ON DISTILLATION TRAY WEIR LOADINGS

By

Michael R. Resetarits & A. Y. Ogundeji

Fractionation Research, Inc.
424 S. Squires St
Stillwater, OK 74074

Paper No. 3a
Distillation Symposium
AIChE Spring Meeting
Tampa, FL
April 28, 2009

Unpublished
Copyright Fractionation Research, Inc.
The AIChE shall not be responsible for statements or opinions contained in its publications
 KEYWORDS

distillation trays tray hydraulics outlet weirs

weir loading weir loading criteria picket fence weirs spray fluidization

blowing

ABSTRACT

Distillation tray liquid rates are usually described via a term known as “weir loading.” A weir
loading on a tray is simply the volumetric liquid flow rate per unit horizontal length of outlet weir.
The last 50 years of distillation tray literature proposes to tray designers a very wide, and confusing,
range of “weir loading criteria.” It is often unclear as to whether the high-end criteria are referring to
the points at which flooding occurs, or, the points at which the number of flow passes need to be
increased. There are also low-end criteria which refer to spray fluidization points or points at which
picket fence weirs need to be used. This paper explains the utility of the weir loading variable and the
associated literature criteria.

Impact of Weir Loadings on Tray Hydraulics

Engineers who design distillation trays always heed outlet weir loadings. The majority of such
engineers measure weir loadings via the units of “gpm per horizontal inch of outlet weir” or
“gpm/inch.”

At low and very low weir loadings, spray fluidization or “blowing” often occurs. In such cases,
designers often specify picket fence outlet weirs or “weir blocks” to artificially increase weir loads,
and avoid fluidization, and increase vapor-liquid contact times. Some designers regard 4.0 gpm/inch
as a low weir loading. Some regard 2.0 gpm/inch as a very low weir loading. The primary focus of
this article, however, is high rather than low weir loadings.

At high weir loadings, Figure 1 shows some of the possible deleterious hydraulic results as
follows:
1. Large crests and high froth heights and “jet flooding”
2. Crests that are so large that those crests do not fit into downcomer mouths, and “choke
flooding” occurs
3. High froth heights sometimes yield excessive pressure drops, and “downcomer backup
flooding” occurs

Some distillation engineers contend that both choking and excessive downcomer back-up both lead to
excessive froth heights, and so, all floods end up being jet floods.

Many distillation tray hydraulicists avoid high weir loadings. They employ weir loading criteria
which they sometimes adhere to strictly. One cure for high weir loadings is increased column
diameter. More often, multi-pass trays or counterflow trays (e.g. UOP MD or Shell HiFi) are
selected. Engineers faced with high weir loadings on 1-pass trays, select instead 2-, 3- and 4-pass
trays. There are very significant disadvantages to multi-pass trays as follows:

2 
 
 1. Reduced tray efficiencies
2. Increased tray costs (more complicated drawings and more tray parts)
3. Increased horizontal ring and vertical bolting bar costs
4. Longer tray installation times

Because of these disadvantages, tray hydraulicists minimize the number of flow passes whenever
possible - and when their weir loading criteria will allow.

It should be noted, however, that flow pass minimization is not without peril. Large weir
loadings and long flow path lengths can lead to large froth height gradients, with larger froth heights
at tray inlets than outlets. These gradients can yield inlet weeping. They can also yield vapor
crossflow channeling. Nevertheless, flow pass minimization has appreciable benefits.

Literature Weir Loading Criteria

Generally, the Glitsch, Koch, Nutter and Norton tray design manuals (3, 4, 5, 6) have withstood
the tests of time.

Particularly, the flooding correlations of those manuals have successfully predicted tray flooding,
with a reasonable degree of accuracy, for many columns for many years. Those manuals, and other
references, have offered up weir loading criteria, i.e., weir loading maxima. For example, the Glitsch
manual (3) first states the following:

“Usually, a smaller tower diameter can be obtained by using multipass trays to

hold liquid rates below 8 gpm/wfp.”

Later, on the same page, the following is stated:

“If the number of passes is restricted, either by customer preference or by tower

diameter limitations, liquid rates up to 20 gpm/wfp can be and have been used.”

Then, in Glitsch Bulletin 674 (7), the following is offered:

“From a capacity viewpoint, a liquid rate greater than 6 gpm/inch of weir is the
rate at which a larger number of flow paths should be considered.”

The Glitsch manual is not the only literature source that seems to be indecisive regarding weir
loading criteria. Table 1 presents the most easily found of such criteria (3, 4, 6, 7, 8, 9, 10). Table 1
shows that there is no national historical consensus regarding the maximum weir loading at which
hydraulic malfunctions occur or when additional flow passes should be employed.

