Tube reactors, mass transfer coefficients

Understanding the effect of reactor diameter on the volumetric mass transfer coefficient is critical to successful scale up. In studies of a three-phase fluidized bed bioreactor using soft polyurethane particles, Karamanev et al. (1992) found that for a classical fluidized bed bioreactor, kxa could either increase or decrease with a change in reactor diameter, depending on solids holdup, but for a draft tube fluidized bed bioreactor, kxa always increased with increased reactor diameter. 

Here, II,L is the hydrogen solubility, which is assumed to remain essentially constant along the entire length of the reactor. The quantity r.V is the open volume of the reactor, A is the transverse cross-sectional area for the hydrogen transfer (as shown in Fig. 7-32), k, is the liquid-film mass-transfer coefficient at the gas -liquid interface, and a is the gas-liquid interfacial area per unit volume of the open space in the reactor. In a physical sense, Eq. (7-39) equates the mass transfer from the gas into the liquid phase with the mass transfer at the surface of the catalyst tube. The constant C, in Eq. (7-39) is obtained, by using the condition (7-40), as... 

Compared to bubble columns, airlift reactors have better liquid circulation but lower rates of mass transfer and mixing. These rates are enhanced in modified airlift reactors with perforated single or coaxial draft tubes. This enhancement is because of the breakup of gas bubbles into smaller bubbles when crossing perforated tubes. The gas-liquid interfacial area and the gas-liquid mass transfer coefficient increases. Similar effect can be achieved with the addition of packing to the riser. ... 

In Example 10.3.1 we considered the calculation of the mass transfer coefficient in the gas phase of a thin-film sulfonator. A schematic diagram of a sulfonation reactor was provided by Figure 10.5. Now, in the modeling of the reactor, the estimation of the temperature profiles along the reactor tube is very important. An important parameter in the determination of the temperature profiles is the gas-phase heat transfer coefficient. Estimate this heat transfer coefficient at the entrance to the reactor for the same set of operating conditions as specified in Example 10.3.1. 

Table 2 summarizes the mass transfer coefficients determined for this model for batch tubes, stirred batch, and flowthrough reactor coi gurations and shows that the diffusive mass transfer coefficient kd increases in this order of reactor type. However, one would expect the mass transfer coefficient to follow such a pattern as flow is increased. On the other hand, although the rate constants for conventional models based on only chemical reaction can also be fit to data from these three reactor types, rate constants for these models should only depend on temperature, and these variations would not be expect. Thus, coupling mass transfer to reaction appears to provide a more meaningful explanation for the effects of flow on performance, but further work is needed to Mly develop and evaluate this approach. 

Comparison of a single-tube packed-bed reactor with a traditional batch reactor was also published in the case of o-nitroanisole hydrogenation, not for productivity purposes but rather as laboratory tools for kinetic studies (Scheme 9.11) [46]. It was shown that the better efficiency of mass transfer enables the microreactor to obtain intrinsic kinetic data for fast reactions with characteristic times in the range 1-100 s, under isothermal conditions, which is difficult to achieve with a stirred tank reactor. However, the batch reactor used in this study was not very well designed since a maximum mass transfer coefficient (kia) of only 0.06 s was measured at 800 rpm, whereas kia values of up to 2 s are easily achieved in small stirred tank reactors equipped with baffles and mechanically driven impellers [25]. This questions the reference used when comparing microstructured components with traditional equipment, with the conclusion that comparison holds only when the hest traditional technology is used. 

Suh et al. [80] 150 Xanthan products 0.125-0.2 20-33 Production of xanthan in a concentric tube reactor and measured holdup and mass transfer coefficients at various stages of reaction. 

The scant information regarding the influence of internal draft tubes on the liquid side mass transfer coefficient is of conflicting nature [67,70]. The following correlation, due to Bello et al. (71 ], may be used to estimate the volumetric mass transfer coefficient in air lift reactors with a downcomer and riser ... 

