Welding Supports to Pipelines 2

[cingold]

I've been working in pipe stress analysis of pipelines, compressor stations, and pump stations for about 2 years now. A recurring issue that I've run into in this time is welding supports to pipe under the pipeline codes.

Every designer and engineer that's been in pipeline for a while refuses to use anything but clamp-on pipe shoes (or other types of clamp-on supports). I understand that in most situations the codes do not allow welding supports directly to the pipe. I believe Section 834.5 of ASME B31.8 covers this for gas lines.

What I don't understand is when I need to design a line stop into the system to reduce thermal growth. In these situations, a clamp-on pipe shoe typically does not have enough clamping force to keep the pipe from moving through the clamp.

Section 834.5 of ASME B31.8 allows welding a pad to the pipe and then welding a shoe to the pad (I could then add tabs to the steel below and achieve my line stop).

I'm looking for some info/advice from the guys that have been in pipeline for a while. Is there a reason that we should not weld this pad onto the pipe and then a shoe onto the pad to achieve my line stop?

I realize I'm horrible at explaining my meaning, so please let me know if you need some clarity on my question.

Thanks,

cingold

 

[BigInch]

For pipelines operating at a hoop stress level above 20% SMYS, you can weld ONLY fully encircled members (sleeves) directly onto pipe. You can weld whatever you want onto the full encirclement sleeve, provided the pipe below is never becomes overstressed. So weld your guides and stops to the fully encircling sleeves. Weld the sleeves to the pipe using fully encircling welds and NDT all welds attaching directly to welds on the pipe.

From B31.8

Clamps and supports shall be designed such that a smooth transfer of loads is made from the pipeline or riser to the supporting structure without highly localized stresses due to stress concentrations. When members are to be welded to the pipe they shall fully encircle the pipe and be welded to the pipe by a full encirclement weld. The support shall be attached to the encircling member and not the pipe. All welds to the pipe shall be nondestructively tested. Clamps and supports shall be designed in accordance with the requirements of API RP 2A-WSD, Section 3.
Clamp and support design shall consider the corrosive effects of moisture retaining gaps and crevices and galvanically dissimilar metals.

I hate Windowz 8!!!!

 

[LittleInch]

BI is correct, however the section quoted is from A842.2.7 which refers to offshore structures and risers.

834.5 allows supports, anchors etc to be directly welded if the hoop stress at design conditions is less than 50% of SMYS. Above that you need to weld on full encirclement sleeve.

In answer to the OPs question - you can use pads etc direct if the design pressure / hoop stress is low enough.

However, pipeline designers and pipeline codes work generally to minimise wall thickness and hence adding point loads and stresses is not something that is generally acceptable and considered as poor design. Allowing the probably quite high forces associated with a line stop to be attached direct to the pipe could overstress the pipe much easier than if it was designed to a piping code which usually results in thicker pipe more able to withstand these sorts of forces.

You probably also need to think about bending moments if the line stop is not applied equally to the pipe (top and bottom or two sides)


My motto: Learn something new every day

Also: There's usually a good reason why everyone does it that way

 

[cingold]

I agree with both of you. By the way, do you two know each other?

This seems to be an argument I have at least once a month. Maybe I'm not communicating it well enough. If the code allows this, and if my calculations say we won't overstress the pipe then why is welding to the pipe (or to a pad on the pipe) rejected?

I guess I was just curious if there were some reason that's not easy to calculate or directly addressed in any of the codes.

Thanks for the help!

 

[LittleInch]

Yes we do.

Why is it rejected - because pipeline engineers don't like it that way. Most thinking pipeline engineers know that their pipe is relatively thin compared to piping and structural members and are just not comfortable with direct welding of anchors and supports onto the pipe. There have been too many occasions when the construction hasn't matched the design and the pipe ends up being bent or punctured and for a simple bit of extra welding of a full encirclement sleeve it's not seen as worth the risk.

My motto: Learn something new every day

Also: There's usually a good reason why everyone does it that way

 

[BigInch]

Stress concentrations (internal corrosion?) are the biggest reasons not to. It doesn't take much to concentrate stresses in thin wall pipe. It's usually concentrated enough without adding to it. It's usually just about as easy to make a full encirclement as it is to make a 1/2 pad.