Glitsch Manual Capacity Correlation

The almost-famous Glitsch Equation 13 (3) is probably the most familiar, and employed, of all of
the publicly-available means for determining distillation tray flood points. That equation is shown
below:

3 
 
 Vapor rate dominates flood calculations, but liquid rate is included in Equation 13 via the GPM term.

In 2002, Weiland and Resetarits (11) pointed out that Equation 13 can be rearranged algebraically
to yield the following:

The above equation reveals that Glitsch Equation 13 includes (inherently) a weir loading term, i.e.,
weir loading impacts calculated % Flood. This begs the following question: If the Glitsch flood
correlation already includes a weir loading term, why is a separate and independent weir loading
criterion required? Indeed, why does anybody require weir loading criteria?

Nutter Manual Weir Loading Criteria

Mike Lockett’s book entitled Distillation Tray Fundamentals (12) endorses the weir loading
criteria of the Nutter manual (5). Those criteria are shown below:

Tray Spacing (In.) Increase Number of Passes if GPM/In. Weir Exceeds

12 3
15 5
18 8
21 10
24 13

Those criteria are (possibly) unique in that they contend that a tray’s maximum weir loading should
depend on tray spacing. This makes perfect sense. Weir loadings affect crest heights and then froth
heights. Large tray spacings accommodate large weir loadings more easily, and vice versa. But - - - -
shouldn’t vapor rates also impact maximum operational weir loadings? Is there a single vapor rate
implied by the Nutter table, or, are the increasing liquid rates of the Nutter table associated with
increasing liquid rates on a 1–to–1 molar basis?

Assuming that every weir-loading-tray-spacing pair in the Nutter table is associated with a flood
point, vapor rates were back-calculated for every weir loading. Two different flood correlations were
assumed: Glitsch and Nutter. Two different vapor-liquid systems were assumed: depropanization and
air-water. These calculations revealed that, nearly, but not perfectly, one vapor rate was assumed
when the Nutter weir loading table was initially generated. It could also easily be inferred that the
Nutter table was back-calculated from the Nutter flood correlation. If the Nutter weir loading criteria
were back calculated from the Nutter flood correlation, or from Nutter laboratory flood data, then are
the Nutter weir loading criteria redundant?

High Industrial Weir Loadings

What high weir loadings have been handled successfully industrially? The best single source of
such information is probably the paper by Resetarits, Schmude and Morehead (10). Table 2 was
extracted from that paper. Table 2 shows that weir loadings as high as 20 gpm/inch have been
handled. Generally, the tray spacings of Table 2 are large. By assuming that each line of Table 2
represents a flood point, a vapor rate was calculated for each line. Those vapor rates are shown in the
last column. Table 2 now implies that high weir loadings can be handled on distillation trays as long
as the trays spacings and vapor rates are “appropriate.”

4 
 
 The Redundancy of Weir Load Criteria

The primary contention of this paper has been implied several times but has not been specifically
stated until now: The maximum weir loading at which a distillation tray can function depends upon
the tray spacing and the vapor rate. Trays can handle very high weir loadings as long as the tray
spacings are large enough and the vapor rates are low enough. As long as a flood correlation includes
a weir loading term, either directly or indirectly, weir loading criteria are redundant and misleading.

Figure 2 was prepared using Glitsch Equation 13. Two different tray spacings were studied.
Each point on the graph represents 100% of Calculated Flood. At higher vapor rates, lesser weir
loadings are attainable. At smaller tray spacings, lesser weir loadings are attainable. Figure 3 was
also derived using Glitsch Equation 13. Again, each point represents 100% of Calculated Flood. At
higher tray spacings, higher weir loadings are attainable.

Public FRI(SM) Data

Fractionation Research, Inc. (FRI) is a non-profit company that performs distillation research,
especially on trays and packings, for its 70 member companies. FRI data that are more than 30 years
old are usually released to the public via the Oklahoma State University (OSU) library. Those public
data provide additional information regarding weir loading criteria.

Figure 4 was taken from FRI Topical Report 10 (13). This graph provides raw capacity data from
valve trays from FRI’s 4-foot diameter test column. Figure 4 shows the following:

1. Different physical systems have different capacities

2. At higher vapor rates, weir loading capacities are lower

Experienced tray designers are very familiar with such trends.

