Branch connections practices 2

[jalvarez]

ASME B31.3, paragraph 304.3.2 establish "that a branch connection has adequate strength to sustain the internal and external pressure which will be applied to it if:"
And among the "ifs" the point (b) mentions that the size of the branch must not exceed one-fourth the nominal size of the run.
So, any connection that does not comply with this (for example a branch made with a half coupling of ¾" on a 2" pipe must be checked/calculated. The same for a 1" half coupling on a 2" or 3" pipe, and so.
Anyway, I'm seeing that normal practice of many Engineering companies is to use these kind of branch connections without any specific mention or calculation.
Is this an "accepted practice", or each Engineering company have the support of these calculations in their own records, or simply nobody pay attention on this issue?
Regards, sorry if this is confusing

 

[BigInch]

Probably just experience, or they are using a line spec where all the listed pipe, fittings and valves have been checked already for the pressure and temperature class the pipe spec covers. A lot of line specifications are set to use a given schedule for all similar service lines less than X psi and YºF, so small diameter takeoffs probably are good for even a higher pressure than the main.

Going the Big Inch!

http://virtualpipeline.spaces.msn.com

 

[JohnBreen]

Hello jalvarez,

You have not quite correctly quoted B31.3 (however, I do understand your point). Within the limits mandated by B31.3, paragraph 304.3.2, the cited couplings and half couplings provide enough "surplus material" within the zone of reinforcement such that these branch connections provide a adequate degree of structural integrity for pressure design. If the branch-to-run ratio exceeds 0.25, the hole in the run pipe is too great a portion of the circle to be reinforced by a coupling or half coupling.

But just to be complete and correct,

B31.3, paragraph 304.3.2:

.......It may be assumed without calculation that a branch connection has adequate pressure strength to sustain the internal and external pressure which will be applied to it if:

B31.3, paragraph 304.3.2(b):

"The branch connection is made by welding a threaded or socket welding coupling or half coupling directly to the run in accordance with para. 328.5.4, provided the size of the branch does not exceed DN 50 (NPS 2) nor one-fourth the nominal size of the run." (the paragraph continues).

and later in the same paragraph:

"In no case shall a coupling or half coupling have a rating less than Class 2000 per ASME B16.11".

The intent of the Code is clear. It limits these branch connections to NPS 2 and smaller and it limits the possible run to branch ratios. Branch connections that do not comply with these limitations must be qualified per paragraph 304.7.2.

Every piping engineer who designs a piping system is responsible individually to the jurisdiction for the safety of that design (regardless of any company's "standard practices" or "design procedures"). The design firm also is equally responsible to the jurisdiction. The jurisdiction typically relies upon the local building code to specify the minimum requirements for design. The local building codes typically refer to one of the B31 Pressure Piping Codes as their minimum requirement. That mandate gives the referenced B31 Code the "force of law". If a piping design does not comply with the requirements of the local building code, it is unlawful to install that piping system and place it into service within the jurisdiction. In some States and Provinces, that is part of the criminal code and the responsible parties will not only face the cancellation of their engineering license, they also face a jail term.

The B31 Pressure Piping Codes are "living documents". Regrettably, some companies' "standards" do not stay abreast of the changes or developments of the piping codes (we have all seen references in some company's "standards" to the now defunct "ASA" national standards).

In summary, if the design of the piping system does not comply with the building code mandated minimum requirements for design, fabrication, erection, examination and testing, IN EVERY REGARD, it is unlawful to install it and place it into service within the jurisdiction regulated by that building code.

Of course, the stress issues associated with branch connections must separately be addressed.

Regards, John.

 

[BigInch]

John, slow down.

Its not exactly "(regardless of any company's "standard practices" or "design procedures")." There are areas of the world where the prime jurisdictional authority is the company... and their standards. Sometimes the standards reference B31.3, 4 or 8, but many times they do not even do that. On many projects there are some specifications that openly deviate from B31.3, 4, 8, DNV, BS, NO, ISO and many other codes, and yet other procedures on the same project that will combine the most conservative concepts and clauses of all several of the codes. One prime example is the Saudi Arabian pipeline design code that take safety factors from B31.8 and apply them to both gas and liquid pipelines. Quite a difference. In another example of a major South American project we reviewed all the worlds standards and basically took the least conservative of all of them. (Its a big world.)

