Products reactor model

The first commercial fluidized bed polyeth)4eue plant was constructed by Union Carbide in 1968. Modern units operate at 100°C and 32 MPa (300 psig). The bed is fluidized with ethylene at about 0.5 m/s and probably operates near the turbulent fluidization regime. The excellent mixing provided by the fluidized bed is necessary to prevent hot spots, since the unit is operated near the melting point of the product. A model of the reactor (Fig. 17-25) that coupes Iduetics to the hydrodynamics was given by Choi and Ray, Chem. Eng. ScL, 40, 2261, 1985. 

Since it is a conceptual study employing a theoretical reactor model, it is also important to appreciate the limits of this type of investigation. The advantage of the computer investigation over a pilot or production reactor investigation is the obvious cost and time saving over the real reactor experiment. 

The selectivity is 100% in this simple example, but do not believe it. Many things happen at 625°C, and the actual effluent contains substantial amounts of carbon dioxide, benzene, toluene, methane, and ethylene in addition to styrene, ethylbenzene, and hydrogen. It contains small but troublesome amounts of diethyl benzene, divinyl benzene, and phenyl acetylene. The actual selectivity is about 90%. A good kinetic model would account for aU the important by-products and would even reflect the age of the catalyst. A good reactor model would, at a minimum, include the temperature change due to reaction. 

All these steps can influence the overall reaction rate. The reactor models of Chapter 9 are used to predict the bulk, gas-phase concentrations of reactants and products at point (r, z) in the reactor. They directly model only Steps 1 and 9, and the effects of Steps 2 through 8 are lumped into the pseudohomoge-neous rate expression, a, b,. ..), where a,b,. .. are the bulk, gas-phase concentrations. The overall reaction mechanism is complex, and the rate expression is necessarily empirical. Heterogeneous catalysis remains an experimental science. The techniques of this chapter are useful to interpret experimental results. Their predictive value is limited. 

Having set up a model to describe the dynamics of the system, a very important first step is to compare the numerical solution of the model with any experimental results or observations. In the first stages, this comparison might be simply a check on the qualitative behaviour of a reactor model as compared to experiment. Such questions might be answered as Does the model confirm the experimentally found observations that product selectivity increases with temperature and that increasing flow rate decreases the reaction conversion ... 

In Table 17.2, fA (for the reaction A products) is compared for each of the three flow reactor models PFR, LFR, and CSTR. The reaction is assumed to take place at constant density and temperature. Four values of reaction order are given in the first column n = 0,1/2,1, and 2 ( normal kinetics). For each value of n, there are six values of the dimensionless reaction number MAn = 0, 0.5, 1, 2, 4, and °°, where MAn = equation 4.3-4. The fractional conversion fA is a function only of MAn, and values are given for three models in the last three columns. The values for a PFR are also valid for a BR for the conditions stated, with reaction time t = t and no down-time (a = 0), as described in Section 17.1.2. 

In this chapter, we develop some guidelines regarding choice of reactor and operating conditions for reaction networks of the types introduced in Chapter 5. These involve features of reversible, parallel, and series reactions. We first consider these features separately in turn, and then in some combinations. The necessary aspects of reaction kinetics for these systems are developed in Chapter 5, together with stoichiometric analysis and variables, such as yield and fractional yield or selectivity, describing product distribution. We continue to consider only ideal reactor models and homogeneous or pseudohomogeneous systems. 

The TIS and DPF models, introduced in Chapter 19 to describe the residence time distribution (RTD) for nonideal flow, can be adapted as reactor models, once the single parameters of the models, N and Pe, (or DL), respectively, are known. As such, these are macromixing models and are unable to account for nonideal mixing behavior at the microscopic level. For example, the TIS model is based on the assumption that complete backmixing occurs within each tank. If this is not the case, as, perhaps, in a polymerization reaction that produces a viscous product, the model is incomplete. 

A schematic representation of this reactor model is shown in Figure 22.2. Particles of solid reactant B are in BMF, and fluid reactant A is uniform in composition, regardless of its flow pattern. The solid product, consisting of reacted and/or partially reacted particles of B, leaves in only one exit stream as indicated. That is, we assume that no solid particles leave in the exit fluid stream (no elutriation or entrainment of solid). This assumption, together with the assumption, as in the SCM, that particle size does not change with reaction, has an important implication for any particle-size distribution, represented by P(R). The implication is that P(R) must be the same for both the solid feed and the solid exit stream, since there is no accumulation in the vessel in continuous operation. Furthermore, in BMF, the exit-stream properties are the same as those in the vessel Thus, P(R) is the same throughout the system ... 

