Slow CO2 leak

[kevin_nj]

I've got a slow leak, going through a 5# tank in about 3-4 weeks, single (commercial) keg system. From what I've read, this slow of a leak is indicative of a low side leak. I'm about to run out to get a refill now, don't even have enough gas to do a good soap leak test.

Upon disassembly, the only questionable thing I've seen is a lack of a washer at the gas line to coupler connection. Should there be one in addition to the check valve?

Background on the system, I built a kegerator conversion last year but was never able to get temp low enough so I basically abandoned it. Last month I got a new freezer and created a keezer. At that time the tank was empty, didn't think too much since nearly a year passed. Got the system set up and balanced, all was well for 3 weeks or so until last night when I had a somewhat flat beer and noticed the tank was about empty.

Some parts diagrams I've seen online show a washer in addition to the check valve, most do not. Should I add one or just wait until I get back with a fresh tank and see what soap test shows?

Thanks

[KillianBoy]

kevin_nj,
Welcome back, from what I have seen the only coupler that uses a neoprene washer, stock, is the "U" coupler. The other method of gas check is the Thomas Vent Check Valve, it is a 2 part valve.
I have suggested that you use both a washer and gas check, this prevents damaging the check valve, a non-damaged check valve should be fine.
For emptying a 5# tank in 3-4 weeks seems like a extremely slow leak or something else.
So, was this just 1 keg?, how was the pour?, How big of keg? PSI setting?, Did you drop regulator since last keg? And I really hate to ask, and please don't take this the wrong way, how do you know it is empty?
Next post also, check the gas check for tearing, same for neoprene washer at beer line. When you do this and fill tank, try and adjust regulator to 20+, then back to <10 PSI, then again, with gas connected, but not beer.
Normally a 5# tank should empty in 7-10 days, if 3-4 weeks, I see something else wrong.
KB

[kevin_nj]

Thanks KB

I've got a kegco D coupler attached to a 1/4 keg of Yuengling, with a kegco dual guage regulator going to a new 5# tank. On the other side I have 25' of beer line going to a Perlick faucet. Yes, I know 25' is a lot, I had figured I'd need to trim it down after balancing the system, but the pour was good (other than a little slow) so I decided to let it be. I was also so damn happy with a successful system I didn't care about waiting an extra few seconds to pour a pint. I did apply keg lube to the coupler and/or keg seal when I hooked up. I'm running a Johnson controller, with the tip of the probe in a bottle of water (not fully submerged). Controller is set to cut in/out at 36-38 degrees. I've also got an analog thermometer inserted into another bottle of water. That and checking beer temp in the glass with a digital all confirm that my beer is staying pretty consistent at 37+/- 1. 13 psi was giving me a very satisfactory pour, with about an inch or so head. The only odd thing was it seemed like the head would dissipate a bit faster than what I'd see in a commercial environment?

All was good for about 3 weeks (same 1/4 keg throughout), guests to my Memorial Day picnic were pleased to say the least. Last night I pulled a pint with almost no head at all. That lead me to check the gauges, the low side was high, about 18psi, while the high side was very low. Lifting the tank lead me to determine that the tank was nearly empty. I have not dropped the tank or regulator.

With this new tank, I added a washer between the gas line nipple and check valve. I also applied some keg lube to both, I had not done that prior. I then cranked the psi to 30 and soaped up all connections. I could not detect any bubbles, or hear any leaks. Afterwards I dropped the regulator back down to 13 and bleed off the excess pressure.

In a little while I will weigh the tank with a bathroom scale so I can monitor over the next few days to see how much it drops.

[kevin_nj]

I may have found it!

Went to weigh the tank and this time I heard a hiss. Don't know why I didn't earlier. Anyway, it was coming from the tank-regulator connection. Which leads me to next question/statement.... The nylon washer for that connection, should it be loose or pressed onto the regulator? I can't remember how it was when I bought it a year ago. The old one was pressed on, I was unable to get the new one to "press onto" the regulator. That's making me think it is not meant to be pressed on and at some point I over tightened it and damaged it?

Couldn't tell if it is still hissing now as the compressor kicked in and it's too loud. Need to wait for it to kick off again. I did not go "too tight" this time, also used a little keg lube for good measure.

Man I hope it's that connection, and I've fixed it.

[PointPleasantNJBeerguy]

Have to be careful with the nylon co2 gaskets. If you over tighten and deform them they won't mate up properly the next go around..

