Allowable loads on valves from piping

[Tarkjell]

During my years as offshore engineer we have obtained a set of allowable loads from a supplier that our piping system may induce on a valve. Typically welded valves or valves with mechanical hubs have been of concern.
The vendor has been required to allow for moment and forces in the magnitude of what the pipe itself can take. We want to document that the valve assembly is not the weak link in the piping system. But is this necessary?

My questions are:

When a valve is designed to for example ASME B16.34, is it necessary to demand a set of agreed allowables loads from the valve manufacturer?

Is the valve automatically capable of taking the same loads as the adjacent pipe when designed according to the codes?

Hopefully some stress engineers out there can shed light on this topic.

Regards
Tarkjell

 

[Frank1344]

Hi Tarkjell,

To my understanding the main concern about valves are to make sure they are adequately supported and their flanges will not leak.

Both issues can be investigated thru softwares like Caesar II and AutoPipe.

I have not seen in any company's practice like Fluor and Jacobs to submit the loads on the valves to manufacturer for approval.

But,

I should address that the size, rating and process conditions may somem times change the whole story.

Regards,
Faraz

 

[carthago]

My point of view only.


1/- Convert external loads into equivalent pressure: delta-p.

This delta_P = (1.273/G^2) *(F + (Me/G))
G= bolt circle
Me = external Bending moment

Fe= External tensile force if present.
Most of the time Fe is << Me/G

check if:
( delta-P) + (design pressure) < flange design rating as per ANSI B16.5
The trick to minimize Me, is to play/adjust the pipe supports and make some stress runs.
The ideal situation ( for weight case) is to add a support as close to the valve as possible.

Both Me and F can be given by Caesar output , if you add Cnodes (at flanges)to your stress model



2/-ANSI B16.34 States also in its paragraph 6.1 ( globally said ) that :

"additional material/ thickness have to be added in order to add strength or rigidity to the body"
And some manufacturers can give allowable loads based on their own tests..
And for some heavy valves you can have the possibility of ordering them with supports attached/bolted to their bodies.
------------------
I know also (for instance!) that Shell specifies for stem of actuated valves = sizing with a safety factor = > 1.5.

 

[unclesyd]

Way back when I worked for a valve company we had a mechanical test area that mainly checked for repeated operation to ascertain a certain life cycle for various valves.

One function of this area was to check the mechanical function of the valves with different loads, mechanical and thermal, imposed on the valve flanges by piping. This was used if any design change was made in particular style production valve or testing of any possible new design. This test rig had the ability to impose any load or combination of loads on a valve and check for leakage and functionality. I don't know how the applied loads were determined, whether it was a code or internal function.

The only valves tested to a specific requirement were some valves supplied to the old Standard Oil Company. The only part of there test I remember is that no "cheater" could be used to seat the valve. I got to assist at times as I was the lowest rank in the Metallurgical and Mechanical Group.

All flanged valves were leak tested and a certain percentage, small, was pressured tested to design pressure plus some multiple. There were periodic proof tests of certain valves styles and types.

I imagine that with all the styles and types of valves with the different origins that a lot more individual testing is being specified today. I haven’t specifically checked but at a local piping supply company I’ve seen a lot more care taken in the transport and packaging of valves than you would have seen 10 years ago. There appears to be a lot more tags attached to valves than there used to be indicating some additional test or inspection or just better documentation. The next time I’m in there I'll ask if anyone is requiring any extra load testing and if so what type.

 

[carthago]

Hi gents,

Based on my own rxperience ( Shell , Lloyd's,..)
But I AM ALWAYS EAGER TO LEARN..

in principle for a piping system
the weak point shall be :
1/-Rotating equipment load.
2/-Nozzle load
3/-Flange rating/ leaking

Regarding its ratings ANSI B31.34 refers also to ANSI B16.5 ( say The pure flange code)

 

[Tarkjell]

Gentlemen, Thank you for responding.

But I still don't feel I have gotten the full answer to my questions.

Unclesyd is obviously on the track, and I would surely appreciate to learn what the response is when questioning the local piping supply company whether they receive any extra load testing requirements.

We conservatively use Kellogg's equivalent pressure method for a quick flange leakage check.

But what about the load capacity of the in-line welded valve? Or the valve connected to mechanical hubs, especially when the hubs are considered to be stronger than the pipe itself?

Are all valves generally stronger than the pipe, and therefore one does not have to obtain a set of allowables from the supplier?

 

[jte]

Tarkjell-

I can't answer your question directly, but I can relate. Several years ago I got involved in an investigation of a failure of a fast acting valve on a dock line in an environmentally sensitive area. This valve was intended for emergency shutdown of the line in case something happened between the ship and the shore valves.

The fabricator's original design was for a welded in valve. Our piping group felt that a flanged valve would be easier to install and maintain. The fabricator agreed to provide a flange facing and drill and tap the body to accept the studs. After installation of the valve the Coast Guard was called in to witness, among other things, the closure of the fast acting valve. Nothing happened. To make a long story short, it turned out that this valve had very tight internal clearances. My analysis showed that the moment induced in the body through the studs / gasket caused enough deformation to bind the gate.

This, however, was an expensive, semi-custom valve to begin with and its design was modified in a schedule-intensive environment. I've never heard of a valve failure of a more everyday valve which was attributed to excessive piping loads.

jt