oxygen tank death

[Bob DBF]

Just the other day, read in the NYT that the journalists/columnists don't write their own headlines.

It's a small trade journal, not the NYT, journalists/columnists write their own headlines (titles), whether it is actually used by the editor is another story. At least that what a friend tells me who works on the local rag. And in any event, it's the same crowd, different jobs.


Bob

 

[TrimixToo]

The narrative is unclear. The most likely explanation I can come up with is that the valve ejected from a pressurized cylinder while the protective cap was in place. The protective caps are not strong enough to withstand the considerable forces that full cylinder pressure would deliver and would have come off. Due to the rounded shape of the cap it would result in blunt force trauma to whoever it hit.

The most common cause of valve ejection is incompatible threads, either threads that are the same size but that differ in form (as with British vs USA threads), or threads that are a similar but not identical size and pitch (as can happen with certain combinations of metric and non-metric threads). Another possible cause is threads damaged by wear or corrosion.

The industrial packaged gas industry has an excellent safety record overall. If I'm not mistaken there are more fatalities per capita in construction (and many other industries).

Then again the photo makes it look like they don't use caps.

I don't agree the cap was likely in place, or at least fully screwed on, as all the ones I get from my local supplier are screwed down. They tell me it's illegal to transport a tank this size with no a cap installed, which if true I'd assume to be a DOT rule.

I just measured a (spare) CGA246 valve stem one with a ruler, and its diameter is about an inch. There are *two* one-inch holes (or slots with similar surface area) in all in the caps for my small stable of T bottles in the garage, or about double the surface area of a removed valve, and more than enough threads to retain it on the tank if a valve were to fail or come off. The caps are not going to need to hold anything near the tank pressure even if a valve comes out, and if screwed on all the way it is difficult to imagine that one would just "pop off." The same caps are used on tanks rated 6000 PSI as the ones rated 2400 PSI, too, so apparently the DOT and CGA think they are strong enough, and retained well enough.

But...there are at least two different threads used on the caps for DOT (or ICC)-rated cylinders. A cap with the larger one might fit over the threads of the smaller male thread without being screwed down. For sure, there would be nothing like full thread engagement. Having the wrong cap on and transport without a cap on, seem more likely to me, but what do I know?

...just my 2 PSI.

Edit: I got curious enough to measure the wall thickness of a cap (with a blade micrometer), and of a handy piece of an LP steel scuba tank. The cap was .127" and the tank wall was .123" (both nominally 1/8", I suspect).

 

[Bob DBF]

Having the wrong cap on and transport without a cap on, seem more likely to me, but what do I know?

I don't think that's an issue since, in the picture of the warehouse in the article (link below), I did not see a cap on any bottle.
Cylinder Training Services. Oxygen cylinder disembowels driver



Bob

 

[Dan]

I don't think that's an issue since, in the picture of the warehouse in the article (link below), I did not see a cap on any bottle.
Cylinder Training Services. Oxygen cylinder disembowels driver



Bob

So many rusted tanks too. I wonder whether their inspection stamps are upto date.

 

[JackD342]

So many rusted tanks too. I wonder whether their inspection stamps are upto date.

I also doubt there was a cap in place. If there was, it would have required BOTH valve AND cap to tear off in order to have the result described. Certainly sounds like the valve threads failed. But if the driver was doing the "tilt and twist" to roll the cylinder on edge, he may have twisted the valve counter clockwise and loosened it. Tapered threads, right? If so, how much or how little loosening is needed to blow it out? May also indicate that along with no cap during transport, the valve was not installed tight enough?
And, could be loosened valve PLUS bad threads?

 

[caydiver]

Many years ago when I was a kid I was at a bbq at a friends house. All us kids were having a grand old time in the lake and were not around when a 2 liter bottle of soda that had been sitting in the sun unopened popped and the plastic top took out my friends dads eye and did other damage. He was a surgeon who never drove again much less operated. Accidents happen. It is horrible for victims and their families but unless someone with forensic knowledge has first hand access to the scene and conditions at the time of the event, all the questions, answers and suppositions don’t mean anything. It was a big brand soda and the company settled with the victim before the lawyers got involved. The kids grew up just great and the mom and dad remained as charming and wonderful as they were before a freak accident changed their lives. This poor man died and hopefully those depending on him will be made as while as possible and they can reach a place where everyday does not revolve around “how”. Everyone can second guess as much as they want but I truly doubt anyone will find answers on a dive website. I do hope the family and loved ones get the answers they need to give them some peace.

 

[Bob DBF]

So many rusted tanks too. I wonder whether their inspection stamps are upto date.

I wonder if it's even in the US. I've been going into industrial gas outfits since I was a kid with my brandy new oxy-acetelyne torch. I expected to see bottles capped and in corrals or some kind of restraint. Also I noticed the bottles stacked horosontally in the truck, which Is not an approved method of transport in the US.


Bob

 

[kelemvor]

I can't figure out your dolly question. I interpret the dolly as meaning a two wheel hand truck. Does that help?

Yes, that's what I would call a dolly. However, I've never seen one that could be tilted away from the operator. Hence my question. Those two bullet points came from the linked article.

I bought a tank of shielding gas last weekend, so now i've had a good look at how those caps work on industrial tanks. It's kind of too bad threads like that aren't included on scuba tanks. Seems like a great way to protect the valve.

 

[rjack321]

I also doubt there was a cap in place. If there was, it would have required BOTH valve AND cap to tear off in order to have the result described. Certainly sounds like the valve threads failed. But if the driver was doing the "tilt and twist" to roll the cylinder on edge, he may have twisted the valve counter clockwise and loosened it. Tapered threads, right? If so, how much or how little loosening is needed to blow it out? May also indicate that along with no cap during transport, the valve was not installed tight enough?
And, could be loosened valve PLUS bad threads?

There is no way you are going to hand loosen a 3/4" NGT CGA 540 valve that is properly installed. The torque required is huge - like 600+ ft/lbs
Pipe Thread Installation Torque Table Chart | Engineers Edge | www.engineersedge.com
54 in/lbs * 12 = 648 ft/lbs (edit should be divide so 54 /12 = 4.5 but this source is incorrect)
If its not properly torqued, its going to leak long before you move it.

So it most likely broke off, or based on the pictures the cylinder threading corroded to the point of failure.

 

[Wookie]

There is no way you are going to hand loosen a 3/4" NGT CGA 540 valve that is properly installed. The torque required is huge - like 600+ ft/lbs
Pipe Thread Installation Torque Table Chart | Engineers Edge | www.engineersedge.com
54 in/lbs * 12 = 648 ft/lbs
If its not properly torqued, its going to leak long before you move it.

So it most likely broke off, or based on the pictures the cylinder threading corroded to the point of failure.

No. 54 in-lbs * 12 is 648 in-lbs. 54 in-lbs / 12 ft-lbs/in-lbs is 4.2 ft lbs. I don't know where that chart came from that you referenced, but I put most screw pipe at 1 star torque, that is, I tighten it until I see one star. Cylinder valves are at 2 star torque, that is, put the cylinder in a air operated cylinder vice and tighten it until you see 2 stars. Removing valves takes many stars. I can't find an actual torque spec for cylinder valves aside from handle torque.