P.S. Blood brothers.

I hate Windowz 8!!!!

 

[BigInch]

Yes onshore is 50%.
Thanks LittleInch.
Can you tell that I'm one that uses full encirclements at all directly welded supports.

Note: Beware of crevice corrosion when using clamps.

I hate Windowz 8!!!!

 

[cingold]

LittleInch, to answer your question, that's why I'm here

More experienced pipeline engineers and/or designers tell me don't weld on the pipe (pad or no pad). But there's never a reason why we shouldn't weld on the pipe.

One of the more difficult aspects of pipe stress that I'm learning about is knowing what to consider that's not directly covered in the codes or in programs like CAESAR II.

 

[LittleInch]

Unfortunately it is common for people to say do this or don't do this "because that's what the code says", without ever having read the actual code sections themselves.... but because "old Joe" said so back in 1988.

Being able to quote code section numbers and texts will often stump those sorts of comments. I can understand the reluctance to weld pads, but full sleeves should pass and to say otherwise is simply not being conversant with the code wording and just people knowing ther eis an issue about welding without knowing the particular details of what is and isn't permitted.

That is of course before we get into company codes / specifications / directives which often get much more rigorous for reasons people have forgotton about, but trying to get it changed is next to impossible.

My motto: Learn something new every day

Also: There's usually a good reason why everyone does it that way

 

[weldstan]

LittleInch,
You're being too kind!! How about, "That's the way we did it the last time." Over and over again. Or "old Joe's predecessor said it in '68?

 

[BigInch]

Then again, if the method's been working fine for 46 years, I'd have to think long and hard before changing it. I'd probably need a percentage of the savings. After all is said and done

There's usually a good reason why everyone does it that way

... LittleInch
What makes me go ballistic is when somebody that's got no idea why Joe did it that way in '68 changes the company spec to something that isn't even understandable in the language in which it's written.

I hate Windowz 8!!!!

 

[LittleInch]

Yes there is usually a good reason, it is just that sometimes that reason is no longer applicable or wasn't actually correct in the first place.

We all have our pet little issues which annoy us when people do things just because that was what they and their predecessors have done before (slopes on pig traps is one of mine) or someone introduced (anchor blocks e.g.) and then they became adopted as the default mode. The no welding of supports in any instance that the OP refers to is probably one of those, though I would still prefer to see a full encirclement sleeve for any significant force being applied to the pipeline regardless of what the code says, which is after all a code, not legislation, and hence is the minimum action you need to do to design a safe system. Lots of people forget that too... Nothing should take away the responsibility of the engineer to design something better / stronger / safer as he or she sees fit.

My motto: Learn something new every day

Also: There's usually a good reason why everyone does it that way

 

[DSB123]

cingold,
Remember Caesar does not apply Stress Intensification Factors at Supports unless you actively apply them yourself. That's an aspect a lot of Pipe Stress Engineers miss especially when considering Fatigue of piping systems.

 

[BigInch]

What? A software package that does not know everything, do everything, watch everything, design everything, recognize it's eating garbage in?

OK, let's add that one to the list.

I hate Windowz 8!!!!

 

[DSB123]

BigInch,
You may be amazed but the number of Pipe Stress Guys who do not realise "local" stresses and SIF's are not automatically included are out there!!!! Scary!!!!

 

[SNORGY]

It's all "design by Xerox".
What did we do before Xerox?
Etch-A-Sketch, I suppose. A bit archaic, but it could produce some good designs, all of which later proved to be inherently difficult to photocopy.

 

[BigInch]

Oh I've changed a few title blocks on original vellum.
Identical meter systems, corrosion inhibitor injection systems, riser designs, subsea tie-ins, compressor installations, but those were pretty much, if not exactly, identical. Never used an offshore tie-in for an onshore tie-in though. Saw one recently for an onshore manifold design for use in a 20 well pad field gathering system. Walkways were perfectly arranged for above ground piping, so there was apparently nothing else they could do except to build the pipe above ground too. If the pipe was located below ground, with the 20 well pads involved, at least 1000 feet of walkways, crossover stairs, operating platforms and untold number of pipe supports could have been, and were, completely eliminated.

I hate Windowz 8!!!!