Pressure testing piping component pneumatically 2

[A_R]

After 3 attempt pressurizing pipe system to test pressure, it was still leaking at flange connection. Planning to preform another one after some modification in pipe system. Wanted to make sure that we do not put to much stress on pipe and welds.

 

[TGS4]

There is no prohibition or requirement for this. It is not addressed in any ASME Code.

 

[LittleInch]

Agree no code limit but this is not a good idea.

Explain why you want to do this.

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[DekDee]

If you cannot stop leaking during the test what will happen when it is put into service ?

 

[moltenmetal]

If the question is how many times a pressure retaining system may be pneumatically tested to 110% of its design pressure, the answer is as many as you like. There is no prohibition against the pneumatic testing of a component beyond a certain number of cycles to my knowledge, though I will be happy to stand corrected if someone can point to one.

Note that a leaking flange is not necessarily grounds for failure, as long as the flange seals at design pressure.

 

[MJCronin]

AR..

It is difficult to answer complex questions with such little information from the OP

How about a little more information ? Are you pneumatically testing at 10 psig or 1000 psig ?

Size matters too ...Is your system small as a mouse or big as a house ?

MJCronin
Sr. Process Engineer

 

[LittleInch]

What code, pressure and % of SMYS are we talking here?

If it's 31.3 then the test pressure at 1.5 x MOP is still below SMYS, so therefore there should be no material damage.

However if you have a fatigue issue any where in the system then you will be eating into the fatigue reserve by a high stress low number of cycles thing going on.

Normally flange connection leaks you just clean off and re-make or tighten up during the test (a large hammer and spanner never goes amiss....)

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[moltenmetal]

Test is pneumatic, so 1.1 x design pressure rather than 1.5x, and no ratio of stresses for temperature either, so really no big hairy deal. But LittleInch is correct- we're assuming a code (B31.3 in this case) and that assumption may not be valid.