×
INTELLIGENT WORK FORUMS
FOR ENGINEERING PROFESSIONALS

Log In

Come Join Us!

Are you an
Engineering professional?
Join Eng-Tips Forums!
  • Talk With Other Members
  • Be Notified Of Responses
    To Your Posts
  • Keyword Search
  • One-Click Access To Your
    Favorite Forums
  • Automated Signatures
    On Your Posts
  • Best Of All, It's Free!
  • Students Click Here

*Eng-Tips's functionality depends on members receiving e-mail. By joining you are opting in to receive e-mail.

Posting Guidelines

Promoting, selling, recruiting, coursework and thesis posting is forbidden.

Students Click Here

Jobs

Metal seated ball Vvlve leak
5

Metal seated ball Vvlve leak

Metal seated ball Vvlve leak

(OP)
I have experienced following problems with 4" #1500 ball valves - metal seated, top entry, double-block-and-bleed.


4" #1500 Ball valve is with seats that are pushed by springs against the ball. In the middle of ball valve, there is 1/2" hole drilled to ball cavity in which we connected drain ball valve 1/2"#1500.

Here is the scenario...

1. With closed 4" ball valve was pressurized it from one side with 100 bar of natural gas.

2. With 1/2" drain ball valve connected to ball cavity we checked seat leakage by opening the valve and no leak was detected - which means metal seat is tight on that side

3. Using 3/4" bypass line around 4" ball valve we pressurized same ball valve from the other side on the same pressure - and still no leakage

4. 1/2" Drain valve was closed and 4" ball valve opened/closed - 3-4 times, letting gas go inside the valve and valve cavity

5. We tried to open 1/2" drain ball valve again to release the pressure from valve cavity, but after just small opening it was leaking and not stopping
for 5-6 mins

6. We closed back the 1/2" valve and tried to release the pressure after the 4" ball valve on 1" relief pipe down the line to atmosphere but the pressure remained 100 bar - like the ball valve is open

This all very strange to me.

I think it might have something with springs and metal sealing when the pressure enters the ball cavity -
maybe that pressure is causing springs and seats not to function well.

Hope for some help on this...

RE: Metal seated ball Vvlve leak

2
4. 1/2" Drain valve was closed and 4" ball valve opened/closed - 3-4 times, letting gas go inside the valve and valve cavity

5. We tried to open 1/2" drain ball valve again to release the pressure from valve cavity, but after just small opening it was leaking and not stoppingfor 5-6 mins

When you tried Step 5, was the 4" valve open or closed?

Yes, it sounds like the seats might be stuck from relieving cavity pressure.  You said the 1/2" drain valve continued leaking with a small opening, you may need to open it fully and try to "slam" the upstream seat on to the ball.

Who is the manufacturer?

RE: Metal seated ball Vvlve leak

(OP)
When you tried Step 5, was the 4" valve open or closed? - It was closed...

Yes we tried to open it fully, but it was leaking so much.

Manufacturer is from China with API6D certificate.

What we tried in test lab is to leave 1/2" drain valve opened all time and the pressure was applied from the side of the closed ball valve.

Then open and close the valve quickly. As drain valve was opened, pressure was not able to accumulate in cavity. In this scenario valve worked well.

RE: Metal seated ball Vvlve leak

Although the description sounds out of spec, API STD 598 permits leakage in metal-seated valves.  Thd leakage rate depends upon the valve size.

RE: Metal seated ball Vvlve leak

(OP)
I think that iso5208 is allowing 45 bubbles per second for dn100, but what we saw is very high flow from 1" relief pipe.
 

RE: Metal seated ball Vvlve leak

"When you tried Step 5, was the 4" valve open or closed? - It was closed..."

Have you tried draining the cavity with the 4" valve in the open position?  It probably won't change the situation, but sometimes not having the valve "pressure-locked" will let the seat free up and move against the ball.  I would still call it failed though, even if that does cause it to seal.  The seat should not become jammed, as it seems to be doing.

"Manufacturer is from China" . . . well, there you have it.

RE: Metal seated ball Vvlve leak

(OP)
Actually we tried to drain cavity in closed position - I will try today in opened.

What is a bit frustrated is that our engineer was in China for witness testing, after that when valves arrived we tested at our facility, than valves were delivered to site and tested before the installation, than the whole installation was test and except minor problems we did not have any serious one.

The seat testing was done with water 1.1 times working pressure and 1.5 times working pressure for body testing as API6D requires. Manufacturer has API6D certificate.

At the end valves are not functioning ...

If the problem is China - why is API issuing certificates to them?

I have attached seat drawing, maybe you can see something that it is missing.

