Contract maintenance workers often will not replace the tray manways unless the tray manway is adjacent to a tower external manway. They reason that once the tray manways that are visible from the tower manway are closed, there is no way for someone to inspect the other trays. This problem is not just common—it is universal. The maintenance force at the Good Hope Refinery pulled this nasty trick on me at the coker fractionator. Equipped with my crescent wrench, I opened the tray internal manway below the side tower manway. I discovered that the 12 trays below this point had their manways stacked in their downcomers. In 1990, I worked on a project to improve fractionation at the Chevron Refinery crude distillation unit in El Segundo, California. When the tower was opened to implement my design, the tray manways were found lying on the tray decks below the diesel draw tray. The lesson is, inspect each tray and then witness the closure of each tray manway, separately.
At a Gulf Coast refinery, the reboiler thermosyphon circulation could not be reestablished after a turnaround. The tower was reopened and a lessthan-alive contract employee was found stuck in the reboiler draw-off nozzle. At the Good Hope Refinery (when I was the technical manager), we once left a complete scaffold (poles, boards, everything) in the bottom of a debutanizer tower. Rags, hard hats, plywood, and especially plastic bags left in packed columns should be removed from inside draw sumps and downcomers. I know it’s rough on the knees, but crawl across every tray and look into each downcomer. One lost flashlight in a small downcomer may flood every tray in the tower. A rag caught on a vortex breaker in a jet fuel draw box has caused a complete refinery shutdown.
Check the tray clips, tray panels, and downcomer bolting bars. At least the nuts and bolts should be finger-tight. If you find a single loose nut, insist that every nut on that tray be retightened. I will check the tray clips and 10 percent of the downcomer bolting bar nuts for finger-tight.
If a tower has a history of tray deck damage due to pressure surge or high liquid level, the mechanical integrity of the trays should be upgraded. This is done by the use of shear clips, as shown in Fig. 8.4. The use of shear clips is not the best way to improve the mechanical integrity of trays but it is the only effective method to use during a turnaround, when no other plans have been made to mechanically upgrade the tray decks and time is short.
Underneath the tray, where the tray panels are bolted together, there is a narrow vertical strip of steel which is called the integral truss. This truss is not connected to the tray ring. Referring to Fig. 8.4, a small steel bar (4 in by 2 in by 0.25 in) is welded or bolted to the end of the integral truss. This bar is the shear clip which is inserted underneath the tray ring. (Do not weld the shear clip to the tray ring.) When an upward surge of vapor pushes the tray up, the force is transmitted along the length of the integral truss, through the shear clip, to the tray ring, and thus to the vessel wall. Many large diameter towers will already have shear clips. But if your inspection indicates they are not present and tray failure has been a problem, the installation of shear clips is the way to go for three reasons:
• The job can usually be done in 24 hours, while other tower work continues.
• The shear clips can be cut from ordinary 0.25 in carbon steel plate.
• Experience proves they are effective in resisting moderate pressure surges.
The downcomer from the bottom tray is submerged in a seal pan (see Fig. 8.3), to preserve its downcomer seal. I always set the horizontal dimension between the over-flow lip of the seal pan, (dimension y) the downcomer at four inches, so I never have to worry about restricting liquid flow from the bottom tray. This horizontal dimension should be equal to or greater than the vertical clearance between the downcomer and the seal pan floor (dimension ? which is typically, two to three inches). If a deformation of the downcomer reduces the horizontal clearance between the seal pan overflow lip, and the downcomer, the resulting restriction can cause the bottom tray to flood due to downcomer back-up. If the bottom tray floods, flooding will progress up the column. With time, the entire column will flood due to the small restriction in the seal pan. That’s why a detailed trayby-tray inspection is important.
Most trays have outlet weirs devoted to maintaining the downcomer seal. But some trays have inlet weirs too, or inlet weirs, but no outlet weirs. A sketch of an inlet weir is shown in Fig. 8.2. Note the horizontal distance between the downcomer and the inlet weir (dimension x). This distance ought to be equal to or greater than the downcomer clearance—that is, the vertical space between the tray floor and the bottom edge of the downcomer. Unfortunately, a small deformation of the downcomer may push the downcomer quite close to the inlet weir. The resulting reduction in the horizontal clearance between the inlet weir and the downcomer will restrict the liquid flow. This will cause downcomer backup and tray flooding of the trays above.
Often, there is no process reason for the use of inlet weirs, especially at higher liquid rates. Then, the inlet weirs may be removed. But some tray types, such as “Exxon Jet Tab,” trays or total trap-out chimney trays with no outlet weir, absolutely require the use of inlet weirs.
The bottom edge of the downcomer from the tray above should be 0.25 to 0.5 in below the top edge of the weir of the tray below. This is called a positive downcomer seal. Without a positive downcomer seal, vapor will flow up the downcomer and displace the downflowing liquid. This will cause flooding due to excessive downcomer backup.
On the other hand, if the bottom of the downcomer is too close to the tray below, then the “head loss under the downcomer” will be excessive. Typically, a minimum downcomer clearance is 1.5 to 2 in. Too small a downcomer clearance will result in restricting the liquid flow from the downcomer. This will also cause excessive downcomer backup and flooding. Check the correct downcomer clearance on the vender tray drawings prior to the tower inspection.
The vertical edges to the downcomers are bolted to bars welded to the vessel wall. These are called, “downcomer bolting bars.” Gaskets are often used to tightly seal the edge of the tray downcomer to these bars. If the bolts are loose or if the gaskets are missing, vapor will blow into the downcomer and displace the descending liquid. Downcomer backup and flooding may result.
