Help: Cryogenic Test Failure Analysis

[JLee81]

A couple of weeks ago, we tested a cryogenic triple offset valve and turn out that the helium nitrogen gas start leaking above tolerance level at -120C and the the amount of leak got worse as the temperature got close to -256C. Then when the temp. came back around -120C the leak was in the tolerance level again. Aslo the valve worked fine in ambient temp. and above. What do you think is the possible cause of the leak after a certain cold temp? Need your professional inputs. Thanks.

 

[terje61]

Hello JLee81,

This sounds like your seals or other closure members are shrinking at different rates. When the temperature is going down the gab between the two parts becomes larger.

Regards,
Terje

 

[terje61]

Forgot to mention that if you want to solve this you can:
- Use different sealing materials that are less susceptible to shrinkage.
- Change the design so that the parts that are more susceptible to shrinking are on the outside of the parts that shrink less. This way your valve will be more tight during lower temperatures.

 

[gerhardl]

Are you the producer? If not: In addition to terjes good and sensible answer: are you sure the valve is a 'real' cryogenic valve, consructed for this temperature, medium and demanded leakage rates?

Sealing construction and sealing mounting and fastening construction and materials?

 

[JimCasey]

Is this a seat leak or a packing leak?
Either way, as tergew says, you have parts that do not fit as well when they shrink with the cold.

If it is a seat leak, you might be able to supply more closure torque with the actuator. Or check with the manufacturer for a different seating material/arrangement.

If it's a packing leak, graphite packing tolerates thermal extremes better than polymeric (TFE) packings. Live-loading the packing helps with thermal compensation also.

If you specified x leakrate at - degrees, the manufacturer is pretty much responsible to re-do the valve until it meets your needs, unless he took exception to your spec or proposed something different.

 

[NGiLuzzu]

JLee81,
did you use a the helium + nitrogen gas mixture?
What if you try with helium gas only?

I've seen a lot of cryogenic tests, but all with the valve immersed in liquid nitrogen at -196°C... how did you reach -256°C? And, most of all...why?

I agree with the suggestions above: if the seats starts leaking at low temperature, probably it is not a real "metal to metal" sealing designed for such low temperatures...

Hope this helps, 'NGL

 

[belowzero]

Hi Jlee81
As NGL has identified the jump from -120 to -256C is rather a big one, as Nitrogen is a liquid at -196C I assume 100% Helium was used.
What was the cryogenic valve manufacturers test/design temp? Having cryogenicaaly tested more Triple offset valves then I care to name, it would help if there was an 'indication' as to the manufacturer, it would be easier to help you then. There are a lot of questions which would need to be asked before this can be solved, however why are you using a forum and not referring back to the manufacturer.
I don't know any valves above about 1/2" which are capable of -256C. The cost alone to test at these temparatures is astronomical.
B

 

[Gobio]

I´m really interested in tests of cryogenic valves. Could you please give me references about this subject. Any document would be helpful for me.
Thanks

 

[NGiLuzzu]

Gobio,
almost all the Customer Specifications about cryogenic testing of industrial valves assume the form of an amendment and/or supplement to BS 6364 standard, an old one... but in many case the only one applicable, as of today...

You can find one exception to this approach in the last editions of Shell specification (MESC SPE 77/200 incorporating and superseding the "old" MESC SPE 77/306) that arises as a "stand-alone" document.

Moreover an EN 12567 standard for LNG valves exists (but it is still rarely applied and unfamiliar to many Engineering Companies), while ISO/TC 153 Committee is working on a draft about isolating valves for low temperatures (ISO/WD 28921).

For more information, you may want to take a look at the following discussions within this Forum:

- thread408-148535;

- thread408-207994;

- thread182-151721.

May be then you will find worth to start a new, dedicated Thread about this issue...





JLee81,
coming back to the original post, how about torques?

Tripple-offset valves are typically torque-seated, but you don't mention anything about this aspect (as well as about pressures and test directions...)...?





Best Regards to All, 'NGL

 

[Gobio]

Has someone the Shell specification MESC SPE 77/200, ISO/DIS21011 and TC153 for my information?

Thanks, Gobio

 

[NGiLuzzu]

Gobio,
for Shell specifications you should refer to Shell Global Solutions International:

http://www.shell.com/home/content/global_solutions/

ISO drafts instead are not for pubblic circulation until they are released as official standards, but are distributed for review and comment to the Committee Members only... so you may try to ask your national standards Institution (BS, DIN, AFNOR, UNI, JIS, GOST-R, CAS, ASRO, CSA, etc.) which should be ISO Member.


Hope this helps, 'NGL

 

[Gobio]

Thanks anegri,

I have a new question: Are the material requirements indicated in ASME (ASME II and ASME VIII for Charpys..etc) enough in order to design a valve for cryogenic service?

Thanks

 

[Gobio]

Where can I purchase a test bench for testing cryogenic valves?

Thanks

 

[NGiLuzzu]

Gobio,
as far as I know, cryogenic testing is usually performed applying flanged test caps to the valve and submerging the resulting assembly into liquid nitrogen within an insulated tank, and all is confined in a protected, dedicated area (sometimes a hole in the ground...); the above mentioned standards and specifications generally provide for a scheme of the test rig.

I believe that cryogenic testing on a bench may be feasible for very small valves only, nowadays... but I would appreciate other Members' opinion on this aspect.
I also investigated such a possibility, some years ago, withih this Forum: please take a look at thread774-127182.


Hope this helps, 'NGL

______________________________

 

[Gobio]

Anegri,

I also heard about the possibility of testing cryogenic valves pulverizing them with liquid Ni instead of submerging them into it. What do you feel about this?

 

[NGiLuzzu]

Gobio,
if you mean spraying cold nitrogen around the valve, this is usually done in order to test at "low" temperatures (typically -50°C) instead of "cryogenic" ones (-196°C).
Of course, in that case, keeping temperatures steady is more difficult than in case of immersion in liquid.

Hope this helps, 'NGL

 

[Gobio]

Anegri,

All your information helps me a lot. Thanks.

Then the only way of testig valves at cryogenic temperatures it´s submerging them into the liquid Ni within an insulated tank.

What do you know about safety requirements for the tests in these tanks?

What size are you thinking about when you say "very small valves only"? How can we test big valves?

Gobio.

 

[Gobio]

Dear Anegri,

- Do you know about any reference or standard where the test at "low" temperatures (-50ºC) by sprayind is explained?

- What do you think about the safety requirements for both tests "low temperatures by spraying" and "cryogenic temperatures by immersion".

Best regards