Manual Valve - Stem-Disc Separation

[jrpower]

I don't know valve failure mechanisms all that well, and I'm trying to understand the failure mechanism of Valve Stem-Disc Separation (sometimes called a plugging failure in Probabilistic Risk Analysis (PRA)), in particular for Manual (hand) valves.

Is this failure mechanism plausible for manual valves? Internet searches only seem to show it mostly occurring in MOVs, and if so does this mean this is because of the torque a motor can produce versus that of a hand valve)?

What causes this type of valve failure?

What types of valves are susceptible to this type of failure (gate, ball, needle, etc)?

Does the size of the line affect this mechanism?

Is this a valid failure mode in steam, high pressure water systems, low pressure water systems, clean compressed air/inert gas systems?

Any insights would be great and appreciated! Thank you, Jill

 

[MikeHalloran]

Only butterfly valves have a stem and a disc to separate.

Other types of valves have different parts with different names that correspond topologically to the butterfly valve's disc, and they can all suffer stem separation, no matter the type of actuator.

It sounds like you are trying to analyze failure modes that you do not understand, in mechanisms with which you are unfamiliar, looking toward some future process wherein types of mechanism will be included or excluded from consideration for some mission by some other party, based on your work.

... which frankly sounds like a recipe for making inappropriate/erroneous design/application selections in a very organized way, by persons of minimal skill, all fully backed by documentation, and all wrong.

No offense intended toward you; you were probably assigned to just get the job done.

You will be able to do a better job, once you have accumulated a pile of broken valves, at least large enough to cover your desk, and have done a detailed failure analysis on each one.

I know; that's not in the budget or the schedule.
Welcome to engineering.



Mike Halloran
Pembroke Pines, FL, USA

 

[jrpower]

You are correct, this is not my area of expertise and it is a failure mechanism I don't fully understand (thus no offense taken), and thus I thought I would ask you guys who are more familiar with valves. I'm a 20 year veteran Engineer of my industry, but again not with valves, and in my industry we have used the terms disc and stem, regardless of the type of valve, gate, butterfly, ball, globe (let's write this off to semantics since I think you guys understand what I'm discussing here). I'm just trying to understand failure mechanisms so I can better analyze component failure rates which is my area of expertise--a data analyst & number cruncher, not a valve expert. To help explain where I'm coming from, and as an insight to valve stem-disc separation, my company had a case a number of years ago where we had a motor operated gate valve where the stem failed which caused the "disc" to drop and block flow in the line, but keep in mind this was an MOV, not a manual (hand-operated) valve.

I'm just trying to find examples of this type of failure, and if anyone has had any experience with this type of failure and can provide some insight into the failure mechanism, i.e. vibration, corrosion, manufacturing defect, etc. Also, is this type of failure only in running systems or can this type of failure ever occur in say a standby (backup) stationary system?

So, I'm just trying to gain some insights (for my own gratification and to be a better data analyst) on manual valves and the susceptibility of manual valves to have the "stem-disc" separate. Any help on this topic is appreciated.

Thanks, Jill

 

[MikeHalloran]

In a gate valve, the part that the stem lifts is the 'gate', or often the 'wedge', and occasionally I will admit, the disc. It normally comprises two disc-shaped structures, held or linked or cast together at an angle to each other to form a wedge, that gets jammed into the seat, which comprises two circular flat face areas held at an angle to each other, forming another wedge of sorts.

I think I have replaced a few manually operated gate valves where the stem fractured at its narrowest point, where the end of the stem rides within some sort of horseshoe shaped feature in or on the gate. I could not say exactly why those particular stems failed; my job at the time was just to replace the valves.

Speculating, because they were all installed in a waste process fluid gathering system in a chemical plant, they were certainly exposed to various process wastes. Because they carried a waste stream and were almost always open, the seats were probably fouled with particulates sufficiently so that the gates could not close completely when needed. The normal response to a situation like this by normal plant staff is to apply a big pipe wrench to the valve handle, which doesn't do the stem tip any good. If the particulate is deep, the gate can't close even with the wrench, and if the wrench does manage to close the valve, said wrench will also likely be needed when it comes time to open the valve, again doing the stem no good.

Gate valves are used because they're cheap, neglecting the costs of repair and replacement.
They survive best when used on clean fluids, and when cycled with modest frequency. They tend to stick closed if not cycled occasionally (again, bringing out the big pipe wrench).

The above is nowhere near a comprehensive exposition on gate valves.

Perhaps you can connect to a stream of broken valves somewhere?

