|
17 Critical Questions
1. Q: How do you calculate tray capacity?
A: Capacity of perforated trays is often plotted as a function of percent hole area. Actually, the capacity of a perforated tray is not much affected by hole area unless the lack of hole area increases pressure drop and downcomer backup to unacceptable values. Eg, if a perforated tray has sufficient hole area to limit dry tray pressure drop to a reasonable value - ca. 2" to 3" liquid at 80% flood - the perforated tray will have the same capacity as a valve tray. A bubble cap tray cannot be designed to have as much hole area as a valve tray and will, therefore, have less capacity.
2. Q: What about flexibility?
A: Flexibility is a function of pressure drop and number of flow paths. If it is necessary to reduce the hole area of a perforated tray to meet certain flexibility limits, the perforated tray will have less capacity than a valve tray. If flexibility is not a consideration, the perforated tray is probably the correct choice unless there are conditions such as extremely low liquid rates which may favor valve or bubble cap trays.
3. Q: How do you cost trays?
A: A standard valve tray will normally cost ca. 20% more than a standard unblanked sieve tray with 0.50 inch or larger hole diameter. A standard bubble cap tray with 4" diameter caps will cost ca. 300% more than a valve tray. If the cap size is increased to 6", the cost differential decreases to ca. 200%. These values are affected by the number of trays involved, tray diameter and materials of construction. For this comparison, perforated and bubble cap trays are assumed to be carbon steel, and valve trays are assumed to be carbon steel tray floors with alloy valves. Costs are based on thirty 9'-0" diameter two-pass trays.
4. Q: How many flow paths should be specified?
A: The minimum number of flow paths consistent with capacity requirements should be specified. This will result in the highest tray efficiency and flexibility at the lowest cost. 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.
5. Q: How do you calculate diameter?
A: After the tray type has been selected, valve tray diameters may be calculated using the Koch-Glitsch design manual. Various technical articles have been written on design and sizing of perforated trays. In addition, many companies have established procedures for sizing this type of tray. In the absence of established procedures, the diameter of perforated trays can be calculated using the valve tray sizing procedure. Bubble cap trays may be calculated by the method of Davies or Bolles.
6. Q: What if existing columns are too small for new operating conditions?
A: In certain applications the use of Koch-Glitsch Grid® should be considered. Koch-Glitsch Grid is fabricated from panels 15" wide, 60" long and 2 1/8 " deep. They are placed in the column in a predetermined orientation. Grid has been used in the pumparound zones of crude columns and FCC fractionators. Grid and Pall type rings are used in GRC beds (Grid/ring combination) in vacuum columns preparing feedstock for FCC fractionators. Grid has about 50-75% more capacity than trays and may have an application in revamp projects.
7. Q: Are there considerations of distribution independent of trays?
A: Consideration must be given to the correct distribution of vapor and liquid regardless of the type of tray selected. Becker and Bolles have studied the effect of maldistribution on large diameter multipass trays and methods which could be used to offset or prevent maldistribution. One feature of the multichordal or sweptback downcomer which was not stressed is the improved distribution of liquid to the tray below. This reduces the area of stagnant liquid pools, thereby increasing efficiency and flexibility at reduced rates through decreased leakage.
8. Q: Does top tray feed demand special care?
A: For columns fed onto the top tray, and when the feed stream contains vapor, special design features are desirable. Eg, parts of the tray subjected to abnormally high forces should be strengthened. Also, the feed pipe should be anchored to the tower wall. Columns such as amine regenerators frequently flow vapor and liquid in slugs to the top or feed tray. Severe hydraulic pounding and tray damage can occur if the feed tray has not been properly designed to withstand this force. For normal fractionators, it is sufficient to deposit the reflux behind a 4-6" high inlet weir or to use a false downcomer. No other precautions are usually necesssary.
