Can alternative gases be used to test water pipes for leaks?

[TSN79]

I work for a plumbing company, and we recently put up some copper water pipes (~30mm) in a newly constructed building. The buidling isn't finished, and since it's below 0 degrees we can't test the pipes with water. So, we though about doing it with air instead. Some people tell me that this should never be done because it's simply dangerous, while others think it's fine.

One question is how much pressure should be used, and for how long? With water we would test it for a couple of hours with 7-8 bars. Pressure during use is about 5 bars. Someone told me that air would perhaps leak out since air molecules are supposedly smaller than those of water. If so, how much (and how fast) should I allow to leak out before believing it to be some kind of bad joint somewhere...?

 

[TVP45]

Pressure testing with water is fairly safe unless you're very close to the leak. Water is only slightly compressed at a pressure of 8 bar and loses it's energy quite quickly as it expands.

Air is a different story since it is greatly compressed at 8 bar; the rupture of a pipe can send shrapnel flying.

Check your codes for what is allowed, but I think you might look into helium leak testing.

 

[brewnog]

I would strongly advise that you don't pressure test water pipes with air. The test pressures needed to prove soundness are dangerous.

I'd break the pipework up into sections with isolation valves, water-test all those you can without freezing the pipes; then look for another means as suggested by TVP, or wait until temperatures rise.

How about testing with hot water?

 

[Q_Goest]

Hi TSN,
Although testing with water is safer, it isn't necessary. What you're asking if I'm not mistaken, is simply leak tesing using air. This isn't a pressure test. There's a difference. A "leak test" is done at no more than 110% of design pressure to see if there are any leaks, whereas a hydrostatic pressure test (such as per ASME BPV Code) is a structural test done to verify suitability of welds, material, etc...

Per ASME B31.3 (Piping Code), paragraph 345, a pneumatic test in accordance with para. 345.5 may be used when "the owner (of the piping system) considers a hydrostatic leak test impractiable".

Para. 345.5 notes the hazardous release of stored energy when compressed gas is used, and suggests a number of steps to reduce risk, including ensuring there is no chance of brittle fracture. Copper pipe is not subject to brittle fracture, but any cast iron or steel fittings might be suspect at very low temperature - although 0 C or 0 F isn't generally considered very cold for even cast fittings.

When testing with air, you should have some type of safety relief set at 110% of test pressure, unless you can be sure there is no chance of overpressurizing the piping.

Para 345.5.5 regards "Procedure" and states that the pressure shall be gradually increased to 1/2 test pressure and a leak test performed. The pressure can then be raised in increments, "holding the pressure at each step long enough to equalize piping strains" (or about 1 minute). The pressure should then be reduced to the 'design pressure' before examining the leakage ...

We do this kind of thing all the time at about 400 psi for copper pipe installations used on cryogenic systems. The leak test is preferably done using a bubble leak detection fluid such as "Real Cool Snoop" liquid leak detection fluid.
http://www.swagelok.com/search/find_products_results.aspx

This fluid won't freeze in cold weather.

 

[stewartcs]

Use a glycol and water mixture to prevent freezing. I'd used at least 30% glycol, and test in sections like brewnog mentioned.

Helium is a bit extreme since the molecular size is so small, i.e. you might not get a good test with He but you probably would with water.

 

[Q_Goest]

Hi stewart,

Use a glycol and water mixture to prevent freezing. I'd used at least 30% glycol, and test in sections like brewnog mentioned.

This would work, but I would never recommend such a thing. Glycol is poisonous, so it should never be admitted to a pottable water system. There are dead ends and parts of such systems that will be very difficult to clean out.

 

[stewartcs]

Oh yeah, I forgot it was a potable water system! Thanks for pointing that out!

 

[wysard]

Glycol. Ouch! But it raises a good point.

Any solute in water lowers it's freezing point. Salt for instance is cheap, and not poisonous. It is useful if temperatures are near but below zero. Salt reaches maximum saturation in water about the -20 C point or so and is really not practical below say -10 C.

Otherwise, you might well want to go with the leak test.

 

[graphico]

I installed new PEX tubing for underfloor radiant heating. Both feed and return have not been hooked up and I plan to air test them.. Would 40 lbs be enough? There are no fittings or splices as of yet and I'm quite sure we did not nick any tubing during the process.

 

[Danger]

Glycol is poisonous, so it should never be admitted to a pottable water system. There are dead ends and parts of such systems that will be very difficult to clean out.

Maybe Scotch instead?
Just thinking out loud.

 

[Averagesupernova]

Let me point out something for those of you who recommend to avoid air testing. If NO testing is done, or testing is done with water under different circumstances, is the valve just opened allowing the pipes to slowly fill? If so, air will be trapped in various parts of the system with the same hazard mentioned previously in this thread.

 

[Q_Goest]

I installed new PEX tubing for underfloor radiant heating. Both feed and return have not been hooked up and I plan to air test them.. Would 40 lbs be enough? There are no fittings or splices as of yet and I'm quite sure we did not nick any tubing during the process.

Leak test as a function of operating pressure. If it operates at 30-35 psi then 40 psi is reasonable.

Maybe Scotch instead?
Just thinking out loud.

Let me know when you need help with your water pipes Danger. I'm there!

 

[graphico]

I tested my 1500 feet of PEX (5 runs) at 70 PSI with air and no leaks.

 

[etbrown4]

I'm aware that professional leak testers will pressurize a leaking water line with helium or hydrogen/nitrogen 5/95 and use an suitable electronic sniffer, to find a leak.

I'm curious to know if it might be practical to use ammonia, argon, nitrogen, or other inexpensive gasses in the alternative.

The reason: all local gas suppliers says helium is short worldwide and they will not sell to a new customer... Helium/nitrogen 5/95 is somewhat expensive locally at about $60 for 40 cf.

Does anyone know if alternative gasses will work and if there are affordable detectors which will sniff these gasses emanating from an underground or under slab leak?