RE: Metal seated ball Vvlve leak

The forgings and castings for most manufacturers are made in China, India or Brasil.  The machining and assembly locations may differ among the manufacturers.  Most clients are increasing their QA requirements largely due to these issues.  Some clients are very satisfied with the shop performance in China.

RE: Metal seated ball Vvlve leak

So, this valve has been tested previously, with natural gas, and worked fine, but now at start-up it's doing this.  Is that the case?

If that is the situation, it makes me wonder if there is sticky condensate or pipeline trash that has gotten in and made the seat stick in its bore.  Welding rod stubs and weld slag cause tons of problems.

If the valve has not been tested with natural gas previously, only with water, I will opine that the different flow forces of the natural gas do not overcome a design or manufacturing problem which causes the seat to bind.

RE: Metal seated ball Vvlve leak

(OP)
Valves were only tested with water. During startup with natural gas, problem is started. Actualy, once high pressure fluid enters valve cavity, problem starts. If the pressure is 50 bar, valve is working ok. If the pressure is 100 bar seats are like stuck and closed valve is leaking very much.
One thing we will try - to use stronger springs, but I am not sure that it can be solution.

RE: Metal seated ball Vvlve leak

It worked at 50 bar.  Then you went up to 100 bar and it leaked.  Did you go back down to 50 bar after that?  If so, did it work or still leaked?

I'm curious, is that a pig receiving valve?

RE: Metal seated ball Vvlve leak

(OP)
It worked again on 50 bar.

RE: Metal seated ball Vvlve leak

(OP)
No it is just regular ball valve on gas storage facility.

RE: Metal seated ball Vvlve leak

(OP)
We now tried the other same size/rate valve.

From one side it leaked on 15 bar. We than tried on 10 it was good. Again on 15  and it leaks.

After that we tried on the other side same valve, and it was good on 10, 20, 30, 40,50 bar and on 65 it leaked. Than tried again on 50 and it was good. Again on 65 and leaked again.

It looks like on some pressure seats are moving away from ball and can not come back to seal, like it is bind.

Now, why is from one side holding only 15 and from the other it is holding 50 bar is strange.
 

RE: Metal seated ball Vvlve leak

I'm beginning to think this is just a seat problem.  Like you, I at first thought the seat was getting in a bind and not returning to contact after the body cavity was pressurized.  But when you said that it worked at 50, then leaked at 100, then when you went back down to 50 bar it worked, that makes me wonder how the seat would have come unbound.  Now you told us another identical valve leaks at 15 bar on one side and 65 on the other.  This situation sounds like the seats won't seal consistently for gas, irregardless of cavity pressure.

What operating temperature was specified when the valves were chosen?  I assume it was high because of the metal-seated arrangment.  Are you at that operating temperature when testing?

RE: Metal seated ball Vvlve leak

(OP)
Temperature is up to 40 deg C. Valves are metal seated due to very polluted gas as it is for underground storage facility and for high pressure - 275 bar operated pressure.

We tested with nitrogen from bottle, so pollution is not the case for our our testing.
 

RE: Metal seated ball Vvlve leak

OK, it is not low operating temperature causing a problem.  It is not pipeline pollution.  The fact that the second valve won't reseal at 15 bar on one side doesn't seem as if it would be caused by the seat binding when pushed back into its bore by cavity pressure.  On the other seats which worked at 50 bar, but not above, the binding seemed to be a likely scenario.

If you take the valve apart to change springs, do you have a way to precisely measure the ball diameter?  I'm curious to see what the ball-to-seat mating shows.

RE: Metal seated ball Vvlve leak

Leaking from one side at lower pressure. Possibel cause: difference in geometry of forces from this side: less exposed area on seat sealings for pressure of seat sealings against seat and (area x pressure) (or) difference in other mechanical forces if pressurized from this side?

OK at 50, not OK at higher, OK at fifty again: deflection or non permanent (automatically reversibel within limits) deformation of mechanical parts of seat sealings? Caused by reasons as above?

Solution? Stronger sealing material or strengthening of seat sealings by altered seat-sealing geometry?

RE: Metal seated ball Vvlve leak

(OP)
@tr1ntx

We always first test seat just with pressure from one side in ball closed position and without pressure in valve cavity and on air 6 bar and water 250 bar, seats are tight - no leak.

The problem starts once pressure enters valve cavity.

As ball is in one piece with stem we measured ball diameter change on the place where seats are and ball is accurate to 0.00 millimeter - looks like perfect. But stem has 0.05 mm error in coaxial sense. It is above 0.03mm which manufacturer designed.

My opinion is that the purpose of springs is only to cover these errors in geometry.

@gerhardl

Do you think that with stronger springs this can be fixed?


We just tested third valve. From one side it leaked at 15 bar, from the other working ok at 90 bar. At moment, we have no option to try above 90 bar. Maybe tomorrow.   

