Overfilling LP Steel Tanks -- How bad is it?

[Cave Bum]

Wow. This is horrible. Sorry to hear this.

Question, if the tank was not yet being filled, how much PSI was still in the tank? Was it o2, and was there any fire related to the event?

It was incredibly devastating to the owner. You can see that in his eyes any times he talks about it. I had only met the lady, her name was Fran, a couple of times getting fills, but a loss like this effects us all, even divers and fill station employees who never met her.

To answer your question, I don't remember what I was told about existing pressure in the tank, but typically you don't refill O2 tanks until they've 500-800 psi and yes it was 100%. Yes there was combustion associated with this explosion and there was evidence of scorch marks around the hole on the picture of the tank that the owner showed me.

 

[Gombessa]

Wow, what a horrible event. My condolences, and it's somewhat worrying to hear that this event (dropping an O2 tank on its valve) isn't exactly a fill-station accident either, sounds like it could happen anywhere, anytime.

I would like to point out however, than explosive compressive shock sounds like a much different type of failure than slowly and systematically exceeding the rupture pressure of the tank. I.e., it's very possible that going from 4500 to 5000 psi over 5 minutes will produce a split or tear failure, whereas going from 500 to - 15,000 psi over 0.1 second on the same take will result in a completely different type of failure that isn't the basis of the particular DOT standard.

 

[Cave Bum]

...I would like to point out however, than explosive compressive shock sounds like a much different type of failure than slowly and systematically exceeding the rupture pressure of the tank...

I would agree but also we've had failures in the old 6351-T6 material that propagated along stress cracks and, as I understand it, those also fractured, creating shrapnel. When that material was first approved nobody had any idea that prolonged storage under pressure would have that effect. The designers and DOT thought it was safe and would fracture in a controlled manner. To the dismay of us all, time proved that wasn't true. If I remember correctly, nobody has been killed by a 6351-T6 tank failure, but people have lost hands or arms. Correct me if I'm wrong. Additionally, I don't think we've seen that kind of damage to personnel in any steel tank failures, which is why I'm saying we don't hear about them the way we do the failures in aluminum tanks.

Aluminum tanks, I'm sure, were designed to fail in a way that exposed fill station operators to the least risk, but over time we've found that the material reacts differently from what was originally thought.

Steel 72's are thinner walled (about 1/8" I think) and they do fail more often but, as far as I've been told, they don't expose the fill station operator to as much of a threat.

None of the fill stations down here, that I'm aware of, will fill steel 72's above 3000 psi, which is still over their rating.

 

[WD8CDH]

What alloy was that tank?

 

[Cave Bum]

What alloy was that tank?

I'm not sure but I think I remember being told it was a Catalina which of course means it would have been 6061-T6 but don't quote me on that, it's been 4 years since this was related to me and the accident was 2 or 3 years before that.

 

[WD8CDH]

The 3AA low pressure tanks have test pressures that are 5/3rds the service pressure. In contrast, the test pressure for most special permit steel tanks is 3/2 the service pressure. So assuming each type of tank had the same 3000 psi service pressure, the 3AA tank would have a 5000 psi test pressure and the special permit tank would have a 4500 psi test pressure.

So one reason is that the safety margin is less for special permit tanks.

Another reason is that they do not have the same lenght of time in service.

A third reason is the steel alloys are different and, unlike 3AA steel tanks, heat treatment of the alloy is usually involved.

So different materials, different engineering/safety standards and less service history all add to to be the reasons why no one significantly overfills HP steel tanks. Now, I do get 3700 psi ot so fill with my 3442 psi tanks when I take them to N FL, so there is a bit of an overfill, but nothing major - but then again an HP tank is already pretty efficient from a weight/volume/bouyancy perspective.

HP tanks actually have about the same ratio of service pressure to test pressure as 3AA tanks if you consider the 10% overfill allowed on "good condition" 3AA tanks. (e.g. a 2250 psi 3AA tank: 2250 x 5/3 = 3750psi, at +10%: 2475 x 3/2 = 3712.5 psi, only 1% less) so roughly the same safety factor.

A 3000psi 3AA tank at +10% would be better compared to a 3300psi HP tank. A 3300psi special permit HP tank would be hydro'd at 4950psi, only 50 psi less than a 3000psi 3AA tank. Or a 3130psi 3AA tank at +10% would be compared to a 3442psi HP tank.

3AA tanks are heat treated too.

 

[rjack321]

I'm not sure but I think I remember being told it was a Catalina which of course means it would have been 6061-T6 but don't quote me on that, it's been 4 years since this was related to me and the accident was 2 or 3 years before that.

I am sorry your friend lost an employee, tank ruptures are certainly violent events.

But AL tanks do not shatter into a millions peices like you claim. They rupture by splitting from the crown (not always but typically). Although the design specifications require 1 piece, there are typically 2 or 3 pieces at most along with the valve - no hand grenade fragments.

From:
49 CFR Sec. 178.46 Specification 3AL seamless aluminum cylinders.
5 (c) 5(ii) Three samples must be pressurized to destruction and failure
may not occur at less than 2.5 times the marked cylinder service
pressure. Each cylinder must remain in one piece. Failure must initiate
in the cylinder sidewall in a longitudinal direction. Rate of
pressurization may not exceed 200 psig per second.

 

[halocline]

Incorrect. The shop where I worked, prior to when I started, had an AL40 Deco tank explode while it was hooked up to the O2 whip but not yet being filled. Sadly, the fill station attendant was killed. It was dropped from waist level and landed on the valve. The explosion, in addition to killing the attendant, completely destroyed the front of the shop. The valve was found about a year later out in the woods about 1000 feet from the shop. The pictures I've seen indicate the tank fragmented. There was a hole in it with missing material.

No offense, but there's no way you can use this story as an example that AL tanks explode into shrapnel when over filled. You're talking about a tank filled with 100% O2 (highly flammable) that was destroyed by sudden trauma. Couldn't it be that the valve breaking simply generated a spark and ignited the O2? In any case, it had NOTHING to do with overfilling.

I'm still waiting for some documented proof that LP72s have repeatedly ruptured due to dive shop overfills.

 

[Gombessa]

Rate of pressurization may not exceed 200 psig per second.

I think therein lies the difference. The regulations govern failure in fill conditions, when workers are most at risk. Failure due to instantaneous, explosive reaction sounds like a different, and more catastrophic, beast. I wouldn't use one situation to infer behavior in the other.

 

[Mark Michaud SELAUSAR]

Not true, The Faber high pressure tanks (and all other HP tanks) are made with thicker walls and/or stronger material than their lower pressure versions. They are absolutely different tanks not just re-labeled tanks.

I think you misundertood me. I did not ay the HP and LP are the same, but it may have sounded like that. Years ago, I was reading an article (if I still had it I would post the link) that stated the "same" tanks that the US rates at 2400 are rated at over 4000 in Europe. This was when Genesis had the only hp bottles that were widely used. The newer HP bottles are a whole other animal.

My point was that there seems to be no real evidence that filling lp (Faber) bottles to a higher pressure is reckless or dangerous. It is commonly done and has been since 1999 (when I started cave diving) and maybe before. I also agree that the hp bottle is gonna replace the lp's over time because there is no need to find someone to fill the lp's higher than rated.