Table 3 summarizes capacity/flood data that were collected from five sets of sieve and valve trays
(13, 14, 15, 16, 17). All of those data were collected at 24 inch tray spacings. The table shows the
maximum weir loadings that were studied using those trays. To achieve each of those weir loadings,
the vapor rates needed to be sufficiently low – as shown in the table. Higher weir loadings would
have been attainable – at lower vapor rates or at larger tray spacings. FRI has appreciable film and
video footage showing trays functioning at high weir loadings, including weir loadings in excess of
literature design criteria.

A careful side study of Table 3 shows that swept-back weirs seemingly have a maximum capacity
benefit of “only” about 10%. Straight, chordal weirs certainly create bottlenecks near their ends
where there is little horizontal area for froths to deaerate and enter downcomers. Although swept-
back weirs make sense, the resultant lost bubbling area is significant. Table 3 does not provide a
strong endorsement for swept-back weirs, but such weirs are a valuable tool that should not be
removed from tray designers’ toolboxes.

Most importantly, Table 3 shows that at least five sets of conventional trays have been run at FRI
at very high weir loadings.

5 
 
 Summary

Several literature criteria are available that provide and promote maximum tray weir loadings.
Trays have been operated industrially, and in commercial-sized pilot plants, at weir loadings in excess
of the literature criteria.

Conclusion

Distillation trays can be designed for very large weir loadings, as long as the tray spacings are
large enough and the vapor rates are low enough. A reliable flood correlation (or applicable field
data) minimizes the utility of maximum weir loading criteria. Such criteria will never go away,
however, especially because high weir loadings can sometimes lead to vapor crossflow channeling
and inlet weeping.

Acknowledgement

The authors thank the staff of Process Consulting Services, and especially Mr. Scott Golden and
Mr. Tony Barletta, whose comments initiated this study of weir loading criteria.

The authors also thank the past and future staffs of FRI for the many hours of difficult work that
went into every single distillation data points.

Nomenclature

AA Bubbling area, ft2

CAF Capacity factor of Ballast trays after correcting for foaming (Graph of Reference 3)

Cb Capacity factor , based on tray bubbling area, ft/s

CFS Vapor rate, actual ft3/s

Vload Vapor load, CFS , ft3/s

ρL Liquid density, lb/ft3

ρV Vapor density, lb/ft3

FPL Flow path length, inch

GPM Liquid rate, U.S. gallons per minute

wl Weir loading, U.S. gpm/inch

wfp Width of tray flow path, inch

6 
 
 Bibliography

1. Kister, H. .Z., Larson, Kirk F. and Madsen, Poul E., “Vapor Cross-flow Channeling on Sieve
Trays: Fact or Myth?”, Chemical Engineering Progress, p. 86-93, November 1992.
2. Resetarits, M.R. and Pappademos, N., “Factors Influencing Vapor Crossflow Channeling”,
AIChE Annual Meeting, Reno, NV, November 8, 2001.
3. Glitsch, Inc., Ballast Tray Design Manual Bulletin No. 4900-Fourth edition, Dallas, TX,
February 1984, pp 13 and 33
4. Flexitray Valve Tray Design Manual, bulletin 960-1, Koch Engineering Company Inc.,
Wichita, KS, 1982, p 8
5. Nutter Engineering, Nutter Float Valve Design Manual, Rev. 1, Aug 1981, pp 10
6. Norton Chemical Process Products Corporation, Valve Tray Design Manual, 1996, pp 4
and 6.
7. Glitsch, Inc., “17 Critical questions and Answers About Trays, Column Internals and
Accessories”, Bulletin 674, 1986. P 1.
8. Kister,Henry Z., Distillation Design, McGraw-Hill, NY, NY, ISBN 0-07-034909-6, 1992,
pp 288 – 294, 340.
9. Kister, Henry Z., Distillation Operation, McGraw-Hill, NY,NY, ISBN 0-07-034910-X,
1990, pp167 – 169 and 175 – 177.
10. Resetarits, M.R., Schmude, D.J. and Morehead, P.W., “Designing Crossflow Trays for
High Weir Loadings,” AIChE Spring Annual Meeting, New Orleans, La., March 10-14,
2002.
11. Weiland, R.H. and Resetarits, M. R., “New Uses for Old Distillation Equations,” AIChE
Spring Meeting, New Orleans, La., March 11-14, 2002.
12. Lockett, Michael J., Distillation Tray Fundamentals, Cambridge University Press, New
York, NY,1986.
13. Keller, G.J., Richards, K.J. and Winn, F.W., “Koch Flexitray”, FRI Topical Report 10,
August 1956.
14. Keller, G.J., Winn, F.W. and Yanagi, T., “Glitsch Ballast Tray”, FRI Topical Report 15,
September 1958.
15. Keller, G.J. and Yanagi, T., “Glitsch Ballast Tray, Type VI”, FRI Topical Report 19,
October 1959.
16. Yanagi, T. and Keller, G.J., “Nutter Type B Float Valve Trays”, FRI Topical Report 31,
July 1964.
17. Yanagi, T., “Glitsch V-O Ballast Tray”, FRI Topical Report 41, December 1967.

