Keg King One-Way/Check Valve destructive investigation (with pictures)

[takai]

Following on from my thread below (which upon re-reading made little sense) I present the destructive investigation of a KK one way / check valve.

I had a chat with Juri at KK when i was getting a new CO2 cylinder and asked him the same question, but he wasnt sure on the design of the check valves either. I had a suspicion it was just a ball bearing in a sleeve that sealed under pressure, and mentioned that I couldnt check without destroying one. He happily volunteered a candidate for destructive testing as they had some old ones that had an 8mm nipple on them rather than the newer 6mm nipples. Thanks goes to KK for the donation of this poor fella... lets call him Bob.

When i got Bob home his nipple was easily stripped off and the ball valve actuation arm removed. However, his ball valve (now now) was less cooperative and required drilling out to remove:

*note* if you are drilling one of these out then absolutely wear eye protection the ball is hardened and tends to fracture rather than cutting cleanly.

With Bobs ball removed it was easy to see the other nylon seat that seals up the other side:

I figured i could drive out the innards with a drift, but I wanted to see if there was any channeling inside the chamber to allow for pressure equalisation, so i chopped Bob in half... sorry Bob.

Unfortunately in this process Bob's smaller ball (after all he isn't Hitler) skittered away across the bench when the end cap ejected itself:

We can see from Bob's autopsy that he a pretty simple affair using just a spring to hold the (missing) ball bearing against the end seat (silver):

The spring is a pretty basic unit with barely any compressive force and so should crack quite easily under 1psi or so. However, this also means that under situations of almost equal pressure between the manifold and the keg you can have the ball floating in the chamber with only minimal pressure holding it against the seat. This ball bearing on aluminium seat contact isn't overly pressure tight and so will bleed pressure backwards across the check valve if the pressure differential isn't high enough.

I think this is what is leading to my kegs slowly depressurising as the SodaStream very slowly bleeds the CO2 in the system (after i have turned off the CO2 cylinder) and this slow change in pressure is enough to bleed out two of the kegs on the manifold (strangely the third remains sealed).

There are better designs than this around, but this will be the way the majority of check valves work, and so it is imperative to keep pressure on both sides of the valve or bleed the manifold entirely.

Thoughts on possible solutions?

 

[Mr B]

You could try glue but I dont think it will hold the pressure.

I think Bobs a gonner Takai h34r:





(No solutions soz)

 

[DJ_L3ThAL]

Put pressure in reverse direction on ball valves to confirm if they seal correctly with the check valve part.

 

[DJ_L3ThAL]

One other thing and part of why I never use any KK products anymore, is during stressful time consuming leak finding their bball valves on the manifolds leak out of the face where the barb seals to the valve body. Theres no oring or seal, its a metal and metal seal which is horrendous for gas, shocking design.

 

[takai]

Metal on metal is par for the course with floating ball check valves. From the exploded design diagram the JG valves are identical. Some will use conical seats, but the majority are this design
From what I can tell the black/red KK push fit check valves are a better design for low pressures, using convoluted rubber diaphragms rather than a ball and seat.

What would make this better is a higher pressure spring. But that would also increase crack pressure as a side effect.

 

[DJ_L3ThAL]

Loads of ball check valves I use at work have a Teflon or similar seat for the ball to seal in gas applications. Metal on metal (of same type) is poor practice with gases unless it is specifically softer on the seat, doubt the KK valves have this thought in their design.

Anyway back to your problem. I'd not be concerned about the check valve not sealing when pressure is in equilibrium across it, as it's not really the function of a check valve to seal off in this state. You want it to seal with liquid or a higher pressure gas coming from the downstream side which by sounds of it will happen. You say you have a known leak upstream, being the likely root cause of your issues I'd address that in the first instance. Otherwise if you are accepting the upstream leak as normal, you should just accept the pressure loss in the kegs as normal too. I'd also submerge the entire manifold into water with pressure up at say 2 bar and any minor leaks will be quite apparent then. You'd be surprised just how many leaks those crappy manifolds have. I personally confirmed this with two manifolds and eventually just bought a higher quality one which still had a couple tiny ones that I fixed myself with Loctite thread sealant.

 

[takai]

Its not so much that pressure is in equilibrium but rather that it is a very small pressure differential which will allow leaking. The manifold isnt leaking at all, you can blame Sodastream for that, and it has been rectified with a ball valve on the connector for the Freedom One adaptor. However, while it may not be an issue for you, for those of us who have live yeasts and bacteria in a keg having it cross contaminate into another keg isnt desirable, and this is what the one way valves will effectively allow.

As for the seal design, the majority of gas seals in food grade check valves are not PTFE or viton seated, they are simply ball bearing on aluminium or stainless seat.

Also FWIW your anti KK sentiment is getting really old, the vast majority of their items are fit for purpose, and they are more than willing to listen to feedback a la this thread.

 

[DJ_L3ThAL]

Your source of pressure is the bottle/sodastream bulb. Sure if there was significant flow at the leaking bottle you would temporarily have a differential back across the check valve, but it should seal from this. If it was passing you would find the pressure would eventually equate the whatever the pressure drop caused by the leak was. I't sounds like the best attention you can have on your system is to look at your ball valves and manifold under water to confirm there are no leaks.

If you find that yeast can travel with CO2 from keg to keg that would be amazing.

Not sure where your "majority" comes from with food grade check valves, we sell them as a distributor at work and I see this stuff everyday, teflon is extremely common seal for gas applications.

My anti KK sentiment? I'm merely providing my feedback on actual experience. Get off your fanboy high horse this is a forum to help people, of which I am trying to help you. I fell into the trap of buying cheap stuff that appeared at face value to have all the necessary features, only to waste hours troubleshooting, returning goods and then shopping for better quality items. I'm not brand bashing only providing my experience to hopefully guide others who don't wish to waste time down a more hassle-free path.

Anyway, good luck with the leaks.

 

[technobabble66]

I'd suggest if the back flow is important to prevent then pull out your check valves and put in proper, gated, ball valves. If that's too expensive (I'm a tight arse so appreciate that's a fair chunk of cash!) then I can only assume it's not *that* important. It might be you're slightly over engineering/analyzing the problem whereby the solution could be fairly simple. There's no way you'll definitely definitely get 100% perfect seals on those one way check valves 100% of the time given your specific conditions. You need to go to something that can definitely do that 100% of the time - ie: something like a manual ball valve.

Fwiw, I know people who have had both good and bad experiences with KK. Mine has been good, but many have experienced significant problems with faulty KK equipment. I happen to know DJ has had every piece of equipment he's obtained from KK to be faulty (except for some keg jackets) & spent a vast amount of time wasted searching for the causes of various problems to eventually determine its from these bit of faulty equipment. Considering this he's fairly measured in his statements really.
I think they're great in many ways, but I appreciate that sometimes the quality of their gear is less than ideal.