Co2 leak, how do you know?

Homebrew Talk - Beer, Wine, Mead, & Cider Brewing Discussion Forum

Help Support Homebrew Talk - Beer, Wine, Mead, & Cider Brewing Discussion Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

brewNYC

Well-Known Member
Joined
Oct 4, 2017
Messages
144
Reaction score
52
I’m trying to figure out if I have a small co2 leak. I kegged my beer and stuck it in the fridge with the gas on 20psi to cold crash and start carbing. The next morning, I stuck my head in the fridge to check the gauge and felt lightheaded. I turned off the gas (valve at the tank closed, regulator open). Came back 8 hours later and the tank gauge showed zero (no surprise) and the keg pressure gauge showed 15psi. After 36 hours with the gas off, the keg pressure dropped to 5 psi. Last night, I turned the gas on, sprayed everything with soap or starsan (no bubbles), then closed both the tank valve and the regulator valve. It held 600psi overnight, so I guess I’ve ruled out leaks in the high-pressure side.

So, either I have a slow leak on the low pressure side, or I don’t have a leak and the lightheadedness wasn’t from co2. Any good way to figure this out, since I’m not getting any bubbles when I spray everything down?

I assume the drop from 20 psi to 5psi with the gas shut off for 36 hours was just the co2 being forced into the beer. I have 2.5 gallons of beer in a 3 gallon keg, so not too much headspace.
 
It is possible to have a slow leak and not be able to see any bubbles by spraying it. I had to submerge everything to see any bubbles. I took one of my old fermenter buckets, filled it with water, and put the lines all the way to the regulator valve in the water. Then sat and watched for bubbles. A very small stream of bubbles started soon after at one of the hose barb connections. But, taking it out and spraying it didn't show bubbles.
 
It is possible to have a slow leak and not be able to see any bubbles by spraying it. I had to submerge everything to see any bubbles. I took one of my old fermenter buckets, filled it with water, and put the lines all the way to the regulator valve in the water. Then sat and watched for bubbles. A very small stream of bubbles started soon after at one of the hose barb connections. But, taking it out and spraying it didn't show bubbles.
Hmm, good to know. Do you crank the pressure up before you submerse, or just leave it like it is?
 
What kind of clamps/connections are on your gas lines? IME, worm clamps suck for being 100% leak free. The only clamps I've used that have zero leaks 100% of the time are oetiker clamps (they clamp 360 degrees, zero gaps). I use those for everything except the lines from my glycol chiller at the chill coils (those are PTC).

If you put the fittings together for your gas side (like a manifold), you could have a small leak there. I've found that you need to wrap more pipe tape than you might think on the threads to get a solid seal. Luckily those are usually "one and done" type setups. A while back I did have small leaks in a gas manifold upon arrival. I took it apart, applied new pipe tape, put it back together and the leaks were gone.

If you have shutoffs on the gas lines, you can also try going through all of those (assuming a manifold) to see which one is the offender (if not all of them).

I would also check the keg poppets and other seals to make sure there's no leaking there.

BTW, the easiest way I found to check a gas manifold is to submerge it in water, make sure there's no trapped bubbles from doing that, and then watch it when you apply CO2 to the inlet. Obviously you'll want to close all the valves on the manifold first. You can do the same method for at least some of the tubing to fitting connections.
 
What kind of clamps/connections are on your gas lines? IME, worm clamps suck for being 100% leak free. The only clamps I've used that have zero leaks 100% of the time are oetiker clamps (they clamp 360 degrees, zero gaps). I use those for everything except the lines from my glycol chiller at the chill coils (those are PTC).

If you put the fittings together for your gas side (like a manifold), you could have a small leak there. I've found that you need to wrap more pipe tape than you might think on the threads to get a solid seal. Luckily those are usually "one and done" type setups. A while back I did have small leaks in a gas manifold upon arrival. I took it apart, applied new pipe tape, put it back together and the leaks were gone.

If you have shutoffs on the gas lines, you can also try going through all of those (assuming a manifold) to see which one is the offender (if not all of them).

I would also check the keg poppets and other seals to make sure there's no leaking there.