Gas stream containing a reacting chemical species (A) is passed through a bubbling fluidised bed catalytic reactor, at a rate of 350 m /h. The reactor has fine catalysf parficles present in the dense phase, which occupies 74% of the bed volume. The density of fhe catalyst in the dense phase is 89 kg cat/m. The reaction is first order with rate constant k = 5 X 10 m /(kg cat)(s). The bubble side mass transfer coefficient is reported as (k i) = 1.1 s. The diameter of the reactor tube is 50 cm. Calculate the height of fhe fluidised bed reactor required to achieve 90% conversion. 

This type of reactor consists of two parts, that are essentially different. In the venturi tube, a very effective gas/liquid dispersion is obtained, with unusually Itigh values of bubble holdup, specific surface area and volumetric mass transfer coefficients. However, its volume is small. In the tank itself, considerable coalescence will occur, but the volume is of course much larger. Both zones are effective with respect to mass transfer. In recent papers by Dirix and Van der Wiele (1990) and Cramers et al. (1992), the feasibility of this reactor type, which is still relatively new, is demonstrated convincingly. It follows from these studies, that for large gas/liquid reactors, that require a large surface area despite a relatively low gas flow rate, the venturi-loop reactor compares favourably with a stirred tank. However, for a reliable scale-up of this type of reactor a lot of experimental work has to be done, preferably under realistic conditions. 

A falling film reactor is essentially a vertical cylinder, where liquid flows downward in a thin film along the wall, and gas flows in the core. Relatively high mass transfer coefficients are obtained both in the liquid and in the gas phases. The liquid phase can be cooled effectively via the wall. Therefore this type of reactor is preferred for very rapid exothermic gas liquid reactions. There are two variations, one consists of a tube bundle, the other consists of one cylinder with a rotor. 

In the present study, the gas-liquid mass transfer coefficients under the prevailing experimental conditions were estimated using the correlations of Akita and Yoshida (17). The radial dispersion coefficients within the catalyst tubes were measured as a function of gas flow rate in a glass model of the reactor. The details of these measurements will be released in a subsequent publication. Based on this knowledge and an estimated value of the catalytic kinetic constant for solvent hydrogenation at the highest temperature and pressure conditions examined in this study, the minimum gas flow rate required to permit neglecting mass transfer resistances was obtained. Ibis value of the gas flow rate was used in all experiments. 

Van de Vusse [1] pointed out that selectivity with respect to I increases with an increase of the mass transfer coefficient (k ). In light of this observation, we have developed a new reactor of cyclonic type in which, due to strong centripetal forces on the gas bubbles, a very high k is realized [2]. This paper deals with the selectivities obtained in sulfonation of benzene with sulfur trioxide. Both neat benzene and benzene diluted with 1,2-dichloroethane were used. This reaction was selected as a model reaction for industrially important aromatic sulfation (e.g. detergents). We studied the reaction in three reactor types that greatly differ in mass transfer characteristics, i.e. in a stirred ceii reactor (low k ), a co-current gas-liquid tube reactor (intermediate k ) and in the cyclone reactor (high k ). 

Reactors with Field tubes should as a rule have a small diameter (400 mm), because a Field tube ensures that the heat is efficiently withdrawn from the whole surface of the apparatus only if the size is small. The Field tube is situated axially in the fluidised layer of contact mass and longitudinally selects the layer from top to bottom. There is cold tap water circulating in the tube (river water can not be tolerated, because it forms scale deposits in the tube and reduces the heat transfer coefficient). 

Extensive experimental determinations of overall heat transfer coefficients over packed reactor tubes suitable for selective oxidation are presented. The scope of the experiments covers the effects of tube diameter, coolant temperature, air mass velocity, packing size, shape and thermal conductivity. Various predictive models of heat transfer in packed beds are tested with the data. The best results (to within 10%) are obtained from a recently developed two-phase continuum model, incorporating combined conduction, convection and radiation, the latter being found to be significant under commercial operating conditions. 

---