;---

I also worked in a factory where we made the tiny ball valves used in refrigeration service kits. The balls and stems were machined from a single piece of brass. The only stem to ball separations I witnessed happened when the girls running the hose crimping machines lost their concentration and inserted the ball valve into the hose crimping dies and started a crimp cycle.



Mike Halloran
Pembroke Pines, FL, USA

 

[Danlap]

Hi Jill,

Fully align with Mike in this regards. Your question is about failure mode of detached Stem/spindle and closure member (Ball for Ball valve, Gate/wedge for gate vlv, Plug for Plug and Globe vlv, and or Disc for Butterfly and Check vlv).
Stem and Closure Member linked together by many method:
- Casted along, some small sizes commodity Globe valve use this
- Threaded
- Threaded and then Pin
- Slotted (some sort of Groove prepared on the closure member and linked against the “male” slot by the stem)
- Etc., however above are the most common
Major failure mode is due to fatigue fracture or stress which are mostly concentrated in the “linked” area. Name a few: poor workmanship, too sharp edges, corrosion on thread, pin located too low and or too big on pin diameter, eroded by medium/alien parts. And I would say most caused by cannot handle the big load requires to operate valve (since the valve are stuck / requires higher friction to operate the valve compare to its stem / CM strength design).

Ball valve:
Before ISO 7121 and or BS 5351 revised in 1986, no mandatory requirement for Ball valve to implement Anti blow out stem/bonnet design. Have seen some Ball valves produced before this period experiencing hand lever falling-out

Globe valve:
If not installed correctly (reverse flow / over the Plug) and not respecting the Opening as per design, would likely prone for vibration. May result fatigue on this linked area.
And many others.

Gate valve:
Align with Mike, most cases, shall Gate valve fails recommended to be replaced with new. Gate valves fail in so many way you could imagine with respect to (sticky) erosive/corrosive medium, cavity problem, differential pressure, using cheater bar when operating the valve (the most common problem whole world), etc.
Below is my 5 year old double expanding Gate valve manually Operated which were operated 2 times a yer. Heat, over torque, sticky medium cause this. The stem is perfectly fine

Seen some cases with other type of valves as well with different failure mode. Never experience cases with Plug valves though

Good luck
Regards,
MR


All valves will last for years, except the ones that were poorly manufactured; are still wrongly operated and or were wrongly selected

 

[gerhardl]

Hi, jrpower!

There is always a risk that something goes wrong, including 'disc separation' (of any valve) including manual operated valves and standby systems.

If you construct and select a valve correctly the valve will have a 'natural' lifespan depending on 'natural wear' for the given fluid, pressure, flow, temperature, viscosity, particle type and content (if any), maintenance, piping layout, etc. etc. If your valves are at a higher rather than 'minimum' quality for the given service, you have maximized the theoretical lifespan.

The only sensible way to make a formula for any failure, is to have a statistical significant number of experienced failures where all factors above are approximately equal, as pointed out by MikeHalloran.

The main problem is however that the main cause of valve failures (95% plus?) is wrong 'operation', including wrong decisions on the points above. (The rest is breakdown after an expected and known natural life-cycle.)

For your 'disc-stem separation': I have seen this failure a number of times, but only caused by overuse of force.
Mainly (one or more in combination):

- Valve is not regular controlled, partially moved (once a periode (year?)), greased (if relevant), wrong for the fluid, wrongly installed; and will get stuck.
- End switches, stops, or adjustments (especially for actuated valves) are wrongly mounted, used or adjusted.
- Extended force larger than the valve is constructed for is used to open or close the valve.
- For manual valves the extra force comes most often use of extended levers. Worst cases (this is real!) One meter pipelength lever for a 100mm valve pushed by hand. Worst case for a larger valve: extended lever pushed by forklift.

A risk analysis is done to prevent and improve things. Proper routine for everything above, controlled by a valve engineer, is a must before you even think about a risk analysis. Unskilled operators and engineers without proper knowledge gives a 100% risk that something will go wrong in the end.

Thumbs up: Your 20 years experience shows, searching information is the best way to start anything!

 

[jrpower]

Thank you guys so much for the information, these are the insights I was hoping for.

A couple more related questions:

Have the failures you've seen always been in running/flowing systems or have there been cases where the failure was found when the plant went to start a standby or backup system that had been stationary for a while, i.e. the valve failed while sitting?

Have you seen any of these types of valve failures in compressed clean air air systems?

Again thank you.

