Does air go bad?

[Oddsnends]

Wracking my brain, took my inorganic chemistry a few years ago...

If I recall correctly,
Iron + O2 gives Fe(II) species
Iron + O2 + H2O gives you Fe(III) species

The bridge example above is referring to an Fe(II) species: a tightly adhering iron oxide coating. The nightmare scenario referenced in the DAN fatality refers to Fe(III) species (aka. rust).

The key is this:

O2 + 4 e- + 2 H2O → 4 OH-

Anyone with chemistry knowledge knows this is an "Oh %#$" situation. You now have to worry about acid/base reactions.

You get redox reactions like: 2 Fe2+ + 0.5 O2 → 2 Fe3+ + O2− in presence of H2O

And equilibriums that favor: Fe(OH)3O2 and other Fe(III) species

The Key here is that rust is not passive like iron oxides are. It does not form a protective layer. It degrades continuously causing pitting. So yes you could produce enough rust to consume most oxygen in the tank.

The key here is you need BOTH water and oxygen. IF you add some electolytes (read salt water) the reactions will be catalyzed at an even faster rate.

 

[bleeb]

That particular tank contained several ounces of rust-colored water, and the tank itself had extensive rust and pitting throughout its inner surface."

Is this a typo? Less than 1% O2 and contributing it to rust? Something doesn't add up.

I appreciate you posting this because I asked for it, but it doesn't quite make sense. Everyone who was involved in the chemistry discussion, including a PHD physicist, agreed that it would take at least 2lbs of rust to bring air down to 1%.

FWIW and I know this isn't conclusive, but 2 lb of steel is about 8 cu.in., which IF spread evenly (the accident report said "throughout") over the entire inner surface of an LP85 (7.25" OD x 24.5") would be roughly 0.013" thick, which doesn't sound like all that much.

 

[EastcoastdiverLLC]

Is this a typo? Less than 1% O2 and contributing it to rust? Something doesn't add up.

I appreciate you posting this because I asked for it, but it doesn't quite make sense. Everyone who was involved in the chemistry discussion, including a PHD physicist, agreed that it would take at least 2lbs of rust to bring air down to 1%.

Not sure Jeff,

That was a cut and paste directly from the DAN report.

Not a PHD, but I am a dive professional...lol...

Gary

 

[Oddsnends]

Found it:

The action on the left is resulting in pitting, so you can't just consider a thin coating over the entire surface (what some of you refer to as flash rusting?). In the presence of water (especially with salts or low pH) your cylinder will continue to get "eaten away."

For anyone who has made a galvanic cell battery, remember what happened to the anode?

Also, wouldn't just a little free rust risk a clog in that narrow tube thingy on the valve assembly?

 

[cummings66]

Wasn't there a diver in Canada last year who died of this as well? I'm thinking I read somewhere where he got down to about 3% in his steel cylinder. I'm certain I read it somewhere recently, I just can't find it.

At any rate, regardless of theory, it has happened and was proven to happen in an old (paper) study. Too bad it was never put on the internet for all to read. I myself would not use the air until a visual and if needed a hydro was done. It's not that expensive and worth the piece of mind.

 

[jefffalcone]

Wracking my brain, took my inorganic chemistry a few years ago...

If I recall correctly,
Iron + O2 gives Fe(II) species
Iron + O2 + H2O gives you Fe(III) species

The bridge example above is referring to an Fe(II) species: a tightly adhering iron oxide coating. The nightmare scenario referenced in the DAN fatality refers to Fe(III) species (aka. rust).

The key is this:

O2 + 4 e- + 2 H2O → 4 OH-

Anyone with chemistry knowledge knows this is an "Oh %#$" situation. You now have to worry about acid/base reactions.

You get redox reactions like: 2 Fe2+ + 0.5 O2 → 2 Fe3+ + O2− in presence of H2O

And equilibriums that favor: Fe(OH)3O2 and other Fe(III) species

The Key here is that rust is not passive like iron oxides are. It does not form a protective layer. It degrades continuously causing pitting. So yes you could produce enough rust to consume most oxygen in the tank.

The key here is you need BOTH water and oxygen. IF you add some electolytes (read salt water) the reactions will be catalyzed at an even faster rate.

I dropped out half way through my junior year, so I didn't get to inorganic. Thanks for the clarification.

 

[Luis H]

This has been an interesting thread. It is great to actually see what seem to be some actual facts and not just some wild opinions.

From the chemical reaction stand point I have nothing to add since I am not a chemist…just an engineer.

The self sealing rust limiting steel alloys such as CORTEN (proprietary name) or ASTM 588 are very effective as they do not tend to pit even under a wet environment and to a lesser degree in a salt water environment. There rust tends to provide excellent protection against further corrosion.

Steel tanks are not made of ASTM 588; most of the 3AA steel tanks are made out of ASTM 4130, which is a Chrome-molybdenum steel. Chrome-moly steels are better against corrosion that plain carbon steels, but they are selected due to their ease of working before heat treatment and strength properties after heat treatment.


The only high corrosion study in Scuba cylinders that I am aware of is from the University of Rhode Island. I have the test report at home, but some of the details that I remember are as follow:
• They tested both steel and aluminum cylinders
• They put about a 1/2 liter of salt water in some of the tanks and a 1/2 liter of fresh water in the other.
• Tanks were filled to max working pressure
• The temperature was maintained to simulate max tropical condition

I can’t recall the O2 levels at the end of the test, but the test had to be cut short because the corrosion in a couple of steel cylinders was starting to compromise the structural integrity of the cylinder, in about a six month period. The rust around the internal water line was creating pits deep enough that they feared the cylinders could catastrophically fail at any moment.

Most of the corrosion occurred at the air to salt water interface and the conditions were intentionally the worst case for the test purpose.


Personally, I recently bought a steel 72 tank that the air in it was 40 years old. I tested the air in it for O2 content. After 40 years it had a perfect 21% O2. The tanks had a bit of rust, but not bad. Obviously it wasn’t progressing. The tank only had about 1200 psi. If it was full I would have dived it after checking the O2 content…how many people could say they dove with 40 year old air. I bet it was fresher back then.

 

[divechilly]

Thanks to all for sharing your ideas!