A Year of CO2 Leaks, Please Help

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.

R. Alan Boyle

Member
Joined
Jan 22, 2020
Messages
8
Reaction score
0
As you gathered from the title, I've been fighting CO2 leaks for about a year, the whole time I've had the Kergerator.

Usually, the pressure holds just fine for a few days or a few weeks, then all at once it completely drains over night/day. I never get more than one 5 gallon keg out of a 5lbs bottle and I these days I get much less.

I have checked for leaks using the soapy water method and also by holding parts under water while under pressure.

I have replaced the regulator with a Taprite T752HP.
I have replaced the lines with Accuflex 5/16th tubing.
I have replaced the main oring on the kegs.
I have new kegs in good condition.
I use Red Oak Tools stainless steel ball locks on one keg and generic white and black plastic couplers on the other.
I use Ferroday Ball Lock conversion couplers on my Sankey when I tap a store keg.
I use very tight worm gear hose clamps.

After a year of dealing with it I'm at my whits end. I bet its something dumb that I'm over looking because I can't find other people having the same problem.

Please help me.
 
Last edited:
As you gathered from the title, I've been fighting CO2 leaks for about a year, the whole time I've had the Kergerator.

Usually, the pressure holds just fine for a few days or a few weeks, then all at once it completely drains over night/day.

I have checked for leaks using the soapy water method and also by holding parts under water while under pressure.

I have replaced the regulator with a Taprite T752HP.
I have replaced the lines with Accuflex 5/16th tubing.
I have replaced the main oring on the kegs.
I have new kegs in good condition.
I use Red Oak Tools stainless steel ball locks on one keg and generic white and black plastic couplers on the other.
I use Ferroday Ball Lock conversion couplers on my Sankey when I tap a store keg.
I use very tight worm gear hose clamps.

After a year of dealing with it I'm at my whits end. I bet its something dumb that I'm over looking because I can't find other people having the same problem.

Please help me.

Unlikely, but...

You're positive the CO2 isn't just running out? How many kegs, and what size CO2 tank?
 
Yes, positive. This last time it held steady for 3 days before emptying. This is the shortest yet but I never get through more than one 5gal keg and usually less. Fair question though. I'll add that to the description.
 
Do you have any equipment you can borrow from someone and swap out? It sounds as if it could be anything that you haven't replaced, even if purchased new.

Also, does your CO2 source fill your tank or do they exchange? If the former maybe your tank is faulty.
 
My CO2 source exchanges. I don't know anyone local who kegs. If I did, what would you recommend swapping out to isolate the issue. At this point I have tons of spare parts. Have you ever heard of an intermittent leak? The thing that is most baffling to me is that I can take a photo of my regulator every day at it will be the same and then it just empties.
 
My CO2 source exchanges. I don't know anyone local who kegs. If I did, what would you recommend swapping out to isolate the issue. At this point I have tons of spare parts. Have you ever heard of an intermittent leak? The thing that is most baffling to me is that I can take a photo of my regulator every day at it will be the same and then it just empties.

I couldn't guess at what it might be, other than it's probably not anything you replaced.

Are all your taps usually being used? I'm wondering if you can try out different combinations of equipment, one at a time. So set up one keg and see if it happens. If it doesn't, try another keg (with different disconnects). If it happens with that keg, then you've narrowed it down. Then swap out one of the disconnects. If it then works, it was the disconnect you swapped. If not, try the other...etc.

But as far as the regulator. That high pressure gauge, that's supposed to tell you how much is left in the tank, doesn't seem to work as a general rule. I have the same experience, and I've heard others have it too. It's fine, indicating your tank is full, then suddenly your tank is empty. So even if you fix your main problem I expect that will keep happening. In a word, that gauge is bullsh*t.

As for an intermittent leak...I don't know. Things can seal up good when the pressure is higher, then loosen when it's lowered. Perhaps small pressure drops (for whatever reason) are causing a loosening of, say, the o-ring around one of the keg lids, or a poppet.

Clearly this problem has captured my imagination.
 
