Fire Water Main Leaking Water But Not Air

[Battlebreaker]

At our construction jobsite we have approx. 1,000FT of 8" fire water from the city main running into our existing building. The existing line was insatlled in 1971. We installed a new backflow device as required by the city at the start of the line. The fire department required us to pressurize the existing line to 200PSI for 2hours. We failed. Over the course of 8 hours the pressure gauge dropped from 200 to 50PSI.

When we called a leak detection company and they pressurized with helium and air to 110PSI, it held for 26 hours until we had to release the air and put water back in. We tried the water test again and failed.

How could it hold 110PSI of air but not 110PSI of water? If there is air trapped in the line when we do our (Hyrdo) water test, could the air be compressed, causing the gauge to drop? I would think it would leak air faster than water, but the opposite is happening. I'm at a loss.

Leak detection guy says he can't find the leak if no air is leaking. FD will not accept air test.

Thanks.

 

[Averagesupernova]

The pipe is leaking air too. The difference is that a pipe full of water (incompressible) will lose noticible pressure with the most miniscule of a leak. Like a couple of drops. But a gas is very compressible so a large volume of it can leak quite a lot relative to the liquid before it is noticed on pressure readings.

 

[boneh3ad]

Going off of what Averagesupernova said, do you have access to gauges that are more accurate and can register changes in pressure of, say, 1 torr or less? That might help you out.

 

[S_Happens]

Hydrotesting correctly can be difficult. Are you able to blind/isolate to test just the new installation or certain sections at a time, or are you having to hydrotest the entire system? When valves are located inside the hydrotest it can really be a pain because even tiny packing leaks cause problems.

You have a large volume of piping, with a small leak. It is for this reason (along with significantly improve safety) that water is used vs a gas and that the FD won't approve the air test. To put what has already been said very simply, it's not that the same amount has to leak out to get the same pressure drop. It's exactly the opposite. Only a tiny amount of water has to leak out to change the pressure whereas with air, a very significant amount has to leak out (think drops of water vs enough air to fill the entire volume of piping at atmospheric pressure).

 

[Yuri B.]

Sadly the line obviously being under the ground (not seen). Otherwise, I may be wrong however, but there should be additives for the water, very clearly seen in the light of an ultraviolet lamp.

 

[Battlebreaker]

Thanks everyone for your response. This makes sense. We are going to keep the air in the system through the weekend to see if it drops by Monday. 110psi of air input 8AM on 5/10/2012. 110psi still holding at 4:30PM on 5/11 (1,000ft x 8" line)

If there is a pin hole, I wonder how long it would take for the gauge to drop 5 to 10psi of air. I will see what it looks like on Monday.

Meanwhile building will be under fire watch since there is no water.
Thanks.
Ed

 

[256bits]

Thanks everyone for your response. This makes sense. We are going to keep the air in the system through the weekend to see if it drops by Monday. 110psi of air input 8AM on 5/10/2012. 110psi still holding at 4:30PM on 5/11 (1,000ft x 8" line)

If there is a pin hole, I wonder how long it would take for the gauge to drop 5 to 10psi of air. I will see what it looks like on Monday.

Meanwhile building will be under fire watch since there is no water.
Thanks.
Ed

You already know there is a leak, so why re-pressurize with air??