Ball Valve Seat Damage 2

[DesmoFan]

I have a water system with a large manifold feeding 10 3 inch pipes. Each pipe off the header has one 3" ball valve, about 1 ft of pipe and a nozzle. During normal operation we bleed water through the valves and nozzles and then close the valves. The system is pressurized to 180 psig and when needed, all the valves are commanded open. When complete the valves are commanded closed under full flow conditions. This is the second valve we have damaged in a short time and the seat damage is in the same exact location on both valves. The damaged seat is on the inlet side and is the part of the seat that is exposed when the valve is approx. 90% closed/open. Each nozzle is sized to flow about 500 GPM. See attached picture. Any ideas on what is causing this would be greatly appreciated.

 

[tr1ntx]

So what we're seeing is the O-ring extruded through the clearance on the seat OD, correct? Is that seat bent inward? It sort of appears so. Or is it worn thin there?

What is the seat material? Looks like PTFE or Nylon, or I guess it could be UHMW. What brand of valve is it? What pressure class / rating are they?

Thinking out loud here . . . when the ball turns so that its spherical OD moves off the seat and the bore is over that area, there is no longer a physical restraint holding the seat in place.

The problem MIGHT be that the o-rings are of too low a durometer and when the seat deflects a little, the o-ring is pushed through. Nah, that doesn't seem likely at only 180 psi. Do you know what the o-ring material is? And its durometer?

The seat might not be physically strong enough and is deflecting due to the pressure under/behind it at the brief time period when the ball bore is over it but the spherical OD is still in contact with the very outer portion of the seat, i.e., prior to breaking the seal and relieving the differential. It could be deflecting enough to allow the o-ring to be pushed up, and then when the bore moves over the seat's outer edge, things just pop out of place. Without knowing the materials and the dimensions/clearances, I'm just speculating.

There might just be too much clearance around the OD of the seat, but from the picture, it doesn't look like that is the case.

How often are these operated? If it is a softer thermoplastic seat material, you might be getting "stick-slip" where the seat is adhering to the ball as it sits, then as the ball begins to turn it "smooshes" the seat out of shape. People sometimes refer to this problem by saying "it's trying to pop the seat out of the pocket". That might be what I'm seeing in the photo when I asked if it is bent inward.

I'm assuming this is a "floating ball" valve, it looks to be. Perhaps it is just dimensionally incorrect and when closed allows that upstream seat to move too far out of its pocket. Is it a Chinese valve?

Well, there are a few ideas and things I've seen before. Without more detailed information, I can't be more definite, so:
-soft/very weak o-ring
-too large of a clearance
-weak seat due to wrong material or improper design
-plastic-to-metal adhesion leading to deformation
-improperly designed/manufactured valve assembly

 

[tr1ntx]

RE: "due to the pressure under/behind it at the brief time period when the ball bore is over it but the spherical OD is still in contact with the very outer portion of the seat"

I was thinking in terms of a trunnion. I should have said "still in contact with the very inner portion of the seat".

 

[DesmoFan]

The valve is a Valworx 521124

Pipe Size: 3" NPT female
Port Size: Full Port, Orifice 3.00 inches, Cv 873
Working Pressure: 300 PSI WOG
Working Temperature: 0-180ºF (0-300ºF with optional valve mounting kit)
Valve Body: Brass
Ball: Chrome plated brass ball
Stem Seals: PTFE (Teflon) with double FKM (viton) O-rings
Ball Seats: PTFE (Teflon) energized with FKM (viton) O-ring backing
Valve Pressure Rating: 600 PSI WOG
Valve Temperature Rating: 366º F maximum
Cycle Time: Approx. 1 second per 90º rotation
Air Pilot Pressure: 80-120 PSI
Air Connections: 1/4" NPT (plus NAMUR pad for solenoid valve)
Actuator Seals: NBR
Actuator Temperature: -4 to 180º F

 

[DesmoFan]

These valves are not used very often maybe once every 3 months for less than a minute. Yes the seat is bent inward.

 

[gerhardl]

This is fairly typical of a floating ball valve, happening when staying open with only a tiny fraction of opening.

Regardless of the relatively low pressure before the valve, you will at startup or when the amount able to get out of the nozzle out is large compared to volume getting through the valve, have atmospheric pressure after the valve. Under theese conditions you will have underpressure (vacuum)over the seat/sealing, probalby also cavitation.

The large pressure difference along the seat perimeter (full water pressure, down to suction at the opening exposed sector) will do things to your valve.