BTW, the easiest way I found to check a gas manifold is to submerge it in water, make sure there's no trapped bubbles from doing that, and then watch it when you apply CO2 to the inlet. Obviously you'll want to close all the valves on the manifold first. You can do the same method for at least some of the tubing to fitting connections.
Hmm, good to know! Yeah, I’m using good old worm clamps from the hardware store. With that said, I’m using 1/4” hose, which is a pretty darn tight fit. I figured the clamps were just extra insurance.

The regulator is an out-of the box deal. I’m a little nervous about dunking it - I’m sure all the valves and such would be ok, but I don’t want water in my guages.

mostly just ticked that the whole “spray it with soap” thing seems to be widespread nonsense.
 
Place a container of water to submerge the fittings below the gauges. Like washing your undercarriage. ;)

The "spray it with X" works, IF the leak is large/fast enough. Slow leaks can be more tricky.

I've had even "tight" tubing on barbs leak before. Well, when I had worm clamps in use that is. Since I switched clamps, zero leaks at those points.

Is the regulator at least a quality brand? I've been using Taprite for all my CO2 and nitro mix tanks/bottles and not had any issues there. At least not ones that a simple rebuild kit didn't resolve.

Since it sounds like you had no leaks when the feed from the regulator was turned off, it has to be a fitting/connection down stream from there. Are you using swivel nuts for the tubing connections or is everything barb fittings?
 
If you disconnect the CO2 source from a keg that is not fully carbonated, the pressure in the headspace will drop as the beer absorbs CO2 from the headspace. This is normal, and sounds like what you observed. No leaks required.

Brew on :mug:
 
if you have a somewhat acurate bathroom scale. you can always take the reg off, weigh the cylinder, hook it all back up. then weigh it again in a couple days?

i THINK that would work....i keep my co2 bottle on a scale 24/7....i know i have a leak when it drops more then an ounce in a day.
 
A
If you disconnect the CO2 source from a keg that is not fully carbonated, the pressure in the headspace will drop as the beer absorbs CO2 from the headspace. This is normal, and sounds like what you observed. No leaks required.

Brew on :mug:
Agreed, I’m working off somewhat anecdotal evidence here- meaning I stuck my head in the fridge and got dizzy. Dropping to 5 psi in 36 hours doesn’t mean I have a leak, and tells me that if it is leaking, it’s slow
 
If you remove the keg from the system, charge it, turn off the tank, you'll know (within a few hours) if you have a CO2 leak. Just note what your gauges read when you start this test. Let it go at least overnight to see IF you have an actual leak or not.
 
Place a container of water to submerge the fittings below the gauges. Like washing your undercarriage. ;)

The "spray it with X" works, IF the leak is large/fast enough. Slow leaks can be more tricky.

I've had even "tight" tubing on barbs leak before. Well, when I had worm clamps in use that is. Since I switched clamps, zero leaks at those points.

Is the regulator at least a quality brand? I've been using Taprite for all my CO2 and nitro mix tanks/bottles and not had any issues there. At least not ones that a simple rebuild kit didn't resolve.

Since it sounds like you had no leaks when the feed from the regulator was turned off, it has to be a fitting/connection down stream from there. Are you using swivel nuts for the tubing connections or is everything barb fittings?
It’s northern brewers co2po single body regulator. Not top-notch, but I’d assume good enough to be fairly reliable.

yes to swivel nuts. Here’s a photo, cause it’s worth…
F5F9A779-AEA4-4185-865D-77A1EFB2FCD4.jpeg
F1F96528-B0A4-4DF4-860B-4195C6928663.jpeg
 

Attachments

  • E94F5899-7FE7-4826-8634-E7DB31540F09.jpeg
    E94F5899-7FE7-4826-8634-E7DB31540F09.jpeg
    1.1 MB · Views: 2
If you remove the keg from the system, charge it, turn off the tank, you'll know (within a few hours) if you have a CO2 leak. Just note what your gauges read when you start this test. Let it go at least overnight to see IF you have an actual leak or not.
This only works with an empty keg, or a keg of beer/water that is already carbonated to a level in equilibrium with the pressure you supply.