I do feel for you as i have lost a few tanks of gas and kegs of beer and i was not happy. Do your kegs hold pressure? I always test mine before filling after that. If the leak its often worn seals in the posts.
 
I have two taps. You are right that I should Isolate them to narrow down the problem. How many 5 gallon kegs do you think I should be getting out of a 5 lbs CO2? I're read 5 half barrels but I don't know if its a linear conversion.
 
I do feel for you as i have lost a few tanks of gas and kegs of beer and i was not happy. Do your kegs hold pressure? I always test mine before filling after that. If the leak its often worn seals in the posts.
After I pressurize the keg I do the soapy water test on the main seal and the posts. Is that what you mean?
 
If you're simply dispensing you should get a good 20 corny kegs through on a fill.

Threaded fittings aren't going to suddenly leak. Flare fittings with proper gaskets that have been correctly tightened, for instance.
Otoh, things that aren't threaded and that may be moved could be suspect.
For instance, I think I would try to test the keg disconnects to see if applying pressure to the side causes the post O-ring to leak.
That's actually a way to determine if a pin lock keg post was fitted with a ball lock post O-ring (pin locks use fatter O-rings).

I recall at least a couple of HBT folks that were plagued with sudden loss of CO2 that eventually tracked the cause down to the wrong post O-rings that with a slight "cocking" of a gas QD would quickly drain the gas cylinder...

Cheers!
 
The high pressure gauge will drop suddenly when the cylinder is empty, that's normal behaviour. It is a pressure meter not a volume gauge. As long as there is some compressed gas it will read at pressure.

On the leak, does it happen only once you connect a keg or even with no keg connected? I agree the post o ring as a possibility. When you close the keezer do the pipes get bent/crushed? This could cause a leak that you don't see with the keezer open. You could spray soapy water everywhere, close it, then open to check after a minute.
 
CO2 is stored in liquid form. It's a property of CO2 (and other gases) that it can maintain liquid form at room temperature at a relatively low pressure. At any time the bottle will be filled with a certain amount of liquid CO2 (very dense) and a certain amount of gas (in the headspace) at the equilibrium pressure. The high pressure gauge measures the pressure in the headspace and as long as there is liquid gas available will always read the same (temperature dependent) pressure. That is because as you use up the gas more liquid will evaporate to replace the gas you used up. This changes suddenly as the liquid runs out. Since the gas cannot be replaced through evaporation any more the pressure will start decreasing very quickly until it completely runs out. This does not by any means imply that your system has sprung a "sudden leak". The leak was indeed continuous but not detectable through pressure alone until the liquid CO2 has completely run out. To actually assess the leak you should weigh the bottle at regular intervals and see if it gets any lighter and by how much. The weight lost is the actual amount (in weight) of CO2 that has leaked out between measurements.
If I were you I would assess the leak with everything connected using the weight method and then start disconnecting elements sequentially until the leak stops. So first thing you will disconnect any keg(s) and see if the weight holds steady. If it doesn't then set the regulator to 0 PSI (to simulate removing the gas line and its adapters) and do the weight check again. If you get to this point and still have a leak then it's probably either the regulator itself or its connection to the bottle. Just to be 100% sure I would still close the shutoff valve on the bottle and do the weight check one more time to rule out a defective bottle. This method is much more reliable than the soapy solution method since it can detect even extremely slow leaks.
 
How are you tightening up your worm drive o-rings, screwdriver or small socket? I find that with a screwdriver these are still sometimes loose.

On that note, does anyone have a good recommendation for a torque value on these?
 
If you are losing that much gas, perhaps leave C02 unhooked until you figure it out. Once beer is carbed, if you turn on gas only when drinking, it works OK. Not as elegant a solution as fixing the problem, but will save on CO2. This is coming from someone with more than a dozen 1980's C kegs and fittings ( I really should get new fittings, but I like the old metal ones).

I'd suspect the keg posts and fittings first. Maybe try applying a little food grade grease like CIP film to O rings, and make sure they are in good shape.
 
