Residential Gas Line Pressure Hold

[Mr.T]

General Question

I have about 150 total feet of newly installed black iron gas pipe w/ various joints and tee's. Mostly 1" line w/ 1/2" branches. Nothing is attached to any appliances or devices of any kind. In pressure test stage, so is not energized with gas yet. The system is currently isolated from the existing, energized system at one end by a ball valve and then to the various branches with ball valves and caps.

My question is that during pressure testing I was chasing a VERY small leak. Required pressurizing to about 90 psi to be able to finally find it. Now, to be safe, I have pressurized again to 90 psi and it is holding well but over multiple days it does bleed down a little (After a week it is down to about 30 PSI).

Is there any level of expected bleed on an effectively closed system like this? Should I consider the gauge or the schrader valve as possible sources (they are not expressing any signs of leakage when using leak check). Any other thoughts?

 

[Jeff H Young]

Soapy water is my way to find leak. we use10 pounds for 10 minutes or 15 I can tell you there is some cheating that goes on in construction . loosing 60 pounds in a couple days not good . on a 10 pound test a few pounds over night is usually ok , is the piping exposed or buried all over in impossible areas? I do high pressure tests to find leaks and then drop to 10 to see how long it holds

tempurature change causes fluxuation

 

[Sponsor]

 

[James Henry]

You've caused your own problem. Gas lines only need to be tested at 50 psi max. The higher you pressurize the pipes the more leaks you may cause.

 

[John Gayewski]

Only test 10 psi for 15 min.

I understand pumping it up to find a leak, it's actually the easiest way.

Yes schrader valves leak. During a test schrader valves should be valved off. Any rubber gasket would be suspect for very slow leakage.

 

[Jeff H Young]

i put spit on end to check schrader valve, no bubbles its good every thread is suspect plus if you done miles of screwed pipe then youve had leaks that werent on the end mid length on weld seam . fittings crack, sand holes in casting on fittings

 

[Mr.T]

Thank you to everyone for all the fast feedback. Some responses:

- I'm using Blue Monster leak check fluid
- I only pressurized to 90 psi because I was struggling to find the original leak. Ended up being at a union. (Ultimately I was able to remove the union entirely)
- The lines and joints themselves should be able to handle much more than 90 psi w/out leaking so I don't think I caused a new leak pressurizing the system that high....I hope.
- I am using a pancake compressor, so the air itself being input will be quite warm and likely moist. So I am taking in to account temperature change and cooling of the air in the lines as part of the pressure drop.
- The loss of 60 PSI was over a span of just over a week. Still unacceptable I would think.
- I tried another experiment and was able to isolate the gauge entirely so that ~95% of the system was between ball valves with the gauge just after the last ball valve. I lost 17psi over 22 hours in the isolated section. I'm sure part of that was due to cooling of the air inside the lines and the fact that the ambient temperature also dropped. But it still seems like that amount of loss would be too high

- I'm going to switch gauges to a 15 psi gauge and test 10 psi and see what the drop is over a 24hr period. Stay tuned!

 

[John Gayewski]

A smaller gauge will show a drop more clearly. That's the main reason to use 10 psi for 15 min. The other reason is temp. If your gonna do a longer test you'd want to start on the morning and come check as the space warms up, you should see a pressure gain.

 

[Jeff H Young]

good plan , good joints dont leak at 90 psi so you will not damage piping with 90 psi . weld pipe requires 60 pound test So Im not sure why a lower amount is tested on screwed pipe because thats code is my reason a pound or 2 in 5 hours might not be bad but a pound an hour is bad . cant you get to all the joints? A trick a guy showed me in testing with soapy water is a heavy amount of soap maybe 3 pts water to 1 soap whippit up reall soapy and apply with acid or small paint brush

 

[Fitter30]

Black pipe is rated to 150lbs. Most residential gas natural gas systems 7" water column = 1/4 lb. Meter side up to 2 lbs regulated down to 7" before meter. There is a big difference between diy bubbles and made for leak detection. Diy bubbles unless you see the bubbles blowing up can miss it. Store bought should leave a foam at the leak then can rechecked looking at it. Gas line in a building shouldn't very much one exposed to the sun can with a long run. Malleable black fitting can leak so if still have a leak spray the complete fitting.

