4

Context

I am adding one branch with two sub-branches to my gas line to supply a fireplace and a range stove.

I am testing the branch for leaks. I have a cap on one side of the branch, a valve shut off at the fire place, and placed a gauge at the stove output to measure pressure drop. All joints have been generously spread with gas sealant.

Test

  • I started the test early afternoon for 4 hours.
  • Pressure at 5psi with my nail gun compressor and a 15psi gauge on the gas line.
  • Dropped 1/10th of 1 psi in about 4hours (ie 0.1 psi).
  • Temperature and outdoor pressure were stable.
  • I cannot find any leaks on the branch and two sub-branches.

Questions

  1. Does gas pressure hold over time or does it lose pressure? (I am thinking of a balloon that loses pressure over time).
  2. Is it possible to lose pressure without leaks? If so why? (I have thought about climate change i.e. outside temperature and pressure)
  3. For leaks, I am using a liquid leak detector found at my local supplier. What better techniques to use to detect small leaks? (I tried soap and water also)
  4. Are small leaks ok?
Toby Speight
  • 346
  • 2
  • 11
Max
  • 595
  • 13
  • 2
    dropped 1/10th but of what 1psi or the total 5psi? – Solar Mike Feb 08 '23 at 15:48
  • 3
    Pressure will change with a change of temperature, but you said temperature was stable. Would probably require a decent drop in temperature to make a difference, more of 20 degree change. Would give it more time to see if it is a temp difference from compressor to line. If pressure still drops then a leak to be found. – crip659 Feb 08 '23 at 15:48
  • Where are you located at? Many counties, townships, etc., don't allow non professionals to mess with their gas lines. Check with your gas company. They will probably check and test at little or no charge. – JACK Feb 08 '23 at 15:48
  • 2
    Most "liquid leak detector" solutions are soap, water, and glycerin (helps with bubble persistence.) – Ecnerwal Feb 08 '23 at 15:54
  • Thank you. @SolarMike not sure if I understand your question. What do you mean by "dropped 1/10th but of what 1psi or the total 5psi?" 1/10th of 5psi is 0.5psi. – Max Feb 08 '23 at 15:57
  • 1
    So you are saying it dropped by 0.5psi. – Solar Mike Feb 08 '23 at 15:59
  • Yeah, we used to add glycerin when we were kids to make giant bubbles. I should have thought about that one. Thank you @Ecnerwal – Max Feb 08 '23 at 15:59
  • I got my 1/10th wrong. Thank you for pointing this out @SolarMike and Ecnerwal. I meant 1:10 of a 1 psi, so yeah 0.1 psi not 0.5psi. – Max Feb 08 '23 at 16:02
  • @JACK His other questions mention drywall so he almost certainly lives in the US where the vast majority of home owners can mess with their own gas lines without a permit. – Navin Feb 09 '23 at 20:35
  • Not needing a permit does not make it wise to DIY, though. Gas is one of those things that I'm sure I *could* do and am not about to touch. – keshlam Feb 09 '23 at 23:51
  • @JACK the job site is in California. – Max Feb 10 '23 at 04:15
  • @keshlam this job has a permit and is professionally inspected. – Max Feb 10 '23 at 04:16
  • ... Ah. I was assuming it wasn't since you were asking how to read the pressure test. My error. – keshlam Feb 10 '23 at 05:08

3 Answers3

10

1/10th PSI (1:50 change) might well be temperature-related (the compressor heats air as it compresses it, and that cools down to room temperature when in the pipe) but simply waiting longer should confirm that. 1:10 change (0.5PSI) is more significant than expected without a drastic temperature change.

A balloon is porous (at the scale of helium molecules) and helium leaks out through it. A gas pipe should not be porous at the scale of air or gas molecules, so any continuous drop in pressure is a (perhaps small) leak somewhere. But changes (both down and up) from temperature differences do happen.

So if you check tomorrow and it's down more, you have some sort of leak. If it's pretty much the same, you don't.

You may need to figure a way to check the closed valve at the fireplace.

Small leaks are Definitely Not OK in gas pipes.

Not really home-owner friendly, but the combination of filling the pipe with helium and using a detector that can sense helium is a "better leak detection" method for troublesome leaks. Generally the bubbly stuff and a stable gauge reading are adequate, as that's an expensive rig you generally need to hire with an operator.

Ecnerwal
  • 174,759
  • 9
  • 212
  • 440
  • 2
    I retested for 4 hours in the morning, and this time the pressure went up due to temperature raising. If I leave overnight, it should drop because nights are cooler. So it is a bit tricky to leave for a long period of time while temperature is fluctuating. – Max Feb 09 '23 at 04:59
  • I have checked several times all connections, no bubbles. – Max Feb 09 '23 at 04:59
  • 3
    @max maybe you have to measure pressure at the same time of the day to eliminate temperature problem. Maybe dead early in the morning? – vasin1987 Feb 09 '23 at 12:28
  • 1
    What's with the *4 hours* thing? Did you not leave it pressurized overnight and just check the reading the next day? The pipe should hold pressure indefinitely, so leave it alone (pressurized) and if it falls over the course of a day or two, it's leaking. – Ecnerwal Feb 09 '23 at 14:09
  • 4 hours is the timeframe I could test during which the outside temperature is stable. If I test for 48 hours, I need to start the test with an outside temperature that the weather will cycle back to within that timeframe. The temperature between day and night these days in CA is significant enough to change the pressure in the pipe to over 1psi, making the reading really hard. That said, I will do it again with a thermometer and a 48 hours timeframe. Thank you :) – Max Feb 09 '23 at 16:29
  • 2
    If you simply check the pressure regularly while testing, you'll either see a downward trend, or variation around a stable value, as the temperature varies. – Ecnerwal Feb 09 '23 at 16:49
6

Physics!

PV = nRT

P is pressure (5psi), V is volume (static [if you don't have a leak!] - the amount of gas your length of pipe holds), nR is also constant. T is the temperature (in Kelvin).

So, pressure is directly correlated to temperature, and to lose 0.1psi from 5 (2%), would require your temperature to drop by 2%. 20C is 293K, so 2% is likely to be less than 6C or 11F. If you think your temperature is changing that much, you probably don't have a leak... I wouldn't risk my life on it though.

Auspex
  • 176
  • 3
  • Cool. High-school chemistry! How about atmospheric pressure, no effect? – Max Feb 09 '23 at 16:31
  • 1
    @Max Beats me! Atmospheric pressure can't have any effect if you _don't_ have a leak, but I don't know if it could mask a leak by making the pipe's pressure change seem totally related to temperature change. This is why I wouldn't risk my life on it! – Auspex Feb 10 '23 at 11:35
4

No , it should not leak. there may be pressure changes caused by temperature as answered. A small leak in a confined space like a wall can eventually cause an explosion as the gas accumulates and mixes with air.

blacksmith37
  • 8,143
  • 1
  • 14
  • 30