We have a case of 24 header (12 bays) air cooler with 2 inlet nozzles (8”) and one outlet nozzle (6”) per header (split headers). The headers have thrust blocks in between. The end header displacement was estimated to be around 115 mm (along the header direction) in the initial stages of the project and the vendor agreed to maintain this gap.

Problem: Due to some installation error or by a last minute design mess up by the vendor the actual installed gap in all headers (including end header) is only 45 mm. If I try to input this 45 mm max displacement limit in the calculation I have expansion stress failure at extreme piping header tees. In this situation what could be the possible solutions? The ones we are exploring include….

Pl. note all the piping and air cooler is installed and hydro tested. This problem was found just weeks ahead of startup.

1)Cut all the thrust blocks and make the header move as individual units. Still to clear expansion stress I need minimum 60 mm gap at end headers.
2)Remove the guide (tube bundle direction restraint) on the headers and make them free along the tube bundle direction. This is giving good advantage.
3)Modify support locations on the piping (relocating axial stop etc). This option needs too many support modifications.

Options 2 and 3 was checked without removing thrust blocks fully but by just removing the bolting b/w thrust blocks so that they have 20 mm gap to slide like a hinge.

We have 3 times API 661 nozzle allowable, so the nozzle load is not an issue. Another point here is there is a huge difference b/w ope temp and design temp. 187 C and 270 C. The 270 C case may occur occasionally but we don’t have any further info on duration etc. The failure is in design case only.

Respected forum members kindly opine.

Is there any responsibility for this issue for anyone outside of the equipment vendor? In the US, this would be entirely his problem. I presume you will make sure that the heater sees 270 C as a part of the acceptance test. If there's a problem, won't it surface then?

As far as the problem goes, physics is what physics is. If there is not enough room to clear the expansion of the hot piping, something is going to give. Most likely, it will be the outer wall that the headers psss through. Are you sure that there is not enoguh flexibility in the air cooler housing to make the problem go away?

If the air cooler housing can't flex sufficiently, then one or more of the axial restraints must be removed selectively, or else physics will remove one randomly and, very likely, noisily. Without seeing drawings of the equipment, it's kind of hard to give concrete advice or a rational recommendation.

It's very hard for me to believe that the equipment vendor has made this sort of mistake and does not understand the concequences. Another possible thing you need to look at - if the air cooler gets hot, will the support steel expand enough to increase the axial clearance for the headers? If that's the case, then you have no problem.

I realize nobody wants to spend the money to fix this right now, but if there is a problem somebody is going to have to fix it eventually. Better for someone to bite the bullet up front and fix the problem with construction trades than to have to shut down an operating plant and fix it with maintenance personnel (who are usually not much good at construction even though they think they are).

I repeat, make sure your acceptance test gets this unit up to the max rated temperature, and keeps it there long enough that if anything is going to break, it breaks right then. That should (if the contract with the equipment vendor is any good at all) make sure that the vendor fixes the problem before he turns over the equipment.

is the 45 mm gap the total gap( i.e +22.5 to -22.5)???

..you may look at the possibility of presetting the header to a position so that you get the total gap in the slot for your analysis..

No the 45 mm gap is on either side. Presetting option is not feasible at this stage.

SAMAbdul,

You're getting some good advice from Craig and Neelam.

Personally I vote for pre-setting the Air Cooler headers. Yes, you will have to cut some slices out of the piping, weld up again and re-hydro but then it's fixed. You say that's not feasible, but the other options don't seem to get you what you need either.

Construction may be upset about having to rework something they thought was complete but as Craig said, that's a far better option than taking the risk of a leak and an emergency shut-down and fix of an operating unit. If it comes to that, you'd better have some good answers for the client as to why you let it happen.

You haven't said much about the inlet piping configuration but it must be pretty flexible if thrust blocks work and cutting them loose helps your gap situation. With that being said, is you piping configuration stiff enough to slide those header boxes without over stressing the pipe? I recommend keeping the thrust blocks even if they have to be modified or replaced to work with the other changes you make.

You said that early in the project the cooler vendor agreed to maintain these large gaps. What happened between then and now? My experience has been that I have to review and comment on the approval drawings and also review all of the following issues and the certified drawings. If the gaps aren't addressed on the drawings, I address it and demand that they show this information on the drawings.

Air cooler vendors usually don't like large gaps because too much air escapes around the sides and reduces the efficiency of the unit. To minimize this, when I have large gap requirements, I have the header pre-positioned to the side towards the center of the entire arrangement and locked in the pre-position location at the factory with removable locks. I also make the vendor show this on his drawing.

Then the piping is designed to match the preposition nozzle locations and I also have the iso note the air cooler prepositioning. If the pipe doesn't fit, hopefully they'll check to see if the air cooler header is prepositioned like the iso says.

The Stress Engineer or Lead Stress Engineer has to follow these things through all the way or it won't happen and they'll be in the situation that you are now.

Have you talked to the air cooler vendor about this situation?

No matter how difficult it is to do, right now is the best time to correctly fix your problem. The window of opportunity is closing fast.

That's my 2 cents.