Codes and Standards for leak testing valves 12

[pipewelder1999]

What industry codes and standards are used for leak testing of valves?

Any information is helpful along with any links to informative sites.

Thanks

 

[cranekiran]

There are many:-

API(AMERICAN PETROLEUM INSTITUTE)598,MSS SP68,AWWA C-504,ANSI B16.34(IT GIVES THE TEMPERATURE VS PRESSURE RATING)

NOW,

For eg:-

you have a butterfly valve of 300 pressure class then how will you get the pressure rating,shell test pressure,seat test pressure?

you first confirm that the valve's pressure testing is as per API 598.Then go to ANSI B16.34 300# class to check the pressure rating for the particular material(body and disc).Under that you have diff. pressures for diff. temp. ranges,Confirm your medium's min to max.temperature then you can get the line pressure for your valve.Go back to API 598 and see the shell test rating(it will be 1.5 times the rated pressure)and seat test pressure(it will be 1.1 times the rated pressure).
There are lots of standards but the design always begins with the pressure class you want the valve to be.

 

[valvetech]

Here are Three of the more common ANSI Classes for valve leakage and the standard and test procedures.

Class III 0.1% of rated valve capacity
Air or water @ 50 - 125 F. 45 - 60 psig or Maximum operating differential pressure whichever is less


Class IV 0.01% of rated valve capacity
Air or water @ 50 - 125 F. 45 - 60 psig or Maximum operating differential pressure whichever is less


Class VI 1" Port Diameter
0.15 ml per minute or 1 bubble per minute Air or Nitrogen @ 50 - 125 F 50 psig or Maximum operating differential pressure whichever is less

Simplified for most popular leakage classes. For complete specifications refer to ANSI/FCI 70-2

 

[pipewelder1999]

Thank you both for your help.

 

[valvetech]

The general understanding is that the ANSI Standard for Control valve leakage allows for a specific percentage of the rated valve capacity to pass through the seats when closed. The variation in standards is from 0.15 ml per minute (1 bubble) to 0.5% of rated valve capacity from Class VI to Class II, based on specific conditions of test.

You also may want to keep in mind the following. If a valve is tested and passes under the standard it is rated to lthat ANSI standard. But if you operate it in the field at out side the exact test conditions the valve will perform differently. But it is still rated to that shutoff class. Example:
A valve tested in the factory on Air or Nitrogen at 50 - 125°F at 50 psi, and does not pass more than one bubble in one minute, will pass the ANSI test to meet the Class VI standard.
Put that valve in the field, with air at 300 psi at 200°F and the valve may pass more than one bubble per minute. Does that valve meet the Class VI standard? Yes.

The point being that when a customer purchases a valve with an expected shutoff rating of ANSI Class VI, they must under stand that under service conditions the valve may or may not perform to the standard.

 

[Burdy]

Just to add to these 'Industry' standards.
British Standard BS 6755 Pt1 specifies both pressures (including test durations) and leak rates or calls up related standards for the ambient pneumatic and hydrostatic testing of valves.
There are essentially two categories of basic designs, which are then controlled, these are 'resilient' or 'soft' seat design and 'Metal' seats. Valves are used for either control or shut off services. 'Soft' seated valves are generally specified as 'bubble' tight i.e. Zero leakage. Where as metal seats generally have a greater leakage rate, these are referenced in the individual design standard and other referenced standards i.e. For a butterfly valve BS 6755 Pt 1 calls up BS5155 the design standard which then in turn calls for BS 5146 Pt1 another test standard.
A valve for Shut off or isolation service will be treated differently to that of a control valve, as the latter doesn't generally perform both (although a large proportion have the ability to do both).
Have a look at thread408-45258 within this site; cranekiran lists all the class leakages, stating what the applications are as well.
Of course there are all the 'standards' used within the individual corporations such as SHELL and MOBIL, which can specify their own leak rates (although pressures generally remain the same).

 

[JAlton]

For Leak Testing of Pressure Relief Valves, ASME Sec. VIII, Div. 1, Para. UG-136(d)(5) requires either a published spec from the Manufacturer or API 527 for Seat Tightness Test and Acceptance Criteria.

Manufacturers typically recommend API 527. Years ago, API 527 only addressed Closed Bonnet Safety-Relief Valves. However, the Standard was recently revised to include Open Yoke Designs as well. The Standard specifies the Test Criteia and allows more leakage for Bellows Designs because the bellows may adversely affect alignment anf thereby Seat Integrity. Soft (Resilient) Seated PRVs are tested for Zero Leakage at 95% of Set Pressure. Liquid Service PRVs are tested for zero drops of liquid for 1 min at 90% of Set Pressure.

