Help Me Diagnose Steam Ball Valve Seat Failure 1

[TiCl4]

We have actuated ball valves, 2" NPS, in 70-90 psig (~315-335 F) steam service. The valve is a floating ball design, 2-piece, full port. It does not have any slow open/close mechanisms. Below are some pictures of the seats after only 6 months of service. These valves are actuated, on average, probably 50-100 times a week. The seats are graphite-filled PTFE. The damaged seat is on the downstream side of the valve. I'm not sure of the orientation of the damaged part relative to the seat.

The valve manufacturer insists this looks like an overtemp/pressure problem, but at a boiler setpoint of 70 psig I fail to see how these even approach the T/P limits of the material. The highest I've seen the boiler is 90 psig. The seat appears to have some erosion on the inside face in the failed area, but the entire failure - what with the seat looking like it's pushed up towards the ball - almost looks like it was pinched by the ball. Can any of you with a more experienced eye help me out?

 

[TBP]

I'd bet that you have a pocket of condensate forming upstream of this valve while it's closed. When the valve opens, the first thing that it passes is a jet of hot water that is also busy flashing into steam.

Do you have a drip-leg and steam trap ahead of this valve?

 

[TiCl4]

TBP - we do, but it’s very old. This is at the end of the line for the header. I’ll check out the condition of the drip leg/trap next week. What would cause the seat to turn upwards like that?

 

[gerhardl]

If there is no damage to the balls itself (no cavitation, hairline scratches, dirt, abrasion damage etc.) the problem is, as TBC says, probably waterhammer/hot water flashing. Check steamtraps and layout/placement for water drainage.

 

[bcd]

Are you sure this is the downstream seat? It looks very much like an upstream seat that is pushed into the waterway of the ball port as a valve opens and then the deformed portion is cut-off by the ball as it closes on the inner surface. I just don't see how a downstream floating ball seat can be pushed upwards towards incoming flow & pressure when the valve opens.

 

[TiCl4]

BCD,

I can't say that I'm 100% sure. However, the other seat appears in good condition, and the valve was leaking by. Due to this, I assumed that the misshapen seat was the downstream seat, as that is the sealing surface. If the seat that is still in good shape were the downstream seat, the valve should not have been leaking by. Perhaps what I call the "good" seat is actually bad too? A picture of the other seat is below. The area in focus is only portion of the seat that appears even slightly damaged and is showing only very minor pitting/scratching.

 

[gerhardl]

If you have even a hairline crack in the sealing or ball surface on a steam valve either steam, condensate or water will eventually pass, might expand at lower pressure, downstream or when caught in room behind sealings inside the valve, and even give cavitation. Difficult to see which sealing is upstream, but in my opinion the damage is caused by leakage over seals.

The list of possible causes are long, and it might be a combination of two or more causes.

Most likely are:
Unsactisfactory layout, dimensioning, filtering and drainage.
Dirt, residue in fluid and on seals/ball giving abration.
Unprecise closing, giving microscopic openings left.
Weakness by age or chemicals (cleaning or preservation media in pipeline?).
Unsuitable ball or sealing material. (Actual temperature and pressure might be higher than allowed)

Good luck!

 

[TiCl4]

All,

I investigated the steam header near the point of the valve failure. The header condensate trap had an isolation valve that had been closed as part of a previous LOTO and had never been reopened. There was certainly buildup of condensate upstream of the valve due to this. Due to the MTBF for this valve of ~6 months, I won't know for a while if this was the culprit. However, thank you all for your suggestions - I'll be monitoring this valve for quite some time.

 

[bcd]

I'd say the deformed seat is definitely the upstream seat. I have seen the same damage several times before. The deformed section interferes with rotation of the ball. It catches the ball port as it tries to pass by. Valve would struggle to fully open and close.

 

[PeterIgg]

I would concur with bcd, this is an upstream seat. As the ball moves from the closed to open position, this part of the seat beconmes unsupported by the ball and if pressure gets behind the seat it can push this part of the seat out. The small grooves (notches) that you can see in the circumference of the seat are there to prevent this by relieving (by providing a flow path) the pressure behind the seat, but if they get blocked they will not function adequately.
What sort of operating speed do you have? < 3 sec? The shorter the time the seat is not fully supported (it is only fully supported when fully open or closed), the better.
These floating ball valves work by the upstream pressure pushing the ball onto the down stream seat, the higher the pressure the more sealing force you have. At 70 psi you have very little, the factory low pressure air seat test is usually the hardest to pass and that is done at 80 psi! A combination of the heat and the mass of the ball may be allowing the ball to drop slightly (assuming you have the stem vertically up) which can lead to slight seat leakage
Generally speaking you are better off with a trunnion mounted ball in low pressure applications, but in steam you then have the issue as to what seals to use in the seat carrier. And, of course, they are far more expensive! The reality is that you know it will happen and you just need to have a replacement in store, depending on what it costs to shut the system down. Of course, if you have really deep pockets, you can go with a rising stem ball valve, this is the service they were made for.

 

[danschwind]

Can't help with your particular problem, but somebody way more knowledgeable than me in this sort of stuff once told me to avoid using ball valves in steam service due to this issue.

Maybe consider changing valve type for plug valves?

Daniel
Rio de Janeiro - Brazil

 

[TBP]

danschwind - There is nothing wrong with ball valves in steam service. You just need to select the correct valve for the service conditions and code requirements. Then you need to have the steam system designed/installed/maintained correctly - and (one or more) of those is where the wheels usually come off.

 

[danschwind]

That seem an awful lot of things to juggle when deciding the type of valve, TBP

Titanium, what is the manufacturer of your valve?

Daniel
Rio de Janeiro - Brazil

 

[TiCl4]

Danschwind,

We use Jomar valve - model 2020D. These valves are quick open/close. I haven’t been the happiest with this valve, but I’ve quoted out other valves (Worcester from Flowserve), and the seat materials were the same as what we use. The valve, too, was mostly the same except it was a bit over 2X the price. Maybe it has better internal design/machining.

The Jomar valve also has the advantage of not needing a mounting bracket for the actuator, saving on height of installation.

 

[danschwind]

Titanium,

I stand by my opinion that ball valves are kinda problematic in steam service, even though I agree with TBP that they can indeed be used provided everything is correct.

My suggestion, if ball valves are to be used, is to try sourcing them from known "steam suppliers", e.g. Spirax Sarco. In the US I reckon you probably have other options as well.

Hope you are able to fix this issue - I can imagine how frustrating it must be.

Daniel
Rio de Janeiro - Brazil

 

[Dom2411]

I fully agree with the statement of PeterIgg;
From the perspective of a valve manufacturer I can only estimate that the quality of the valve you've been using is pretty poor (just checked the model).
The sealing surface on the upstream seat looks pretty bad - I guess it has always been looking like that since it only works as a kind of bearing in a
floating ball system (only downstream seat is sealing like you've already mentioned before) and therefore should look "unused".
If the low pressure seat testing was really performed by the manufacturer, at least the downstream seat should have been of much more better quality, otherwise
the LP seat test could have never been passed (did you receive any 3.1 certificates for that kind of LP seat testing btw?).
Just for my informaton: Are these simply drop in seats or are they fixed in a groove?
Have you meanwhile found out any news?


Dominik Reule
Global Project Engineering Manager

http://www.as-schneider.com