Prevalence, causes, and prevention of oxygen fires

[Michael Guerrero]

Anecdotal reports suggest that:

1. Fires, explosions, equipment incidents where combustion is believed to have occurred, burns, etc., are common in diving equipment using >40% oxygen.

I think I need to better understand what you consider to be common, as using my definition I don't believe this to be true. Of the many tech divers I've dived with, I've only known one to suspect he had an incident by borrowing a regulator that was supposed to be clean but was dirty. I haven't had an issue after 3 years of diving with O2 and doing all my own O2 cleaning of my regs and tanks.

 

[NAM001]

Anecdotal reports suggest that:

1. Fires, explosions, equipment incidents where combustion is believed to have occurred, burns, etc., are common in diving equipment using >40% oxygen. These take place despite widespread understanding of the risks, by people who believe they are using safe practices.
   
   
2. Most of these incidents do not appear in DAN reports because they do not occur during the dive itself. Those that occur during a dive are usually not fatal, and are not reported to DAN for that reason.
   
   
3. Oxygen is widely used for medical treatment, welding, and other industrial purposes. While fires etc. do occur with these uses, they are extremely rare on a per-cylinder-handled basis (with the exception of extremely common fires caused by individuals smoking cigarettes while using home oxygen)

I'd like to start a discussion on how we can improve the safety of this area of technical diving, and would welcome your thoughts.

What is needed to discuss. I haven't read the things you have but i have worked decades with O2. Any fires i ever knew about were not O2 fires. they were O2 fed fires. it was trash burning and the O2 made it burn hotter. There was a few and very few fires of O2 it self as listed on a submarine. I have been looking for those pictures a long time and can not find them. When the space capsule burned on the launch pad I believe they did studies after wards that said less than 60% acted as air and over 60% it acted as pure O2. One can make what they want from that but I take it as over 60% IS THE THRESHOLD FOR HAVEING high O2 envoronment related fires. A fire is a fire the more O2 the hotter but what diffrence is there between a 2500 degree fire in a 21% atmosphere and a 3500 degree fire n a 100% atmosphere. either one will do you in. For me it is the threshold of ignition when talking about O2 concentrations. I dont hold a lot of credibiity is the reported dangers of O2 as presented. The always show a rag with an O2 bleed on it igniting with no match. Never seen it and and have never been ale to recreate the claim. I do not deny they exist. I have spent years n spaces with pure O2 bleeds to them and people hanging their cloths on the discharge nozzles. Never a fire. Of course we clear these behaviors as fast as they can be reported. fires need 3 things we all know. the less quality of one then the higher quality of another component must exist to get a fire. IE damp material then you need more heat or more O2 environment to get the same fire. I am told that O2 itself will burn. I also have never seen that. but then aluminum will burn as a fuel also under the right conditions. For everything I have seen O2 has never been a fuel but a darn good accellerant.

 

[kensuf]

If you have an IP creep or something happen to the 1st, your whole deco bottle is gone until it can be serviced. This will probably cancel your dive day depending on logistics available. A quick 1st stage swap with a spare from your bag is a lot easier.

I love the idea of those integrated valve/regs especially for bailout but its just not practical for me.

This is why I don't rely on color coded hoses / second stages for my gases, but instead rely on the bottles being properly analyzed and labeled.

 

[Colliam7]

ILuck? O2 isn't "that bad" unless you really get all the pieces of the fire triangle in place? NASA had years of successful shuttle launches, until they didn't. They also lost the crew of Apollo 1 to an O2 fire decades before that but after many successful Gemini launches.

Agree. And, that was my point, in using the example of the fill station I referred to, and the general lack of awareness in many shops of reasonable procedures.

The problem, possibly what started 2airishuman thinking about this, is the catastrophic nature of the results when there IS a fire. Apollo 1 was an example of a catastrophic outcome. The event I referred to in an earlier post (and linked to a news report) was another.

In that latter case, the injuries were severe, but no deaths occurred. It was not altogether clear that it was more than a case of (very) bad luck. But, the shop that had cleaned the involved cylinder for oxygen service, and filled it, was held liable (and is no longer in business). And, one of the reasons is that they had no documentation of the procedure they used for cleaning equipment for oxygen service. As I understand it, they 'said' they did it, but had no written procedure to document what they did.

So, where does that leave us (to paraphrase a line in the 'myth' article)? As a shop person, filling cylinders with O2, or even blending high (>50%) percentages of O2, I want to stack the odds in my favor, so I want the cylinder and valve cleaned for oxygen service. That is not a guarantee of the outcome, but it is one of the contributing variables that I can influence. As a diver, notwithstanding any appreciable evidence of a problem, I will have my deco regs (and cylinders) cleaned for oxygen service. I can't prove it makes a difference. I can't say it is more than a CYA procedure (legally and personally) on my part. But, it is a relatively small price to pay.

Thanks for the follow-up comment on the 'nitrox VIPs'. I understand the procedure (black light inspection) although I have never used that particular term. And, I certainly don't see value in performing that for cylinders being filled with pre-mixed enriched air <40%. But, such inspections may have a place. I 'grew up' in a shop where we did A LOT of blending, and cleaned a considerable number of cylinders / valves and regs on a periodic basis. We had a blending board that was superb in design (by a staff member who was an engineer in his day job), construction, and functionality. We did some things that others considered ill-advised (as one example, we would boost 100% O2 to 3000 psi fill pressure, when other shops would only fill to supply pressure of 2400 psi). Unfortunately, that shop closed after the untimely death of the owner at a young age from a non-diving medical event. But, about a month before his death we (the owner, the staff member I mentioned, and I) held a meeting with the head of PSI/PCI to discuss what we 'should' be doing to mitigate risk when using 100% O2. And, one of the things we agreed to consider was to periodically inspect cylinders, that had been previously cleaned for oxygen service and had only been filled in our shop, rather than automatically re-clean them. We were going to get together again in 2 months to make a decision. But, the owner's death and the closure of the shop cancelled those plans. Still, there was sentiment among some members of our staff to move in that direction, simply because cleaning cylinders and valves is a tedious process.

