Trouble Shooting Distillation Columns

Part 3:

Trayed columns

Mark E. Harrison,

Tennessee Eastman Co., and John J. France, Glitsch. Inc.*

' For information about the authors, see p. 123 of the first article o: this series. .Marc.". 19S9. p. 116.

Overall, trayed columns operate via countercurrent liquid and vapor flow, with staged contacting for heat and mass transfer. Each stage is often modeled as a mixer-separator, with liquid-vapor contacting followed by the separation of liquid and vapor for transport to, respectively, the stages below and above (Figure 1). This description is somewhat oversimplified because the liquid actually flows across the tray, creating a concentration gradient.

FIGURE 1. Liquid and vapor flows are not altogether countercurrent in a trayed distillation column

As in a packed column, gravity is the driving force for liquid downflow. and pressure differential moves the vapor up through the trays. The surface area for mass transfer is provided by the bubbles and droplets generated by the injection of the vapor into the cross-flowing liquid. This differs from what occurs in the packed column, where the wetted packing provides much of the surface area for mass transfer.

Trays normally are designed to operate at liquid flowrates of between two and eight gal/min per inch of outlet weir length. At lower flow-rates. significant liquid entrainment into the higher tray and downcomer (relative to the total liquid flow across the tray) can backmix the achieved separation — reducing the apparent efficiency (Figure 2).

Liquid flowrates higher than 12 gal/ min per inch of outlet weir increase the liquid crest over the weir, adding to the liquid level on the tray. A higher level raises the tray pressure drop and boosts the tendency of liquid weeping through the vapor flow openings. Vapor flow is bounded on the low side by that required for efficient liquid-vapor contacting and weeping prevention. Conversely, a higher vapor flowrate hikes tray pressured drop, abetting entrainment. and may eventually lead to column flooding.

Trays can operate in a continuous vapor regime (with small droplets of liquid dispersed in the vapor above the trays) or in a continuous liquid regime (with bubbles dispersed in the liquid). Although operating rate is also a factor, the former condition is characteristic of vacuum columns and the latter of high-pressure and high liquid flowrate columns.

Where flooding can begin

In a packed column, the apportioning of column cross-section for liquid and vapor flow can vary to suit the relative liquid and vapor loading. In a trayed column, these flow areas are fixed by how the design engineer allocates downcomer and transfer area. Whether flooding starts in the transfer area or in the downcomer will depend on which area is limiting. Tray flooding is often differentiated into jet flooding and downcomer flooding.

Jet flooding occurs when the vapor passing through a tray generates a froth that carries excessive liquid into the tray above. This rapidly increases the tray pressure drop and the liquid recycle across the tray. The higher pressure drop and liquid flowrate backs liquid higher into the downcomer until liquid spills onto the tray above. The liquid continues building, flooding in order the higher trays. This is called jet flooding because the flooding is initiated by the jetting of liquid from the tray transfer area (Figure 3).

Jet flooding results from more energy being transferred from the entering vapor into the liquid on a tray than can be dissipated in the space above the tray. Most common in vacuum and low-pressure columns (in which the vapor flowrate is much higher than the liquid flowrate), jet flooding can also occur on trays having a low percentage sieve-hole area or a small number of valves or bubble caps per tray transfer area. The higher vapor velocity resulting from the reduced contacting area can initiate jet flooding.

The variables that primarily affect jet flooding are the tray vapor area, tray spacing, and vapor flowrate and density. Among the less influential variables are tray geometry, type of contacting medium, liquid flowrate and other liquid and vapor properties.

