Mehanical Clamp repair design for a leaking pipe....Asme viii or B31.3 1

[psotil]

I am looking into the design a mechanical repair sealing clamp for a leaking pipe.
The PCC-2 refers to the original construction code (b31.3) for design parameters. But the data in b31.3 refers to seamless pipe and doesn't account for the fact that the clamp is:

- made up of 2 halves of a pipe
- each half has 2 lugs welded on to it
- bolted together and not welded
- sealant is used to close any gaps between the pipe and the clamp

I have been able to find answers re the bolts and lugs in asme viii bpv div 1. But I haven't found any clarification the required shell thickness.

Please help?

 

[LittleInch]

This usually comes from the vendor, if anyone actually asks for it. Mostly they are fairly basic calculations based on a whole pipe and then looking a the forces on the bolts.

These things tend not to be designed per se, but qualified by experience in use and performance testing to get a pressure rating.

The part that can't be designed is your "sealant is used to close any gaps between the pipe and the clamp "

There have been some FEA analysis done on grouted shells to prove their effectiveness in re-inforcing inner pipe defects, but most of those are not rated or can be used for through wall defects.

A clamp should only be a temporary fix until you get it repaired properly, but needs to be a suitable rating. Buy one from a reputable supplier and don't try and make one yourself.... Someone like this

http://www.plidco.com/product-line/clamp-sleeve

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[jte]

Well, it depends... Obviously there are many vendors who sell off the shelf clamps. But there are also custom designed clamps, for whatever reason I tend to be more involved with those than with off the shelf clamps. There can easily be circumstances with plant piping where an off the shelf clamp made for straight run or even a single elbow just don't work for the piping layout in question.

For a B31.3 application, I'd tend to use Section VIII Div. 1 design. Two reasons: First, I'm a vessel guy before I'm a piping guy. Second, VIII-1 has a higher design margin against ultimate tensile and usually provides more thorough design guidance for some of the details. Most of the time I don't find myself actually designing mechanical clamps (PCC-2 Part 3 stuff), more often I'm reviewing the work of a leak seal contractor's engineering arm.

As LittleInch points out above, FEA can be used (in which case I'd say VIII-2 Part 5 methods with VIII-1 design margins), and the sealant itself is difficult to engineer. I've never been provided the actual "recipe" for the sealant by a leak seal contractor. Strictly speaking, PCC-2 places no limit on a mechanical repair in terms of time. Jurisdictional issues may govern. But my experience with bolted boxes leads me to agree again with LittleInch: Don't depend on a bolted clamp to be reliable for the long term, particularly if thermal cycles are involved. The hardest ones are on steam condensate return lines with intermittent temperature spikes.

One final thought: Don't get too "enthusiastic" with the quantity and injection pressure of the sealant. If you calculate a volume of 10 liters and have injected 15, you have probably collapsed the carrier pipe.

 

[psotil]

Many thanks LittleInch and jte.

I am looking into the maths side of this and the practical side. I want to be able to confidently say that A standard is more appropriate than B standard. Like you Little inch I am a part of a team that reviews a contractors work and so I have been assigned this task by my senior engineers as I am new and need to familiarise myself with this area.

I have read in the post construction code (pcc-2) for repairs on piping refers to the original construction code to determine the design parameters but I just don't agree that the assumptions are accurate in the original construction code b31.3 for a split clamp that is bolted together around a leaking section of pipe.

By installing a clamp, it is in effect containing the line contents that is leaking from the pipe and therefore it is better classified a pressure vessel in my view. But I have yet to see anything written in any standard to support this.

any ideas?

 

[jte]

psotil-

Out of curiosity, where are you located? Your profile (click on your user id) shows a location code that I'm not familiar with.

Regarding PCC-2, there is no such thing as the "Post Construction Code". The document is simply the second one issued by ASME's Post Construction Committee, it is a standard, not a code.

As a young (sorry, "less experienced" to keep the HR folks happy!) engineer I was focused on vessels. Thus I have a bias towards Section VIII. Note your logic (which I've used many times in the past), specifically the notion that a pipe transports, a vessel contains a fluid. This leak box is there to contain a leak, not to transport. Then, consider that Section VIII Div. 1 allowable are typically lower and the fact that U-2(g) is pretty much where you'll wind up regardless of whether you say you are using B31.3 or Section VIII-1. End result: I've never had anybody tell me that using a Section VIII Div. 1 approach is not safe.

Neither VIII nor B31 will explicitly cover this, don't bother looking. As you said, PCC-2 will in most cases refer to the original code of construction (if one can be identified). But by meeting VIII-1 you inherently meet B31.3. With very little extra cost in terms of materials or fabrication. It's hard to "go wrong" with a VIII-1 design.

 

[psotil]

Thanks jte, I have been asked not to reveal any info that could potentially reveal my company's identity. Apologies, my manager was very clear.

I agree with you about using asme viii over b31.3, I was looking for a well justified explanation which I was hoping to find in the standards but that didn't help.

This was something an end customer had come back to my team with in the past. Where they were insisting that the PCC-2 be adhered to and hence wanted the shell thickness calculation from b31.3 to be used. Without telling me how this was actually dealt with, my manager had asked that I determine the best way to respond in this situation.

Many thanks for your time

 

[jte]

psotil-

No worries about the location, just an odd code ("EN, US") which to me is more of a language code rather than a location code. For what its worth, my location as shown in my profile is accurate, though it is well known that the location algorithm is not nearly 100%. And I understand about company confidentiality and don't go out of my way to reveal mine. I understand those who have reasons for identifying themselves, and that's fine as well.

I'll be perhaps a bit repetitive: If I were in your shoes, I'd provide the client calc's which indicate that the design complies with B31.3 stress values. It's just that the calculated stresses are also compliant with VIII-1 limits. Unless there is some weight or size issue (and this sometimes happens), then everybody is satisfied.

If you are providing services to my company (and for this purpose, my location code is entirely irrelevant), and if the issue is elevated to my attention, then I think you know what the guidance would be!

For what its worth, I don't envision any "clarification" on the issue within PCC-2. The PCC Subcommittee for Repair and Testing (which is responsible for PCC-2) is not discussing any changes of this nature. If such a request were to come up, I think it would be shot down. If a request for interpretation were to be submitted, I suspect that the main focus of conversation would be how to reword the inquiry such that a "yes" or "no" answer could be provided - if it isn't seen as "consulting" in which case a standard response to the effect of "The committee does not provide consulting services" would be issued.