Page 44 - Gas Purification 5E

34 GasPunification

The design of downcomers for sieve plate columns is reviewed by Biddulph et al. (1993).
These authors present the following four rules of thumb for downcomer sizing based on
years of experience:

1. Use a velocity of 1.6 ft/s (0.5 ds) for liquid flow under the downcomer (based on unaerat-
ed liquid).
2. Use the same velocity for liquid flow under the downcomer and liquid flow on the tray to
assure a smooth entry.
3. Keep the head loss due to the underflow clearance to no more than 1.0 to 1.5 inches of hot
liquid.
4. Mow adequate residence time in the downcomer for the disengagement of vapor; 3 sec-
onds for a nonfoaming system and 6 seconds for a foaming system.
Spmy Contactors

Spray contactors can be categorized into two basic types: (1) preformed spray, which
includes countercurrent, cocurrent, and crosscurrent spray chambers; spray dryers; cyclonic
spray devices; and injector venturis, and (2) gas atomized spray, which consists primarily of
venturi scrubbers. Many commercial spray systems use more than one type of spray contac-
tor and often combine sprays with trays or packing.
The correlations developed for predicting the performance of tray and packed towers are
not generally applicable to spray contactors because of fundamental differences in the con-
tact mechanism, particularly with regard to “a,” the effective area for mass transfer. In spray
contactors, the contact area is related more to the number and size of droplets in contact with
the gas stream at any time than to the configuration or volume of the contact chamber. Since,
in most spray devices, these values are determined primarily by the liquid flowrate and the
pressure drop across either the spray nozzles or the venturi throat, it is not surprising that
attempts have been made to correlate spray system performance with power consumed in the
operation. Such a correlation was originally proposed by Lunde (1958), and a plot that
includes his data is reproduced in Chapter 6 (figure 6-18). Roughly, the correlation indi-
cates that to realize two transfer units (overall gas), for example, the total amount of power
required is

Contactor hp/lOOO scfm
Venturi scrubbers .................................................... 2.0
Crosscurrent sprays with mesh ............................... 1.0
Spray cyclones ........................................................ 0.5
Packed tower (3-in. htalox Saddles) ...................... 0.2
Spray tower ............................................................. 0.1

Theoretical correlations have been developed for predicting mass transfer rates for both
the gas and liquid phases with droplets of known size. Unfortunately these correlations are of
little value for design because the droplet size is highly variable and uncertain with common-
ly used equipment. The problem of design is further complicated by backmixing in the gas
phase, which is significant in most spray chambers.
Venturi scrubbers, ejectors, and most spray chambers are, at best, single stage contactors.
Cocurrent contactors also fall into this category. The theoretically ideal performance of such
units is to produce gas and liquid products that are in equilibrium; actual hardware can only