RE: Metal seated ball Vvlve leak

OK, now with more information, we see that it is not a seat seal problem.

In addition to covering small errors in geometry, the springs provide the initial sealing force so that there is not leakage as pressure begins to rise from 0.

Stem-to-ball axial alignment: 0.03mm (design) versus 0.05 (measured), that is a large percentage variation, but in terms of the actual distance, I wouldn't think 0.02mm would be a large enough error to cause leakage.  Like you said, the springs should make up for that.  But I don't know the spring rate or how much deflection they have.

Stronger springs should help, although at only 15 bar, the seats shouldn't be binding in the first place.  Stronger springs might make it work, but they might just mask the real problem.

Then again, perhaps the springs are designed too weak in the first place and are thus incorrect for the valve assembly.

RE: Metal seated ball Vvlve leak

(OP)
In original there were 12 springs.
We added 12 more but those 12 are actually spring in spring so total of 36 springs.

The result is that the valve that was holding 15 and 50 bars from different places are now holding 20 and 60 bars.

It is step forward but very small. Beside that valve was much more difficult to open/close.

Another thing I want to point out.

When valve in closed position is pressurized from the tap in ball cavity in the middle of the valve - seat are opening at 5-6 bar with original springs used. This is normal behavior if you look at seat image from my previous posts. Diameter where seat is touching ball is smaller than diameter where is the O-ring so pressure in ball cavity is easily opening seats once force due to pressure is higher than force in springs is.

I think it is the design problem as you can not trap pressure in ball cavity. What I want to say that is if you pressurize valve in half opened position, than close the ball and let fluid from each side go out from the valve, fluid should be trapped in ball cavity, but it leaks out until pressure drops to 5-6 bar. This is happening with valve that are good if pressure is applied from the side and you check leak on the other side.

RE: Metal seated ball Vvlve leak

Just an observation on checking seat leakage via the cavity - you do need to take into account the volume of gas in the cavity. I would hazzard a guess that the volume of gas in the cavity would be around 2 - 3 litre. 2 l of gas at 100 bar is quite a lot. Is the pressure of the gas at the outlet still as great after 5 - 6 mins as it was initially?

RE: Metal seated ball Vvlve leak

(OP)
We have gauge connected to cavity and the nitrogen connected to other side and in to the same cavity.

Valve is working like safety valve - once pressure rise to certain level, pressure opens against the spring.

This is bad behavior in my view, bat that scenario will never happen in exploration so it might be irrelevant.

Anyway in the past we tested valves with testing fluid trapped in cavity, but with this valves it not possible to trap testing fluid for more than 5-6 bars.
 

RE: Metal seated ball Vvlve leak

as posted earlier:
"Manufacturer is from China" . . . well, there you have it.

I'm not trying to be a jerk about it, only to pass along what experience has shown me.  Although the management of American companies, and European companies too I guess, have been in a mad rush, tripping over themselves to give the Chinese manufacturers all our hard-earned manufacturing technology in order to get items at a lesser cost, they don't always get it right.  I've seen many cases where an important detail has slipped through the cracks, been "lost in translation", or just been ignored because it was too costly or unavailable.  I've also seen more than a few instances where a Chinese "spin-off" company from one of the export parts manufacturers attempts to produce the complete product, but does so quite poorly.  Crime doesn't always pay, I guess.  (Not just China, India, Romania, etc. too, though I've seen it more from China.)

RE: Metal seated ball Vvlve leak

zorans,

Do you ask the valve manufacturer for this issue? What is their response?

RE: Metal seated ball Vvlve leak

(OP)
Yes, I am in contact with them.

They have no explanation. I asked them to come and see what is going on but still they are hesitating about it. I also send them a video of out testing, but they answered that testing method is not good - which I agree it should be better, but we can not test real flow conditions in our lab. If they come and I take them to storage facility they will get real impression.

We had in the past bought many valves from them and always been able to fix if something is wrong with valves. What we did is always buy more items than needed than use them like reserve.

Once we also have been manufacturing ball valves, but with raw material being more expensive then their final product, we closed that production and now concentrating on testing valves before shipment and eventually fix malfunction valve.
But with this metal seated valves we have hard time to find the reason of leak and it is also for us impossible to repair or replace some parts as very accurate machining is required, which we do not have unfortunately.

RE: Metal seated ball Vvlve leak

2
Stepping in on this, though I realize it won't solve the problem:

The original poster mentioned that the vendor tested the valve with water, but that it's now leaking with air.  My agency won't allow leak tests to be conducted with water when the final product has to be gas leak-tight (which is not a "zero" leak rate, but a "less than xxx sccm").  The different size molecules will allow a water test to pass, where it will fail the gas test.

Fixing it will probably come down to precision machining, unfortunately.