Brew on :mug:
 
If you remove the keg from the system, charge it, turn off the tank, you'll know (within a few hours) if you have a CO2 leak. Just note what your gauges read when you start this test. Let it go at least overnight to see IF you have an actual leak or not.
Well, I did charge it, close the valves and leave it over night with no leaks, so I know it’s not on the high pressure side. I guess doing the same thing with the valve open /keg disconnected would tell me if it’s a hose or fitting. Combine that with dunking the keg in a bucket, I should be able to figure it out. Good tip.

try legging, they said - so much easier than bottling!
 
I meant charge the gas SYSTEM, leaving the keg disconnected from the CO2 line. That way the keg is NOT in the mix, removing it from the equation. IIRC, this is "do I have a CO2 leak" resolution 101. Basically, with no kegs connected, you're checking the integrity of your connections.

I would still advise ditching worm clamps anywhere you have them. Oetiker clamps are 10000x better at clamping without leaks.
 
I meant charge the gas SYSTEM, leaving the keg disconnected from the CO2 line. That way the keg is NOT in the mix, removing it from the equation. IIRC, this is "do I have a CO2 leak" resolution 101. Basically, with no kegs connected, you're checking the integrity of your connections.

I would still advise ditching worm clamps anywhere you have them. Oetiker clamps are 10000x better at clamping without leaks.
Yeah, I misunderstood what you were saying.

This method rules out leaks up to the disconnected QD. It won't find leaks due to a bad fit of the QD to keg post, or any leaks in the keg itself. Testing with an empty keg does an end to end test, but uses about 80 g (~ 3 oz) of CO2 if you test at 15 psi (or 40 g if the keg is filled with air to start.)

Brew on :mug:
 
IME, seeing if the keg has a leak can be done after you check the gas system (minus keg). Usually you charge to a set pressure (I typically use 10psi for this) and leave it at least overnight, or for a day or three. You then connect it back up to the CO2 feed to see IF it pulls more in. If it does, and the keg was completely empty, then you have a leak in the keg someplace. I'd then charge the keg to a higher pressure level and see if the leak shows itself. Although I can't recall ever having a keg leak that wasn't pretty evident. Such as the lid wasn't aligned correctly when you sealed it. That's why many of us don't lock the lid before hitting the keg with ~10psi getting it to seal before locking it down. Poppets will typically get stuck open/depressed, which are also easy to spot. But those won't leak when the QD is on.
Take a wrench and make sure the gas and liquid posts are tight. If they are, chances are you don't have leaks there. If the PRV in the keg lid isn't somehow stuck open/leaking, then that would pretty much rule the keg out. Unless all the o-rings are butt-nasty and need to be changed.
 
Ok, so the system minus the keg held pressure overnight - meaning the leak is in the keg somewhere. Sounds like it’s time to dunk, and probably re-set the lid.

now that I’ve p
 
Ok, so the system minus the keg held pressure overnight - meaning the leak is in the keg somewhere. Sounds like it’s time to dunk, and probably re-set the lid.

now that I’ve p
I missed the post where you said how you determined the keg is actually leaking. The only symptoms I have seen so far are pressure drop on an uncarbonated keg when the CO2 supply is turned off. This is normal until the beer is fully carbonated. The closer you are to full carbonation, the lower the pressure drop when shutting off the CO2.

Brew on :mug:
 
Last post got cut off.
So, going forward, I think I’ll cut off the gas at the tank when I don’t have a keg on, but not bleed out the leftover pressure in the regulator. If it holds pressure, I’ll know that there are no leaks in the regulator or gas lines before I hook up the next keg a few weeks later. After I seal the keg lid, I’ll just dunk the keg in a bucket to look for leaks (no point testing an empty keg, since setting the lid seal correctly is part of the equation). No more pointless soap and water tests!
 
I missed the post where you said how you determined the keg is actually leaking. The only symptoms I have seen so far are pressure drop on an uncarbonated keg when the CO2 supply is turned off. This is normal until the beer is fully carbonated. The closer you are to full carbonation, the lower the pressure drop when shutting off the CO2.

Brew on :mug:
Anecdotal and questionable evidence :) I stuck my head in the fridge to check the guages and just about passed out. Maybe a co2 leak. Or perhaps my blood has acquired a taste for alcohol and now rushes to the side of my body closest to large quantities of beer, depleting my head of oxygen. That would explain a lot…

whatever the case, I have now revolutionized the way I check for leaks.
 
Back
Top