Checking for oxygen bottle leaks

[crb]

I recently changed the oxygen bottle on my gas welding setup, and felt a significant leak onto the back of my hand from the coupling between the regulator and the oxygen bottle. This turned out to be a fault with the oxygen bottle itself, which was replaced by the supplier. Being a bit hard of hearing, I would not have heard any hiss (even if there was one), and it was lucky I wasn't wearing my welding gloves at the time otherwise I would have been none the wiser about the leak. The question arose in my mind about routine checking for leaks (or at least when changing bottles in the future). My supplier suggested that water could be used, but cautioned against adding washing-up liquid; he did not know of any commercial leak-detecting liquid. Any suggestions? CRB

[Patrick]

Something like this might be of use;

[Grunff]

What was his reason behind "don't add washing up liquid"?

When I worked in a lab, each lab had its own bank of cylinders (oxygen, nitrogen, sometimes argon, and sometimes acetylene). Next to each bank was a dedicated washbottle used for testing connections whenever a new cylinder was fitted. This contained water with a squirt of washing up liquid. Always worked.

[Andy Dingley]

Use weak detergent solution in water - but the detergent should be something like Teepol (BOC's own recommendation) that's _just_ a detergent, not the typical "washing up liquid" that's thickened with salt. I use Ecover (from the kitchen), same as I use to wash the car (and for much the same salt-free reasons).

If you're stuck, drool on it (ask Dr Drivel for expert instructions). Don't spit on it, you can't tell if they're your bubbles or the cylinders'.

Be very careful in looking for leaks around the high pressure oxygen side. Also keep the cylinder valve covers in place and check for dirt before connecting. There's a ridiculous pressure in those cylinders and it's enough to cause accidents you wouldn't imagine where possible.

[Jeff]

and don't let anything oil based near it & don't use ptfe tape

Regards Jeff

[Paul]

You can use PTFE tape the oil free kind, and snoop is the commerical name for a leak detecting fluid.

Paul

[Gizmo]

Do those oxygen bottles not have to be pressure tested / crack tested in the same way that scuba tanks have to be tested ? And marked accordingly

[Dave]

Your regulator will have a cylinder contents gauge; ensuring it's output is turned off, fit the regulator to the new cylinder, turn on the cylinder valve and observe the gauge indicates a full cylinder, turn off the cylinder valve and watch the gauge - if it starts to fall you have a leak. Of course this may be either the regulator or the cylinder connection - to narrow it down you'll need some leak detector fluid. This should really be just a plain detergent (BOC recommend 'Teepol'). As with all high pressure cylinders it is very important to avoid anything oily or oil based coming into contact with the valves; this includes hand creams and barrier creams!

Dave

[Dave]

Yes, they are tested - the test pressure should, along with other information such as date of manufacture and water capacity, be stamped into the neck of the cylinder. There is usually a plastic ring around where the valve meets the cylinder body indicating when the next test is due.

Dave

[Dave]

In the aerospace industry, we were told to use Hederol (sp)

I don't know if that helps you.

Dave

[mrcheerful]

most plumbers places sell aerosol leak detector that is safe for any gas (Mark Vittow), or a very dilute washing up solution is the most commonly used.

The leak you describe is often caused by the taper insert in the outlet, it comes loose easily, you 'can' tighten it with an allen key, but getting them to change the bottle is better.

mrcheerful

[crb]

Thanks for the suggestion. I already have a can of that, and have successfully used it on natural gas installations, but it does not specifically mention oxygen, so I was wary about using it. I have asked the manufacturer/distributor for their opinion - no reply yet. CRB

[crb]

He didn't give a specific reason, only that "oxygen is funny stuff". Perhaps he was just being super-cautious, as he said he might use washing up liquid himself, but he wouldn't advise anyone else to, and recommended just plain water. CRB

[Ian Stirling]

Washing up liquid has enough organic content that it could concievably burn in high pressure oxygen. (practically anything will) Once the detergent-water falls off, a film of water/detergent will be left, and it'll leave a thin film of 'pure' detergent, once it dries.

While this is _probably_ not a risk, especially if you use it at the bare minimum concentration, where it only reduces the surface tension, without any extra, putting films that could concievably burn on a surface that'll be exposed to 3000PSI oxygen is something that is generally only done after extreme thought.

[Andy Dingley]

So what ? _Fingerprints_ will burn in high pressure oxygen. But there's little fuel mass, so it's no big deal. Same for detergent traces

- the salt worry is about corrosion of the fittings.

PTFE tape cant be used on high pressure oxygen because of its cold flow behaviour - it can literally be squeezed out of the joint. For tapered threads after the first reducing orifice (such as the gauges) it's OK. Fluorine compounds aren't really interested in burning in mere oxygen.

[James]

This stuff called Micro leak detector from refrigeration technologies . It's used in hvac to find the smallest of leaks is condensers, coils. Line set... I use only that or dawn dishsoap only.. it's like 10 bucks for a big spray bottle and a little goes a long way with this stuff... it will be your best friend.. you just wipe it off after checking for leaks