[kevin_nj]

Yeah, the more I think about it that leak had to be newly created when I hooked up the refilled tank. I can't believe that I could have missed that earlier. I'd also expect that leak to have drained the tank much faster.

So I'll continue to monitor and weigh the tank over the next week or two.

[PointPleasantNJBeerguy]

Should be able to get at least the fiber co2 washers at your co2 filling spot

[KillianBoy]

kevin_nj,
You really don't have measure beer/air so many times, beer should be sufficient.
From time you turned gas on to keg to Memorial Day, did you adjust regulator or touch dial? How many days before MD?
Depending on Analog/Digital controller, it may not be submersible. I really don't want to seem nit picky but when you say "Controller is set to cut in/out at 36-38 degrees.", is incorrect it is either cut-in OR cut-out, not both, depending on your differential I'd say cut-in.
Understand beer will still pour if you run out of gas, though very slowly, think of it as a can of beer that you re-seal after you drink a couple of sips, it should stay carbonated a few days. Your long line is causing more problems than it solves, 10 feet should be the longest.
KB

[kevin_nj]

KB

I had just typed out a fairly long detailed reply, which vanished when I hit post and the system told me I was no longer logged in.

Here's the short version:

I know the system is either cut-in or cut-out; I don't recall which way I programmed it. I think it is set to cut-out at 36. I know I have it set to a 2 degree differential since I'm reading liquid temp, not air temp. To me it does not really matter if it is set to cut-in or cut-out, unless I am missing a benefit of one over the other. The temp probe is not submerged to the point that the water level goes above the metal onto the wiring.

I had the system balanced at 36-38 degrees, 13 psi to a 2.6 v/v Yuengling by around 5/17 and had not adjusted anything since then. I'm curious what problems other than a slower pour the longer beer line is causing?

[KillianBoy]

kevin_nj,
Last post in your previous thread explains disappearing post:
"If you cut and pasted text to new post, I've had the same problem, don't know what it is but best to just write in the new post box, you could try and wait till the auto save function rolls by a couple of times and see if it posts right."
After posting and you find end of post disappeared, edit post and paste again, it should hold the second time, just delete (backspace) original post.
You don't need to go into any lengthy response, "don't care if cut-in or out, my differential is 2" would have been enough. Yes, with 2 diff, it doesn't matter if cut-in or out.
But with that short of diff and sensor partially in water, the compressor may run long and stay off long and may have a unpredictable on/off cycle.
Again not sure which model ETC you have, if the same digital one sold at MM, cut-in/cut-out isn't programmed, it is changed by altering a jumper. IF the one from MM, probe is submersible.
If probe is submersible, then put whole thing in water and keep diff at 2, if not leave whole probe out of water and set diff at 4+. IF probe is in air, diff at 6, setpoint at 38, if cut-in, beer should be at 35 degrees, if cut-out than beer should be at 41, this is why I asked about diff, it makes a difference. If you have at 2 and wholly in water, it should be fine, if probe out of water I'd go longer diff and confirm whether cut-in or out
If you touched regulator the week before MD (wiped condensation off gauge or adjusted), this small action may have loosed the connection enough to cause leak at connection. If you didn't touch at all. then still a mystery, a high pressure leak will empty tank in 3 days easy. If adjusted on 5/17, could be you loosened connection that day and the keg lost all carbonation on MD, then you had flat beer the next day.
By having that long of beer line will cause massive resistance, can cause less CO2 in beer at glass, it might effect temperature of beer. Installers will say never go longer than 10 feet of 3/16 line, there is no positive of going a longer, especially that long. 10 feet is the longest you should run, period.
KB

[kevin_nj]

KB

Yes, I have had problems posting before, and as like before I am creating the post within the forum, not outside and doing a cut and paste. The auto save has activated while typing the post as well.
As to the detailed posts, that's just how I am. I'm a Systems and Technical Trainer by profession, and being detailed goes along with it.

Regarding the cycle times, I have a data logger which tells me the cool down cycle is lasting an average of just over 30 mins; and a warming cycle lasting on average 4-5 hours. So, while not a refrigeration tech, I'm not concerned with those cycle times. I am using the digital controller with the programming jumpers.

Fair enough on the beer line length, once this keg kicks I will consider shortening it. My bigger concern now is the CO2 leak. I'm not going to worry about line length until I've gotten the leak resolved.

