CO2 leak, but only after two days

[physics911]

I have, what appears to be, a leak in my CO2 system, but only after two days, not before. Allow me to explain.

To start off with, I have drenched the system with Starsan and water multiple times. I found a few minor leaks when I first built it, but none after the first day or two. After I started noticing the "odd" leak behavior, I submerged all the parts, including pressurized kegs in water, and watched each component for at least 5 minutes and saw no bubbles.

Here is what I am seeing, and note, this has happened at least three or four time since I built the keezer a couple of weeks ago.

Equipment
10# tank -> Taprite dual gauge regulator -> 4-way manifold -> 4 pin lock kegs -> 4 Perlick 525SS taps

I'm waiting on my brew batches to be ready, so this is all done with empty kegs.

The times below are approximate, but pretty darn close.
At time 0 I pressurized the empty tanks to 31 psi (I've done the same procedure at different psi) and fully closed the valve on the CO2 tank.
At the end of the first 24 hours, dead on 31 lbs
At the end of the first 36 hours, dead on 31 lbs (leave for work)
At the end of 48 hours, pressure has dropped to about 25 lbs
By tomorrow morning, if history repeats, the pressure will be down in the teens.

Also note, at times 0, 24, and 36, the high pressure gauge stays steady at around 700 psi. At time 48 it too begins to fall, and eventually goes to zero as the low pressure gauge drops.

There are no other people or animals in the house, nor anything else which might impact the system while I'm at work. The keezer is not even cooling right now, so no mechanical vibrations. Nor am I fiddling around with the connectors or anything else during the test time.

Again, this has happened multiple times. So over the weekend I pulled the posts off the kegs and keg lubed everything, liberally. I also replaced the post rings with silicone rings specifically for pin locks. I did this because when I disconnect the gas in lines from the kegs, and pressurize only the supply hose, manifold, and gas lines with disconnects, so far I haven't noticed a leak after the 48 hour period. This leads me to believe it is somewhere between the gas in post and the tap.

Again, also, I have submerged every part, except the regulator, under water and found no bubbles.

What I have not done so far is submerge at low pressure, though I have used Starsan at low pressure. Nor have I let it sit pressurized for more than about 60 hours with the gas lines off the keg.

Seems odd to me that it has held a very steady pressure for about two days, and then began to drop noticeably after that, and done this multiple times.

Input is appreciated.

Thank you,
Chris

 

[day_trippr]

Closing the cylinder valve leaves at least a couple/few cubic inches of volume occupied by CO2 at tank pressure. That would translate into some multiple of that volume at 31 psi. I don't know if it's linear but I bet it's close.

It is much more likely that you actually have a steady leak somewhere than a leak that spontaneously manifests after 48 hours...

Cheers!

 

[physics911]

day_trippr,
I don't mean to imply it is specifically spontaneous, but when I say it is dead on steady before it drops, I mean dead on - not even a half a needle change. This morning when I left for work it was exactly where it has been, at 31 lbs, since Monday. When I got home this evening it was at 26. So far I haven't touched it this evening. The high pressure gauge is at 0.

I'll find the leak for sure, eventually, but I'm puzzled as to what is cause this behavior.

Thank you,
Chris

 

[day_trippr]

So imagine a system leak that emits a small bubble once every five seconds. Add the volume of those bubbles over 48 hours. Then take the internal volume of your regulator body and cylinder coupler, multiply that by the tank pressure before you closed the tank valve, then divide the result by 31. That would represent the effective volume of the reservoir of pressurized CO2 at the initial condition.

It wouldn't be surprising if the latter volume accounted for the leak over that period - and then when the reservoir of CO2 drops below your keg pressure, you'll start seeing the system pressure drop...

CHeers!

 

[physics911]

Thanks for the input.
At this point I am need to get more detailed with the elimination of elements.
I'll pressurize the system up to a known amount and then disconnect the gas in lines from the kegs. After a few days I should know for sure whether or not the leak is on the gas supply side. If that holds up, I'll reconnect the kegs one at a time and see if there is a big pressure drop. I know I'll lose a little pressure by reconnecting, but if I let them sit for several days, if it is a keg leaking it should drop enough to throw up a flag.

Thank you,
Chris

 

[brewprint]

I watched a video once about a leak a guy couldn't find.

He ended up submerging the top of tank.

There's a valve at the top of the tank that cannot be checked otherwise.

I suggest shutting the gas off and disconnecting the regulator. Then submerge the entire top of the tank to the tank itself. You're leak may be there.

 

[raouliii]

I agree with day_tripper in that you likely have a continuous slow leak. Don't get hung up on the fact that it doesn't impact the low pressure gauge reading until two days after closing the tank valve. There's enough high pressure between the tank valve and regulator to create the situation you report.

I'll throw out a few leak related tips: Many tanks have to be fully, tightly, opened to prevent leakage at the valve stem. There should be a seal at the connection to the tank. Any metal to metal contact flare fittings should have a nylon washer installed. I even use washers for flare connections at keg disconnects. IMO, oetiker style clamps are better for tubing to barbed connections. In addition, barbed connections are best clamped with the tubing warmed for a minute in hot water. All NPT (pipe threads) fittings should have teflon tape applied.

 

[physics911]

As you have pointed out, you are correct, of course. The pressure between the tank valve and regulator was enough create two days of "steady" low pressure.

Anyway.....

All components listed below are new from kegconnection.com

I know it is contraindicated, but no amount of starsan and water was revealing the leak, so I took the tank, regulator, hoses, manifold, and disconnects and submerged in the bathtub. There is a bubble coming out from behind the regulator shroud about ever 4 minutes or so. I've rotated and agitated the whole system underwater several times to help remove any trapped air behind the bonnets, so I'm pretty confident the 4 minute bubble is a leak, not trapped air, considering I so no other bubbles and the gauge continues to drop.

The regulator is a Taprite dual gauge, and the bubble is coming from behind the black shroud, near the low pressure gauge. I'm sure it is not coming from the low pressure or high pressure fittings into the body of the regulator.

This is a bad regulator right?

Anything I can do to disassemble and tighten components instead of replacing?

I've already removed the gauges and output barb, cleaned the connections, re-tefloned, and tightened thoroughly; and no bubbles from any of those connections, nor anywhere on the manifold or disconnects. .

Thank you,
Chris

 

[iijakii]

I'm not familiar with the taprites, but "behind the black shroud" -- so the actual body here?

Not coming from any port/connection? Sounds like a bad regulator to me.

 

[physics911]

That's right, from behind the black shroud on the body of the regulator you have shown.
I've basically spent all night sitting beside the bathtub watching this thing, after a lot of other testing. I've performed the testing with the manifold shutoffs open and closed, and the only place I am getting a bubble is from behind the shroud.

 

[day_trippr]

The black thing is called a bonnet, and it traps the diaphragm against the regulator body. If the diaphragm isn't making a perfect seal against the body it will indeed leak.

You could remove the bonnet and diaphragm, clean the sealing rim on the body, then lube that rim and the matching sealing surface of the bonnet before reassembly. Good chance that will fix things up...

Cheers!