9. Q: Will all liquid feed streams contain vapor?
A: Intermediate feed streams should be checked carefully for an estimate of possible vaporization. Feeds normally considered to be all liquid frequently contain vapor as a result of pressure and/or temperature change before entering the column. If vapor can be present or if the feed is at a temperature greatly different from the feed tray, the feed should not be introduced into the downcomer. Circulating reflux is an exception here, since it can be colder than the liquid in the downcomer without fear of vaporization. If vapor is present in the feed, it is recommended that a wider tray spacing be used at the feed tray. The increase in tray spacing depends on the quantity of vapor present, but 6 to 12" is usually sufficient. Generalizations cannot be given for feed streams which are predominately vapor.
10. Q: How do you orient seal pans and nozzles?
A: Orientation of the seal pans below the bottom tray downcomer should be set so reboiler vapor does not impinge directly on seal pan overflow (this could entrain liquid to the bottom tray). When one must orient the reboiler return nozzle so vapor impinges on the overflow, the nozzle should be extended into the column and a tee or other device added so vapor enters parallel to the edge of the seal pan.
11. Q: Are there special considerations for the bottom tray?
A: If space between the bottom tray and the liquid level is insufficient, vapor might cause excessive liquid to be entrained to the bottom tray. This can damage the trays. Insufficient space below the bottom tray is responsible for 50% of the problems occurring in the lower part of columns.
12. Q: Should I use reboiler baffles?
A: A preferential reboiler baffle is often used below the bottom tray as a means of directing the liquid from the bottom tray to the reboiler. If one is used, it should be oriented perpendicular to the axis of the seal pan in the case of multipass trays. Otherwise the vapor must break through a curtain of liquid to enter the bottom tray. This could cause excessive entrainment and possible maldistribution. If a baffle is used to direct the liquid to the reboiler, it should be oriented parallel to the 90 - 270 axis for single-pass and parallel to the 0 - 180 axis for two-pass and four-pass designs.
13. Q: Are sumps always needed?
A: Product draw sumps are required in several types of columns. These may be for partial or total draw of the liquid from the downcomer above. If the total flow from the downcomer is to be withdrawn, a positive seal is recommended to prevent vapor from flowing up the downcomer. A total draw sump with a positive downcomer seal will generally require additional spacing above the draw tray as well as below. In addition, it may be desirable to use a leak-resistant type of tray above the draw tray for better operation at reduced rates.
Certain guidelines have been developed for the design of product draw sumps:
- In order to maintain interchangeability of tray parts, the sump width must not be greater than the width that a downcomer in that location would have.
- Normal sump depth is 1 ½ - 2 times the nominal nozzle diameter for partial draw sumps.
- If the sump depth exceeds 40% of normal tray spacing, additional spacing should be provided.
- If the downcomers are sloped and extended into the sump, the downcomer area should not be decreased to less than 50% of the top area.
In some cases it is possible to increase the width of the lower portion of the sump. In other cases it is preferable to use a nozzle at each end of the sump. When the nozzle diameter exceeds the downcomer width, nozzles at each end of each draw sump should be considered for four-pass trays and for large column diameters.
Draw nozzles should not be designed only on the basis of velocity. Entrance head should allow ½ - 1" clear space between liquid level and the top of nozzle.
14. Q: Should chimney trays be used?
A: Chimney trays may be used as draw trays, transition trays and/or protection against leakage. Use a chimney tray if residence time is required for pumping, start-up or other reasons. Since this type of tray frequently maintains a fairly high level of liquid (and consequently a tremendous weight), special consideration should be given to its design. One consideration is the placement of the draw nozzle. A flat chimney tray can be used with the nozzle located at tray floor level, or a portion of the tray floor can be lowered to form a sump and the nozzle located at the sump floor. Both designs require the same liquid head to force the design flow rate through the nozzles. Therefore, locating the nozzle in a sump lowers the liquid level on the chimney tray by an amount equal to the head requirements. This reduces the weight which the tray must support, but has little effect on residence time since that portion of the liquid depth corresponding the the head requirements should not be considered as residence time in most cases.