A fluidized-bed reactor consists of three main sections (Figure 23.1) (1) the fluidizing gas entry or distributor section at the bottom, essentially a perforated metal plate that allows entry of the gas through a number of holes (2) the fluidized-bed itself, which, unless the operation is adiabatic, includes heat transfer surface to control T (3) the freeboard section above the bed, essentially empty space to allow disengagement of entrained solid particles from the rising exit gas stream this section may be provided internally (at the top) or externally with cyclones to aid in the gas-solid separation. A reactor model, as discussed here, is concerned primarily with the bed itself, in order to determine, for example, the required holdup of solid particles for a specified rate of production. The solid may be a catalyst or a reactant, but we assume the former for the purpose of the development. 

Figure 23.8 Flow/kinetics scheme for bubbling-bed reactor model for reaction A(g) +. -> product(s)...

Many studies on the modelling of esterification, melt polycondensation, or solid-state polycondensation refer to the reaction scheme and kinetic data published by Ravindranath and co-workers. Therefore, we will examine the data sources they have used over the years. The first paper concerned with reactor modelling of PET production was published by Ravindranath el al. in 1981 [88], The reaction scheme was taken from Ank and Mellichamps [89] and from Dijkman and Duvekot [90], The kinetics for DEG formation are based on data published by Hovenkamp and Munting [60], while the kinetics for esterification were deduced... 

The increase in efficiency between the first- and second-generation reactors was attributed to less water in the feed and lower operating temperatures. Reactor models indicated that the major source of heat loss was by thermal conduction. The selective methanation reactor lowered the carbon monoxide levels to below 100 ppm, but at the cost of some efficiency. The lower efficiency was attributed to slightly higher operating temperatures and to hydrogen consumption by the methanation process. Typical methane levels in the product stream were 5-6.2%. ... 

In 2000 two major petrochemical companies installed process NMR systems on the feed streams to steam crackers in their production complexes where they provided feed forward stream characterization to the Spyro reactor models used to optimize the production processes. The analysis was comprised of PLS prediction of n-paraffins, /xo-paraffins, naphthenes, and aromatics calibrated to GC analysis (PINA) with speciation of C4-C10 for each of the hydrocarbon groups. Figure 10.22 shows typical NMR spectral variability for naphtha streams. Table 10.2 shows the PLS calibration performance obtained with cross validation for... 

An advanced cracking evaluation-automatic production (ACE Model AP) fluidized bed microactivity unit was used to study the catalyst and feed interactions. The fluidized bed reactor was operated at 980°F (800 K). Every feed was tested on two different catalysts at three cat-to-oil ratios 4, 6, and 8. Properties of laboratory... 

Mechanisms of reactions are important for industry because they provide information useful for optimizing catalyst and reactor conditions. The study of reaction mechanisms in industry cannot stand alone as it can in academia. Mechanistic studies are funded to solve plant problems, to decrease operating costs, and to improve product quality. There is a wide variation in industry in the amount and type of mechanistic research funded and the timing for such research. Mechanistic research on chemical reactions is most easily justified when it is focused on the development of commercial products for a company. Often the results of mechanistic studies are not published but used instead in reactor modeling. The second reason is that competitors would obtain the information at no cost. 

The C,-fraction of naphtha crackers is used as a feedstock in the Mitsubishi fluid bed process for the production of maleic anhydride. This process was commercialized in 1970. Many data related to this process including the catalyst screening, laboratory experiments, pilot plant design, reactor behaviour and the development of higher selectivity catalysts may be found in the patent literature (12-17).The patents thus give a nearly complete picture of the scale-up process. The data have been used in the present investigation to test the fluid bed reactor model. 

Heat transfer studies on fixed beds have almost invariably been made on tubes of large diameter by measuring radial temperature profiles (1). The correlations so obtained involve large extrapolations of tube diameter and are of questionable validity in the design of many industrial reactors, involving the use of narrow tubes. In such beds it is only possible to measure an axial temperature profile, usually that along the central axis (2), from which an overall heat transfer coefficient (U) can be determined. The overall heat transfer coefficient (U) can be then used in one--dimensional reactor models to obtain a preliminary impression of longitudinal product and temperature distributions. 

The completely mixed model succeeds in representing part of the experimental data and predicts that at industrial conditions the reactor is open-loop unstable. Initiator productivity decreases are accounted quite accurately only by the second reactor model which details the mixing conditions at the initiator feed point. Independent estimates of the model parameters result in an excellent match with experimental data for several initiator types. Imperfect mixing is shown to have a tendency to stabilize the reactor. 

For reactor design calculations it is necessary to know the total devolatilization rate as well as the species production rates. Therefore, one needs to include in the reactor model all the reaction rates that are available for the devolatilization of the particular coal. Kayihan and Reklaitis (8) show that the kinetic data provided by Howard, et al. (5,6) can be easily incorporated in the design calculations for fluidized beds where the coal residence times are long. However, if the residence time of pulverized coal in the reactor is short as it is in entrained bed reactors, then the handling of ordinary differential equations arising from the reaction kinetics require excessive machine computation time. This is due to the stiffness of the differential equations. It is found that the model equations cannot be solved... 

---