The concept of tray stability can apply to bubble caps and be used as an alternative method to determine the minimum efficient capacity of these devices. A new stability correlation for bubble-cap trays is proposed and checked against FRI data
The bubble-cap tray has been in the market place as a distillation device since the early 1800s . It has been used extensively on distillation tray equipment worldwide as a highly efficient vapor/liquid-contacting device. Many people have examined the maximum capacity of bubble caps [2, 3], but very few have studied turndown. A U.S. Patent from 1987  endeavored to enhance the turndown capability of bubble-cap trays to as high a turndown ratio value of 15:1. It is the intent of this article to establish a turndown criterion for bubble-cap trays such that the tray designer can maintain efficient operation at the lowest loads. A relationship is established, based on operational data, that shows what the minimum efficient vapor load can be on bubble-cap trays.
Bubble-cap tray and turndown
The stability of sieve trays, movable-valve trays and fixed-valve trays was discussed in Ref. 5. The theory presented there is that there is a relationship between the upward force of the vapor to a tray that will balance against the gravitational force of the liquid (froth) depth and maintain good efficient vapor distribution onto the tray. Maintaining good vapor distribution should enable good tray performance. A “stability” term was introduced that was basically a modified Froude number. Bubble-cap trays were excluded from the discussion at that time because the liquid is not balanced against the vapor on such trays. The liquid is prevented from weeping through a bubble-cap tray by the physical presence of the bubble cap risers.
However, questions have arisen about the turndown capability of bubble-cap trays. Normally the topic of turndown is not an issue, since trays that have a loss of efficiency (and theoretical stages) at turndown can easily be compensated by an increase in reflux ratio (heat duty) to the tower. However, there are times (in highly heat-integrated units) when there is little or no extra heat available to overdrive the tower when operated at turndown. In these cases, the issue of efficient turndown of bubble cap trays becomes extremely important.
With the typical sieve or valve tray, the prevalence of weeping can be one way to determine the minimum operating vapor load to such a device. As mentioned above, the author has offered an alternate idea to weeping on how to determine the turndown of such typical trays with the introduction of the “stability” concept:
η = stability factor
ΔPDRY = inches of liquid
HS = Hydrostatic head of liquid on the tray, inches of liquid
This stability concept should, in theory, be applicable to bubble-cap trays with some modification. The theory is that to maintain good efficient operation of a bubble cap tray, the vapor must be flowing through a high percentage of the bubble cap’s risers.
One modification to the concept of Equation (1) would be that the hydrostatic head of liquid has to have the height of the slots on the bubble caps removed. The vapor that is introduced to the liquid on a bubble cap tray does not have to “fight” against the full depth of liquid (or froth) on the tray to be able to flow through the cap. The other modification to Equation (1) is the definition of dry tray pressure drop itself. Dry tray pressure drop is a simple theory that has been explored by the author in other articles [ 6–8].
ΔP DRY = inches of liquid
VH = Hole velocity, ft/s
ρV = Vapor density, lb/ft 3
CV = Orifice coefficient
ρL = Water density, lb/ft 3
gC = Acceleration of gravity = 32.174 ft/s 2
If one assumes that the four critical areas of a bubble cap are equal (riser area, reverse area, annular area and slot area) or are at least limited by the riser area, then the dry-tray pressure drop through the riser can be used to establish the ∧ P DRY in Equation (2) above. The C V for smooth-bore risers can be assumed to be 0.43. With this, Equation (2) reduces to the following:
V R = Riser velocity, ft/s
The data to test the validity of this theory were established by Fractionation Research Inc. (FRI) in Alhambra, California between the years 1956 and 1968 [ 9]. The data were taken in an industrial-size tower (48-in. dia.) at total reflux and show the tray efficiency versus vapor load. There were 125 data sets taken during this time period with numerous compounds operating from deep vacuum to 500 psia. Figure 1 shows an example of one such data set with cyclohexane/ n-heptane operating at 24 psia pressure. One can see that where the tray efficiency drops from about 80% to less than 50% as the C B falls below 0.1 ft/s, that this would be an indication of minimum efficient operation. (Note: The C-Factor, C B, is a vapor load factor based on the bubbling area.).
A review of the available data was carried out to determine suitable cases to validate the theory. The data sets from FRI included an indication of weeping from the trays. Normally a bubble-cap tray should not weep, because most trays are either welded in or fully gasketed. The FRI trays were neither, since operators have to change out the trays frequently to be able to test more trays quickly. The weeping observed was present at the tray seams and at the tray support rings in the FRI column. Recent work at FRI has shown that even minor wall weeping or leakage can cause a significant loss of tray efficiency at low vapor rates. Due to this, data exhibiting signs of weeping were excluded.
Most of the tray data were for bubble caps 4 in. in size. Of the 4-in. caps reported, there were two styles; slotted and “tea cup.” There was a comprehensive set of data from the “tea cup”-style caps that covered a wide set of pressures. In addition, by examining data from the “tea cup” cap, which is the industry standard today, all subjectivity with regards to slot shape and size is eliminated from the discussion.
Examination of the data did lead to some subjectivities. Namely, what is a tolerable reduction in tray efficiency at turndown? Figure 2 shows what was picked as the turndown limit of efficient operation (which was about a 10% loss in tray efficiency) for a set of data at 165 psia. The tray geometry examined here is reported thoroughly in Ref. 9. The inner diameter is 47.75 in., distance to outlet weir from tray centerline is 15 in. and the flow-path length is 30 in. The outlet weir height is 2 in. There are 37 bubble caps per tray, each being 4 in. in size and having a 1⁄4-in. gap under the cap. The downcomers are sloped, and there is a recessed inlet pan. The outlet weir length is 37 in. Table 1 shows the examination of the two data points surrounding this choice of minimum tolerable operation in Figure 2. The value of the stability between these two data points is about 0.2.
