Membership questions? Log in issues? Email info@brewersassociation.org

Author Topic: Testing for a CO2 leak  (Read 22022 times)

Offline epb

  • 1st Kit
  • *
  • Posts: 11
Testing for a CO2 leak
« on: April 16, 2010, 07:16:12 pm »
I've had some trouble finding a CO2 leak in my system lately and am trying to test everything out to make sure I've found it.  My first question is about my internal pressure.  I've heard from a lot of sources that internal pressure doesn't matter, but I've noticed that it's been below the "Order CO2" line on my regulator when I've had a leak.  Right now, it's in the fridge and around 450 psi.  Yesterday, it was around 500 psi, but has been in and out of the fridge since then.  Assuming that my internal pressure is OK, what I've done is I've hooked my my regulator to a keg and then turned off the CO2 at the tank and at the regulator.  I've heard that if I leave it overnight and the serving pressure remains the same (it's currently at 14 psi), it's likely I don't have a leak.  I just want a second opinion.  Does that sound right?  I've already sprayed everything with soapy water and cannot find a leak so I'm wondering if there's anything else I can do.  I appreciate the help.  Thanks.

Offline rjharper

  • Brewmaster
  • *****
  • Posts: 646
    • Angry Scotsman Brewing
Re: Testing for a CO2 leak
« Reply #1 on: April 16, 2010, 10:11:06 pm »
As someone who's run a chemistry lab for almost a decade, I've dealt with my fair share of compressed gas leaks.  You are correct that if you pressurize a closed system, it should hold pressure even if the supply is turned off.  So yes, if you connected an empty keg, took it to 14 psi, then closed the valve on the tank, then it the line should hold that.  Now if the keg is full of beer or water, thats another matter (it might absorb the CO2).

Leak detecting is a slow and painful process.  If your line doesn't hold pressure, this time re-pressurize the keg, and close the valve on the regulator instead.  Continue to work your way down the line in a methodical process.  You're also correct that at fridge temps the high pressure dial is useless.  Note the tare weight on the tank, then go weigh it on a set of bathroom scales and note the difference - I use that process to gauge how much gas I have left.

Offline epb

  • 1st Kit
  • *
  • Posts: 11
Re: Testing for a CO2 leak
« Reply #2 on: April 17, 2010, 07:47:53 am »
I just checked and my serving pressure is still at 14 psi.  I turned on the gas at the regulator and it immediately went into the keg and the serving pressure dropped to about 13 psi.  The keg is full and as you pointed out, the beer might absorb the CO2.  The way I see it, I didn't learn anything from this experiment other than there's no leak between the regulator and the CO2 tank. 

In regards to the internal pressure, it sounds like that's all dependent on the temperature of the CO2 tank as well as it's tare weight.  Is that accurate?  I ask because I'm still trying to make sense of why it would have dropped with my original tank when I got a leak.  The tank was normally at about 550 psi and as soon as I hooked up another keg, it dropped to about 450 overnight and then further down later in the week before finally hitting 0 when the tank was empty.  I got another tank, thought I found my leak so I hooked it up and it immediately went to about 450 again for the better part of the week before hitting 0 when the keg was empty.  Now I'm on my third tank and it's around 450 again so I'm really paranoid here.  Does anyone know the science behind it all and explain why I've seen these particular readings?

Offline a10t2

  • Official Poobah of No Life. (I Got Ban Hammered by Drew)
  • *********
  • Posts: 4696
  • Ask me why I don't like Chico!
    • SeanTerrill.com
Re: Testing for a CO2 leak
« Reply #3 on: April 17, 2010, 08:05:57 am »
The tank pressure depends *only* on temperature, as long as there's some liquid CO2 left. Which is the case until the tank is about 90% empty. If your regulator is reading 450 psi, you're either nearly out of CO2, the tank is extremely cold (about 24°F), or the gauge is defective.

Disconnect everything except the regulator, let it sit for a couple hours, and see what the gauge reads. Regardless, if you're running out of CO2 in a week you have to have a leak somewhere. (Assuming you have a 2.5 lb or larger tank.)
Sent from my Microsoft Bob

Beer is like porn. You can buy it, but it's more fun to make your own.
Refractometer Calculator | Batch Sparging Calculator | Two Mile Brewing Co.

Offline epb

  • 1st Kit
  • *
  • Posts: 11
Re: Testing for a CO2 leak
« Reply #4 on: April 17, 2010, 08:28:19 am »
The tank pressure depends *only* on temperature, as long as there's some liquid CO2 left. Which is the case until the tank is about 90% empty. If your regulator is reading 450 psi, you're either nearly out of CO2, the tank is extremely cold (about 24°F), or the gauge is defective.