Under most conditions rectangular chimneys are more economical than round ones. If rectangular chimneys are proposed by the tray fabricator, they should be accepted since this reduces costs.
A chimney height of 6 - 12" is normally adequate for low liquid flow. 12 - 18" is usual for high liquid flow if the draw nozzle is located in an inlet sump. All chimneys should have hats located a sufficient space above the chimney to give a peripheral area of 1.25 times the chimney area. The hats should extend at least 1" past the chimney on all sides and should be turned down slightly (15 ) angle starting 1" from the edge of the hat, if rectangular) to prevent liquid from running back underneath the hat and down into the chimney. If a leak-free design is required, the inlet sump should be seal-welded and gasketing used on the chimney tray floor. A center sump is preferred over a side sump, as some flow restriction can occur at very high liquid rates due to the shorter weir lengths of side sumps.
15. Q: What accommodations must be made for flow path changes?
A: When the number of flow paths is changed a transition tray is usually required. The transition tray design has been simplified in recent years and can be used for the vast majority of cases.
16. Q: Do crude towers require special treatment?
A: Atmospheric and vacuum crude columns need a considerable space in the flash zone to handle the large volume of vapor present and to prevent entrainment to the tray above. It is recommended that a tangential helical baffle or vapor horn be used at the feed nozzle. Minimum spacing of 30" above the baffle and 36" below the baffle should be adhered to. The feed should not be permited to enter at the cone even if a helical baffle is used.
17. Q: Is special attention needed during start-up?
A: In order to avoid damage to equipment, several of the common causes of problems will be noted here. These are well known but frequently overlooked:
High liquid level in the column
If the liquid level is too high, vapor velocity causes massive entrainment to the bottom tray. This entrainment may even be in the form of waves of liquid. This constant buffeting of the tray loosens bolting and can dislodge the tray.
Too rapid liquid drainage
When a column floods, the liquid level may rise to a point several feet above the bottom tray. If the liquid is withdrawn from the bottom of the column at a rate greater than it can flow down through the trays, a vapor gap will be formed below the bottom tray. The presence of a vapor gap below the bottom tray, however small, imposes on that tray the weight of all liquid above it and could cause immediate failure. This is fairly common during start-up when the column becomes flooded and the immediate reaction is to lower the level as rapidly as possible.
Steaming out a column
Many columns are operated on steam-water prior to start-up in order to check out instrumentation. Caution must be exercised to assure that steam is not condensed during this operation. Condensing steam can result in a downward-acting differential pressure across a tray which may exceed the mechanical strength of the tray and cause failure.
Water in a column
Many trays are installed and tested for tightness and may even be seal-welded to insure against leakage. As a result of the tray leakage tests or from washing out the column during the shutdown procedure, water may remain in the column. If the feed to the column is extremely hot, this water left in the column may be vaporized instantaneously, causing an increase in vapor rate. Liquid may also be lifted into the trays above. The net effect is damage to the trays.
Pressure surges
All columns should be designed so any change in pressure causes an upward flow of vapor instead of a downward flow. In the case of vacuum columns, the valve used to change from vacuum to atmospheric pressure should ideally be located near the bottom of the column. If the temperature or contents in that zone prohibit location of the valve in that area, the valve may be located at the top of the column. However, its size must be restricted to ½ - ¾". A larger valve may permit too large a quantity of air or inerts to enter the column. If trays have a liquid level on them, the vapor will not be able to flow downward through the tray freely. This can create a high pressure differential and cause failure. Similar situations have been noted where rupture disks have been located in the lower part of the column at pressures above atmospheric.
Back to Top
Other Questions
1. Q: What is flooding?
A: Flooding is an expression of the capacity of the fractionating device in a column. Flooding is evidenced by the holdup of liquid in the column. This increased liquid holdup is denoted by the increased pressure drop across the tray. As pressure drop increases with an increase in either liquid or vapor flow rate, the liquid level in the downcomer increases. Ultimately, the liquid level will reach the tray above, and if the condition persists, the whole tower fills with liquid. If an attempt is made to increase either the vapor or liquid rate when approaching the flood point, the other rate must be reduced. A flooded column is very unstable and exhibits a reduced efficiency.