Either you have to control the valve movement, real full open/full closed operation, improve the ball valve construction,or probably best: change to regulating valves that are able to regulate and can withstand the cavitation conditions.

The false operation may becaused by careless operation, rongly adjusted closing elements, damaged or poor construction, particles or firm objects in the stream, damaged seats due to causes above;.. or perhaps a wish to regulate the single nozzels as a feed-stock may have hydraulic conditions giving higher flow at some orifices, lower at others.

 

[superweo]

Gerhardl is right.
A control valve should be more appropriate for this service.

Regards

 

[ione]

You have received valid inputs above (especially by gerhardl).
The problem is that ball valves should be operated fully open or closed as they are not supposed for control applications. They were originally conceived as on-off valves without much care to throttling characteristics. The partially closed configuration with local increased velocities could be the responsible for the seat extrusion.

Another aspect that would worth the effort to focus one’s attention on is ”cavitation”, because damages related to this phenomenon could be much dangerous than seat damages. Cavitation could be an occurrence in valve operated partially closed. Some manufacturers have developed “cavitation vs valve opening degrees” graphs showing safe areas where the valve can be operated without risk of cavitation and subsequent damages. Ball valves are definitely the least indicated for almost-close service. One final note. Do not underestimate the fluid temperature, as it is not a trivial factor. At higher temperatures cavitation vapour pressure increases and cavitation is more likely to happen.

 

[DesmoFan]

Thanks to everyone for the valuable input, we do use these for open/close service but the closing/opening time is fairly long, about 12 seconds. With our current pnuematic setup we can't currently improve that. I was thinking of replacing the ball valves with butterflys. Thoughts?

 

[tr1ntx]

OK, I see what it is from looking up Valworx. When I saw that O-ring it made me think of a seat seal like a trunnion ball valve would have. They're just using it like a spring beneath the seat.

Also, the way I read it, you are not operating the valve in a throttling mode, i.e., partially open. Correct?

With an air-op valve and a cycle time of 1 sec, you're not getting cavitation damage or vapor vacuum suction or anything like that. Either the seat material is too weak for the application or it is an inherent problem with that valve design which allows the fluid to get behind the seat. Well, it might not be a design problem, but rather a manufacturing problem that allows too much clearance on some of them.

The seat material being too weak can take 2 forms: either it is just so pliable that the flow catching on the back edge can push it out of shape, or that previously-mentioned "stick-slip" issue is causing the deformation. I see rust on the seat face in the photo. That's what made me think of the stick-slip in the first place. Do you know if some of them are reluctant to open and then pop free suddenly?

Delrin (acetal homopolymer) or nylon seats might fix the problem, if you can get them for that valve.

 

[DesmoFan]

No throttling but the cycle time is over 10 seconds, not 1 second. Thanks

 

[tr1ntx]

I saw the 1 second time in the specs you listed, and I'm pretty sure I saw the same on the Valworx website. Why the discrepancy, not enough air or are they sticking?

 

[ione]

Have the closing procedure been slowed down because of surge related issue?

 

[DesmoFan]

The valves are commanded closed and then when we want water we just vent the "muscle" air to the actuator and let the actuator springs open the valve but it takes a while to get about 10-12 seconds to full open

 

[tr1ntx]

OK, I understand now, you have them set-up so that the actuator "fail" position will open the valves. Well, the 10-12 second opening time is the main cause of the seat failure (too long of a residence time at partially open), but you should be able to overcome that with stiffer seat material.

On a side note, unless the vent is small and restrictive, the spring should cycle the valve faster than that, I would think.

 

[DesmoFan]

Thanks tr1ntx, yes the vent system is restrictive

 

[gerhardl]

Commonly restrictions through solenoid valves or vents on actuators could be restrictive: better to use a solenoid valve with high cv and then restrict (within allowable limits on the solenoid valve) with throttle valve on the solenoid, or incorporated in solenoid in/out pipelines.

(Remember that a 1/4" solenoid valve could have an actual through-bore from (normally) about 0,8mm up to 4-5mm, which would mean a lot of difference).

Better to have real and faster on/off with ball-valves than change to BFL. BFL could give more problems and could more easy stick and give higher torques than ball-valves.

Regulating valves, as stated before, gives the ideal solution.

Apart from a suspectively long opening/closing time that indicates higher torque or more restrictive air supply than
ideal, (or too small/weak actuators) the high temperature will add to sealing 'floating' problems.

Carbon reinforced sealing materials possible? Check valvetypes and constructions for steam/condensate purposes!