PD5500 - Pressure Vessel Repair Proposal

[barrind]

We have a vessel which is corroding internally. Apart from the general overall loss in wall thickness there are a few quite small areas were deeper pits have formed, and have been mapped with a corrosion mapping technique. In the past one or two of these have developed through wall thickness cracks resulting in a leak, which have been repaired with an insert patch. With a view that the other pots will eventually leak in the same way, we are looking at a possible temporary repair method allow the vessel to continue operate whilst a full insert patch repair can be planned.

The method we are looking at is essentially fitting a branch over the pinhole leak when it occurs, with a blank flange, to effectively seal off the leak. I.e. the branch would become pressurised.

I am trying to assess this using section 3.5.4.1 in PD 5500, assuming it will act as a set-on nozzle. There will be no reinforcement of the shell or the branch so I need to confirm the thickness of the shell as is, and the 4" sched 40 pipe we propose to use as the "nozzle" are sufficient according to the code.

Since there is no actual opening in the vessel, as it will only be a pinhole leak, I am asking how would the d, defined as "the mean dia of the nozzle, or the bore of the opening in the shell" be used in the calculation?

Section 3.5.4.3.3 which details the procedure to arrive at a required thickness of the nozzle uses a factor "rho" to influence this. This equation for rho uses "d". However, since there is no actual opening in the shell, "d" can either be the mean dia of the nozzle, or effectively zero, given the definition for d. One will give rho=0 and the other not. A rho of <0.1 is defined as the SHELL not needing any reinforcement, which is good, but I need to show the unreinforced pipe is also ok. Note, the shell is 11mm thick in its corroded condition, and sched 40 pipe is 6.02mm, design pressure is 2.1 bar.

Before I go any further, can anyone offer any advice on how to interpret the code given the type of nozzle I am trying to assess? Am I barking up the wrong tree here, since there is no opening in the shell, does this even need assessing as a nozzle, given the 4" branch itself only needs to be 0.7mm thick to withstand the design pressure on its own?

 

[jte]

Hi barrind-

The type of repair which you are describing is something I've done several times. When I describe it to people who are not familiar with it I tell them to think of it as a hot tap but without cutting the hole. I am not familiar with PD 5500; I'm focused on ASME. For a similar circumstance with a Section VIII vessel, I'd chat with material engineers about what the short term corrosion rates are and see if I can conservatively take credit for any remaining thickness under the nozzle after the pinhole forms (don't forget "external" corrosion on the shell after break-thru).

I've actually chatted with some Section VIII committee members about an interpretation which would address this type of issue but haven't followed through. But my driver for being interested in this is completely removed from post-construction activities.

I'll throw out a couple other ideas which you may find more appealing, depending on the depth of the pit or more precisely, the remaining wall thickness at the pit. First, you can perform an external weld buildup. Since you are already talking about the "hot tap minus the tap" repair, you are clearly comfortable with welding to the exterior, so like I said, it comes down to thickness at the pit. Take a look at ASME PCC-2 Article 2.2 for some ideas.

The second idea was one I first heard in the late 1990's from a now famous personality in the vessel and piping engineering community: Before doing the external weld buildup as above, lay down a pass or two of corrosion resistant alloy to stop the corrosion. Then do the external weld buildup that you'll take credit for when running calc's.

 

[metengr]

The above approach is identical to a welded leak box repair but instead you are using a nozzle design to effectively create another pressure boundary. The ASME PCC welded leak box has come up at NBIC meetings because this is obviously a method for in-service vessels only and not new construction, which is the scope of ASME B&PV Code, Section VIII.

I believe this method, if recognized in the NBIC at some point, would be an alteration versus repair because of design calculations that would be required to ensure safety.

 

[barrind]

Thank you for your suggestions. Yes, PD5500 is also intended for the design of vessels rather than checking in service modifications. The weld overlay idea sounds interesting as a preventative measure. I guess you would not recommend this as a course of action after a leak has occurred, since there would be a through thickness crack, which I guess would have to be fully removed prior to building back up with weld, making it potentially a more arduous task compared to placing a nozzle over it.

jte, you say you have done this type repair yourself, do you mean you backed this with with ASME VIII code? Can you point me to the relevant section as I do have a copy.

In the meatime, since I can't use the PD5500 design code to fully back the type of nozzle I plan to use in my original Idea, I am going to do an FE model of it this week and assess it that way.

 

[r6155]

See ASME PCC-2 Article 2.12 Fillet weld patches (my edit is 2008).
Regards
r6155

 

[metengr]

No, this would be a welded leak box and not a fillet welded patch. The OP inquired about using a set on nozzle with a blind flange, aka welded leak box.

 

[r6155]

@ metengr
I know the difference, thanks. Fillet weld patch is my proposal only, I consider better design than a welded leak box because it´s easy and calculations are simple. My opinion only.
Regards
r6155

 

[jte]

barrind-

The VII-1 calc's aren't particularly interesting - the only issue is how much available area in the shell inside the new nozzle do you take credit for.

If you look up a PCC-2, be sure to get the 2015 edition. The 2008 has seen more than a few improvements.