Backmixing

This section will analyze how the velocity profiles, axial mixing, and residence time distribution are related. It will be shown why simple conveying screws have poor axial mixing capability. New mixer geometries that are specifically designed to improve backmixing will be discussed. 

---

Backer laminates Backing materials Backmixing Backset... 

Effect of Different Decrees ofNlixing. Once Eqq is evaluated on the basis of A/, and N- using equation 87, it has to be translated into Ey by considering the degree of mixing on the tray. It is obvious that for a small plate with completely backmixed Hquid,... 

In the case of unmixed vapors between the plates, the equations, being implicit in Ey, have also been solved numerically (112). The results depend on the arrangement of the downcomers and are not too different numerically from equation 93. In reaHty, however, the Hquid is neither completely backmixed nor can the tray be considered as a plug-flow device. 

Fig. 28. Relationship between Fqq and at different degrees of Hquid backmixing (4). Curves represent different Peclet numbers. From top to bottom...

Fresh butane mixed with recycled gas encounters freshly oxidized catalyst at the bottom of the transport-bed reactor and is oxidized to maleic anhydride and CO during its passage up the reactor. Catalyst densities (80 160 kg/m ) in the transport-bed reactor are substantially lower than the catalyst density in a typical fluidized-bed reactor (480 640 kg/m ) (109). The gas flow pattern in the riser is nearly plug flow which avoids the negative effect of backmixing on reaction selectivity. Reduced catalyst is separated from the reaction products by cyclones and is further stripped of products and reactants in a separate stripping vessel. The reduced catalyst is reoxidized in a separate fluidized-bed oxidizer where the exothermic heat of reaction is removed by steam cods. The rate of reoxidation of the VPO catalyst is slower than the rate of oxidation of butane, and consequently residence times are longer in the oxidizer than in the transport-bed reactor. 

Fig. 8. Theoretical residence time distributions A, combustor style approach to plug flow B, turbulent bed (100% backmixed).

Fig. 11. Catalyst cooler configurations where (a) is backmix cooler and (b) is flow-through catalyst cooler (69).

Dried product is collected in either cyclones or baghouses depending upon the product-particle size. When primary coUection is carried out in cyclones, secondary collection in a baghouse or scrubber is usually necessary in order to comply with environmental regulations. A rotary valve is used to provide an airlock at the discharge point. Screws are utihzed to combine product from multiple cyclones or large bag-houses. If required, a portion of the dried product is separated from the main stream and returned to the feed system for use as backmix. 

Axial Dispersion Backmixing in bubble columns has been extensively studied. An excellent review article by Shah et al. [AIChE... 

Liquid Dispersion Spray columns are used with slurries or when the reaction product is a solid. The absorption of SO9 by a hme slurry is an example. In the treatment of phosphate rock with sulfuric acid, offgases contain HF and SiF4. In a spray column with water, solid particles of fluorosilic acid are formed but do not harm the spray operation. The coefficient /cl in spray columns is about the same as in packed columns, but the spray interfacial area is much lower. Considerable backmixing of the gas also takes place, which helps to make the spray volumetri-caUy inefficient. Deentrainment at the outlet usually is needed. 

RTD, gas phase Nearly plug flow Backmixed Backmixed Narrow... 

Tray efficiency 0 j is supposed to represent a measure of the deviation from equilibrium-stage mass transfer assuming backmixed trays. However, the estimate of tray efficiency requires accurate knowledge of the equihbrium vaporization constant. Any deviations between the actual equihbrium relation and that predicted by the database will be embodied in the tray efficiency estimate. It is a tender trap to accept tray efficiency as a true measure of the mass transfer hmitations when, in fact, it embodies the uncertainties in the database as well. 

Mixing of product and feed (backmixing) in laboratory continuous flow reactors can only be avoided at very high length-to-diameter (aspect) ratios. This was observed by Bodenstein and Wohlgast (1908). Besides noticing this, the authors also derived the mathematical expression for reaction rate for the case of complete mixing. 

Recycle reactors at that time were called Backmix Reactors. They were correctly considered the worst choice for the production of a reactive intermediate, yet the best for kinetic studies. The aim of the kinetic study for ethylene oxidation was to maximize the quality of the information, leaving the optimization of production units for a later stage in engineering studies. The recycle reactors could provide the most precise results at well defined conditions even if at somewhat low selectivity to the desired product. 

A continuous flow stirred tank reactor (CFSTR) differs from the batch reactor in that the feed mixture continuously enters and the outlet mixture is continuously withdrawn. There is intense mixing in the reactor to destroy any concentration and temperature differences. Heat transfer must be extremely efficient to keep the temperature of the reaction mixture equal to the temperature of the heat transfer medium. The CFSTR can either be used alone or as part of a series of battery CFSTRs as shown in Figure 4-5. If several vessels are used in series, the net effect is partial backmixing. 

Backmixing due to particle distribution in dense and dilute phases. 

---