2. Q: What causes flooding?
A: Correctly designed fractionating devices flood due to the vapor and liquid rates reaching those rates for which the device was designed. Incorrectly designed trays may flood due to a) insufficient downcomer entrance area, and b) excessive pressure drop due to excessive liquid and vapor loadings. Flooding may also be caused by foaming, plugging, mechanical obstructions or a change, usually sudden, in the operating conditions of the tower, ie, loss of vacuum, loss of reboiler, etc.
3. Q: What is jet flood?
A: Jet flood occurs when the vapor velocity through the liquid bed or tray becomes sufficient to aerate the liquid to such a degree that the space between trays is full. This aerated bed causes an increase in the pressure drop and impedes the flow of the liquid into and through the downcomers. It is called jet flood because of the "jetting" action of liquid spray from tray to tray.
4. Q: What causes downcomer flooding?
A: A downcomer may flood due to insufficient entrance area, insufficient downcomer volume, insufficient escape area from the downcomer or excessive tray pressure drop.
5. Q: Why does Koch-Glitsch recommend a maximum percent of flood of 80%?
A: This value gives a conservative 25% safety factor to allow for any inaccuracies a customer might make in calculating loadings and for those in the design equations. The industry generally employs 80% flood as a standard for the maximum operating range for normal design. Most systems show a decline in tray efficiency at rates exceeding 80% flood. This results In extra energy use or a product of lower quality.
6. Q: Why are so many downcomer sizes used?
A: Koch-Glitsch uses downcomer sizes in certain standard increments (0.75" for single pass trays) to minimize cost to customers, but frequently receives specified sizes from customers. Some customers permit adjusting these to Koch-Glitsch's standard widths and others do not. A rule of thumb for single pass trays is 1" downcomer width per foot of column diameter. Due to the wide variety of customer needs and materials handled, Koch-Glitsch makes many non-standard widths available.
7. Q: What is the purpose of recessed inlet sumps under the downcomer?
A: Recessed sumps are used for very high liquid rates as this permits an adequate escape area under the downcomer without resorting to taller overflow weirs. (A taller overflow weir can reduce capacity, by increasing pressure drop and downcomer backup.) Recessed sumps should not be used as a means of maintaining a downcomer seal for low liquid rate services, because of the difficulty in preventing leakage from the sumps. An inlet weir would be more appropriate. Recessed inlet sumps should not be used in those services where plugging could occur.
8. Q: What is the purpose of anti-jump baffles?
A: In simulator tests, two-pass trays flooded at 80% of the rates where trays with baffles flooded. For the tests, the center downcomer width was eight inches. The width of center and/or off- center downcomers at which the vapor could blow the liquid over the downcomer is not known. Koch-Glitsch currently accepts 16 inches as the width where baffles are not required.
9. Q: Why do some tray manufacturers not use downcomer baffles?
A: Koch-Glitsch always uses them. Others may not believe in their value. We have been advised that other manufacturers would furnish them after start-up of the columns if such proved necessary. This ignores the cost of a shutdown and evades the question of who pays for the baffles and the cost of installation.
10. Q: What is the difference between splash baffles and downcomer anti- jump baffles?
A: Splash baffles are used in low liquid rate service. Their function is to prevent liquid from being "blown" over the overflow weir. Anti-jump baffles are used to prevent liquid from blowing to the opposite flow path on multi-pass trays. Splash baffles are located just slightly in (towards the active area of the tray) from the overflow weir. They are commonly used in conjunction with notched overflow weirs. Their height must exceed the spray height for full effectiveness. The Koch-Glitsch Ballast® tray design manual has excellent illustrations of these features.
