Ominous tanks leak?? Help

[T. Bix]

Super super super easy fix. Call John at Northeastscubasupply.com and get a valve rebuild kit for those valves...... now here's where you have to be careful!!! OMS valves are a little different in a couple of areas so make sure you let him know or even send him pics so he gets the correct parts to you. Buy some extra burst discs while you're at it, if you pop one on the road it can be a bit challenging to find them at most LDS's.

 

[halocline]

The leak is from the isolator valve, not the crossbar.

This leak would allow both cylinders to drain. If both cylinder valves were open, then both cylinders would drain equally through the leak.

However, if one of the cylinder valves were closed, then such a leak would drain only the cylinder with the open valve.

Hmmmm? Seems to me the isolator is working fine. He closed it and it saved the gas in the tank that has no leak while the leaky tank is draining....exactly what the iso valve is supposed to do. Or did I read something wrong?

Anyhow, getting back to the valves, the o-rings in the valve are likely all downstream of the valve seat itself, so the most likely thing is the burst disc or the tank neck o-ring. In fact, that's all it could be unless you have a leaky valve seat, which is pretty rare. One other possibility is a leak on the manifold crossbar, but those usually have 3 o-rings in series, so it's almost unheard of to have a leak there.

If you want to eliminate the possibility of a leaking valve seat, just put a regulator that you know does not leak on the post, then store the tanks for a while with that post open (iso still closed) and see what happens. If the leak stops, the problem was the valve seat, if not, burst disc (probably), tank neck o-ring (maybe a bit less likely) or crossbar. (very unlikely.)

If it were my doubles, probably first thing I'd do is drain the offending tank, replace the burst disc, open the iso to equalize, then close it and store, check back in a few days. No more leak? Off you go....

All the other fixes require dismantling the doubles, and at that point I would agree that it's time to rebuild the valves.

Edit: looks like you probably already took care of it.

 

[kotik]

Folks... look at the video he posted. The leak is coming from the isolator valve, it's not a burst disk or tank neck o-ring that's leaking. I agree that one of the tank valves must have been closed for only one tank to have drained down.

It looks like several posts have been made by those who see the problem, but that still recommend changing the burst disks. Is the thinking that an overpressure has blown a seal on the isolator, so the burst disk must not have burst when it should have?

 

[Dsix36]

When he shot the video, were both tanks valves off? Were both tank valves on?

It is possible the he has both a leaking tank valve and a leaking isolator valve.

 

[kotik]

When he shot the video, were both tanks valves off? Were both tank valves on?

It is possible the he has both a leaking tank valve and a leaking isolator valve.

If a leaking tank valve were turned off, there would be bubbles.

If a leaking tank valve were turned on, there would be a LOT of bubbles :blinking:

---- Edit ----
Now that I just wrote that, the mystery of only one tank dropping pressure arises again.

Ermaclob, by any chance did you have a regulator on one post and the tank valve open on that side, but the tank valve closed on the other tank?

 

[halocline]

Folks... look at the video he posted. The leak is coming from the isolator valve, it's not a burst disk or tank neck o-ring that's leaking. I agree that one of the tank valves must have been closed for only one tank to have drained down.

You're right, I totally missed the video. I bet that there are valve stem o-rings on one side of the iso valve that are exposed to tank pressure even when the iso is closed. I've actually never rebuilt one, only tank valves.

It never would have occurred to me that there would be stem o-rings or seals exposed to tank pressure when the iso is closed, but if it's closed in the video, it looks like that's the case. Not a very good design IMO.

 

[kotik]

If a leaking tank valve were turned off, there would be bubbles.

If a leaking tank valve were turned on, there would be a LOT of bubbles

Hmmm... now I see the flaw in my understanding of what we see in the video.

