Tie-in on expansion loop

[SWMechanical]

I am doing a piping design. One of the branchs has to come off from the expansion loop of the header. Does anybody know if it is good to have a tie-in on the expansion loop?

Thanks,

 

[BigInch]

NOT! The line is probably moving the most that it ever does at an expansion loop. Also usually full of bending stresses.



"If everything seems under control, you're just not moving fast enough."
- Mario Andretti- When asked about transient hydraulics

http://virtualpipeline.spaces.msn.com

 

[MJCronin]

SW...

Why must the new branch line come of from the expansion loop... Are there not better places to make the branch (near an anchor or termination ??

Did an MBA tell you this..????

The proper way to evaluate a new, significantly sized branch line is to re-perform the stress analysis with the new piping included in the model.

-MJC

 

[JohnBreen]

SWMechanical

The gentlemen who posted above have expertly explained the problem. It would, for the reasons they cite, be very BAD PRACTICE to put a branch connection in the expansion loop - DO NOT DO IT!!!

Regards, John

 

[pennpiper]

I vote with John Breen,
DO NOT MAKE THE TIE-IN ON A LOOP!!!!

 

[worrell]

What if there is no choice but to add the line to the expansion loop due to interference with other lines or other weird spatial problems like that? There are only so many things that we can control (and only so much $ in the budget). What do you do then?

 

[JohnBreen]

worrell

A BAD engineering practice is a BAD engineering practice for a reason. You have an obligation to the owner. The owner has loaned you his trust to give him an acceptable job - that is what he is paying an engineer to do.

So, you go with your supervisor and sit down with the owner and explain what his options are and give him an accurate assessment of the consequences of putting in a branch connection that will be a fatigue issue. This is NOT a decision that you make for the owner. If the owner is truthfully aware of the situation and still insists that you do something that is bad practice you document it (send the owner the minutes of the meeting with your fatigue calculations and you recommendations against the bad practice). You mail a copy of the minutes to yourself to get the postal date on it and you put it in your file. Then you do the best you can to ameliorate the condition by conservative design and you have very thorough NDE of the welds done. Then you make sure not to walk by that branch connection when the system is in service (you don't owe anyone your life).

Then you get on with your life.

John.

 

[rconner]

Not professing any expertise in such an issue, but just curious (at least if the tie in must be here, and at least if the line could be shut down for awhile) if it might be possible to replace the loop with some sort of straqight or in-line expansion joint(s), then make the tie-in close to where they want it though in a straight piping section that is somehow insulated (by new anchorages etc.) from subsequent "bending (or other) stresses" or loop movements?

 

[BigInch]

John,

I don't see how one could continue with a design, if you know its bad practice, even if you have advised the owner of the possible consequences and have sent a registered letter to yourself. Actually I think it would be illegal for a PE to knowingly continue with such work, and/or sign & seal it if s/he thought it was so. Isn't it better to draw the line in the sand in those kind of situations?

"If everything seems under control, you're just not moving fast enough."
- Mario Andretti- When asked about transient hydraulics

http://virtualpipeline.spaces.msn.com

 

[SWMechanical]

Thank you all for your help.

What if the tie-in is close to the loop?

 

[BigInch]

How close is close???

Its still going to have high bending stresses, flexures and axial movements. Those bending moments don't disappear for a way.

Buy some more pipe. Run along closer to an anchor.


"If everything seems under control, you're just not moving fast enough."
- Mario Andretti- When asked about transient hydraulics

http://virtualpipeline.spaces.msn.com

 

[SWMechanical]

It is about 2.5m away from the loop. I ran CAESAR on the piping. It passed the test. Does it mean it is OK?

Thanks,

 

[BigInch]

No, it just means Caesar doesn't know any better than to make a tap on a loop.

Kinda' like when my hydraulics program says its OK to have fluid veloities of 200 ft/sec.

"If everything seems under control, you're just not moving fast enough."
- Mario Andretti- When asked about transient hydraulics

http://virtualpipeline.spaces.msn.com

 

[zuccus]

I think all of you guys are assumming that this expansion loop is going to move quite a bit. What if in the center of the loop it didn't move at all or the whole loop just didn't move very much.

Also the line coming off the tap could be very flexible and/or designed to rest in an unstressed place while in hot/cold service.

I think we need a little more information.

I still don't think it is a good idea, but I bet a good analysys of the siuation would let you truly understand the consequences.

Zuccus

 

[BigInch]

Then it would seem the loop is/was redundent.

Look, we can only tell you its not going to make the best idea list this year.