Downcomer flooding refers to flooding initiated liquid from the downcomer. Several factors contribute to where the operating liquid level is in a downcomer. The dominant one is the liquid backup that balances the pressure drop across the tray. (This same pressure drop exists across the downcomer.) The equivalent clear liquid level on a tray below, which is influenced by its outlet weir height and the weir's crest of flowing liquid, generates a "static" liquid backup into the downcomer. The head loss caused bv the liquid flowing through the restricted space between the bottom of the downcomer panel and the seal pan also contributes to where the level is in a downcomer. Thus, the total liquid head from downcomer backup is the sum of the tray pressure drop, the liquid and the head loss from the liquid flow under the downcomer. The liquid entering the downcomer is aerated, however. and this lowers its density, increasing the liquid height in the downcomer above that equivalent to the clear liquid head. If any of the foregoing contributors level on the j becomes excessive (for instance, too tray be- > much head loss under the downcomer or aeration due to foaming), liquid will back up in the downcomer and spill onto the tray. This will increase the tray pressure drop.

FIQURE 2. Sieve tray performance at various relative liquid and vapor flowrates

FIQURE 3. Sequence of jet and downcomer flooding in a trayed column

FIQURE 2. Sieve tray performance at various relative liquid and vapor flowrates

FIQURE 3. Sequence of jet and downcomer flooding in a trayed column

Tray Efficiency Distillation Column

cnGinccmnG KRTORC

and back up the downcomer level even \ further, flooding the trays in an up- j ward progression.

Pressure drop and efficiency

Because its violent method of generating mass-transfer area (compared with ; the packed column) occurs at the expense of pressure drop energy, a trayed column operates at a higher pressure drop per stage than does a packed columns. For example, based on ! a tray efficiency of 70% and a 21-in. tray ! spacing (a HETP" of 30 in.), typical I trays will operate at a pressure drop of i about 4 to 11 mm Hg per stage, com- i pared with 1 to 4 mm Hg per stage for a I 30-in. HETP packing.

Tray pressure drops range from 2Vi to 8 mm Hg, depending considerably on I the liquid and vapor loadings, tray type, ; outlet weir height, and design pres- j sure-drop limitations. Operation at very ! low tray pressure drop can result in liquid weeping, whereas operation at a pressure drop higher than 8 to 9 mm Hg per tray will often bring on flooding. The liquid level contributes to the tray pressure drop even at low vapor flowrates. In the packed column, however, the liquid contribution is not significant, except at high liquid or vapor flowrates — i.e.. with increased liquid holdup on the packing.

Tray efficiency depends on tray geometry, liquid and vapor properties and operating conditions. There is no reliable method of predicting efficiency. Data from similar columns (including pilot-plant units) afford the best basis for estimating tray efficiencies. Columns distilling hydrocarbons often operate at Murphree efficiencies near to, or above, 100%. In many process services, tray efficiencies of between 40% and 80% are realized.

Types of contactors

The most common tray-column vapor-and-liquid contactors are sieve, valve and bubble-cap trays. Since about the 1950s, sieve and valve trays have virtually displaced bubble-cap trays, which are more expensive and offer lower capacity.

The bubble cap's more-positive liquid trapping confers good operating-rate

'The heicht of packinc chat makes a «eparation equivalent ;o one tneoretical piate — a measure of man-transfer eificiencv.

Sieve trays are less expensive than bubble-cap or valve trays because they are the simplest to make flexibility because liquid cannot normally leak through them. Bubble-cap trays often can operate at vapor flowrates as low as one tenth of their maximum capacity without significant hydraulic or efficiency problems (Figure 4).

On the negative side, the capacity of a column equipped with bubble-cap trays is generally about 15% less per unit transfer area than that of a column having sieve or valve trays, because the height of the bubble caps makes the tray spacing narrower. Additionally, the caps tend to obstruct liquid flow-across the tray at high flowrates. A liquid level gradient across a bubble-cap tray can result in the maldistribution of vapor flow toward the outlet edge of the tray. This can cause premature jet flooding. In extreme cases, the vapor flow may be reduced to nothing on the inlet edge of the tray, allowing liquid to flow back through the cap riser. Other shortcomings are the higher pressure drop due to the more-com-plex vapor flowpath and a greater tendency to foul and collect solids.

Sieve trays are the least expensive because they are simple to make. Holes or orifices, typically lA in. to 1 in. in diameter, are punched in tray panels to provide a hole area of from 5% to 15% of a tray's total surface area (Figure 5). Sieve trays offer higher capacity with lower pressure drop and entrainment than do bubble-cap trays.