ASME Sec. VIII, Div. 1, Para. UG-136(d)(4) requires PRVs over 1" inlet size, when designed for discharge into a closed system, ie. Closed Cap or Packed Lever, to be tested in the Secondary Pressure Zone for Leakage at any possible leak point, ie. Body/Bonnet Gasket, Bonnet/Cap Gasket, Body/Nozzle Connection, etc. This test is to be conducted with a minimum of 30 psig or at the expected backpressure, whichever is greater. This test is typically referred to as a "Back Pressure" Test.

PRVs tested on Steam are expected to exhibit no visible leakage when viewed against a dark background at 90% of Set Pressure.





J. Alton Cox
President
DeLuca Test Equipment

http://www.delucatest,com

 

[NGiLuzzu]

Just for information, for those interested in European market, the relevant standards are:
DIN 3230, Part 3 (1982);
BS 6755, Part 1 (1986);
EN 1349 (2001), sub-clause 6.2;
IEC 60534-4 (1999), Amendment 1 (1986).
The last two apply to control valves and are pratically equivalent to FCI 70-2.
Regards, 'NGL

 

[NGiLuzzu]

Sorry, yesterday in my list of standards I forgot:
ISO 5208 (1993).

I agree with BURDY about the need to distinguish between stut-off and control valves, between "resilient" or "soft" seats and metal-to-metal ones.

Bye, 'NGL

 

[nfrihart]

What is the difference between BS 6755 Part 1 and BS 6755 Part 2? Thank you.

 

[NGiLuzzu]

Nfrihart,
BS 6755-1 (1986) is more general, as it regards "standard" TESTING OF VALVES - SPECIFICATION FOR PRODUCTION PRESSURE TESTING REQUIREMENTS: it is officilally superseded by BS EN 12266-1 and BS EN 12266-2 standards.
BS 6755-2 (1987) instead regards "special" TESTING OF VALVES - SPECIFICATION FOR FIRE TYPE-TESTING REQUIREMENTS and is still active.

Bye, 'NGL

 

[nfrihart]

NGL: Thank you very much for the information, it was very helpful.

Neil

 

[Khusan]

Hello,
I jsut want to know and want the whole explainatory information about Leakage Classes in Vlaves, I only know there are 6 classes but how they are divided , I want to know that..

 

[NGiLuzzu]

As Burdy said above, just have a look at Thread408-45258 within this site, where cranekiran lists all the class leakages...

 

[LSThill]

Team Members

Paulin Research Group Houston Texas 1
Flanged Joint Liquid Leakrate Predictions - DRAFT
May 26, 2003 (Rev. 2)
A.W. Paulin Jr.
Paulin Research Group

http://www.palun.con

go to download

Leonard@thill.biz

http://www.thill.biz

 

[Brayman]

Just a point BS6755 Part 2 has now been superseded and replaced by BS EN 10497 2004.

 

[PEHasan]

Valve testing - API 598

 

[JAlton]

Recent ASME Revision For Leak Testing of Pressure Relief Valves, ASME Sec. VIII, Div. 1, Para. UG-136(d)(5) used to require either a published spec from the Manufacturer or API 527 for Seat Tightness Test and Acceptance Criteria. Now you have a third option. The new 2004 ASME inserts the phrase "or another specification acceptable to the user" into UG-136(d)(5). This means an owner/uder may agree to a wet papaer towel or soap bubble test ro even an audible test. Many PRVs installed above a rupture disc have little need for superior seat tightness. Its a thought.


J. Alton Cox

http://www.delucatest.com

 

[cliff6361]

I am amazed at the lack of knowledge in this important area by many of the responders. There are many standards on the street that outline the testing requirements for pressure testing valves and other pressure retention devices. The key is the purchaser of valves. He/She should have knowledge of these standards and specifiy what it is he/she will accept.

The basic industry standards are:
ASME B16.34 (Pressure and Temperature ratings and test pressure rules)
MSS SP 61 (Good step by step guide to use to develope procedures from and provides good quantitative seat leakage acceptance)
API 598 (Provides test pressure, test sequence, hold times and seat leakage acceptance)
API 6A and D (Simular to all of the above)

Other standards will referance the above or outline testing by adding to the basic requirements as outlined in the above standards.

All above standards require each organization performing pressure testing formulat and write their own test procedures to meet the stated requirements.

One pit fall to avoid is trying to clasifiy standard on-off valve seat leakage is to specify a Classes as defined in ANSI/FCI 70.1 standard. This standard provides test methods and espected leakage rates applcable to control valves. If one takes the time to read and understand this standard they will find there is no class that provides zero leakage but many will argue different.

 

[jsummerfield]

Clif6361 makes a good point that the control valve leak test standard in not a block valve standard. I lack the latest FCI 70 standard that incorporates the IEC requirements. However, control valves are not block valves. Control valves leak. Usually we expect that block valves have zero leakage but that is not always true either.

API 598 and other standards mentioned address zero leakage, depending upon the size and soft vs metal seats, etc. This can be clarified in the valve data sheet for a small soft-seated valve by phrases such as API 598 Tested - ZERO LEAKAGE.

John