11. Q: When are picket fence weirs used?
A: If the liquid rate is high enough not to require splash baffles, but not high enough for a stardard weir, picket fence weirs would be used. For example, liquid rates of 0.10 gpm/inch of weir or less would suggest splash baffles. Liquid rates of 0.10-1.00 gpm/inch of weir indicate picket fence weirs.
12. Q: When are notched weirs useful?
A: Notched weirs are of very limited value except when used with splash baffles. The spray height on a typical tray will be 12"-18". Thus, most of the liquid entering a downcomer does not flow over the weir but is blown over the weir.
13. Q: What about inlet weirs?
A: Inlet weirs are excellent for providing a positive downcomer seal for low liquid rate service. In addition, they are more easily made leak-free than recessed inlet sumps. If their height is excessive, the liquid flowing over the inlet weir will cause leakage through the inlet row(s) of valves or perforations.
14. Q: When is tray leakage detrimental to the operation of the tray?
A: A tray can leak 10% of its liquid without significant loss in efficiency, provided the leakage occurs randomly from the tray.
15. Q: What is the meaning of unit reference?
A: Unit reference is the design vapor rate expressed as the percent of that rate required to have the tray fully active (or in terms of a Ballast tray, having all the Ballast units fully open).
16. Q: How is unit reference used?
A: Minimum values of 40, 60 and 80% have been set for one-,two- and four- pass trays, respectively. If the minimum rate specified by a customer would give a unit reference less than these values, corrective measures are used by Koch-Glitsch. Included in corrective measures are a) blanking, b) omitting units, or c) using a selected number of heavy units.
17. Q: Some manufacturers use 14 and 16 gauge units on alternating rows. Why does Koch-Glitsch use 16 and 12 gauge
units, since this is more expensive ?
A: Studies show that the difference in weight between 14 and 16 gauge units is insufficient for reliable design. The weight differential between 12 and 16 gauge is adequate to give reliable results. FRI tests for some valve trays show a substantial loss of efficiency at low vapor rates. Each had 14 and 16 gauge valve units. This did not occur during any of Koch-Glitsch's valve tray tests as the units were 12 gauge equivalent in all cases.
18. Q: What determines the flexibility of a tray?
A: Several items affect the flexibility of trays. Among them and their effects are the following:
a) Liquid rate. The higher the liquid rate, the greater the leakage rate that can be accepted. However, the percent leakage that can be tolerated remains the same.
b) Pressure drop. A high pressure drop gives a higher unit reference.
c) Number of flow paths. The fewer the number of flow paths to keep active, the greater the flexibility.
d) Tray spacing. Large tray spacings permit designing with higher pressure drop and higher unit reference.
e) Foaming. Non-foaming systems can be designed to operate at higher pressure drop and unit reference than foaming systems.
19. Q: Which Koch-Glitsch tray has the greatest flexibility?
A: Highest flexibility can be obtained by slotted bubble cap trays and the Koch-Glitsch A-1 Ballast tray. The top of the slots on a bubble cap tray represent less submergence than normally used on a valve tray, which permits a more even bubbling area at low rates. The wide stance of the Ballast plate of the A-1 unit permits it to operate in the tilted position. In the tilted position, the unit passes less vapor than if it were fully open. All other type valve units are a) shut, b) open or c) moving from one position to the other. As a result, at comparable vapor rates, the A-1 Ballast tray has a larger area active than other standard valve trays.
20. Q: Which has the highest capacity, a valve or sieve tray?
A: If flexibility is ignored, a sieve tray has the same capacity as a valve tray.
21. Q: If a sieve and valve tray have the same capacity, why are sieve trays not better?
A: A sieve tray having a hole area equal to 14% of the active area has the same capacity as a valve tray. The efficiencies at 80% flood are "approximately" equal. At 50% flood, the efficiency of the sieve tray is significantly less than that of the valve tray. The major limitation of sieve trays is this lack of turndown because of the tendency to weep at lower rates.