With their larger holes, sieve trays resist fouling better than do bubble-cap and valve trays. Because there are no valves or caps to corrode, corrosion is less of a problem, although it may enlarge the openings. The main drawback to sieve trays is the reduced turndown, because the holes can weep liquid at low vapor flowrates. At very low vapor flowrates, all the liquid entering the tray may pass through without overflowing the outlet weir, an occurrence known as dumping. Because of this, the turndown of a sieve tray is often limited to about 50% of its design rate. Consequently, sieve trays are most of ten used when turndown is not critical, i Valve trays represent a compromise I between bubble-cap ana sieve trays, i providing greater operating flexibility i at a cost only slightly higher (Figure 6). Valve trays offer about the same capacity and efficiency as sieve trays, but i tend to weep less. Although not as ! weep-proof as bubbie-cap trays, valve i trays can be turned down to about 20Tc i of maximum capacity without weeping ! because the valves begin to ciose at low | vapor rates, reducing the tray's open area. Special valve designs provide additional turndown. Because of this flexibility, valve trays are often chosen for services involving variable liquid and vapor flowrates.

Dual-flow trays do not have down-comers. being designed so that upflow-ing vapor and downflowing liquid pass through the same orifices. This type of tray is less expensive than others, but

Distillation Alcohol

is rarely used because of low efficiencies and limited operating flexibility-. The principle advantage of such trays is. with its large openings, the more-troublefree handling of dirty and fouling liquids. Most other tray types represent a modification of the sieve, valve or bubble-cap tray.

Details of the transfer area

A tray consists of a liquid-inlet, a heat-and-mass-transfer. and a liquid-outlet zone.

Liquid flows under the downcomer and into the inlet zone, where a few inches of the tray area is intentionally void of contactors, so that vapor will not blow into the downcomer. Excessive vapor flow through the downcomer could reduce its ability to deaer-ate liquid flowing into it. This dead zone is more needed for valve and bubble-cap trays than for sieve trays because vapor issues from the valves and caps with some horizontal velocity. A short bar is sometimes installed to deflect this horizontal vapor flow from the contactor. to ensure that vapor will not enter the downcomer.

The outlet zone extends from the last row of contactors to the outlet weir. Contactors are also absent from this zone, to start the disengagement of liquid and vapor before the overflow enters the downcomer. This helps to reduce the disengagement load of the downcomer and often lessens the tendency of the outlet row of contactors to blow liquid over the outlet weir into the downcomer.

The remainder of the tray is active in contacting vapor and liquid. It is from this area that jet flooding and liquid weeping originate.

The downcomer typically takes up a segment of a column's diameter but can be a pipe, box or other shape. Spacer tabs are often installed between the bottom of the downcomer and the seal

Downcomer Tower Installation

pan to minimize the chance of the downcomer panel being incorrectly installed or dislocated. In some cases, commonly with pipe downcomers, the outlet weir may be located some distance away, to provide a dead, or quieting, zone between the weir and the pipe, which disengages considerable vapor from the liquid and vapor before the frothy mixture overflows into the downcomer.

Flow arrangements

The most common tray liquid-flow arrangement is the single-pass cross-flow. It is the least prone to vapor and liquid maldistribution (Figure 7). However. a column's vapor capacity rises as a function of the square of its diameter, whereas the liquid flowrate across any chord length only increases proportionally to. the diameter. Multipass trays serve to lower the liquid flowrate across a column, especially large-diameter one.

Two-pass trays alternate between a center downcomer and two side-down-

Cross Flow Single Pass Sieve Tray


Bubble-cap trays offer a high turndown ratio but the capa take up a lot of space

PKHJRK 5. Sieve trays, such as this single-pass one (downcomer to the left) consist mainly of a panel full of holes in the transfer area

FKMJRK 6. Valve trays (such aa the four-pass one shown upper left) represent a compromise between the bubble-cap and sieve trays

cnGjrrecflinG FCATURC

' How a good turn eliminated hydraulic hammer. 1 The ease of the noisy feedpipe i Problem: A water-hammer-type pounding at the col-| umn feedpoint was violently shaking column and the connecting piping. The column was operating at about only ! 30% of design rate.

