Buckled Internal Shell under Half-Pipe Jacket: FEA Options for FFS? 2

[TiCl4]

All,

A recent API 510 internal inspection of our pressure vessel has revealed some distressing information. The PV has a half-pipe jacket covering most of the sidewall. At some point in the past, internal pressure of the jacket exceeded MAWP by a fair margin, buckling the internal wall that is between the weldment points of the half-pipe. Not that it matters now, but I believe this was due to a PSV being installed backwards, something I noticed and had corrected several years ago. The normal process for this jacket is cooling tower water, and the inlets and outlets both close when cooling is not needed during a reaction phase. The trapped water then heats up, as it is still exposed to the hot reactor (150-200 F). Normally a thermal relief provides pressure protection, but this particular reactor was had the mis-installed thermal PRV.

Oh well. Here we are.

The "buckled" wall is uniform. Anywhere there is a half-pipe, the vessel wall is "popped in", forming a more "full pipe" jacket rather than half-jacket.

Anywhoo, the MI inspector is going to recommend the reactor be taken out of services as PV. I have asked for an FEA model of the reactor to determine if the failed state of the reactor can still retain the original MAWP and MAEP of the inner vessel and the MAWP of the jacket, essentially following an API 579 FFS review. I have read, but never seen substantiated, that half-pipe jackets provide stiffening to the shell of the vessel. ASME VIII is silent on this matter, meaning FEA must be done to prove this out.

Questions:

[ol 1]
[li]Has anyone seen FEA completed for a half-pipe jacketed reactor, specifically looking at MAEP and the benefit the jacket provides for stiffening?[/li]

[li]Looking at the picture below, does anyone have a gut feeling on what may result from an FEA analysis? The PV itself was modestly rated - approx 25 psig @ 400 F MAWP and F.V.[/li]

[li]Would this type of buckling reduce the vessel's MAWP?[/li]

[li]Unlike a full conventional jacket, which can crush a pressure vessel, it doesn't appear that the half-pipe is capable of fully collapsing a vessel. What would the failure mode be for excessive pressure in the jacket, then? The jacket alternately handles 60 psig CTW and 75 psig steam. It was originally rated at 150 psig @ 365 F.[/li]

[li]Following from the prior question, what is the main risk of continuing to run CTW or steam in the jacket? The pressure vessel normally operates at atmospheric pressure - the pressure rating is simply for emergency relief design.[/li]

[li]Non-engineering question: How screwed is this vessel?[/li]
[/ol]

 

[TGS4]

1. Yes. We've done that type of work, too.
2. Gut feel says its a solid "maybe". What's the access like to get a high-quality laser scanning of the vessel interior?
3. Probably not.
4. This is a non-catastrophic buckling situation (a snap-through buckling, if you will). My bigger concern will be cracking on the attachment weld of the half-pipe to the vessel shell on the outside.
5. Not sure. That may have to be a discussion with your jurisdiction having authority.
6. Depends. What's the replacement cost (including lost-production during the outage) the replace the vessel. That might tip the scales one way or the other regarding doing FFS.

 

[r6155]

Out of service NOW!!!

Regards

 

[TiCl4]

As an addendum, would UG-80 requirements for out-of-roundness come into play here? If a FEA analysis, as part of a API 579 / ASME FSS review, indicates the reactor could still retain its original MAWP and MAEP, but has deformations in excess of UG-80 requirements, would ASME VIII still disallow continued service as a pressure vessel? Or does an engineering analysis under FFS review supersede that provision?

Edit: A laser scan would be okay, TGS4. There are two sets of internal coils that would mean we'd have to take a lot of shots in the lower 1/3rd of the vessel to get a scan of the wall, but it would be doable.

 

[TGS4]

ASME Section VIII, Division 1 is for original construction only. So, UG-80 is no longer relevant.

 

[r6155]

Please, those who like FEA or some particular calculations, try to understand that this equipment is SCRAP !!

 

[TGS4]

r6155 - with all due respect, on what basis do you make that assertion? It has nothing to do with liking FEA or similar. I just think that an evaluation to determine fitness-for-service is warranted. That will include extensive inspection. I’ll trust the math, thank you very much.

 

[r6155]

@ TGS4
Do you think this equipment can continue to function?