22. Q: Why or when should a bubble cap tray be used?
A: Bubble cap trays are a good choice where leakage from one zone to another is detrimental. They should also be considered for low liquid rate service with pressure drop limitations.
23. Q: What size bubble cap should be used?
A: If the number of rows of caps are such that a larger size will still result in six rows of caps, the larger size should be considered. Consider a diameter of 7'-0". The number of rows of caps will be about 22, 12 and 8 for 3",4" and 6" caps, respectively. Because of the difference in cost, the 6" caps would be recommended even though the 3" caps appear to have a greater efficiency.
24. Q: What about a tea-cup cap?
A: Slotted caps are preferred for low liquid rate trays with pressure drop limitations. Tea-cup caps are suitable for other services. Slotted caps are recommended for large diameters due to beam camber, out-of-levelness, etc.
25. Q: What about the Koch-Glitsch V-O tray?
A: The Koch-Glitsch V-O tray is comparable to a sieve tray in performance. The V-O tray can be adapted to a wide range of conditions by altering the rise of the unit. For those towers where more than one perforating pattern is required, the V-O is normally more economical than sieve trays. The V-O unit is an ideal tray to use at the top section of an atmospheric crude column. Monel® trays are usua!ly used In this area. The reflux is frequently treated with ammoniacal solutions which are very corrosive to Monel® due to its copper content. Any movement of a part, such as found in a valve tray, results in removal of the protective film and gives rapid corrosion. Since the V-O tray does not have moving parts, it is an excellent choice. It is a better choice than sieve trays as the panels normally are compatible with the Ballast trays in the area immediately below.
26. Q: Why does Koch-Glitsch specify rectangular or square stacks on chimney trays?
A: Rectangular risers can be formed so that the walls are an integral part of the tray floor. This eliminates most if not all material waste and reduces cost. In addition, only the vertical seams need welding, whereas round risers require welding of the vertical seam and welding to the tray floor also. Several small risers give better vapor distribution than one large riser and are therefore recommended. On occasion, the use of several small risers permits a reduction in the space required for installation.
27. Q: Why does Koch-Glitsch insist on draw sumps instead of permitting the draw nozzle to be flush with the chimney tray floor?
A: Draw nozzles can be quite large - maybe 12"-14" in diameter. Generally, the depth of liquid required to provide the head necessary for the liquid flow rates is about equal to the nozzle size. Thus, this depth is not usable at design flow rates -only at reduced flow rates. This depth of liquid imposes a loading up to 40 Ib/sq ft on the chimney tray and requires a taller riser height. The use of draw sumps reduces mechanical strength requirements, lowers the riser heights and may permit a shorter tower.
28. Q: Will liquid flow down the risers of chimney trays in the event of upset conditions?
A: It is questionable if liquid will flow down the risers at either normal design vapor rates or upset conditions. Koch-Glitsch experience indicates that, more often than not, the liquid will not flow down the risers. Chimney trays should be provided with downcomers if this condition is likely.
29. Q: What vapor feed entry is recommended for atmospheric and vacuum crude towers and FCC fractionators?
A: The vapor horn is probably the best choice. Next best is a helical baffle with tangential entry. Koch-Glitsch prefers the latter, as it is very simple and reasonably economical. Experience indicates that the feed entry can not be in the cone if the column is to operate satisfactorily at 70-80% flood.
30. Q: What about mist eliminators?
A: Mist eliminators have high removal efficiency over a wide range of operating conditions. The efficiency does decrease at low vapor rates which creates a desire on the part of most designers to reduce the diameter of the pad. Unless the pad is in a service where the mist (entrainment) is independent of the vapor rate (a pad located above a spray header, for example) the diameter should not be reduced since the mist is nil at low vapor rates and a decrease in efficiency is not apparent. Some manufacturers show removal of particles 8 microns diameter and larger. Frequently, the particles smaller than 8 microns are equal volumetrically to the particles larger than 8 microns. This could be critical in sulfuric acid plant applications.