Troubleshooting: The location of the noise suggested a problem with the feedpipe. A check of the design drawings indicated that the feedpipe and feed sparger might be somewhat oversized, especially at the 30% feed rate. The sparger discharge-orifice velocity was calculated to be less than 1 ft/s. The feed was cool and far from its bubble point, so flashing in the sparger could be ruled out. One postula-tion was that feed liquid was running out of the upstream ; orifices, allowing vapor to enter the feed sparger through ' open downstream orifices, and that the condensation of this vapor in the feed sparger was causing a hydraulic hammering.

Corrective action: One solution might have been to i plug some of the orifices to raise the discharge velocity to j several ft/s. However, to keep velocities below 6 ft/s at j design feed rates, the following remedy was implemented: ! the feedpipe was turned so that the discharge orifices were on top of the pipe; this ensured that the sparger remained full of liquid at low feedrates; additionally, a deflector bar was installed above the orifices to keep feed from impinging on the tray above.

Outcome: The hydraulic hammer was eliminated.

Only an inspection disclosed the problem: The case of the top-flooding column

Problem: After several months of operation, an amine j stripper in a natural gas treating plant designed to remove i C02 from a rich amine solution became hydraulically unstable, surging reflux into the overhead accumulator.

Troubleshooting: The surges and instability indicat-j ed flooding. The bottom product continued to meet specifi-1 cations, suggesting that the flooding was starting near the top of the column. The relatively small pressure-drop rises at the time of surging reinforced this conclusion. Foaming became suspected, but the addition of antifoam agents brought little improvement Raising the tower pressure also produced marginal benefit. The trouble-shooter decided to inspect the tower. Opening the column revealed that the top few valve trays were coated with a 1-inch thick layer of black crystalline material, which analysis established to be mostly iron. Because the tower did not show signs of corrosion, it was concluded that the material was being carried into the plant by the natural gas stream.

Corrective action: The material was found to be readily soluble in an acid solution, so the trays were washed with acid.

Outcome: After being cleaned, the column operated as expected; Washing the column with acid once a year was found to be sufficient to maintain the column j capacity.

Debris in the downcomers: The case of the clearances made too narrow

Problem: The carbon-steel valve trays of a naphtha splitter were being severely scaled, often limiting capacity. At each turnaround, the trays had to be cleaned, a difficult and time-consuming task. To minimize scaling and facilitate cleaning, the decision was made to replace the valve trays with stainless-steel sieve trays. Upon startup with the i sieve trays, the column turned out to be hydraulically ■ unstable. The overall pressure drop was erratic and high, ; and the column did not properly separate the key components. Only reducing the boUup rate to a fraction of design rate kept the tower stable.

Troubleshooting: The column pressure drop indicated flooding. Because the pressure drops in the stripping and ! rectification sections could not be gauged separately, it was i not possible to locate the floodpoint. The drawings revealed the following: the flow from reboiler jetted directly into the ; seal pan; an inlet weir had been placed on the feed tray; the ! weir height and clearance under the downcomer had been j lowered to provide greater turndown (calculations indicated the 1-in. downcomer clearance specified was more than j adequate); and a reboiler overflow weir had been raised to within seven inches of the reboiler return nozzle, and this was possibly limiting the disengagement space.

