Testing Requiremts of ANSI B31.3 vs Testing standards of API-598 1

[valveman30]

I have seen many conversations related to valve testing requirements of API-598, MSSP61, API-6D and ANSI B16.34. Additionally I have seen as many comments related to testing requirements complying to ANSI B31.3. However, anyone, please tell me where in ANSI B31.3 does it dictate testing standards that supercede accepted standards of testing for API-598, MSSP61 or ANSI B16.34?

 

[BigInch]

It's not in B31.3 The code requires only what it requires and those other things you have seen may be specified in addition to the B31.3 requirements, or not, at the discretion of the buyer.

We are more connected to everyone in the world than we've ever been before, except the person sitting next to us. Lisa Gansky

 

[valveman30]

Thank you Big Inch for your reply, I have reviewed many threads with your comments concerning ASME B31.3; this discussion has been on going for a long time. To get to the point, WHY would any engineer state he does not care what API-598 states as required testing of his steam piping! His following is ASME B31.3,B31.4 and B31.8. In all past reviews and to my understanding, ASME-B31.4 acknowledges the specification of design per API-600. Hence forth, per API-600 design, accepted testing criteria is to API-598 and ANSI B16.34. Any testing beyond the stated requirements of API-598 & ANSI B16.34 must be stated in the customers request and purchase order. Expectations of a valve to be exposed to high pressure testing beyond the stated testing time of API-598, will have an adverse effect on the valve packing, function and service life. Yours any any comments regarding extended testing of a standard API-600 design valve is welcomed and encouraged.

 

[BigInch]

I'm looking at B31.3-2008 (still in the process of updating) and there is no reference to 598 in that. There is one to 599.

B31.4-2006 is referencing 598 under Appendix C. In the large print it says,

"NONMANDATORY APPENDIX C
PUBLICATIONS THAT DO NOT APPEAR IN THE CODE OR
NONMANDATORY APPENDIX A BUT MAY BE OF
INFORMATIONAL BENEFIT" (The red bold is my emphisis.)

B31.8 no reference to 598/599/600 in the 2007 edition.

To my knowledge, or lack thereof, these are specs issued to be in conformance with corresponding ISOs.

In the past, whenever we needed extra assurance, valves for pipeline applications were speced at our option to the API 6D spec, "PIPELINE VALVES". So I'm not totally sure why that isn't being used in B31.4 & .8 pipeline applications instead of 598.

Nobody should be using B31.4 & .8 for steam.

I'd have no problem spec'ing valves for extended testing, if it is necessary, however at least some justification should be provided for why it is necessary to order them with above normal testing requirements of the code, whatever that might be. Its probably not too hard to justify, if you want it.

An example might be for offshore service where easy maintenance is impossible, or at the least, very expensive and/or where no leakage is critical to safety or product contamination, etc. I can tell you that it is no picnic to have a valve fail offshore, even in shallow water. The manufacturer's guarantee to replace the valve, only covers the valve parts that failed, NOT the small army of divers, equipment and operators you have to send out there to do it. Not very good when you get back 25,000 on 1,500,000 in total replacement expenses.

Where extra testing is not justified, I view extended testing like burning in your computer for a year before you touch the keyboard. A waste of time and money, but in today's environment, that might be a smaller waste than what you get even if it fails in noncritical service. Its all about "precived risk today" and how big a meal ticket the lawyers can make out of it later, should anything happen. If its on your own property in an obscure location, maybe a plain vanilla valve is OK, otherwise, be cautious. A steam valve in a tunnel with risk to human traffic passing by, off the plant's plot plan... its starting to go critical.

As for why it might not be wanted, well its just as easy to find reasons not to do it. It is extra time and cost... and it was still not mandantory as of the code editions that I just looked at.


We are more connected to everyone in the world than we've ever been before, except to the person sitting next to us. Lisa Gansky

 

[valveman30]

Thank you Big Inch, your information is very helpful and useful. What is your opinion on testing a valve for example API-600 design valve, class 600# at 1950 psi for one hour. The valve is being tested with the piping system; the system will be used for steam service. I contend that the use of a commmodity valve purchased from any supply house or manufacture should not be tested for any durantion of time in excess of the valve's rating. In this case the class 600# valve should not be tested past 1480 psi. The manufacture told me they will not support a warranty for the valve when it is tested beyond the specified time stated within API-598. Additionally they stated that pressure testing the valve beyond the alloted time may have effects on the structure of the valves packing and bonnet to body connections. I support the statement of the valve manufacture, yet I am concerned that if the valve is tested at this elevated pressure for the duration they desire, it may be a safety issue to the plant operators presently or in the future. API-598, states the shell of the valve may be tested at 2250 for 1 minute. The manufacture told me that if the end user wants testing times beyond those of API-598, it must be clearly stated in the purchase order and special procedures will be taken to acomplish this test. This is an issue engineering has ran into on many times, the vendor stating that the valve should not be tested beyond the standards of API-598, API-6D or ANSI-B16.34 unless special conditions have been made in there purchase order and orders of TPI.

 

[BigInch]

I would test the valve's body to at least 1.25 x rating for as long as I wanted, 4, 8, or 24 hours, or as long as I needed to to get a passing pipeline test chart. I wouldn't let a manufacturer tell me that the valve should not be tested past its rating pressure, at least up to 1.25 X, even if I bought it off the shelf. That's equivalent to no safety factors in the design, so if that turned out to be true, I'd rather burst it during a test, rather than wait for later on when in service.

Just between us boys, I've seen valves off the shelf that appear to be exactly the same as valves that came with the specs, but if you order them spec'ed at purchase, the manufacturers don't get any wiggle room. Of course they charge for that, but you're paying for the CYA factor, which can be worth a whole lot more than what they charge.

I'd generally order off the shelf, unless as I mentioned, if its a critical service valve, ESD, most offshore appliations, mainline block and isolation, or some similar situation. For me personally, no need to high spec for the typical in-the-plant valve, unless you can write down some kind of justification for it, but company specs do go overboard at times and you have to stick to them or ask for a variance. Chances are that if a company spec requires them, they had a reason to write it that way, so (if anyone remembers why) you won't get a variation anyway.

We are more connected to everyone in the world than we've ever been before, except to the person sitting next to us. Lisa Gansky