Jill

 

[Danlap]

Hi jrpower,

Have the failures you've seen always been in running/flowing systems or have there been cases where the failure was found when the plant went to start a standby or backup system that had been stationary for a while, i.e. the valve failed while sitting?
Please define stationary for a while? xx months or years?
Being normally close (sitting) for some period of time does not mean the sealing element is not wetted by the medium. The medium will infiltrate through leakage path of sealing element e.g. between disc and seat. And some nasty medium will rest inside valve pocket (cavities) and solidified. Thus causing the valve stuck and add friction to operate.
And for rising stem valve with gland packing, not being operated for some period of time will cause the gland packing losing its permeability / becomes dry and sometimes stiff.
Major factor above and some others added with excessive force by Operator will cause the stem and disc detached while about to be operated. Often found out while in seating position, however the term fail at stationary condition is not always / entirely true.

Have you seen any of these types of valve failures in compressed clean air air systems? Yes, but very rarely. In some big sizes line of compressed clean air system
Due to process: Shall the differential pressure applied too big (between Downstream and Upstream) and vibration (either initiated by the valve or other nuisance vibration nearby). This will originate the fatigue and or over stressed.
Again, added with human factor by means of excessive force, you will end up with the same problem.

Most common tools operator bring on field are spanner and cheater bar. It can move / turn most hand-wheel however with other consequence.


Regards,
MR

All valves will last for years, except the ones that were poorly manufactured; are still wrongly operated and or were wrongly selected

 

[jrpower]

Thanks. I've seen the cheaters used on the larger valves, not really on the smaller ones, and not on the compressed air systems. On the compressed air systems I've seen a few valves sometimes harder to turn than others on the larger header lines, but again haven't seen a cheater used on those, but good to know that the occurrences are rare.

Yes the valves that I'm asking about would have been sitting in the open position and the potential for them to have failed while the line is sitting in standby just sitting waiting for the line to be needed and flowed when necessary. The frequency of the standby systems getting flowed through spans all ranges, some lines are tested every 3-6 months and others can go 1.5 years or longer in some cases. And in a few cases like in the fire protection lines, sometimes never if near say the sprinkler heads, the line may have never been flowed at all. And in most cases for these standby systems, these lines are indeed wet, not dry lines.

 

[jrpower]

Also, what about needle valves and their "linked" area as discussed above in MRs first post? And what are your thoughts on it's susceptibility? Thanks.

 

[Danlap]

Hi jrpower,

I assume your talking about small size Needle valves. Not like other experts here (which I am not), I don't have an experience with big size needle valves up ranging from 2" to 20".
as for small Needle valves, shall broken then suggested to be replaced.
The spindle and Plug is usually as one entity. Either forged (to shape) and or machined afterwards. The benefit of this technique is that the mechanical properties along the spindle and Plug are relatively the same (less residual stress). And small needle valves are designed a tight fit way, therefore less prone to vibration..
Since it is regulated more frequently, Small needle valves are more prone to wear on its sealing element and internally leaking/ unable to regulate the flow as per desired. It is more common rather than gets disintegrated between the Plug and spindle.
Google the cross section drawing of needle valve for better overview.

Regards,
MR

All valves will last for years, except the ones that were poorly manufactured; are still wrongly operated and or were wrongly selected

 

[gerhardl]

On compressed air lines:

In my experience every manager having any problem with any valve or piping component being used for compressed air, swears that they have completely clean lines. Filters and water drainage points are correctly placed. The valves can in no way be exposed to particles from air, dust , oil (if any used), particles from pipelines, sealing parts or any rests of filings, welding, soldering or cleaning materials etc. etc.

Unfortunately this have in all cases proved to be wrong.

Three main lifetime problem points:
1. Startup: Wrong layout, mounting, residues in pipeline.
2. After repair: Wrong layout of new lines or replacement, wrong mounting, residues in pipeline, dirt residues in older pipelines pushed into other parts.
3. Re-compressing lines that have been out of use for longer periodes (years).

Preventing: Correct layout and correct type of filters and drainage points, with correct maintenance. Care in procedures and cleaning in any work on the pipelines.

 

[jrpower]

Thank you guys for all of your expertise and insights; it has really given me a much better understanding not only into this type of failure mechanism, but also into valves, in general. This will help me in my overall thought process in the work I do in risk analysis. Many in my area have no practical working knowledge of components; thus making incorrect assumptions which leads to either a lot of uncertainty in the analysis or applying the failures where it is unwarranted, but by understanding the details such as those you have provided, I get a more realistic view on what I am analyzing and hope to share this with others. So again, I'm am thankful to you guys for providing such great information to an outsider.

Sincerely,

Jill