The reflux rate was lowered until the column became hydraulically stable. Raising the reflux only a little made the column unstable. This sensitivity to liquid rate suggested a liquid flow-path problem, such as downcomer flooding. Feed to the column could be hiked if the total reflux was left unchanged. This indicated flooding in the rectification j section. Suspecting a problem with downcomers in the j rectifying section, the troubleshooter checked to see if the design tray pressure drop could be a significant contributor to downcomer backup, and found the pressure drops to be reasonable. Unable to attribute the stability problem to a definite cause, the troubleshooter recommended an inspection. The inspection revealed that the clearance under the downcomers in the rectifying section ranged from % inch to % inch, vs. the 1 inch specified. Additionally, some of the downcomers were completely filled with loose scale and fiberglass tray gasketing material. The clearance problem was mainly attributed to scale left on the tray support rings during the installation of new panels that raised the height of the tray panels. Thus, tray improper installation restricted downcomer clearance, which limited the liquid throughput and caught the larger pieces of scale and loosened tray gasketing.

Corrective action: The travs were removed, the support ring cleaned, the trays reinstalled, and the gaskets omitted. The inlet weir on the feed tray was also left out. To eliminate potential problems at higher rates, the reboiler return was deflected away from the seal pan, and the reboiler overflow weir was lowered.

Outcome: Restarted, the column operated properly at design capacity. The switch to sieve trays reduced the frequency of tray cleaning to remove scale.


comers. Thus, on one tray, liquid flows from two siae-downcomers to a center downcomer: on the next tray (up or down), liquid from the center down-comer splits into flows to both of the side-downcomers. This arrangement makes it possible to increase the outlet weir length per unit of transfer area and reduce the liquid flowrate across the tray. It is occasionally used in vacuum columns to iower the tray pressure drop by reducing the height of the liquid crest over the outlet weir.

Three- and four-pass trays are used less frequently because of their tendency to distribute vapor and liquid poorly. In columns in which the liquid flowrate is low. reverse-flow trays having a single chord section divided into down-comer and seal-pan area are frequently installed to reduce the total downcomer area without resorting to extremely narrow ones (Figure 7. bottom).

If liquid and vapor flowrates are expected to differ (e.g., be higher above feed tray than below), the design engineer will often arrange different column diameter sections for differences in the flow pattern, such as changing from single-pass to two-pass trays at the feedpoint. Such variations in diameter or flow require a specially designed transition tray. These transitions are often mechanically complex, and their installation must be carefully evaluated for possible hindrance to liquid or vapor disengagement.

Inlets and outlets

A feed that does not flash is usually delivered through a perforated pipe to the entrance side of a tray. This ensures that the feed will be mixed well with the liquid entering the tray. Good mixing is particularly important in columns having few trays. A deflector plate is often installed to absorb the feed's velocity energy (see box, p.130).

At a vapor feed, tray spacing is often increased by an amount equal to the feed-nozzle diameter, with the nozzle set half way between the trays. If the nozzle is located closer to the lower tray, the entering vapor may interfere with the mass of liquid and vapor above the tray. If placed higher, the vapor expanding into the tray above could initiate flooding.

Flashing feeds are either fed

Two Pass Tray

Figure 7. Single-pass crossflow tray (top left) is the most common; downcomer in two-pass tray (top center) alternates betwen ends and center; flow from central downcomer splits in the four-pass tray (top right); baffle in reverse-flow tray (bottom) divides downcomer and liquid area through a perforated pipe or into a specially designed box or pipe, to allow the liquid and vapor to separate without excessive entrainment. Two-phase feedpipes should be designed to avoid slug flow. A vapor or a flashing feed should never be introduced into, or near to, the entrance of a downcomer. Vapor fed into a downcomer or liquid feed flashing in a downcomer will reduce downcomer capacity.

liquid sidedraws are more convenient in trayed columns than in packed columns because downcomers provide a ready place for collecting liquid. Often, the downcomer seal pan is recessed to improve the disengagement of vapor from the liquid. Sidedraw piping must be sized for self-venting and the gravity flow of saturated liquid, to avoid flashing in the line. The initial horizontal run of gravity sidedraw piping should be free of valves or other flow restriction.

Jet vs. downcomer flooding

The troubleshooter can sometimes distinguish between the two types of flooding on the basis of a column's response to changes in liquid and vapor loading. One useful test is to increase the column feedrate while monitoring the column pressure drop and the reflux rate required to maintain the separation. Continue doing this until the signs of flooding are detected, then reduce the feedrate to avoid flooding.

