Have a vendor who quoted me 'bubble tight shutoff' for several full bore ball valves.In-line pumps No mention even of FCI 70-2 (Classes I thru VI for control valves - some ball valve companies do) or even API 598 on his literature. I looked through all the specs and can't find a definition of the Bubble Tight Shutoff term. Magnetic pumps(I've 'googled' it to death.)
If can relate the term to a 'spec/standard/test' at all ?

Bubble tight means different things to different people, if your buying a valve from these guys specify the test you want to conform to there are several classifications that are used for valve tests as you have mentioned. Go back and have them confirm they conform to your requirements. That said my experience is bubble tight is synonomous with class V test level (but it wouldnt stand up in a court).Suber-water pumps

The term "bubble tight seal" is very old.  My understanding is that it originated from a test where a valve was installed in an air system with an open end pointing up.  Some soap was poured into the open end and air was applied to the other side.Screw pumps  The valve passed if no bubbles formed in the soap.  I'm not sure if this is true or not, but it was the only explanation I ever heard that was at least plausible.

Bubble-tight is a relevant term if applied to a certain set of conditions.  It must be explained that way.  For example, if you do an seat leakage test with air as permitted by API 598, ASME B16.34, MSS SP-110, and other standards, you can place a layer of water over the top of the seat being tested and look for free flowing bubbles.  No bubbles during the test duration means it was "bubble-tight" and the valve passes the test.  You will waste a lot of time testing tight sealing valves if you try to measure minute amounts of leakage with flow meters.Self-priming pumps

It is important to recognize that the factory test is strictly for quality.  Are all the parts installed properly, are they manufactured correctly, etc.  Never try to correlate the factory test with proof of performance.  there are way to many variable and assumptions involved that the results are not vaild.

MORE NEWS