 

[EngineerEPC]

Dear TiCl4,

Thats a lot of popping in. I thought it was internal jacket, then i read the post carefully.
Best wishes for next path

different topic ===

We had experience of using laser welded jackets on outer shells of brewery tanks having diameters in the range of 4-5 metres and shell thickness in the range of 6-8 mm. (MOC = SS304)
these laser welded jackets were pressurized to bring them to shape (around 80 barg pressure), we were worried about damage to main shell, however no damage.the laser welding of jacket shell to main shell results in penetration into the main shell also, giving stiffenes to it.

Regards

 

[Trestala]

Probably proof testing this vessel is an option (in addition to the FEA and FFS) if it is verified there are no cracks on the weld?

 

[r6155]

the MI inspector is going to recommend the reactor be taken out of services as PV

What I say is exactly the same as the inspector.

Regards

 

[FMJalink]

@TGS4,
I have to admit not to have experience with pressure vessel regulation in the USA, but your statement "ASME Section VIII, Division 1 is for original construction only. So, UG-80 is no longer relevant." seems a bit strange or just on a specific point in the discussion.

To my knowledge, compliance with ASME VIII div 1 has to be shown after modifications to an ASME div 1 stamped pressure vessel. The photographs in this thread indicate massive plastic deformation of the vessel shell at the half pipe jacket area. In axial direction surely the tolerances used for the full vacuum calculation will be exceeded. More on the positive pressure: The shell wall at half pipe jacket has plastic deformation and will therefor not have not the same material properties as had the virgin material as used with the design. Also there can now a lot of stress be expected at the half pipe weld to the vessel shell, which is different from the original design.

To me this appears to be more a modification to the pressure vessel than a deteoration or limited errors in the fabrication.

Investigation whether the pressure vessel can be continued could be undertaken. I do not have high hope there, looking at the photographs. Maybe the section with the half pipe coils can be replaced.
Kind regards

 

[TGS4]

@ TGS4
Do you think this equipment can continue to function?

I don't know. I haven't performed the necessary evaluation. I haven't seen any inspection records. I'm not willing to condemn (or approve) a vessel at first glance - that would be irresponsible of me as a professional.

It is up to the owner to determine whether the cost of such an evaluation including all of the necessary inspection) is worth performing, considering that there is a non-zero probability that the evaluation could definitively condemn the vessel. They would need to weigh the cost of such an evaluation vs the cost of a replacement, and the lost opportunity cost while the vessel is out-of-service, and the schedule of it all. Risk-reward calculations here.

If it were my vessel, and the replacement cost were < $100,000, then I'd junk it and replace it. That metallurgy looks expensive, though, so I have no idea about the replacement cost. Could even be a long-lead time metallurgy, which complicates the risk-reward calculation.

Taking it out of service pending an investigation is certainly the safe and proper approach at this time.

 

[TGS4]

FMJalink - when you speak of modification, are you referring to damage incurred in-service, or are you referring to purposeful modification, like a repair? If it is damage (of which the situation in this thread is a prime example), then the appropriate approach is to use a fitness-for-service approach, such as that described in API 579-1/ASME FFS-1. If it is repairs, then one would NBIC. This certainly not an error in fabrication, which would, in fact, be covered by the Code of Construction.

You are correct that the material has suffered from a change from the original plate material - it has undergone significant plastic deformation, which has the possibility of leading to cracking. Hence my recommendation to perform a fitness-for-service evaluation, which will necessarily include a large amount of inspection, including surface examination looking for cracks.

 

[FMJalink]

TGS4 - Thanks for the explanation. Appears there is some difference in how things are to be handled in the USA and over here in the EC, though in basis very simular. I did mean a damage as here applicable. Most likely a NoBo would that not see as a normally possible damage, but as a change/modification to the pressure vessel, though unwillingly.
Here also a fit for service check is performed, often using the API Std 579, but normally the design code stays the basis and fitness for service investigation is basis for the "temporary" solution with monitoring of the situation.

 

[GD2]

I am looking from a different perspective. The reactor has now turned into a corrugated vessel. The material strength of the side wall has increased and can take more compressive loads.
However, I will be concerned with the pressure loss inside the reactor and the longitudinal stiffness. If the pressure drop is acceptable, longitudinal bars can be attached to take the longitudinal forces.
I would go for a FEA with the reactor new shell configuration.