Plot the reflux rate and pressure drop against the feedrate. The reflux

rate rising nonlinearly before signs of flooding appear (indicating a higher rate is required for efficient separation) suggests a loss of efficiency from liquid entrainment prior to jet flooding. Flooding detected without such signs of lost efficiency or a sudden steep rise in pressure drop indicates downcomer flooding (Figure 8).

This test may be less effective when the vapor loading is not uniform in the column. In such a case, jet flooding can begin on a few trays and develop into column flooding without a detectable loss of separation. Such an occurrence can lead one to false diagnosis of down-comer flooding.

Causes of jet flooding

Excess loading, fouled trays, restricted transfer area and poor vapor distribution, improper feedpipe and tray installation. and foaming constitute typical causes of jet flooding.

Excess loading — As with a packed column, the troubleshooter should have some expectation of the maximum capacity or pressure drop of the column, as designed or previously operated. If the column is limited by flooding or pressure drop to a feedrate close to the designated column capacity, the troubleshooter should evaluate the column's hydraulic performance by means of available correlations. Because of the empirical character of flooding correlations, a predicted value of 80% flooding may correspond to actual flooding flowrates.

Fouled trays — Any restriction to vapor flow accelerates the velocity of the vapor as it enters the liquid (see box, p. 130). This increases both pressure drop and entrainment. Many-fouling substances can be detected in laboratory distillation apparatus.

Other restrictions to vapor flow mav include the fol

Distillation Trays Operation
FIGURE 9. Corrosion has destroyed the valves on the feed side of this tray

lowing: mats left from the previous entry into the column, fallen panels from upper trays, dislocated baffles, or forgotten blanking strips installed to prevent weeping during a previous low-rate operation.

Restricted transfer area — Poor vapor distribution across a tray can result in local areas of high entrainment, and initiate jet flooding. Uneven vapor distribution can also occur when one area of a tray has a disproportionate amount of missing or degraded contactors, dislodged trays, or missing manwavs (Figure 9). High liquid gradients across bubble-cap trays will distribute vapor flow disproportionately toward the outlet side. Vapor flow can also be deflected bv internal

Distillationi Tray
PIGUM 8. Test reveals whether flooding is jet or downcomer i

structures, such as support beams.

Improper installations — Vapor or flashing feed (the result of volumetric expansion, flowpath deflection or velocity gradient) may interfere with the normal liquid and vapor traffic in a column. The inlet of a feed containing vapor (such as reboiler return i must be located properly below trays. Incorrectly interchanged tray panels of different design can increase vapor velocities, pressure drop and entrainment.

Foaming — By hindering the disengagement of vapor above a tray, foaming can create excessive frothing and entrainment.

Stacking the downcomers

Common causes of downcomer flooding include: excessive liquid flow, restrictions. inward leakage of vapor, improper feed introduction, unsealed bottom-seal pan, and foaming.

Excessive liquid fiov: — The cross-sectional area of a downcomer and the clearance under it limits its capacity. By comparing design or demonstrated liquid flowrates with actual flowrates, the troubleshooter can evaluate whether downcomer design is limiting column capacity. A troubleshooter who suspects a design limitation should check the columns hydraulic performance.

Restrictions — Blockage of the flow area under a downcomer will back up liquid. Debris (e.g., tools and loose tray parts), deposits (e.g., sediment), and dislocated internal structures (e.g., tray panels, feedpipes and baffles) cause such restrictions. Common is a downcomer panel installed so as to limit the liquid outlet clearance r

Distillation Tray Design

FIGURE 10. Collapsed valve trays (above) and missing caps (center) totally undermine column efficiency

Dislocated internal structures can also hinder flow into a downcomer. stacking up liquid on the upper tray. ____

initiating jet flood- X

ing or hiking the tray pressure drop and backing up liquid into the next higher downcomer.

