Spring Supports - near rotary equipments 2

[clooney]

Hi guys,
here is the situation:
The process plant construction is completed. All equipments, pipes and relevent supports and Springs are in place. PLANT NOT IN OPERATION YET. At this stage, supports are basically bearing the Dead Load of the piping system.
The inspector arrives at the site and wants to verify the compressor nozzle alignment with connected piping before he OK's it for start up. Upon unbolting the pipe flange from compressor nozzle, the pipe flange disengages from the nozzle and pipe moves upward by 60mm about 2.4" (by the way, the horizontal misalignment was insignificant). The inspector does not approve of this large of a gap between the pipe flange and the compressor nozzle.

Noteworthy:
i) The nozzle is oriented vertically upward.
ii) The vertical pipe run (40" n.p.s.)connected to the nozzle is approx. 3 meters (10') long and is supported by Double Can Spring Supports.
iii) The Horizontal run immediately after the vertical run is also supported by a single spring hanger.
iv) The contractor has removed the lock pins from these springs (i.e.; the springs were not locked during this inspection).

My questions:

1)Should the springs be locked during this inspection at the preset cold load? Or is it O.K. to remove the lock pins.

2) Is this 2.4" vertical gap pretty significant?

3) How far would you go (percentage-wise)with field adjusting the spring supports if the design cold loads differ from the actual (as-built)cold loads. Note that the springs were delivered locked at the preset cold loads specified by the designer.


Thanking you in advance,
C

 

[bigskyguy]

As per UPC code, all piping shall be installed so that there is no strees on fittings or the pipe itself, UPC also has a detailed pipe hanger spacing spec...If you adjust your pipe springs before piping is used (filled) so that their is no stress on joints and pipe, when the system is charges the spings should absorb when weight differance equally, giving you a stress free system. Keep in mind that flexes are needed at any anchor point (pumps, coils, exchangers,ect)...

 

[CRG]

This is an awkward question to answer without a lot more information.

Typically, the purpose of a spring supports is to minimize stress on the piping system and connected equipment due to changes in pipe loads from pressure, temperature, and other changing external loads. With this in mind, is the vertical preload intended to minimize the stress on the piping system and the connected equipment, and will it accomplish this as adjusted?

If yes, does the resulting nozzle loading exceed the manufacturer’s limits?

If no to the original question, make the adjustment to minimize the effect on the piping system and connected equipment.

Who employed the inspector and whom does he represent?

Does the inspector understand the design parameters for the piping system and how it will change the loading at the nozzle during operation?

I recognize that I have not answered you questions. I believe the answers to your posted questions are that it depends. I would recommend that you have the engineer of record review the installation and then make a decision as to the adequacy of design and installation. Document the inspection with an explanation for preload and pass the information on to the inspector.

 

[abeltio]

Some piping systems are designed to have what is called "cold spring"... a calculated stress imposed on the system so that when the system is at the design load the load on the nozzles do not exceed the code limits, e.g. NEMA, API etc.

the inspection procedure should have been agreed prior to the inspection...
and the inspection starts with the approval of the design calculations during the engineering phase.

applying this line of thought for inspection in case of a welded joint (e.g. on a steam turbine) will this inspector require that the joint is cut and then rewelded?




saludos.
a.

 

[clooney]

agreed CRG. I know it is awkward and I was really looking for a generally applicable reply. It is almost impossible for me to provide you with all the design calculation details and give you a perfectly clear picture of the situation.
What I know is that the design calcs (stress anlysis) have been checked and approved. The external loads on Nozzles have also been approved by the equipment vendor.
Can you or anyone else tell me what is common practice in performing these alignment checks in the field? Are the springs supposed to be unlocked? Ofcourse if the springs are unlocked, they might have been preloaded such that upon unbolting the pipe from the nozzle, the spring would pull the pipe up. I suppose I will have to eventually get a hold of the CAESAR II (pipe stress analysis program) model to verify the situation at hand.