Disconnect everything except the regulator, let it sit for a couple hours, and see what the gauge reads. Regardless, if you're running out of CO2 in a week you have to have a leak somewhere. (Assuming you have a 2.5 lb or larger tank.)

Thanks.  Let me give you some more back story on the issue and see what you think.  I had actually left the CO2 on and connected to a keg for about 9 hours.  Before doing this, it was at room temp and the internal pressure was at 700 psi.  After sitting in the fridge all day, it dropped to just over 500 psi.  I took it out wanting to see what would happen if I left it at room temperature again, but also turned off the gas in an effort to test a leak in my tubing.  Overnight, the internal pressure remained the same as well as the serving pressure.  I turned the gas back on and the internal pressure immediately jumped back up to 700 psi while the serving pressure remained at 14.  The way I see it, my tubing is OK.  That's when I put it back in the fridge, hooked it up to the keg for a half-hour and then turned the gas off.  When I got home, the serving pressure was still 14 psi, but the internal was down to 450.

With all that in mind and the fact that this is brand new 5 lb CO2 tank that's only been connected to a possibly leaking keg for 10 hours total, is it likely I've already used up 90% of it?  I'm going to go disconnect my keg and leave the gas on and see what happens.  Should I also turn the gas off at the regulator so it doesn't make it's way into the tubing?

Edit:  I just turned the gas back on and my serving pressure immediately jumped up to 17 psi.  I don't know why it would be that high.  I didn't have the locking nut twisted tight so maybe that's it.  Don't know if this means anything at all, but FYI.
« Last Edit: April 17, 2010, 08:34:07 am by epb »

narvin

  • Guest
Re: Testing for a CO2 leak
« Reply #5 on: April 17, 2010, 11:17:12 am »
If your regulator is reading 450 psi, you're either nearly out of CO2, the tank is extremely cold (about 24°F), or the gauge is defective.


If your tank is on the hump in the back of a mini-fridge, it's sitting right up against the cooling element, so it will be very cold.  My high pressure gauge usually reads about 450 - 500, even when full.

Offline epb

  • 1st Kit
  • *
  • Posts: 11
Re: Testing for a CO2 leak
« Reply #6 on: April 18, 2010, 08:43:23 am »
If your regulator is reading 450 psi, you're either nearly out of CO2, the tank is extremely cold (about 24°F), or the gauge is defective.


If your tank is on the hump in the back of a mini-fridge, it's sitting right up against the cooling element, so it will be very cold.  My high pressure gauge usually reads about 450 - 500, even when full.

It is and I noticed it's quite cool back there.

I checked my readings last night after turning off the gas and disconnecting the kegs and it was at 14 psi.  I let it site overnight and it was down to 12 psi.  I turned on the gas to see if it would increase and it didn't.  I then let all the pressure out through the blowoff valve and then turned the gas back on.  This time it went up to 15 psi.  I don't get it.  Maybe my regulator's flaky. 

If I do have a leak in the tubing, submerging it probably shouldn't be too hard.  I know I'm not supposed to put the regulator itself underwater so maybe just get it to where it connects with the tubing and leave all the dials above water?  Do I leave the gas on the entire time?  How quickly would I see a leak?

I'm probably just being super-paranoid at this point, especially since I just put the tank on the bathroom scale to weigh it.  The tare weight was 3.5 kg which comes out to about 7.7 lbs and the weight on the scale was a little over 12.  Pretty much exactly where I want it to be with a full-ish tank.  I really should just stop worrying about it at this point, but I've wasted so much money on CO2 and want to be absolutely sure I'm leak-free.

Offline a10t2

  • Official Poobah of No Life. (I Got Ban Hammered by Drew)
  • *********
  • Posts: 4696
  • Ask me why I don't like Chico!
    • SeanTerrill.com
Re: Testing for a CO2 leak
« Reply #7 on: April 18, 2010, 12:36:04 pm »
Forget about soapy water, submersion, StarSan, etc. Just buy a bottle of this stuff: http://www.northernbrewer.com/brewing/gas-leak-detector.html It will last forever, and over the life of the kegerator, probably pay for itself.
Sent from my Microsoft Bob

Beer is like porn. You can buy it, but it's more fun to make your own.
Refractometer Calculator | Batch Sparging Calculator | Two Mile Brewing Co.