Liquid backup can flood a total collection tray if sufficient liquid is not removed from it. For this reason, such trays or downcomers are often designed with internal overflow protection. Similarly, downcomers downstream of partial sidedraws may not be sized to handle the liquid flow if the sidedraw is stopped or restricted.

Vapor leakage — Poor assembly or corrosion damage can let vapor leak into a downcomer. This flow can hinder the flow of liquid into a downcomer or cause excessive aeration. Vapor jetting under the downcomer from closely located contactors (especially valves or bubble caps), or vapor entering downcomers that have been bowed outward into the tray transfer area, represent additional routes by which vapor can leak into a downcomer. An inadequate liquid seal also allows vapor to flow into a downcomer. A weir that is not level or too short (or any deficiency that causes the tray liquid level to be too low) can unseal a downcomer.

Improper feed introduction — Vapor or flashing feeds that enter into or above downcomers can promote flooding. A liquid feed that contains low-boiling components can flash in the downcomer after mixing with the liquid flowing off the tray.

Bottom seal-pan — Difficulties arise with the downcomer from the bottom tray and its seal pan. A high liquid level in the column base will cause liquid to back up into this downcomer. Similarly, liquid and vapor returning to th reboiler and impinging on the exit of the bottom seal pan will restrict liquid flow from the downcomer.

Foaming — Liquids that foam require longer than usual residence times in downcomers. to disengage. Unexpected foaming will excessively aerate ! the liquid in the downcomer, causing j excessive downcomer backup. j

Insights into inefficiency |

Some common causes of efficiency problems include: collapsed trays, liquid entrainment, liquid bypass (caused by damaged or poorly assembled trays, weeping and dumping), and uneven liquid and vapor distribution.

Collapsed trays — trays are a common cause of low efficiency (Figure 10). This can often be seen as low pressure drop through a section of a column, because the vapor flow is meeting little resistance. However, accumulated liquid upon collapsed tray panels can give an opposite indication. Another sign of

FIGURE 10. Collapsed valve trays (above) and missing caps (center) totally undermine column efficiency missing trays is a constant temperature (i.e.. no fractionation) at a column section.

Entrainment — Back-mixing via liquid entrainment reduces efficiency. A noticeable loss of separation at a high throughput rate is often a precursor to jet flooding. Improved separation at lower throughput or reflux confirms the likelihood of an entrainment problem. An entrainment-induced loss of efficiency can occur well below the flooding rate if the entrainment is high relative to the flow of liquid across the tray. This occurs most often when the column liquid rate is low.

Liquid bypass — Short-circuiting liquid flow across a tray bypassing vapor contact diminishes tray efficiency. This may be caused by damaged or poorly assembled tray panels that allow excessive weeping. Low vapor flowrates and missing contactors will also cause weeping . Weeping usually subsides with higher flowrates. Thus, improved separation at higher throughput may indicate a weeping problem.

Uneven distribution — As does liquid bypassing, poor liquid distribution undermines liquid and vapor contacting. Among the most common cause of inefficient distribution are outlet weirs that are not level. Poor tray blanking (such as installing wide blanking strips parallel to the liquid flow) can create liquid paths unagitated by vapor contact. Deflected flow, missing contactors (Figure 10) and a liquid gradient across bubble-cap trays can cause vapor maldistribution. ■

Reorinted from CHEMICAL ENGINEERING Mau 10S0 mnu««hi ioao hu i— ,.„.►, *m

Part 4: Auxiliary equipment

0 0


  • liviano
    How to stop flooding in a distillation column?
    6 years ago
  • barry
    Do risers under bubble caps leak or weep on trays?
    5 years ago
  • Simone
    How to troubeshoot distillation column?
    4 years ago
  • aatifa
    How to fix tray in distilation column?
    1 year ago
  • harding
    How to control weeping in distillation column?
    8 months ago
  • Asphodel
    How to avoid flooding in distillation column?
    5 months ago
  • hayley
    How to avoid weeping in distillation column?
    5 months ago
  • negisti
    What causes burping in a distilation column?
    1 month ago

Post a comment