Keezer CO2 leak...or is there?

General Back Story:

I have built a keezer with 5 taps.
I use a 5# bottle.
Primary reg is directly connected to bottle has two gauges: Tank & Line
Primary reg is connected to secondary 6 body reg.
Each body has its own reg.



So right after I built it I got it up and running and was happy...that is until about 2 weeks later when the bottle was empty.
I blew that off as not realizing how much CO2 was needed to carb up the 5 kegs...plus I force carbed a few others.
One bottle swap and another 2 weeks later...another empty bottle.

So now I realize I have a leak somewhere.

Out comes the soapy water sprayer. I find nothing.
Next comes the baby powder and smoke to see if I can see any currents. Nothing.

Anger ensues and time passes.

So I decide over the July 4 holiday to finally get this running.
I get all new hoses as I figure I could have a pinhole leak somewhere in them. I dismount the secondary and take it apart and reassemble.

Now to make sure I have no leaks...I take the entire assembly and submerge it in a big ass tank.

Leak found...it was in one of the pressure relief valves on the secondary.
Replaced and retested...nothing.

Perfect....well maybe.






Here's my concern:

I charged the lines but turned off the tank. The tank gauge says around 55 and the Primary's line gauge says around 52. No problem there.

However I left it charged, with the tank off, and NONE of the keg lines "open" (so no gas dissolving happened there)...come morning the primary's tank gauge only read 30.

I "was" very confident there was no leaks when it was underwater...but if that was the case, why did the gauge change overnight?

I did not however take note of the Primary's line gauge before I recharged the system and went to work (where I'm at now).


Could the tank gauge change like that just by sitting there?

Anyone have any advise they can give me here?


Thanks!
Jay

If the 55, 52, and 30 readings are PSI for CO2, then I believe your pressure is too high for the beers. Most serving pressure is 9-12 PSI. Tank pressures are usually read in the hundreds of PSI and will quickly drop off on the low side when the tank is empty. If I have the time to carbonate over several days I'll put the regulator at my desired pressure and let it go. As it carbonates and chills more CO2 goes into solution. Even if your beer is carbonated at room temp, when it chills it will absorb more CO2. Since it is also possible to over carbonate at any given temperature, it sounds like that may be the case. What happens when you pour? Too much foam?

You CAN get leaks past the tank valve when it's closed, (it should be fully open to not leak...), so the "shut the tank off and wait" leak detection method is not foolproof.

shortyjacobs said:

You CAN get leaks past the tank valve when it's closed, (it should be fully open to not leak...), so the "shut the tank off and wait" leak detection method is not foolproof.

Click to expand...

Perfect.
That's what I was hoping for.

Still not a guarantee one way or the other....but with any luck you could be fine.

Was the initial pressure reading before putting your gear back in the keezer?

Example: did you charge the lines when everything was warm and at room temp then move to the keezer and wake up the next morning to find the psi dropped?

I freaked out the first time I put my tank in my keezer to see it drop from 800psi down to 500 psi. I thought I had a major leak but realized the temp change caused the pressure to drop.

I'm hoping it is just the tank valve leak as I haven't turned my keezer back on yet.

The gauge definitely goes back up to 55-60 (or the green Full area on that one) when I open the value again.

I'm going to pressure up the kegs tonight and I guess I'll see in a week or two if I have to take it all apart again.

The regulator could leak. Pressurize the system with 30-40 PSI and close the tank valve, and the valve downstream from the regulator. Wait 6 hours and if the pressure drops on the regulator gauges, there is a leak somewhere in/on the regulator/tank connection. If this passes then repeat the process with the valve from the regulator open, and the tank valve closed (system pressurized to 30+) Wait 6-12 hours and check the gauges. If the pressure of the gauges has dropped, there is a leak downstream from the regulator. It could be anywhere. If that's the case, Check and tighten every fitting, if you're unsure about any then disconnect them and reconnect them with new teflon tape, using 2 wrenches to get the fittings as tight as you can.

TheBroonery said:

The regulator could leak. Pressurize the system with 30-40 PSI and close the tank valve, and the valve downstream from the regulator. Wait 6 hours and if the pressure drops on the regulator gauges, there is a leak somewhere in/on the regulator/tank connection.

