API 607

[joeswoes]

I want to submit a ball valve for fire testing per API 607. As the water is trapped inside the closed ball during the test, and thermal expansion can lead to a bomb, wouldn't the ball require an upstream vent and thus be a uni-directional product ? Also, our valve is non-metallic. Metallic valves are allowed by API 607 to be welded to the mating pipe to eliminate leakage at the connections. This, however, is not an option for us. API 607 does not allow insulation anyway on the valve during test, so insulating the valve-pipe connection is also not an option.

Any thoughts from the forums valve fire-test experts ?

 

[NGiLuzzu]

joeswoes,
may I ask which edition of API 607 standard are you testing to?

(Please take a look at thread182-133010, thread408-131095 and thread809-133009 within this website).

Thanks in advance, 'NGL

 

[joeswoes]

current, I assume to be 5th Ed.

 

[NGiLuzzu]

joeswoes,
I've got ISO 10497 (2004) standard in my hands, but - as far as I know - it is equivalent to 5^th edition of API 607 (and the paragraphs numbering should be the same, too).

Please refer to clause 5.3.2.8 which says that a pressure relief provision must be "incorporated in the system (...) to protect against potential rupture of the valve if it is designed such that liquid can be trapped in the cavity".
See also Figure 1, details number 14 and number 16.

Hope this helps, 'NGL

 

[bcd]

You mention the valve is non-metallic. I assume this means the seats and seals, and not the body. If just the seats and seals, heat the valve as quickly as possible to melt the seats which will minimize the increase in cavity pressure. For pipeline seal,on flanges, use sprial-wound graphite type gaskets (Flexitallic). For thread end, use graphite tape and graphite based lubricant and tighten as much as possible without splitting the connection. This will minimize the external leakage.

bcd