Girth Flange Gasket Leaks due to Vessel Lifting or Handling 4

[VFK]

An existing BEU type heat exchanger, 1200 mm ID (~47") and weighing 10,000 kg (~20K pounds), needs to be removed from an offshore oil rig and brought to a workshop on land so that the tube bundle, weighing 5700 kg and 2000mm long, can be replaced. The reason this scope cannot be performed on the oil rig is the bundle puller, supplied to the client at project handover for performing mechanical handling of the tube bundle offshore, is in disrepair.
* The shellside is CS 150# and service is low pressure steam and the tubeside is DSS 600# and hydrocarbon condensate service.
* The tubesheet has kammprofile gaskets (1250 mm OD, 15mm wide, SS316 4mm core with 0.5mm graphite both sides) on the shellside and tubeside and sit in recesses between the tubesheet and girth flanges.
* The bonnet studs are B7's, 28 off, 2 1/4" x 1000mm long plus 8 off collar studs of the same diameter. They will be tension-tightened and the target bolt assembly stress will be limited to 70% of bolt stress yield but the final figure will be assessed to PCC1 with Sfmax derived per the WRC-538 method. The OEM under-cooked the original bolt target stresses and the client dervied new loads with the help of a flange analysis expert.
* The plan is to hydrotest the new tube bundle with the bonnet installed with the production gaskets and production bolting. The workshop is proposing to not remove the bonnet post-hydro (there are nozzles to drain/dry/inspect the bonnet). We would then perform a leak test on nitrogen to prove the tubesheet gaskets are not leaking prior to releasing the heater from the workshop. We are not anticipating over-stressing of the bolting, hence why we intend to use the production bolts.
* The heater's lifting points are a pair of rated lifting trunnions on both the shell and on the bonnet and the centre of gravity is at the tubesheet. So it is inevitable for the bonnet to see lifting loads.
* The heater might be lifted at least five times after workshop release and will be shipped to the facility. Best case, we could get it down to two lifts by transporting it in a container.
Before I speak to an engineering consultancy, I would like to refine my understanding:

1. Is it best practice to replace girth flange gaskets post-hydro? I'm aware typically gaskets experience not-insignificant relaxation during the hydrotest. Other than corrosion/contamination issues, are there any other reasons to replace the gaskets? We intend to request the N2 leak test @ 90% of design pressure post-hydro as a control to prove the gasket is tight. I'm anticipating many will be reluctant to disassemble and reassemble the bonnet due to the large qty, size and tensioning requirements for the bolts.

2. If the workshop leak test is a success, is it plausible the gaskets may leak after the heater is installed due to loads experienced during handling/transportation? Is it credible the root cause for the leak could be the girth flange gasket(s) being "disturbed" or seeing localised crushing due to the longitudinal bending moment imposed on the bonnet during lifting (200 kNm [5 tonne x 2 x 2000 mm]) and/or due to transport acceleration loads?

3. Hypothetical scenario: after the heater has been installed the tubeside gasket leaks on the final N2 test in commissioning; troubleshooting the leak is unsuccessful and the decision is made to replace the tubeside gasket; the shellside gasket did not leak. In this case it would be preferable to remove the bonnet but keep the tube bundle in-situ using the 8 off collar bolts, because there is no reliable means of pulling the bundle offshore. Is it plausible that removing 28 of the 36 bolts could cause the shellside gasket to be "disturbed" and may introduce a new risk of the shellside gasket leaking after the bonnet is reassembled?

Sorry for the verbose post, but the devil is in the detail...

 

[VFK]

@BJI and Christine - thank you for the feedback.

Posting back: resolving the lifting moment into an equivalent pressure: 3% of the channelside design pressure, and 60% of the shellside design pressure.
This results in a relative change in gasket stress of 2 for the channelside and 0.02 for the shellside.

In theory the channelside should see no issue from the lift and the shellside we should monitor, but this offers little comfort, because in doing these checks we've revealed some deficiencies in the OEMs target bolt stress method and flange design. Posting for your reading interest only, the OEM's target bolt stress (60% of bolt yield strength) imposes an achieved gasket seating stress that exceeds the kammprofile gasket max allowable stress in API 660 by 50%. Hence the channel and shell gaskets both become more vulnerable to damage during the lift.

Unsurprisingly, our PCC1 calc indicates the target bolt stress should be controlled by Sbmin which is only 40% of the bolt yield strength. Even with this load the shellside gasket sees 20% over the API 660 max allowable gasket stress.

This equipment has laid idle since commissioning so we have little operational feedback on the performance of the flanges. Suffice to say we have some reviews and decisions to make going into this next phase of the equipments service. One option may be to install a slightly wider gasket... In the meantime we'll do some sensitivity checks about assessing the bolt relaxation potential risk at the operating temperature as well given the flange is lagged.

As far as the lifting is concerned, after the hydro we'll retension, lift the heater up/down and then perform the leak test. We'll transport in a container/lifting frame. And offshore perform a leak test asap and have a contingency plan in place to pull the bundle as a plan C. On the plus side the operator has a sound work instruction for reassembling the bonnet to ensure the collar bolts are adequately tensioned.

Thanks to everyone for their help. And for future posts I'll be sure to provide a dwg in my original post.

 

[BJI]

I was suggesting you compare it to the bolt loads, which will be the same for both sides of the joint (unless the TS is threaded). So you are talking about a negligible percentage of the bolt loads for the joint as a whole.

The API 660 maximum gasket stress is relatively low at 40 ksi, I would not be concerned about an increase of up to 50%. You could be making the joint integrity worse by increasing the gasket width, if you don't know why it was selected. Maybe it was an oversight by the OEM, but you did mention a bolting expert was involved at some stage.

There are a number of reasons why bolted joints shouldn't be insulated, bolt relaxation could be one, if you are in the creep range, but likely more of a concern is the decrease in clamping force do to the increased bolt temperatures. This will lead to a decrease in gasket stress, and my give an indication as to why the assembly gasket stress is high.

Good to see you are doing a detailed assessment and considering all possibilities.

 

[GD2]

VFK,
As long as you make the packing sea worthy to restrain from all rocking, tilting, sliding and impacts, you should be fine to transport a fully tested HX.

GDD
Canada