I had the same problem about a year ago. It turned out to be a leaking valve on the distribution block. If the valve was not set in a specific position it would slowly leak.

The trick is how to isolate the leak without wasting a lot of CO2. I did it by isolating each segment of the system but with the CO2 tank in the off position for most of the testing.

Turn on the CO2 tank. The tank pressure gauge will show current pressure in tank and the other gauge will show your outlet pressure. **Then turn off the CO2 tank.** Both gauges should not move if the system is tight. If there is a leak, the gauges will come down but you're only wasting a VERY small amount of CO2. Once you step through the isolation and see the gauges move over a period of time (usually 24 hours), then you can start to narrow in on source.

Here are some isolation suggestions:

1) Turn off the duckbill valve on the taprite regulator so no CO2 is being supplied to your kegs. Turn on the tank CO2 valve to pressurize the regulator then turn off the tank. If the gauges move during 24 hours, your leak could be the regulator. That taprite has a built in washer on the CGA320 connection to the CO2 tank. Be sure it is actually there and in good shape. Over-tightening could damage it. Be sure there is not an additional washer added. If the pressure holds, then check next item.

2) Disconnect all kegs from CO2 posts. Turn on CO2 tank to pressurize the lines, **then turn off CO2 tank**. Turn off distribution block valves (assuming you have valves). Monitor gauges. If you lose pressure it could be the line from the regulator to the distribution block.

3) If you did not lose pressure with the distribution valves off, then turn on each distribution valve watching the gauges on the regulator with the tank turned off. If you see a pressure drop, then your leak is between the valve and the quick disconnect.

4) If the lines do not show a pressure drop, then your leak is likely in your kegs (assuming they are fully carb'ed). Do the same procedure by pressuring the system, then turning off the CO2 tank and watch for pressure drop on the gauges.

Hope this helps.

Report back your findings.

~HopSing.
 
Are you fully opening the valve to the tank on the regulator?

That should be fully open and seated at the open limit, not just partially open to allow gas flow.
 
Isolate as mentioned above.
Turn off the 5lb bottle. Does it leak?
Open tank and charge the system, disconnect from kegs and turn off 5# tank.
That sort of thing. As mentioned above.
 
My CO2 source exchanges. I don't know anyone local who kegs. If I did, what would you recommend swapping out to isolate the issue. At this point I have tons of spare parts. Have you ever heard of an intermittent leak? The thing that is most baffling to me is that I can take a photo of my regulator every day at it will be the same and then it just empties.

when you mentioned "intermittent leak", it jogged my memory of something I noticed on one of my ball locks. When there was any weight or strain on the gas line to one of my kegs, it would leak at the quick disconnect. But only when there was just enough horizontal force on the connection, otherwise it was fine.
 
^That^ is what I was commenting about earlier. Something that isn't "static" being nudged enough to suddenly leak.

Assuming proper post O-rings are used and they're not cut or nicked, the other thing to be aware of: if you've ever forced a gas disconnect onto a beer post it is possible to permanently deform the QD enough that it'll be prone to leaking. Done that...

Cheers!
 
After I pressurize the keg I do the soapy water test on the main seal and the posts. Is that what you mean?

I have never used soapy water as i think you can hear the leaks. I sanitise and pressure mine empty to about 20PSI and leave them until i need to fill them. When i started i did have some leaks but gradually swapped all the seals over and a few poppets. Another source of leak i had was the gas tube going onto the regulator barb. I was moving it a lot and after a year or so it developed a leak. Easily fixed and faintly audible.
 
How's this for a coincidence?...

Just set up a Nitro tap. Had my usual thursday night gang over and went to show it to them. Kaput. Tank empty after my first keg on nitro after just a day or two.

So, clearly, a leak.

At least I have fewer possible sources of the leak than the OP...
 
OP UPDATE:
I found two leaks so far. I dunked the whole QD/Barb/Gas hose into a bowl of water.
1. There was a very slight leak coming from the barb/hose connection - Too small to see with the soap test. So I cut an inch off the hose and reconnected it. This fixed that leak.