The area underneath the downcomer is called the downpour area. If a tray deck corrodes, it often first holes through in the downpour area. This will cause flooding due to downcomer back-up.
The bottom edge of the downcomer will be somewhat flexible in larger diameter towers. If the width of the tower is less than 5 ft, then the downcomer bolting bars prevent flexing of the bottom edge of the downcomer. However, if the width of the downcomer is over 5 or 6 ft, then downcomer bracing brackets (see Fig. 8.1) are required. The bottom edge of the downcomer should be immobilized by attachment to the bolting bar or bracing bracket every 4 to 5 ft, of downcomer width.
Recall that the pressure outside the downcomer is slightly greater than the pressure inside the downcomer. Therefore, a force will push the downcomer toward the vessel wall and reduce the open area of the downcomer. This restriction promotes downcomer backup and flooding. Don’t expect to see this deformation of the downcomer during your inspection. Once the vapor flow through the tray stops, the downcomer will spring back to its design position.
For smaller diameter towers a visual check of tray deck levelness is sufficient. For two-pass trays, a small diameter tower is less than 8 ft.
For single-pass trays, a diameter of less than 6 ft is small.For towers of 10 ft or more in diameter, check for out-of-levelness of a tray check using a carpenter’s laser level, available in hardware stores for about $40. Purchase a level that has short tripod legs. Use the bubble to level up the legs. Set the level on one end of the tower, and check the height of the red beam at the other end and at the center of the tray for out of levelness. As it is often dim and dusty in the tower, the trace of the red laser may be clearly visible. Low points and areas of the tray deck which are out of level can now be easily identified.
The more level the tray, the better the mixing efficiency between vapor and the liquid. Certainly, if the tray out-of-levelness is greater than the height of the weir, tray efficiency will be badly degraded.
Checking for weir out-of-levelness is easy. Set the laser level on the edge of the weir. Using the bubble glass level indicator, adjust the laser level to a true horizontal position. The line of red light compared to the top of the weir will indicate how much of the weir is out of level. A weir that is more than 0.5 in out of level should be re-adjusted. If it is not, stagnant liquid pools behind the higher section of the weir, as described in the prior chapter, will result and ruin the tray‘s efficiency.
Reboilers are sometimes inserted into the bottom of a tower. These are called “stab-in” reboilers. It is not a terribly good idea, because it makes it more difficult to fix a leaking or fouled reboiler without opening the tower itself. However, the “kettle” reboiler, shown in Fig. 7.7, has essentially the same process performance characteristics as the stab-in reboiler, but is entirely external to the tower.
Note that in a kettle reboiler the bottoms product level control valve does not control the level in the tower; it controls the level on the product side of the reboiler only. The liquid level on the boiling or heat-exchanger side of the kettle is controlled by the internal overflow baffle. But what controls the tower-bottom liquid level?
To answer this, let us see how such a gravity-fed or kettle reboiler works:
1. Liquid flows out of the tower into the bottom of the reboiler’s shell.
2. The liquid is partially vaporized.
3. The domed top section of the reboiler separates the vapor and the liquid.
4. The vapor flows back to the tower through the riser line. This is the column’s stripping vapor or heat source.
5. The liquid overflows the baffle. The baffle is set high enough to keep the tubes submerged. This liquid is the bottoms product.
The liquid level in the bottom of the tower is the sum of the following factors:
• The nozzle exit loss of the liquid leaving the bottom of the tower
• The liquid feed-line pressure drop
• The shell-side exchanger pressure drop, which includes the effect of the baffle height
• The vapor-line riser pressure drop, including the vapor outlet nozzle loss
Note that it is the elevation, or the static head pressure, in the tower that drives the kettle reboiler. That is why we call it a gravityfed reboiler. Also, the pressure in the kettle will always be higher than the pressure in the tower. This means that an increase in the reboiler heat duty results in an increase of liquid level in the bottom of the tower.
Should the liquid level in the bottom of the tower rise to the reboiler vapor return nozzle, the tower will certainly flood, but the reboiler heat duty will continue. Unfortunately, reboiler shellside fouling may also lead to tray flooding. This happens because the fouling can cause a pressure-drop buildup on the shell side of the reboiler.
Remember, though, that the increased tower-bottom liquid level will not be reflected on the indicated bottom level seen in the control room, which is actually the level at the end of the kettle reboiler. This is a constant source of confusion to many operators who have towers that flood as a result of high liquid levels, yet their indicated liquid level remains normal.
Figure 7.6 shows a once-through forced-circulation reboiler. Such a reboiler differs from a thermosyphon reboiler in that it has a pump to force circulation, rather than relying on natural or thermosyphon circulation. This extra pump seems rather wasteful—and it is.
The great advantage of forced circulation is that careful calculation of the pressure drop through the reboiler and associated piping is not critical. But as we can see in Fig. 7.6, the operator now has two tower bottom levels to control. Further, if the hot-side liquid level rises above the reboiler return nozzle, the force of the vapor and liquid rushing back into the column will cause the trays to flood, but the reboiler heat input will not be affected.
Most often, forced circulation is used with fired reboilers. If flow is lost to such a reboiler, furnace tube damage is likely to result. Hopefully, this is less likely to occur with a forced-circulation reboiler. Also, the higher pressure drop of a furnace may force the designer to use a pump. Sometimes we also see a forced-circulation reboiler system if the reboiler heat is to be recovered from a number of dispersed heat sources that are far away from the tower and hence a lot of pressure drop has to be overcome.