There are no regulators to be seen, so I assumed both tank valves must surely be turned off since we don't see air erupting from the outlets. But if they were turned off, then of course the isolator valve wouldn't be pressured up, and there would be no bubbles at all. So-- my incredible powers of deductive reasoning tell me there must be at least one of those pressure containing DIN plugs involved that can't be seen in the video. In that case, it would indeed be possible for one tank valve to be closed, one open, and the isolator closed for us to see what the video shows-- it could even be possible for the tank valve that's open to also have a leak itself but not bubble, if the isolator valve leaks a lot worse.

Dsix36, you're right, there's not enough information to positively root out all possible leaks. Who would have thought there could be so many things to verify?

Here's what sounds like a thorough diagnostic check procedure, critiques are welcome--

1. Cap off both tank valve outlets, either with a regulator or pressure containing plug.
2. Submerge the valves in water.
3. Open both tank valves and the isolator valve.

If there are no bubbles, the test is over. This is actually the worst result you could get, since now the only possible cause of the original problem is that an evil genie is stealing some of your pressure, and possibly raiding your refrigerator as well.

If any valves do leak, turn them off. If the leak at that valve stops and no new leaks appear, then again the test is over, and the problem is isolated to that one valve. Same with a burst disk or neck o-ring-- turn off the valve on that side, in case there is a smaller leak somewhere else that might be revealed by closing off the larger leak path.

 

[Fishpie]

If valves are leaking from the handle it's probably the packing around the stem (might be the copper washer).
The quick and dirty fix is to generously lube the o ring / teflon washer(s).
No need to drain tank(s) or disassemble manifold.
Remove handle.
Undo packing nut.
Remove valve stem.
Separate teflon washer/o ring sandwich (replace if you have parts).
Lube with appropriate grease.
Put back together making sure that the valve stem is mated properly with seat and copper washer is in place to seal the packing gland.

 

[Dave1w]

If valves are leaking from the handle it's probably the packing around the stem (might be the copper washer).
The quick and dirty fix is to generously lube the o ring / teflon washer(s).
No need to drain tank(s) or disassemble manifold.
Remove handle.
Undo packing nut.
Remove valve stem.
Separate teflon washer/o ring sandwich (replace if you have parts).
Lube with appropriate grease.
Put back together making sure that the valve stem is mated properly with seat and copper washer is in place to seal the packing gland.

I'd say Fishpie is most probably right. The cylinder valves only isolate the regulator outlet, not the manifold, so when you close them, you do not turn off the cylinder to the manifold, just to the outside world. Reason it would leak from one side and not the other when the isolator is closed is that one side goes to the seat side of the seat plug, and the other goes to the back side of the seat plug, so the O-rings, or in thermo, the copper crush washer around the stem is probably the problem.

Take a look here http://www.xsscuba.com/graphics/valves/manifold_thermo_new.jpg and Exploded View - Manifold Center Bar to see what the insides look like.

If it's a Thermo, they have a copper crush washer and you can sometimes just tighten down the bonnet nut thing a bit more and crush the washer a bit more, but I think most other manufacturers have now moved to o-rings. I would hesitate to do any servicing of the manifold valve with the tanks pressurised, because unlike the tank valves, the manifold isolator valve can see pressure on both sides of the seat plug (take a look at the diagrams above).
So your options are to get a service kit, drain, disassemble the valve, clean replace and lube o-rings and rebuild it, which really actually quite simple, or for peace of mind take it to a shop.

 

[halocline]

There are no regulators to be seen, so I assumed both tank valves must surely be turned off since we don't see air erupting from the outlets. But if they were turned off, then of course the isolator valve wouldn't be pressured up, and there would be no bubbles at all.

Yes it would be. The manifold connects the tanks upstream of the posts, it's always pressurized. The problem is in the design of the isolator. Ideally it should shut off flow between tanks in a way such that the seat seals flow from each tank to the iso valve. Looking at the parts, it's clear that the seat only seals flow from one tank to the valve, and the other tank's gas is sealed by the stem o-rings. The seat does keep air moving from one tank to the other, but I am really surprised to learn that one tank's gas could escape out the iso valve stem o-rings.

On a regular tank valve, you can remove the stem and it's o-rings from a pressurized tank with no problem; the seat stays in place and seals the air.