 

[boulderjohn]

...they had no documentation of the procedure they used for cleaning equipment for oxygen service. As I understand it, they 'said' they did it, but had no written procedure to document what they did.

The school district for which I worked was the first ever large school district to be ISO certified (in the administrative areas). Although that did not last long, while it lasted, I was in charge of internal audits. ISO stressed the importance of having all procedures clearly identified and readily available. If I ran a dive shop, my work area would have the procedures for tank fills and oxygen cleaning prominently posted. I would use something like the PSI-PCI inspection checklist used, with the used list for each tank filed.

 

[rjack321]

Thanks for the follow-up comment on the 'nitrox VIPs'. I understand the procedure (black light inspection) although I have never used that particular term. And, I certainly don't see value in performing that for cylinders being filled with pre-mixed enriched air <40%. But, such inspections may have a place. I 'grew up' in a shop where we did A LOT of blending, and cleaned a considerable number of cylinders / valves and regs on a periodic basis. We had a blending board that was superb in design (by a staff member who was an engineer in his day job), construction, and functionality. We did some things that others considered ill-advised (as one example, we would boost 100% O2 to 3000 psi fill pressure, when other shops would only fill to supply pressure of 2400 psi). Unfortunately, that shop closed after the untimely death of the owner at a young age from a non-diving medical event. But, about a month before his death we (the owner, the staff member I mentioned, and I) held a meeting with the head of PSI/PCI to discuss what we 'should' be doing to mitigate risk when using 100% O2. And, one of the things we agreed to consider was to periodically inspect cylinders, that had been previously cleaned for oxygen service and had only been filled in our shop, rather than automatically re-clean them. We were going to get together again in 2 months to make a decision. But, the owner's death and the closure of the shop cancelled those plans. Still, there was sentiment among some members of our staff to move in that direction, simply because cleaning cylinders and valves is a tedious process.

There are a number of reasons why black light inspections are really not helpful and give you a false sense of security. PSI has perpetuated some of these myths unfortunately as they are not chemists and really are just reiterating CGA guidelines in different words.

1) Not all hydrocarbons even fluoresce under the ~350-400 nm UVa wavelengths of black lights. So you could have massive amounts of contamination and not see it. Massive is a subjective word because you only need 0.5 mg of hydrocarbons in a square meter surface area to no longer be clean (and that's pretty dirty)
Here's an old thread that discussed how most hydrocarbons we care about, including silicon greases and synthetic compressor oils don't show up under the long wavelength UV light from a black light
UV lights for checking tanks for hydrocarbon contaminiation?
And these were massive gram sized blobs of contamination.

2) The cylinder is almost never the point were contamination accumulates to the point of ignition anyway, the valves are the choke point. Once the valve contamination ignites, the cylinder pressure often rises so dramatically that the cylinder actually explodes and/or the neck is so weakened that the cylinder explodes. Pressure goes up in a flash fire waaaay faster than a burst disk could release it, and in an aluminum cylinder its microseconds between cool and melting. You can't see the insides of the valve and once you take it apart you might as well wash it anyway.

3) Thus my recommendation to annually clean valves and biannually clean cylinders - but only those cylinders and valves that actually have to (due to the need for >40%) be partial pressure filled in the first place. Because there is no suitable way to inspect for the types of contaminants at the relevant levels we care about.

 

[Magnus Lundstedt]

I don't think the casual diver takes a blender course, there isn't much of a necessity to know how to mix your own gasses considering most dive clubs (around here at least) will not let you use their compressor, they will do it for you.

In our club most divers do their own fill, half the club's members have the blender card.

I know of many clubs where every member have access to the compressor at any time, to use it the member only need a 5 minutes instruction from some other member.

 

[Scubagaskets]

The 2 most important o-rings when you talk about pure oxygen is 214 and 112.
Scuba Dive Viton®Green Sh90-112 for tank valves
Genuine Green VITON® Sh90-214 - Premium O-rings Provider for Scuba

apart from that at scubagaskets we have a separate viton o-ring category for pure oxygen applications

VITON® Green Scuba O-Rings for pure Oxygen

 

[KevinNM]

The med O2 tanks I was familiar with were 2215# and operational flow rates were low. This may have an effect on the the continued use of the O-rings.

The vast majority of medical O2 problems are due to poor practices in an O2 rich environment, after the O2 is clear of the tank and reg.

There was the series of medical O2 fires that got aluminum regulators banned in the late 90s. I lived in one of the towns that had an ambulance burn up and a paramedic hurt.

 

[KevinNM]

Even oxyhacker recognizes that if you go slow enough and keep the heat side of the fire triangle low enough, you can basically put 100% into a gasoline container. Not really a great idea but you only have to take away one side of the fire triangle. We agonize over the fuel side and give lip service to the heat side. I have yet to see anyone in this 4 page thread even mention heat (except in the context of fast opening valves).

It's not typically heat like you think, though it can be. It's often particles accelerated by the gas flow and striking the contamination.

"In the case of gas velocity, it is not the flow of gas that can cause ignition, but a particle that has been propelled by the gas and impacts the system with sufficient force to ignite. The heat generated may be sufficient to start a fire, depending on the material impacted. Friction from a component malfunctioning or operating poorly can generate heat. Friction between two materials generates fine particles, which may ignite from the heat generated."
http://www.airproducts.com/~/media/files/pdf/company/safetygram-33.pdf