9.0 too

0 121 242 363 484 sos 726 847 966 X>S9 12(0 lbs/ft? hr. vapor velocity

Rgur* 8.16 (Continued) Typical efficiency characteristics of packed towers, (c) Typical efficiency characteristics for wire-mesh structured packings; (d) example of efficiency characteristics, measured for random packing, that deviate from those in ports a and c. (fart d from J. S. Eckert and L F. Walter, Hydrocarb. Proc. & Pet. Ref., February 1964. Reprinted courtesy of Hydrocarbon Processing.)

have the efficiency curve of Fig. 8.16c rather than 8.16a. Here region A-B slopes upward, and point E is not observed. According to the mass transfer models of Bravo and Fair (50) and of Spiegel and Meier (21), raising vapor and liquid loads enhances the mass transfer coefficient and interfacial area (this improves efficiency) but lowers the gas residence time (this lowers efficiency). In corrugated-sheet packing, these counteracting effects are balanced, and efficiency is independent of loads. On the other hand, surfaces of wire-mesh packings are fully wetted due to their excellent wetting characteristics, making the interfacial area independent of loads. The residence time effect outweighs the mass transfer coefficient effect, giving the efficiency curve of Fig. 8.16c.

An alternative explanation, preferred by the author, is in terms of mechanisms postulated by Kurtz et al. (3 la). As liquid rate increases, more vapor is entrained down the bed. This drops efficiency. Because structured packings permit far less lateral movement of fluids than random packings, far more vapor will be carried downward. The vapor entrainment will be most detrimental to efficiency when fluid lateral movement is restricted most. This can be expected with narrow flow channels (e.g., wire-mesh structured packings), at high liquid rates and high pressure.

The efficiency curves described above are somewhat idealized. In practice, efficiency curves generally follow the above principles, but may deviate from the patterns shown in Figs. 8.16a to c. Kunesh (51) states that point E in Fig. 8.16a is not always observed, and that in many cases the curve between points B and F is flat. Figure 8.16d is an example of experimentally measured curves for a random packing that deviate from Figs. 8.16a and b. The hydraulics and mass transfer processes taking place in a packed bed are extremely complex and are not well-understood.

Summary. The above discussion identifies the following regimes for packed column operation.

■ The turndown maldistribution regime (region to the left of A in Fig. 8.16a, not identified from the pressure drop curves): Operation in this region is undesirable because of poor efficiency.

■ The preloading regime (regions A-B and A'-B' in Figs. 8.15 and 8.16a): Most packed towers are designed to operate in this region. Column efficiency is independent of flow rate and column pressure drop uniformly increases with vapor flow rate. An exception is wire-mesh structured packing, where column efficiency declines with an increase in loads (Fig. 8.16c).

■ The loading regime (regions B-C and B'-C' in Figs. 8.15 and

8.16a): Liquid replaces vapor as the continuous phase as a column changes from normal to flooded operation. An excellent illustration of this process was given by Harrison and France (52). Most columns can operate in this region and achieve maximum efficiency, but it is not normally used for design. Some instability ("operational flood") may occur at higher rates in this region (53).

■ The flooding regime (above point C and C' in Figs. 8.15 and 8.16a): This region is characterized by instability, entrainment, and poor efficiency and is therefore avoided.

Was this article helpful?

## Post a comment