Then i reconnected everything;

2. There was a very slight leak at the gas post which only occurred when the QD was connected. I replaced the oring on the poppit and on the post and that appears to have fixed it.

Hopefully, that'll be it but I will give an update when the tank is dead if there is still a leak.

Thanks a bunch for the input everyone, some of your suggestions helped with the troubleshooting.
 
I had the same problem about a year ago. It turned out to be a leaking valve on the distribution block. If the valve was not set in a specific position it would slowly leak.

The trick is how to isolate the leak without wasting a lot of CO2. I did it by isolating each segment of the system but with the CO2 tank in the off position for most of the testing.

Turn on the CO2 tank. The tank pressure gauge will show current pressure in tank and the other gauge will show your outlet pressure. **Then turn off the CO2 tank.** Both gauges should not move if the system is tight. If there is a leak, the gauges will come down but you're only wasting a VERY small amount of CO2. Once you step through the isolation and see the gauges move over a period of time (usually 24 hours), then you can start to narrow in on source.

Here are some isolation suggestions:

1) Turn off the duckbill valve on the taprite regulator so no CO2 is being supplied to your kegs. Turn on the tank CO2 valve to pressurize the regulator then turn off the tank. If the gauges move during 24 hours, your leak could be the regulator. That taprite has a built in washer on the CGA320 connection to the CO2 tank. Be sure it is actually there and in good shape. Over-tightening could damage it. Be sure there is not an additional washer added. If the pressure holds, then check next item.

2) Disconnect all kegs from CO2 posts. Turn on CO2 tank to pressurize the lines, **then turn off CO2 tank**. Turn off distribution block valves (assuming you have valves). Monitor gauges. If you lose pressure it could be the line from the regulator to the distribution block.

3) If you did not lose pressure with the distribution valves off, then turn on each distribution valve watching the gauges on the regulator with the tank turned off. If you see a pressure drop, then your leak is between the valve and the quick disconnect.

4) If the lines do not show a pressure drop, then your leak is likely in your kegs (assuming they are fully carb'ed). Do the same procedure by pressuring the system, then turning off the CO2 tank and watch for pressure drop on the gauges.

Hope this helps.

Report back your findings.

~HopSing.

This is a really good walk-through, thanks.
 
How are you tightening up your worm drive o-rings, screwdriver or small socket? I find that with a screwdriver these are still sometimes loose.

On that note, does anyone have a good recommendation for a torque value on these?
socket driver. I did find a leak here but it was not loose. If anything it was over tightened.
 
Ironically after i posted here i got a leak. Its a bad seal on the hosebarb attatched to the regulator caused by cheap low quality jubilee clips. Glad you got it sorted!
 
I have never used soapy water as i think you can hear the leaks. I sanitise and pressure mine empty to about 20PSI and leave them until i need to fill them. When i started i did have some leaks but gradually swapped all the seals over and a few poppets. Another source of leak i had was the gas tube going onto the regulator barb. I was moving it a lot and after a year or so it developed a leak. Easily fixed and faintly audible.

Just want to add here that while it can be great to keep things under pressure, there is some consideration that should be given to storing things uncompressed so that gaskets are free to expand and be only compressed during use. I'm not saying I do that, I'm just saying it's a thought.
 
Related question:

If I fill keg to 38psi, then close off the tank, then come back an hour later and the gauge reads 20psi, does that indicate a leak (somewhere between regular and tank, obviously)? I was thinking it may just be the gas dissoving into the beer, but it was only an hour, and it's Nitrogen (75%) which apparently doesn't disssolve into the beer easily.

EDIT: Never mind. Gas-in coupler (flair-type) passed the underwater test before, but I just did it again and it failed. Then retightened. Still failed. Then added some teflon tape. It passed this time.
 
Last edited:
Was the beer carbed at all first?
It's difficult to do pressure tests on kegs with fluids that absorb CO2 - and your beer gas is 25% CO2.
If you really need to test a keg it's best done bone dry so as not to futz with gauge observations...