Going the Big Inch!

http://virtualpipeline.spaces.msn.com

 

[JohnBreen]

Hi BigInch,
Your point is well made and I thank you for that. Really, I did not mean to make my posting a "rant" but rather just to state the reasons for the rules as unambiguously as was appropriate. But of course I made my point with my "North America" blinders on and with no regard to how things are done in other areas (albeit I have seen several water systems in Angola comprising wooden pipe). My concern is focused upon unsafe pressure systems that are being built in areas where Codes and Standards ARE mandated and without proper regard to the rules that have resulted from lessons learned in pressure technology (and I guess my work on Code Committees might also influence me).

It is a pity really that all people, everywhere in the world cannot benefit from today's technology but as you point out in some areas, "it is the way it is". One thing that worries me about that is there is so much variance between one set of company standards and those of any other company that there will be no consistency in design. And, to some degree, the cheaper you can make it the lower you can bid.

So I guess the answer to jalvarez's original question ("is this an accepted practice?") is "yes and no". It depends upon the jurisdiction in which the piping system is to be built, but definitely not acceptable in places where B31.3 is mandated. If the question is "will it comply to B31.3?", the answer is "no". If the question is "will it be safe?" the answer is "the rules include a factor of safety of 3 for most applications (but for good reasons)". But getting back to jalvarez's question again, if a company is concerned about documenting calculations (i.e., covering their liabilities), then it would be prudent to design to the letter of an internationally accepted design Code.

Thanks again, BigInch.

Regards, John.

 

[jalvarez]

JohnBreen
I think that I quoted correctly the code, in the necessary extent to be understood. As the discussion went in the direction I liked, fine.
One step ahead:
The usage of half couplings as branch connections is covered by 328.5.4 (d) and the Figure 328.5.4D, sketches (1) and (2).
In any of these two cases I think that could be convenient to have a good fitting between the half coupling and the run pipe, and this could be done machining adequately the half coupling, or using an adequate weldolet, or sockolet or whateverlet. What is the common practice in the US?
...
And talking about codes, in my country you must design following "an internationally recognized code"
What is our practice, as owners? Being specific with the code to be used. But it always happen that the companies have their own "shortcuts", and this is the reason for my original question.
BigInch, to search for the most permissive code... sounds bad, frankly speaking, despite the fact about how is this world.
Best
J

 

[BigInch]

No worries. It was an excellent rant.

You made many valid points and I understand why. I don't enjoy digging in 25 codes to find the "best" way to do things and I would be happy to see the universal adoptation of, for example the ISO 13623 pipeline standard, or any other for that matter,. Whenever a new one comes out, it seems to be just another standard that can be confused, used and abused, or replaced by another one from time to time.

I called on it just based on the slight chance that he's south of the border.

Going the Big Inch!

http://virtualpipeline.spaces.msn.com

 

[NozzleTwister]

jalvarez,

You asked what normal practice is at many engineering companies. What's normal will change from industry to industry and from what was done in the old days compared to now. Most of my experience has been in the engineering of refinery piping and here's what's normal for me.

1. All piping 2" NPS and below will use forged socket-weld or threaded fittings. Where the run is 2" NPS or below the branch will always be a forged tee. I haven't seen half-couplings used except they sometimes show up on off-the-shelf equipment or components.

2. Where the run pipe is 3" NPS or greater, branches 2" NPS and below will be sock-o-lets or thread-o-lets. Full size branches will be welding tee fittings. For reducing branches, calculations are performed and where additional reinforcement is required a re-pad or weld-o-let is specified. Where economics is a concern, weld-o-lets are used for branches up to about 6" or 8" NPS and re-pads used for larger branches. Some clients like all reinforcement to be re-pads and a few others all weld-o-lets. I've seen a 24"x20" weld-o-let, but man that's a chunk of steel and a lot of welding.

3. Of course where no additional reinforcement is required for reducing branches with the branch greater than 2" NPS, pipe on pipe stub-ins are fine for pressure. Your Piping Engineer may request some reinforcement to reduce excessive thermal stresses at the intersection.

4. Also, for robustness, piping 2" NPS and below is normally sch. 80 minimum and 1/2" & 3/4" nipples being schedule 160. Also it's common to see branch piping below 2" to be sch. 160 to the first block valve.

These are what have been normal engineering practices for me the last few years but I know other industries probably do something a little different.


NozzleTwister
Houston, Texas