By the way, the inspector is an auditor sent from the Consulting Engineers (LINDE AG - Munich, Germany) who had the design contract .

thanks,
C

 

[carthago]

I believe pipe stress was performed to check the system under design condition, e.g:
-Maximum temperature and design pressure.
-Minimum temperarture and design pressure.
The purose of the pipe stress is in my opinion:

1/- To check the stresses
1-a/- Primary due to weight and pressure AND also spring loads..
1-b/ Secondary : due to temperature.

2/-Nozzle loads: To check if they are lower than allowable as given by equipment manufacturer or relevant API ( 610 , 617 ,...)
Most of the time the stresses are OK.

It is the nozzle loads which defines the supporting function and spring loads near the equipment.

--------------------

My idea to convince your inspector is to take (or ask your stress man) to model your actual cold situation where:
- there is no pressure
- design temperature is at ambient, means equipment
or their nozzles have no predisplacemt yet.
-Springs are released , means with manufacturer set loads

If you can prove that equipment nozzles are not overstressed/overloaded than its OK.

 

[carthago]

Forgot to tell that..... spring supports shall be UNLOCKED only after :
-Hydrotest
-Before commissioning

 

[clooney]

carthago,
You're right on. That is pretty much where we're headed. I am trying to obtain the original stress file rather than model the whole system all over again. As a first step, I'll have to release the pipe at the connecting node to the compressor nozzle at cold condition and see if the displacements check with what has happened in the field and see where things go.
thanks again.
C

 

[StressGuy]

I think your inspector probably screwed up.

With the travel stops in place, the millwright and piping installers should have been able to get the flanges in alignment. At no time should the travel stops have been pulled during this process. It's one thing to adjust the threaded connection of the rod to can to get the nozzles aligned, that's typical. Once you pull the stops and then try adjusting the threaded connection, you are changing the load setting and that's no good.

Springs are designed to carry an operating load at operating temperature. Being a compressor system, there shouldn't be a liquid component in play, although insulation weight could be a factor if the piping has not been insulated in the field yet.

There is no requirement that an equipment nozzle carry zero load. Responsible vendors provide allowables for a reason - it's impossible for a connection to carry "zero load". Now, with a properly laid out support system, in your case I would expect a spring either directly over the nozzle or very nearby in the first horizontal run, it is possible to "float the flange" with the travel stops in place for bolt up purposes, but not with the travel stops free.

Obviously, we don't have the details. But, I can, and have, seen cases with a hot piping system and a relatively long vertical rise to the elbow requiring that the spring have a heavy cold setting so that the nozzle load stays within the allowable when that vertical rise expands and causes the load to shift from the spring to the nozzle. Unless you are going to use a constant spring, which is a rare thing for these circumstances, you are going to have load shift back to the nozzle. The vendor allowable number is typically a range of +Y to -Y and I've often been faced with having to have the nozzle load near the +Y limit in the cold case with the springs active so I could satisfy the -Y limit with the system at temperature.

That is my gut instinct in this case. The analysis should be checked. If the cold case with springs active shows that the vertical load is within the vendor allowable range, there is no problem aside from having to repin the springs and completely start over with the flange alignment of the nozzle. Your inspector most likely created a big unnecessary delay in your project startup.

And yes, this is a bit of a sore point with me.

Edward L. Klein
Pipe Stress Engineer
Houston, Texas

"All the world is a Spring"

All opinions expressed here are my own and not my company's.

 

[clooney]

stressguy- I am quoting you "....have, seen cases with a hot piping system and a relatively long vertical rise to the elbow requiring that the spring have a heavy cold setting so that the nozzle load stays within the allowable when that vertical rise expands and causes the load to ....". I still haven't seen the caesar model to verify it. But I am leaning more towards your abovementioned senario.
thanks again,
C

p.s. pretty good gut instinct there