Cheers!
 
Was the beer carbed at all first?
It's difficult to do pressure tests on kegs with fluids that absorb CO2 - and your beer gas is 25% CO2.
If you really need to test a keg it's best done bone dry so as not to futz with gauge observations...

Cheers!

It is carbed a little bit. Given that, a bit of a drop in pressure if I turn off at the tank is normal, then (not necessarily indicative of a leak)?
 
CO2 absorption will certainly continue until the beer has reached the temperature/CO2 pressure equilibrium point, which can be predicted and controlled using our favorite carbonation table.

If you simply apply "chart pressure" for your beer's temperature and aim for a typical ale carbonation level (the middle of the range is roughly 2.4-2.5 volumes) it will take close to a couple of weeks for a full 5 gallon corny keg to reach equilibrium.

That would be for straight CO2, of course, but if you're carbing your keg on beer gas it still takes time to infuse the entire volume with the CO2 component...

Cheers!
 
There was a very slight leak at the gas post which only occurred when the QD was connected. I replaced the oring on the poppit and on the post and that appears to have fixed it.

Glad you rooted out a few of the leaks in your system. As a side note, I have two CO2 tanks, a 10# that fits in my fridge and a 20# I use to refill the 10# tank. I came across a second regulator so as a best practice, I hook up my recently kicked kegs to the 20# tank, pressurize the lines, then turn off the Co2 tank valve and keep an eye on the gauges. If I see a drop in pressure, I rebuild the keg including the poppet, dip tube, and lid o-rings.

~HopSing.
 
Glad you rooted out a few of the leaks in your system. As a side note, I have two CO2 tanks, a 10# that fits in my fridge and a 20# I use to refill the 10# tank. I came across a second regulator so as a best practice, I hook up my recently kicked kegs to the 20# tank, pressurize the lines, then turn off the Co2 tank valve and keep an eye on the gauges. If I see a drop in pressure, I rebuild the keg including the poppet, dip tube, and lid o-rings.

~HopSing.

...and possibly gas feed lines, gas diconnects, tank side connections and leaky regulators...
 
...and possibly gas feed lines, gas diconnects, tank side connections and leaky regulators...

Absolutely! Those are tested without any kegs connected in the same way. Pressurize, turn off Co2 tank and monitor gauges. It needs to hold for at least 24 hours. If the gauges drop after 24 hours it's a micro-leak, but I'm not too concerned about it.

Also during the empty keg test, I forgot to mention that the room temp should be fairly consistent. A significant drop in temp could drop the gauges with the CO2 tank off.

~HopSing.
 
CO2 absorption will certainly continue until the beer has reached the temperature/CO2 pressure equilibrium point, which can be predicted and controlled using our favorite carbonation table.

If you simply apply "chart pressure" for your beer's temperature and aim for a typical ale carbonation level (the middle of the range is roughly 2.4-2.5 volumes) it will take close to a couple of weeks for a full 5 gallon corny keg to reach equilibrium.

That would be for straight CO2, of course, but if you're carbing your keg on beer gas it still takes time to infuse the entire volume with the CO2 component...

Cheers!

The chart doesn't say anything about time. Is the underlying assumption that it's "eventually"? So a couple weeks?
 
iirc someone actually took on the attempt to quantify "time" in a carbonation calculator and posted something about it on HBT within the last couple of years. I'd have to try digging for it.

But, in my experience, 5 gallons of under say ~1.015 FG in a corny keg held at 36°F and 12 psi straight CO2 takes a bit over 2 weeks to reach equilibrium. That corresponds to the green line in this chart that @Bobby_M posted in this thread (which might be worth reviewing)...

5970be7036f823f045eaded0b3198501.gif


I mention final gravity because when I used to fully carbonate my 11% imperial stout that finishes somewhere around 1.024-028 it took nearly twice as long just to get to 2.0 volumes. But I switched to beer gas a few years ago so now I only carb it 1.2 volumes...

Cheers!
 
Back
Top