Patricia Lougheed

******

Please see FAQ731-376: Eng-Tips.com Forum Policies: Eng-Tips.com Forum Policies for tips on how to make the best use of the Eng-Tips Forums.

RE: Metal seated ball Vvlve leak

(OP)
After long working hours, endless testing, strong headaches, ... it is now my opinion that the problem is not the design - this design is known as "single piston" and it must be down to material quality and its strength to resist deformations when high pressure applies.

I am not sure if anybody of established manufacturers has developed metal seated ball valves for this high pressures #1500 with "single piston" design approach. If so it would be good for me to know which is it.

RE: Metal seated ball Vvlve leak

I'm pretty sure Vintrol can handle a 4" Class 1500 valve.  E-mail is sales@vintrol.com

RE: Metal seated ball Vvlve leak

First of all, changing or adding springs will not be any help. As you mention yourself the valve is a "single piston effect" seat, and the springs are there to seal at low pressure.
In my line of work we use a lot of MtM seated ball valves and gate valves. Valves designed to PSL3G and up to 10.000PSI (690Bar) and sizes up to 24" for subsea use and 25 year lifetime witout service.
I don't know your Cinese vendor, but we use only European vendors.
The trick with MtM ball valves are the hardfacining of seat and ball. I assume your valve has this, if not you can send the valve back to China.
At our vendors the last step before assembly and testing is the individual lapping of ball and seatring with diamond paste. This is done by hand by skilled craftsmen. If this isn't done right the valve will never seal at gas testing.
Another thing you mention is "pressurized cavity". If the bore pressure and cavity pressure are equal there will be no "piston effect" on the seatring, and the valve will not seal.
Reading this tread I think the problem are the valve quality and design. When you apply pressure to the closed valve the seatring will virually try to "swallow" the ball. And if the seatring are not properly designed, the contact area will move from the initial contact area at zero pressure.
If you want futher help send me the full documentation such as GA drawing with bill of material and FAT procedure.

RE: Metal seated ball Vvlve leak

(OP)
@Nilsoni

Another thing you mention is "pressurized cavity". If the bore pressure and cavity pressure are equal there will be no "piston effect" on the seatring, and the valve will not seal.

------

Does it mean that none valve with single piston effect will seal or just these mine from China?

What I meant was do you know any manufacturer that has Metal seated valve with this kind - single piston effect type - that works?

Another thing we tried last Thursday and was on the facility and not in out testing lab.

We connected to 1/2" hole in ball cavity relief testing pipe 3/4" with opened gate valve to the atmosphere away from operator. Once valve in closed position pressurized from one side, leak was not detected on that 3/4" drain pipe. After that, over bypass valve, it was pressurized from the other side - the leak was not detected. This indicates that seats are good.

While opening that valve - rotating ball from close to open position, flow to atmosphere occurred on 3/4" drain pipe. Once ball in full open position - flow through 3/4" stops and when repeated from open to close position same happened.

This is prove to me that once I eliminate pressure in ball cavity, valve works fine.

This is what you say that if pressure is same in valve and in valve cavity - there is no piston effect, there are just springs pushing seats to close. I am not sure if any manufacturer can overcome this problem.

 

RE: Metal seated ball Vvlve leak

"What I meant was do you know any manufacturer that has Metal seated valve with this kind - single piston effect type - that works?"

Yes.

RE: Metal seated ball Vvlve leak

Zorans,

A few comments which may not help you in this instance but should assist in the future.

1/ If you are purchasing a metal seated ball valve and it is intended for gas service then you should add additonal test requirements to your purchase specification so that the valve manufacturer has to perform a high pressure seat using gas. API 6D has an allowable leakage rate for liquid testing which cannot be correlated to leakage found during testing with gas.

2/ API6D has a series of static pressure tests and again additional functional tests should be specified in between the main body test and individual seat tests.
This helps to prove the repeatability of the valve test results.

I would have two concerns about the seat design: -

1/ The seat is very long and may be prone to tilting and locking, and this could be verified by visual examination of the seat pockets after test.

2/ Seat sealing face appears to be quite rigid and inflexible, therefore any movement in the ball during pressure test may not be taken up by a similar movement in the seat.

You could try lapping the seats to the ball using diamond paste and match the seats to specific sides of the ball. This could hel reduce the variability assuming seat tilting is not the problem.

Let me know how you get on.  

Red Flag This Post

Please let us know here why this post is inappropriate. Reasons such as off-topic, duplicates, flames, illegal, vulgar, or students posting their homework.

Red Flag Submitted

Thank you for helping keep Eng-Tips Forums free from inappropriate posts.
The Eng-Tips staff will check this out and take appropriate action.

Reply To This Thread

Posting in the Eng-Tips forums is a member-only feature.

Click Here to join Eng-Tips and talk with other members!


Resources