The exact same trays were also operated in C6/C7 service at 24 psia (Figure 3). Here the 10% loss of efficiency is very close to the data point at run number 6,489. Table 2 shows the examination of the two data points surrounding this choice of minimum tolerable operation in Figure 3. The value of the stability between these two data points, and close to run 6,489, is about 0.2.
Repeating this procedure for all the other “tea cup” data sets yields an interesting set of results, as shown in Table 3. It appears that there is a constant stability factor number for turndown of 4-in. “tea cup” bubble-cap trays. This value is approximately 0.2, which is considerably smaller than other trays as reported earlier. Ref. 5 shows that most other trays need to stay above a stability number of 0.6 to remain efficient. It makes sense that bubble-cap trays can use a smaller stability number since they should not weep.
Keep in mind that all tray data examined thus far used only a 2-in.-high outlet weir. The question remains, “What happens when a different outlet-weir height is chosen?” There was one set of data examined at FRI using very tall 6-in.-high outlet weirs. The bubble caps for this data set were slotted, not “tea cup.” This data set (runs number 4,094–4,101 operating at 24 psia) was thus examined to see if a stability of 0.2 was reasonable for this unusually tall outlet-weir height.
Figure 4 shows this data set and Table 4 shows the examination of the two data points surrounding the choice of minimum tolerable operation. Table 4 results did employ a reduction of the hydrostatic head due to the elevation of the 1.75-in.-tall slots on the bubble caps. The value of the stability between these two data points is again very close to 0.2. Therefore, a conclusion can be drawn that bubble cap trays can employ a minimum stability factor of 0.2 and that the correlation above can apply to both “tea cup” and slotted caps.
The concept of tray stability can be applied to bubble caps and can be used to predict the minimum efficient capacity of these devices. A new stability correlation for bubble cap trays has been developed and checked against FRI data. Based on this limited set of data it may be concluded that the tray stability correlation result needs to remain above a value of 0.2 for bubble cap trays.
1. Forbes, R. J., “Short History of the Art of Distillation,” E. J. Brill Publisher, Leiden, the Netherlands, pp. 308–309, 1948.
2. Bolles, W. L., “Optimum Bubble Cap Tray Design,” four-part series, McGraw Hill Publishing, New York, N.Y., 1956.
3. Distillation Subcommittee of the Research Committee, “Bubble-Tray Design Manual,” AIChE, N,Y., 1958.
4. Lockett, M. J., Summers, D. R., Smith, V. C., Upchurch, J. C., “High Turndown Bubble Cap Tray,” U.S. Patent 4,711,745, December 8, 1987.
5. Summers, D. R., Spiegel, L. and Kolesnikov, E., Tray Stability at Low Vapor Load, Conference Proceedings of “Distillation and Absorption 2010,” p. 611, Eindhoven, the Netherlands, September 12–15, 2010.
6. Summers, D. R., van Sinderen, A., “Dry Tray Pressure Drop of Rectangular Float Valve and V- Grid Trays”, in Distillation 2001 Topical Conference Proceedings, AIChE 2001 Spring National Meeting, Houston, TX, April 25, 2001.
7. Summers, D.R., Dry Tray Pressure Drop of Sieve Trays, Chem. Eng., June 2009, pp 36–39; www.chemengonline.com/dry-tray-pressure-drop-sieve-trays-revisited.
8. Summers, D.R., Cai, T. J., Dry Tray Pressure Drop of Sieve Trays Revisited, AIChE Spring Meeting 2017 Topical 8 Conference Proceedings, March 27, 2017.
9. FRI Bubble Cap Tray Data Base – Public Domain files, Oklahoma State University Special Collections and University Archives.
Dan Summers is the tray technology manager for Sulzer Chemtech USA, Inc. (1 Sulzer Way, Tulsa, OK 74131; Phone: 918-447-7654; Email: email@example.com). After graduating from SUNY at Buffalo in 1977, he started his career with Union Carbide’s Separations Design Group in West Virginia. He has since worked for Union Carbide Linde (now Praxair), UOP, Stone & Webster (now TechnipFMC), Nutter Engineering and Sulzer Chemtech. He is the author of over 60 papers on distillation and is a listed inventor on three U.S. patents. Summers is a member of FRI’s Design Practices Committee and was the chair of that committee for 12 years. He is also the current chair of AIChE’s Separations Division Area 2a “Distillation and Absorption.” He is a registered professional engineer in both New York and Oklahoma and is a Fellow of AIChE. He was also the recipient in 2016 of the prestigious AIChE Gerhold Award for outstanding work in the Application of Chemical Separations Technology.
This company’s Hybrid Still technology (photo) is designed to separate extremely heat-sensitive fine chemicals and other substances. These systems combine…
With a thorough investigation and proper problem diagnosis, an entire gas plant limited by a deethanizer reboiler circuit was successfully…
Data points from literature are refined into a single correlation defining dry-tray pressure drop in sieve trays The sieve tray…
Membrane distillation (MD) — a thermal desalination technique used for recovering pure water from concentrated salt solutions — is hampered…
This company’s stainless-steel fractional distillation systems (photo) are available for batch or continuous processes, and are designed for critical applications,…