Mechanically Agitated Vessel

The following summary of operating characteristics of mechanically agitated vessels is confined to the data available on liquid-liquid contacting. 

The hold-up and bubble diameter in mechanically agitated vessels are given by the following empirical expressions ... 

Based on the practical history of scale-up, most fermentation processes for alcohol and organic acid production have followed the concepts of geometric similarity and constant power per unit volume. From the above concept, and as a strong basis for translation of process criteria, only physical properties of the process were considered in the scale-up calculation. For power consumption in an agitated vessel, there is a fixed relation between impeller speed, N, and impeller diameter, l)t. The constant power per unit volume, for a mechanical agitated vessel is given by ... 

Mixing times in mechanically agitated vessels typically range from a few seconds in laboratory glassware to a few minutes in large industrial reactors. The classic correlation by Norwood and Metzner for turbine impellers in baffled vessels can be used for order of magnitude estimates of... 

FIGURE 11.2 Mechanically agitated vessel with gas sparging. 

Mixed-flow Mixed-flow Mechanically agitated vessel 10 0.02-0.2... 

From Table 7.5, the mechanically agitated vessel gives the best performance, not only in terms of selectivity and yield but also in terms of the reactor volume. The reactor volumes in Table 7.5 are only indicative as they are based on an assumed hold-up of the gas in the reactor. 

Table 1 reports a wide spectrum of typologies of biofilm reactor upflow anaerobic sludge bed (UASB), fluidized bed, airlift, fixed bed with and without recycle, mechanically agitated vessel, rotating drum and rotating biological contactor. Each reactor is characterized by positive features and drawbacks. 

For gas absorption, the equipment possibilities are generally packed columns plate distillation towers, possibly with mechanical agitation on every plate deep-bed contactors (bubble columns or sparged lagoons) and mechanically agitated vessels or lagoons. Packed towers and plate distillation columns are discussed elsewhere. Generally these... 

Conventional reactor designs have been optimized to create a basis for comparison. The mechanically agitated vessel achieves the highest yield of 74.4%, followed by the bubble column reactor with a yield of 72.9%, and the co-as well as the countercurrent reactors both achieving a yield of 69.5%. 

In recent years attempts have been made to improve the gas-liquid mass transfer by changing the design of the mechanically agitated vessel. Mann et al. (1989) evaluated the use of horizontal baffles mounted near the gas-liquid surface. Horizontal baffles prevent vortex formation, generate less shear than standard baffles, increase gas holdup, and improve gas-liquid mass transfer. The latter two results are due to the rotational flow below the baffles, which causes gas bubbles to move upward in a spiral trajectory and induces surface aeration. For a 12-inch i.d. and 18-inch-tall stirred vessel, they showed kLat to be improved by a factor of 1.6 to 2.3 with 30 to 50% lower agitation power compared to the standard vessel. 

The major characteristic of a polymeric reactor that is different from most other types of reactors discussed earlier is the viscous and often non-Newtonian behavior of the fluid. Shear-dependent rheological properties cause difficulties in the estimation of the design parameters, particularly when the viscosity is also time-dependent. While significant literature on the design parameters for a mechanically agitated vessel containing power-law fluid is available, similar information for viscoelastic fluid is lacking. 

Gas flow has little effect on heat transfer in a mechanically agitated vessel containing power-law fluid. While for turbine stirrers the heat-transfer coefficient for a power-law fluid can be obtained from Eq. (7.7), a more generalized form Nu = a[Re /(m)]2/3 Pr1/3 should be preferred. Here the expression given by Metzner and Otto (1957) for Re /(m) should be used and the viscosity in Prandtl number must be the constant viscosity value at high shear rates. 

In previous sections, we examined the design parameters for gas-liquid, gas-solid, liquid-liquid, gas-liquid-solid, biological polymerization, and special types of mechanically agitated reactors. In this section we present a brief review on available techniques for the measurement of various mixing and transport parameters for a mechanically agitated vessel. Both physical and chemical techniques are examined. 

The physical technique just described directly measures the local surface area. The determination of the overall interfacial area in a gas-liquid or a liquid-liquid mechanically agitated vessel requires the application of this technique at various positions in the vessel because of variations in the local gas (or the dispersed-phase) holdup and/or the local Sauter mean diameter of bubbles or the dispersed phase. The accuracy of the average interfacial area for the entire volume of the vessel thus depends upon the homogeneity of the dispersion and the number of carefully chosen measurement locations within the vessel. 

Figure 10.2. Representative impellers for fluid mixing in mechanically agitated vessels (descriptions are in the text).

Also, data on particle-liquid mass transfer from suspended solids in gas-liquid mechanically agitated vessels are practically nonexistent (R18). However, many studies have been published on mass-transfer experiments in the absence of gas, which give an idea of the magnitude of k. Recent reviews by Nienow (N9) and Blasinski and Pyc (B17, B18) indicate two fundamentally different approaches to the prediction of A s the Kol-mogoroff theory, which implies equal at equal power input per unit volume (B17) and the terminal velocity-slip velocity theory which relates ks to the value that would apply if the solid particle moved at its terminal velocity (H2). As explained by Nienow (N9), the resulting values of A s are approximately the same. Use may be made of the graphical correlation given by Brian et al. (B29). 

---