7 
 
Figure 1: Tray Flooding Mechanisms
Jet Flood Choke Flood

Back-Up Flood

8 
 
 Figure 2: Glitsch Equation 13 Flood Calculations

Figure 3 : Glitsch Equation 13 Flood Calculations (Cb =0.200 ft/s)

9 
 
 Figure 4 : FRI Valve Tray Capacity Data (Topical Report 10)

10 
 
 Table 1
Literature Weir Loading Criteria
gpm/inch
Reference First Quote Second Quote Third Quote

3 8 20

7 6

6 10 13.5 10 - 15

4 10.4 – 11.7

8 7 – 13

9 7 – 13 10 – 13 20

10 25 18.8 20

Table 2
Industrial Design Points

Application Tray Type Tower No. of Tray Spacing, Weir Load, Cb @

Diam., ft Passes inches gpm/inch flood, ft/s

Product Stripper Conv. Valve 9 2 27 19.17 0.228

Deethanizer SUPERFRAC 5 1 24 14.25 0.275

Crude Twr – PA Bi-FRAC 25.5 2 30 15.92 0.271

Crude Twr – PA SUPERFRAC 14.3 2 36 18.17 0.264

VGO Stripper Sieve 9.8 2 29.5 19.58 0.232

Stabilizer SUPERFRAC 10 2 24 17.58 0.234

Deethanizer Sieve 7.5 2 24 14.5 0.265

Debutanizer SUPERFRAC 5.6 2 24 13.08 0.280

FCC Main Frac – PA Bi-FRAC 24 2 36 20 0.246

C3 Splitter NYE 11.5 2 18 12.08 0.250

Stripper Conv. Valve 10 2 24 18.75 0.222

FCC Debutanizer SUPERFRAC 9.5 2 24 18.17 0.228

Crude Twr – Stripper Sieve 16 2 24 13.83 0.272

11 
 
 Table 3

FRI Valve Tray Capacity Data – Maximum Weir Loadings

FRI System Pressure Tray Weir Type Maximum Cb

Topical (psia) Spacings Weir Load (ft/s)
Report (inches) (gpm/Inch)

10 iC4-nC4 165 24 Swept 12.52 0.172

Back

10 C6/C7 24 24 Swept 10.42 0.297

Back

10 C6/C7 4.5 24 Swept 7.02 0.311

Back

15 C6/C7 24 24 Swept 10.5 0.326

Back

15 C6/C7 4.5 24 Swept 9.1 0.356

Back

15 C6/C7 50 24 Swept 10.6 0.270

Back

15 iC4-nC4 165 24 Swept 12.5 0.179

Back

19 C6/C7 24 24 Straight 14.6 0.284

19 C6/C7 4.5 24 Straight 12.6 0.357

19 iC4-nC4 165 24 Straight 17.7 0.178

31 iC4-nC4 165 24 Swept 13.04 0.233

Back

31 C6/C7 24 24 Swept 10.20 0.350

Back

41 C6/C7 24 24 Straight 14.46 0.319

41 iC4-nC4 165 24 Straight 17.54 0.174

12 
 
 WEIR LOADINGS IN DIFFERENT UNITS

GPM/FT GPM/INCH CFS/FT M3/HR/M

22.4 1.87 0.0500 16.7
24.0 2.00 0.0535 17.9
36.0 3.00 0.0802 26.8
48.0 4.00 0.107 35.8
50.0 4.17 0.111 37.3
72.0 6.00 0.160 53.7
84.0 7.00 0.187 62.6
96.0 8.00 0.214 71.5
100 8.33 0.223 74.5
112 9.35 0.250 83.6
120 10.0 0.267 89.4
150 12.5 0.334 112
156 13.0 0.348 116
200 16.7 0.446 149
240 20.0 0.535 179
250 20.8 0.557 186
300 25.0 0.668 224

Notes:

1. CFS/FT * 448.8 = GPM/FT

2. CFS/FT * 334.46 = M3/HR/M
3. Linde range: 0.05 to 0.25 CFS/FT
4. Good guideline for fluidization is 2.0 gpm/inch
5. TR138 If statement kicks in at 4.0 gpm/inch at TS=24in

13