 

[Jeff H Young]

oh yea like I said if youve run many miles of scred pipe a leak can be any where although less than 1 percent are not at a thread or fitting. 35 years I think i used store bought leak detector a couple times OP is using blue monster store bought rarely seen the store bought stuff used Im not a DIY we use dish wash soap

 

[Mr.T]

Update: 3 hours and it is holding strong at 10PSI. Actually went up about 0.25 psi likely due to temperature change. Will continue to monitor. If after 24 hours it holds with no bleed down, I will try 20 PSI on my 30 PSI gauge and give it one more 24 hour run. If that holds I'm calling it good and ready for inspection.

Thank you again for the words of wisdom from everyone.

 

[Jeff H Young]

if it holds 10 you are good just leave it alone maybe lightly tap side of dial with your finger if it hasent dropped at all by morning

 

[Mr.T]

Update:
After 24 hours, it was showing just a hair under 9 psi this morning, but temperatures went down quite a bit over night too. I left it for an extra couple hours and it had begun to rise to just over 9 psi. So I think I am finally good! Thanks again to everyone.

Lastly, RIP Terry. He certainly will be missed. Thoughts and thank you's to his family for helping continue to keep this forum and his memory that is forever part of it, going.

 

[wwhitney]

Update:
After 24 hours, it was showing just a hair under 9 psi this morning, but temperatures went down quite a bit over night too.

FYI, 10 psi gauge = 24.7 psi absolute (as atmospheric pressure is 14.7 psi), so 9 psi gauge = 23.7 psi absolute, and that's a ratio of 96%. Absolute zero is -460F, so if it was pressurized to 10 psi while at 70F, that's 530R (Rankine, absolute scale with the same degree size as Fahrenheit), and from simple ideal gas behavior you'd see 9 psi at 96% * 530R = 509R, or 49F.

Cheers, Wayne

 

[Jeff H Young]

Congratulations thats a good test ! just have it at or slightly high for inspectors around here they often like to bleed a little air off take note or mark with pen or sharpie perhaps walk and look at the piping if its still open or check other stuff return 10 minutes or so later . We usually dont have a test on gas piping untill final plumbing which is crazy but its after all the dry wall cabinets and finish has been set its up to the plumber to check his work or risk tearing open walls at end of job (been there done that it gets ugly and isnt fun)

 

[John Gayewski]

FYI, 10 psi gauge = 24.7 psi absolute (as atmospheric pressure is 14.7 psi), so 9 psi gauge = 23.7 psi absolute, and that's a ratio of 96%. Absolute zero is -460F, so if it was pressurized to 10 psi while at 70F, that's 530R (Rankine, absolute scale with the same degree size as Fahrenheit), and from simple ideal gas behavior you'd see 9 psi at 96% * 530R = 509R, or 49F.

Cheers, Wayne

I came up with 63 degrees based on a simpler expression of Charles law. Meaning the if the volume remained constant the temp would've dropped to 63 degrees.

I guess I'm not clear what your trying to say.

10/9= 1.11111

70/1.11111= 63

Using T1 P1
__=__
T2 P2

 

[wwhitney]

I came up with 63 degrees based on a simpler expression of Charles law.

The pressure and temperature in all the versions of the Ideal Gas law (we used the same version, which you nicely rendered as an ASCII equation) need to be "absolute", i.e. relative to vacuum for pressure, and relative to absolute zero for temperature.

You can see that this is so, because the equation as you applied it would give you 0 gauge pressure at 0F, negative gauge pressure at negative temperatures F, and eventually a gauge pressure below -14.7 psi at a very low F temperature (around -105F)--i.e. negative absolute pressure.

So for the equation you applied, it needs to be (10 + 14.7) / (9 + 14.7) on the pressure side [since atmospheric pressure is 14.7 psi, although if the elevation is high enough, that should be adjusted], and (70 + 460) / (X + 460) on the temperature side [since absolute zero is -460F.]

Cheers, Wayne