31. Q: Should mist eliminators have top holddown grids?
A: Some customers say yes, and use top holddown. Others prefer the pad be dislodged when they become plugged and therefore omit the top holddown. In the absence of customer specifications, Koch-Glitsch recommends both holddown and support grids.
32. Q: What is tray efficiency?
A: Calculations are made which indicate the vapor and liquid composition that would result if equilibrium existed at that particular temperature and pressure. One theoretical plate would accomplish these compostitions. By repeating these calculations until the desired terminal compositions are reached, the total number of theoretical plates, or trays, can be calculated. The number of theoretical trays used, divided by the actual trays, results in the tray efficiency for that system and at those operating conditions. In the case of packings, the height of packing to equal a theoretical plate is used (HETP) to express efficiency. If trays on 24" spacing have 50% efficiency, four feet are required for a theoretical plate. If a packing has an HETP of four feet, the two devices are equal efficiency-wise.
33. Q: Under what circumstances are packings chosen over trays?
A: A general statement might be made that packings are chosen at either of the two ends of the design spectrum: a) very small towers (less than 24" diameter), b) low pressure drop limits and c) services where plastic or aluminum materials would be beneficial. Packings are highly desirable in small towers due to mechanical limitations imposed due to size. In some tower sizes, trays other than cartridge style are impractical, if not impossible, to install. Packings are desirable in high liquid and vapor situations where pressure drop is a primary consideration. An excellent example is the vacuum crude tower, where pressure drop is held to a few mm Hg over the entire tower.
Applications where random packings are generally unsuitable are large diameter towers with low liquid and high vapor rates, the primary limitation being the liquid distributor and not the packing itself. An example would be a glycol contactor. Other undesirable cases would be operations that require high turndown. This is due to the drop in efficiency at low liquid rates. Although this situation can be compensated for by design, it usually becomes more expensive than trays.
34. Q: How do Koch-Glitsch packed column design procedures dilfer from those 01 other manufacturers?
A: For the past 25 years, a widely known packing manufacturer has designed columns using the Generalized Pressure Drop Correlation method. This method assumes all packings behave sufficiently alike that they may be compared on the basis of a pressure drop value and not percent of flood or capacity. Recently the manufacturer has published articles proclaiming the need to improve the design methods previously used. The "new" method employs the use of "C" factor, which is the method Koch-Glitsch has always used. The basic difference is that the Koch-Glitsch method has always designed on the basis of maximum capacity and not a pressure drop rule of thumb.
35. Q: What is KGA ?
A: KGA, as it is used in packing comparisons, is an overall mass transfer coefficient. KGA is usually presented in units of:
Ib moles/(hr) (cu ft) (atm)
Koch-Glitsch uses KGA in absorption studies to determine the relative merits of different packings. When comparing two packings, the one with the higher KGA is a more efficient packing. The relative differences obtained by absorption KGA's do not, however, correspond to the differences found in the distillation mode. KGA is almost never employed to design distillation columns. HETP is almost always used for packed tower design purposes because it is easily related to the equilibrium state calculations employed by most designers.
36. Q: What is the maximum allowable bed depth for a single metal packed bed?
A: There is no conclusive evidence that packed beds may not be infinite in depth, from a process standpoint. There are, however, mechanical limitations due to crushing or ring deformation with deep beds. Koch-Glitsch has provided 2" metal Ballast rings for 30 foot deep beds, but generally recommends 20-25 feet as a maximum.
37. Q: What is the maximum allowable bed depth for a single plastic packed bed?
A: The same mechanical limitations of crushing or deformation exist as for metal packed beds; however, due to the lower yield strengths of plastics, these problems are more pronounced. Operating temperature and chemical environment also significantly affect plastic characteristics. For these reasons, Koch-Glitsch recommends plastic packed bed depths of 10-15 feet.