Click to expand...

As mentioned above, this method of determining whether there's a leak or not isn't reliable. The way that CO2 tank valves are made (double seated), pressure from downstream can leak past the valve when it's closed. This also means that you should open the valve all the way to "back seat" it when in use, as any less than that is a potential risk for a leak past the valve.

JuanMoore said:

As mentioned above, this method of determining whether there's a leak or not isn't reliable. The way that CO2 tank valves are made (double seated), pressure from downstream can leak past the valve when it's closed. This also means that you should open the valve all the way to "back seat" it when in use, as any less than that is a potential risk for a leak past the valve.

Click to expand...

Are you saying you should fully open (unscrew) the main tank valve? I usually just give it a turn or two to open and leave it at that.

tesilential said:

Are you saying you should fully open (unscrew) the main tank valve? I usually just give it a turn or two to open and leave it at that.

Click to expand...

CO2 tank valves are rear seating/back seating valves. There is packing to prevent gas from leaking past the valve stem when the tank is closed or partially open, but over time this packing can develop a slow leak. It probably doesn't matter much for most tanks, especially newer ones, but IMO backseating (opening fully) is good practice. It would appear from the above comments that shortyjacobs feels the same way I do about this.

JuanMoore said:

As mentioned above, this method of determining whether there's a leak or not isn't reliable. The way that CO2 tank valves are made (double seated), pressure from downstream can leak past the valve when it's closed. This also means that you should open the valve all the way to "back seat" it when in use, as any less than that is a potential risk for a leak past the valve.

Click to expand...

That's interesting.. from my experience in plumbing I always turn the valve 1/2 turn away from either fully open or closed because I've found that in most cases, if a valve has been fully/tightly seated (either open or closed) for a length of time (ie 10+ years), the valve freezes and simply will not turn. I often find that on the main shutoff valve in houses.

Back to CO2 leak isolation - if I have the tank valve fully open and backseated, the regulator reads pressure, and I close the shutoff valve downstream from the regulator, the system downstream from the regulator should be isolated and sealed. If the pressure drops, when you open the regulator shutoff valve, you should hear gas transfer into the the lines downstream. If the lines downstream are sealed and have no leaks, then having the regulator valve closed for a length of time and then opening it would not transfer any gas because the pressures have been equal on both sides of the valve.

Right?

I had to dip my connections in water to see the tiny leak. One little bubble every 3-5 seconds.

I disconnected everything, bought a vise for my bench so I could really tighten everything down.

I also used teflon and thread sealer from Home Depot. Now no leaks.

But this doesn't solve any fix for a regulator leak. Also, my LHBS sold me a red plastic washer looking thing to put between the regulator and the tank. Also, don't use teflon tape on the tank threads the regulator screws onto.

I was just going to say, don't forget that little washer between the gauges and your tank... that got me the first time I hooked up my kegerator, that and not using keg lube and the head to my on/off valve wasn't on tight enought... lot played into that leak problem.

sorry to be the first to just say it but: you still have a leak!

All this discussion of valve styles and all that is just theoretical junk. Here's the real deal. If I disconnect all my kegs and take the regulator off the tank, it will hold pressure for at least a full 5 day weekend without losing one single psi.... ie, you have a leak. find it!

Any chance you're using cheap tubing?

Nope...Tubing isn't cheap.


There may still be a leak, but it is holding strong fully charged and installed for the past 3 weeks.

If it lasts the month I'll be fine and chalk up a bottle a month as a reasonable expense.

thats good to hear. I opened my keezer last week to find my co2 empty, after a month or so of using it. I refilled the tank and hooked it up, went through every connection. seemed like all was good so I started hooking up the kegs. good think a bit of soapy water was still on everything, when i was moving the hose around the leak showed up. that was a few days ago and looks like i have it fixed now, or at least the gage still reads the same.
I will have to check my tank valve to ensure all the way open, i have learned through the years to fully open and then give a 1/2 turn back just like stated above. didn't know the co2 bottle sealed on both open and closed. Thanks for the tip on that one.