 C) compared to combustion processes (>800°C). Low-temperature operation permits the use of mechanical devices to control solids mass flux. In combustion processes, the rate is controlled by non-mechanical devices. Fluid mechanics of CFB catalytic reactors and combustors are significantly different, as shown in Table 2. 

Operating Characteristics of Fast Fiuidized Bed Catalytic Reactors and Combustors... 

In fluidized bed reactors, a significant part of the reaction can take place in the freeboard (e.g. some catalytic reactors, fluidized bed combustors). The concentration or hold-up of solids in the freeboard must then be evaluated. 

Gas-phase chemistry has been shown to affect combustion characteristics in catalytic reactors, even in sub-quenching channel confinements and particularly at elevated pressures [5]. To allow for extended parametric studies, gas-phase chemistry has not been considered. In doing so, a conservative estimate of the stable combustion regimes was obtained, since gas-phase chemistry has been shown to extend the stability limits of catalytic combustors [3]. Finally, mixture-average diffusion provided the gas-phase transport model [6]. 

Finally, the currently available 2-D fiiU eUiptic model has provided valuable insight on the fundamental physics of catalytic reactors. The established model could further be used to investigate both the steady-state and transient behavior of a large number of reactor setups, ranging from catalytic microreformers which would produce fuel for micro fuel cells in situ, to advanced combustor solutions for large-scale power generation systems such as the rich catalytic/lean bum concept. 

Somewhat related is a process proposed and demonstrated on labscale by the University of Siegen (Germany). The process is called the (Herhof)-Integrierte Pyrolyse und Verbren-nung (IPV) process and is decribed in detail by Hamel et al.60 In this process, biomass is converted with high-temperature steam to pyrolysis gas in a fixed-bed reactor. The generated carbon from this reactor is led to a stationary FB combustor from which the hot ash is returned to the first-mentioned reactor. The ash works catalytically to reduce the tar content of the gas produced. The gas is further cleaned and conditioned using a scrubber and electrostatic filter from which the catch is returned to the FB combustor. 

An early application of a combined steam reformer/catalytic combustor on the meso scale was realized by Polman et al. [101]. They fabricated a reactor similar to an automotive metallic monolith with channel dimensions in the millimeter range (Figure 2.65). The plates were connected by diffusion bonding and the catalyst was introduced by wash coating. The reactor was operated at temperatures between 550 and 700 °C 99.98% conversion was achieved for the combustion reaction and 97% for the steam reforming side. A volume of < 1.5 dm3 per kW electrical power output of the reformer alone was regarded as feasible at that time, but not yet realized. 

The key interplay of reaction kinetics and transport phenomena in a catalytic combustor must be treated using rigorous reactor models. In the next section, we use a simple model to describe the behavior of a catalytic combustor and to interpret the technology breakthroughs that led to the successful implementation of catalytic combustion to reduce NO in power generation. The model will be kept simple, even though its additional complexities are readily incorporated, because our purpose is to show the main characteristics of a catalytic combustor rather than to provide accurate simulations of expected performance. 

We shall develop next a single-channel model that captures the key features of a catalytic combustor. The catalytic materials are deposited on the walls of a monolithic structure comprising a bundle of identical parallel tubes. The combustor includes a fuel distributor providing a uniform fuel/air composition and temperature over the cross section of the combustor. Natural gas, typically >98% methane, is the fuel of choice for gas turbines. Therefore, we will neglect reactions of minor components and treat the system as a methane combustion reactor. The fuel/air mixture is lean, typically 1/25 molar, which corresponds to an adiabatic temperature rise of about 950°C and to a maximum outlet temperature of 1300°C for typical compressor discharge temperatures ( 350°C). Oxygen is present in large stoichiometric excess and thus only methane mass balances are needed to solve this problem. 

To avoid high-pressure drop and clogging problems in randomly packed micro-structured reactors, multichannel reactors with catalytically active walls were proposed. The main problem is how to deposit a uniform catalyst layer in the microchannels. The thickness and porosity of the catalyst layer should also be enough to guarantee an adequate surface area. It is also possible to use methods of in situ growth of an oxide layer (e.g., by anodic oxidation of a metal substrate [169]) to form a washcoat of sufficient thickness to deposit an active component (metal particles). Suzuki et al. [170] have used this method to prepare Pt supported on nanoporous alumina obtained by anodic oxidation and integrate it into a microcatalytic combustor. Zeolite-coated microchannel reactors could be also prepared and they demonstrate higher productivity per mass of catalyst than conventional packed beds [171]. Also, a MSR where the microchannels are coated by a carbon layer, could be prepared [172]. 

Under normal operating conditions, in which the combustor is sufficiently warm and operated under fuel rich conditions, virtually no NOx is formed, although the formation of ammonia is possible. Most hydrocarbons are converted to carbon dioxide (or methane if the reaction is incomplete) however, trace levels of hydrocarbons can pass through the fuel processor and fuel cell. The shift reactors and the preferential oxidation (PrOx) reactor reduce CO in the product gas, with further reduction in the fuel cell. Thus, of the criteria pollutants (NOx, CO, and non-methane hydrocarbons [NMHC]), NOx CO levels are generally well below the most aggressive standards. NMOG concentrations, however, can exceed emission goals if these are not efficiently eliminated in the catalytic burner. 

---