Figure 35. Typical power consumption relations for triple impeller installation, giving higher horsepower in proportion to the lower impeller.

In regard to tank shape, it has turned out over the years that about the biggest tank that can be shop-fabricated and shipped to the plant site over the highways is about 14 ft (4.3 m) in diameter. As fermentation volumes have gone from 10,000 gallons (38 m3) to 50 or 60 thousand gallons, tank shapes have tended to get very tall and narrow, resulting in Z/T ratios of 2:1,3:1,4:1, or even higher on occasion. This tall tank shape has some advantages and disadvantages, but tank shape is normally a design variable to be looked at in terms of optimizing the overall plant process design.

This leads to the concept of mass transfer calculation techniques in scaleup. Figure 36 shows the concept of mass transfer from the gas-liquid step as well as the mass transfer step to liquid-solid and/or a chemical reaction. Inherent in all these mass transfer calculations is the concept of dissolved oxygen level and the driving force between the phases. In aerobic fermentation, it is normally the case that the gas-liquid mass transfer step from gas to liquid is the most important. Usually the gas-liquid mass transfer rate is measured, a driving force between the gas and the liquid calculated, and the mass transfer coefficient, KqO or l^a obtained. Correlation techniques use the data shown in Fig. 37 as typical in which KqO is correlated versus power level and gas rate for the particular system studied.

Figure 36. Schematic showing gas-liquid mass transfer step related to the other steps of liquid-solid mass transfer and chemical reaction.

Figure 37. Typical KqO versus horsepower and gas velocity correlation; box on right indicates typical pilot plant experiments; box on left indicates typical full-scale range.


Figure 37. Typical KqO versus horsepower and gas velocity correlation; box on right indicates typical pilot plant experiments; box on left indicates typical full-scale range.

If the data are obtained on small scale, then translation to larger scale equipment means an increase in superficial gas velocity, F, because of the change in liquid level in the large scale system. This normally pushes the data toward the left and would possibly result in a lower power level being required full-scale for the same mass transfer rate.

This has to be examined with great care, because any change in the power level will change the liquid-solid mass transfer rate; change the blend time, the shear rate and, therefore, the viscosity of non-Newtonian broth, and could necessitate many other process considerations.

5.3 Mass Transfer Characteristics of Fluidfoil Impellers

Experiments made with the sulfite oxidation technique evaluate the overall KqCi relationship for radial flow turbines and a very typical curve, shown in Fig. 3 8a gives the value of KqO versus power and various gas rates. One will notice that there is a break in the curve which occurs about at the point where the power of the mixer is approximately two or three times higher than the power in the gas stream. A curve taken on similar conditions for the A315 impeller does not show the break point (Fig. 38b), and matching of the two curves shows that at the low end of the power levels the A315 results in a higher mass transfer relationship, while at higher power levels, the R100 is somewhat better. However, with ±20%, which is with reasonable accuracy for these kind of measurement comparisons, the mass transfer performance is quite similar. The difference of performance on a given fermentation application should thus be higher or lower in terms of the mass transfer coefficient and needs to be studied in detail when a retrofit is desired.

One large difference between A315 and R100 impellers in fermentation is the blend time. Every R100 impeller sets up two flow pattern zones. Thus, in a large fermentation tank with three or four R100 impellers there are six to eight separate mixing zones/cells in the vessel. If the A315 impeller is used for the one or more different impeller positions, it sets up one overall flow pattern which gives one complete mixing zone and results in a blend time on a batch basis of approximately one-half to one-third the time it takes on the R100 configuration. It is quite typical in current practice to use a radial flow turbine at the bottom while using a series of A315 impellers (either one, two or three, on the top positions). This has the overall tendency to reduce the macro-and microscale shear rates and also can either increase productivity at the same power level or retain the original productivity at a reduced power level. This, in a fermentation process, is of very great importance economically.

Was this article helpful?

0 0

Post a comment