Today's issue is it appears the balance got thrown off while messing with the CO2 leak. I'm getting nearly half a pint of foam with initially small air bubbles in the line becoming large as it sits. My research leads me to believe this is due to the keg losing carb from the leak, or too low a pressure setting. I've bumped the psi slightly, to 14 to see what that does over the next 24 hours.

[KillianBoy]

kevin_nj,
Sometimes it's good to be detailed but sometimes KISS works better.
I can see how your cycle times are so wild, your probe is half in and half out, if you post model of JC controller, I could tell you how best to set it, the ETC will tell the kegerator when to turn off and on, if the bottle of water is on a metal plate, the plate will conduct all the cold to bottle and won't turn back on till it reaches diff or setpoint. For certain, water may not be the best medium, I have had problems with water containers in kegerator, it is all about mass, a 6 ounce to 1 gallon won't react the same as a keg 1/2 to full keg, I have frozen large Gatorade bottles solid and poured 38 degree beer, if you want to use submersible, use glycol (antifreeze) or beer as a medium.
If it is a small leak at connection, setting PSI high or low won't matter, as long as the regulator gets a pressure higher than setting it will apply that PSI to keg, no matter what. A leak at this connection will empty tank in 3 days. The MM premium regulator uses a fiber washer, the economical ones has a built-in washer probably like yours, if you used a washer WITH the built-in, the "O" ring may be deformed and not work properly, if the fiber washer doesn't fit, a neoprene washer might work in a pinch. I found using a large washer with a small hole works best for me (the Haier regulator won't take the MM fiber washer).
OK, so if you post picture of tank end of regulator I could tell you best way to make it leak free and post model of JC ETC, I could tell you if probe is waterproof and best settings and if cut-in or out, my feeling it has a submersible probe and unless you changed the jumper is cut-in.
I disagree with "balance got thrown off while messing with the CO2 leak.", either low pressure side leak causes balance problems (low pressure side will lower actual PSI getting to keg) or messing with leak, which led to temperature problems which messed with balance. A high pressure leak or messing with this leak (unless you had door open for extended period of time) shouldn't mess with balance.
You did pour well in beginning, I feel when you adjusted regulator on 17th, you knocked the connection loose starting the slow leak ( I wiped condensation off gauge once and caused a high pressure leak) MD party got rid of all CO2 in keg leading to flat beer.
Really don't know if low PSI is your problem, with such long of a line the CO2 maybe coming out of the beer in line and not keg, could be the warmth of keg and environment might be causing balance problems. As suggested before, make all your modifications, close door and wait 18-24 hours before trying beer, if foamy open door, check bubbles, beer temperature and PSI, if something is off, reset and WAIT 18-24 hours before trying again. Resetting anything takes a while to see the effect, and try to do thing one at a time (thank you for waiting to cut line).
KB

[kevin_nj]

Here's some photos.
Everything except the controller is housed in a 17cf upright freezer, perlick faucet is inside as well. Johnson Controller A419; Kegco Premium Pro Series Dual Gauge Regulator (has a hard white nylon washer at tank-side connection). Also photo of temp probe (mostly submerged), bottle contains a mixture of iso alcohol and water and is next to the keg sitting on a piece of MDF (so is the keg). Also attaching a shot of the data logger results. I did not read this as being "wild cycle times" more as being an efficient unit that holds its temp well. The logger is reading air temp, which is why I believe it has a lager temp swing then the 2 degree differential that the ETC is programmed for.

20150603_201532.jpg20150603_201614.jpg20150603_201622.jpglog.jpg

[KillianBoy]

kevin_nj,
Forgot you are using a freezer, they are better at retaining cold than kegerators, usually the cycle is in reverse, if you didn't change any jumpers than it is cooling cut-in, other than cutting line if you don't have forced air, I'd get a fan inside unit, it would help keep temperatures even from top to bottom.
Everything looks good, hope the connection holds and everytime you touch regulator, check for tightness at connection (not overtight as noted). By what I have read about the ETC the probe is water tight, you should drop it all the way in and see how your cycles change.
KB

[kevin_nj]

The manufacturer documentation calls it a "splash proof" sensor; they do sell an alternative submersion at a cost greater than I can justify.

I will probably re-weigh the tank tonight for comparison. I imagine there should be very little difference over 48 hours.

This freezer also already has a fan for air circulation.

http://www.frigidaire.com/Kitchen-Appliances/Freezers/Upright-Freezer/FFFH17F2QW/