38. Q: Can grid-style support plates be used with plastic packings?
A: They are not generally recommended, as plastic packings tend to extrude through the openings and be cut by the thin vertical tines of the supports. The one exception would be grids with expanded metal over the top. This option, however, is not usually recommended because it tends to flood more easily and is usually more expensive than the vapor injection support plates.
39. Q: Why are tangential entrances recommended for certain tower feeds?
A: Tangential feeds, with helical baffles or vapor horns are most commonly found in atmospheric and vacuum crude towers. This type of feed arrangement is preferred when there are very high vapor velocities in conjunction with liquid droplets. The tangential feed entrance causes the vapor to follow the contour of the vessel as it begins to expand and decrease in velocity. A second effect is that the liquid droplets, since they have a greater mass than the vapor molecules, tend to maintain a straight path and collide with the tower wall. This is a desirable effect in that it reduces the entrainment carried to the first section of the tower. In a vacuum tower this is especially beneficial since the liquid droplets in the feed stream are primarily heavy bottoms and metals.
40. Q: How is a flashing feed distributor different?
A: Flashing feed distributors are normally used for two phase feeds that contain a much higher percentage of liquid than might be found in the feed to a vacuum tower, for instance. The flow to the distributor might be considered as slug flow and tends to be very violent. The purpose of the distributor is to absorb and dissipate the momentum of the liquid, and direct the liquid to the distributor tray floor while separating the vapor to overhead.
41. Q: Are packed tower internals less expensive than trays?
A: For towers large enough to install standard trays, packing internals are about two or three times as expensive as trays. For example: a 16'-0" MEA absorber with 2-pass trays was to be revamped for high capacity. Process design dictated the use of either 4-pass trays or 3.5" Ballast rings in SS304. Cost calculations showed loose packing to cost 2.5 times as much as trays. Efficiencies are comparable. Note: the example illustrates the expense of packed tower internals, but does not include the difference in cost of installation.
42. Q: What is the difference between bed limiters and bed hold-downs?
A: Bed limiters are used with metal or plastic tower packings to prevent expansion of the bed at high flow rates. They are designed with openings small enough to prevent the passage of individual pieces of packing. Bed limiters are always attached to the tower wall by means of a support ring or bolting to clips. Bed hold-downs are used with ceramic packings to prevent the upper portion of the bed from fluidizing and breaking up at high flow rates. The hold-down rests directly on the packed bed and relies solely on its weight to restrict bed movement. In the event of a surge or upset condition, the bed hold-down does not have sufficient weight to prevent damage.
43. Q: Why don't other manufacturers recommend the use of bed limiters and bed hold-downs?
A: They take the position that the beds will perform properly under normal conditions and they do not wish to jeopardize their bids with the addition of an unspecified device. Koch-Glitsch, on the other hand, feels that restraining devices are worthwhile, and that we have the responsibility to advise the customer strongly about these safety devices. They are extremely economical in comparison to the damage that might occur as a result of an operating upset.
44. Q: What is the best liquid distributor for a packed tower?
A: There are many options available, each of which possesses particular advantages. These are:
a) Orifice style. Orifice style distributors are gravity flow devices where the liquid flows through small orifices in the tray floor. Orifice style distributors are usually preferred for small diameter towers (less than 4'-0"), clean liquids and where high turndown is not required.
b) Notched trough. This style distributor is also a gravity flow device, but the liquid flows through notches in the side of rectangular troughs. It is generally used in towers 4'-0" in diameter and greater. The notches are particularly suited to high liquid rates, high turndown and liquids that foul or carry solids.
c) Pipe. Two basic distributors are used: 1) Pipe orifice and 2) MuItiple spray nozzle. These styles may be selected if the liquid feed is clean and under slight pressure. Turndown for these distributors can be limited by mist generation and splashing. Spray distributors are best suited to large diameter towers and should have multiple nozzles for best performance. The spray distributor should be avoided in small diameter towers using only one or a few nozzles.
Back to Top
|
|
|