"If everything seems under control, you're just not moving fast enough."
- Mario Andretti- When asked about transient hydraulics

http://virtualpipeline.spaces.msn.com

 

[mayda]

While I agree it is good practice to avoid tie-ins on loops, I also agree with "zuccus": without the whole story, it is unreasonable to flatly state it is a bad design. Keep in mind, piping systems are under varying stresses for their entire length, not just at loops. It could be possible to indavertently put a tie-in at a high stress location, that is why for any line that has any kind of expansion stresses, one should do a model to check it, AND make sure the results from the model do seem reasonable.

Depending on the design and operating conditions, there are likely to be areas on an expansion loop which have little movement and low stresses, and may be just as safe to tie in there as at another location. Basically it comes down to: It has to be analysed.

If, for whatever reasons (besides stresses in the loop), this is the location where the tie-in needs, or is best, to go, it should be analysed and the decsion made based on the results, not just "we never do that".

If we always stuck to "we never do that", we might not have made it to the moon. JohnBreen is right - one has to investigate the risks, and make the ultimate decision maker aware of the risks and consequences. Of course, if the analysis says its going to fail, then we have an obligation to refuse to go ahead with the design. But we do also have the obligation to investigate and determine what is actually the best design, all things considered, for each particular situation, and in some, that could well be putting a tie-in on an expansion loop.

 

[Sid7]

"Depending on the design and operating conditions, there are likely to be areas on an 'expansion loop' which have little movement and low stresses, and may be just as safe to tie in there as at another location. "

!! You sure not talking about the loop itself? Or are you really?

Siddharth
These are my personal views/opinions and not of my employer's.

 

[BigInch]

Ya. Ya. Ya. The question was, "Does anybody know if it is good to have a tie-in on the expansion loop?"

The answer to that question is NO... period.


I have not failed. I've just found 10,000 ways that won't work."-Edison “If Edison had a needle to find in a haystack, he would proceed at once to examine straw after straw until he found the object of his search. I was a sorry witness of such doings, knowing that a little theory and calculation would have saved 90% of his work.- Tesla

 

[mayda]

Sid7:

Yes, I am talking about the loop, but note I said "depending on the design ...". A lot depends on where anchors & guides are located, size of the tie-in, etc.

For example, a simple "U" loop of 6" Sched 40 CS pipe operating at 600 deg. F, 50 ft OAL between anchors, with a 10 ft long "U", e.g. (I don't know if my crude sketch will come out right, but x = anchor, ! = guide, length a = 20ft, b=10ft, c=10ft

|--c--|
b b
x---a---!| |!---a---x

Max. expansion stress at "c" approx. 6600psi
Movement approx. 5/8 inch sideways

This is a relatively low stress, and the movement would not be hard to accomodate depending on the design of the tie-in. Of course, one also has to look at fatigue life, but for a lot of expansion loops may only see one cycle per year at most.


If one were to put a guide at "c", as below, and dlete the guides on the "a" lengths, then max stress at "c" is about 3400 psi, and movement is essentially zero (depending on how evenly the pipe heats up.


|--!c--|
b b
x---a---| |---a---x


I DO agree it is good practice to avoid connections on loops, and I do not recall any that I I have designed that way, BUT, all I am saying is, it is not necessarliy "bad" design, IF some design work is actually put into the decision.

 

[rconner]

If anyone is looking for some sort of politically correct verbiage for which there might be more of a consensus?,

“Tie-ins for various reasons can be an inherently, rather highly stressed locations in a pipeline. Locating them in an “expansion loop”, that is also subject to additional stresses as by movement/bending etc., is not likely to help.”

While my following comment are off the subject, the reported contrast Mr. BigInch notes of Tesla and Edison (I believe both involved in some invention/research, and whom I think I’ve heard talked about variously before) is quite interesting. There is no doubt in my mind that a little (at least “good”) thinking can save a lot of needless experimentation, trials and expense and (and maybe even some embarrassment or worse etc.); however, at least in a few cases where “theory”, “calculation”, and/or what one has been told have been wrong, and in such a case maybe experiments(er), or sort of going through the stack straw by straw, may have some value. At least in the latter case, experiments or trials, or I guess looking a little deeper in the stack for something that may or may not be there, can be worth a good many “expert opinions”.

Over a great many years I have however run across at least a few folks who on the other hand appear to be pretty hard-headed and even to the point of some irritation when confronted by facts (at least as best I think I know them), and it has appeared they like to in effect see for themselves or in some cases even like to bang their head repeatedly against a brick wall. Who knows why they do it, maybe someone once told them something that was wrong, or at least it “feels pretty good once they stop”! I'm sure noone else has ever expereinced this -- everyone have a good weekend.