Page 135 - Applied Process Design For Chemical And Petrochemical Plants Volume II
P. 135
124 Applied Process Design for Chemical and Petrochemical Plants
Table 8-12 (Cont’d)
General Listing and Comparison of the Major Contacting Trays
._ .-
Efficiency Pressure OPrnting
vPeTr?ry Manufacturer Capacity Turn-down WOP Flexibility
Fe Tray Glitsch Same or Similar to Low Good
Figure 8-78 better than Sieve or
** Sieve/Valve Valve
Bubble Cap Norton Good High/High Medium Good
(FRI plain
3 in. and 4 in.)
Fimre 8-79
valves, MR2L Norton Good Medium/Good Medium Good
Figure 8-80
Valve, L Norton Good Medium/Good Medium Good
Caged, MR2 Norton Good Medium/Good Medium Good
Valve, M Norton Good Medium/Good Medium Good
Caged, MR? Norton Good Medium/Good Medium Good
Caged, G Norton Good Medium/Good Medium Good
*Not in wide usage.
**Nye Tray, 10-20% increased tray (over sieve or valve) capacity and good efficiency. More capacity from existing column. Improved inlet area for sieve
or valve tray with greater area for vapor-liquid disengagement.
+Not offered as all inclusive analysis, but as somewhat general guidelines based on the respective manufacturer’s literature description and best inter-
pretation by this author. This Table is not intended to be a decision-making guide, and the author recommends that the engineer discuss and pre-
sent separation requirements to the respective manufacturers. There is no intention to suggest a negative performance by any manufacturer’s
designs or fabricated equipment.
# This specialized Sieve Tray design is of high efficiency and operates with exceptional short tray spacings, sometimes as low as 6 in. between trays.
Compiled from recent manufacturer’s literature. Credit is acknowledged for the use of this material, which is not all-inclusive in this table.
Capacity: Quite similar to sieve tray, as high or higher these trays have the bubbling contact action from valves,
than bubble cap tray from 50% up to 100% design rate either moveable (liftable) or fixed (usually stamped or cut
(varies with system and design criteria). Performance at from the tray floor itself). In addition to the valve trays
specification quality falls off at lower rates. above, there are sieve trays (with multiple downcomers) of
EfJiciency: Usually not quite as high as bubble caps in Union Carbide Corp., Linde Div., and the Turbogrid tray
region of design, but falls to unacceptable values below of Shell Development Co., and Ripple Tray (sieve type) of
60% design rate. Stone and Webster Engineering Corp.
Entrainment: Only about one-third that of bubble cap tray.
Application: Systems where high capacity neardesign Bubble Cap Tray Design
rates to be maintained in continuous service. Handles sus
pended crystal and small solid materials, as well as polymer The bubble cap has been studied extensively and sever-
forming materials. Holes become plugged in salting-out al design recommendations have been presented over the
systems where trays run hot and dry (as underside of bot- years. The most complete and generally applicable is that
tom tray). Good in vacuum or low-pressure-drop design. of Bolles [5]. It should be stressed that proper mechanical
Tray Spacing: Can be closer than bubble cap due to interpretation of process requirements is essential in
improved entrainment. Twelve-inch is average; 9 to 18-in. design for efficient and economic operation. There is not
acceptable; 18 to 30 in. for vacuum. just one result, but a multiplicity of results, each unique to
a particular set of conditions, and some more economical
Pqtmetary Trays than others. Yet at the same time, many of the mechanical
design and fabrication features can be identical for these
There are many special tray designs which solve special various designs.
problems and exceed the capabilities of the conventional The tray and caps operate as a unit or system; therefore
trays. The comments regarding performance are those they must be so considered in design (Figures 863 and 866).
claimed by the manufacturer, see Table 8-12. Due to the public nature of so many design techniques,
Not all tray designs solve special problems, even though the individual engineer has sufficient information to pre-
some may have unique performance features. Most of pare a design that can be expected to perform satisfactorily.