Scuba Tank Explosion - Myth or Reality?

[Drewski]

Curious this thread popped up just when I had a problem.

I generally jam fill my 2400 psi LP Worthington 121s to 3200 PSI. Last summer, I was rushing and left one of them partially uncovered in my car parked in a Hatteras parking lot. Direct sun exposure to the metal tank body in a closed car with the windows cracked about 1".

That evening when I returned, I discovered that the burst disc had failed. I have them disced for 5000psi. I marked the tank so I knew which one it was.

Last week, I took 10 of my tanks in for hydro, the burst disc failed tank as well, all Worthington LP steels with identical fill histories. Guess which one failed? Not only failed, but REALLY failed. 29% overall expansion rating. Worst the tester had ever seen. Clean as a whistle on the inside. All the others were 2-3% for comparison.

Clearly, the tank in question got hot enough to fail the disc. The event lasted long enough to stress the steel.

It can happen...

 

[String]

I have seen burst discs go. Tanks left in the shade in the open air, temps around 45 C (113F). It was really noisy and drilled a nice deep hole in the sand! Suspect the Burst Discs may have been fatigued due to expansion and contraction from heating, cooling, filling and emptying.

A burst disk is a failure point and they are fairly unreliable. Poor quality metals, poor calibration and various other things. They're not legal in many country for good reason.
You're just introducing a weak, potentially unreliable failure point that has nothing to do with the tank integrity.


A standard steel 232 bar (3410psi) tank full at 25c (77f) is going to take a temperature of 174c (345f) to get it up to its test pressure of 348 bar (5115psi) and therefore the limit of what its tested to make sure it can withstand.

Unless your car is on fire you wont get 174c (345f) even with a black painted tank lying in direct sunshine!

I know US tanks are generally rated to lower pressure than elsewhere but if the ratio of working:test pressure is the same the maths remain the same.

 

[cocoajoe]

Volume is constant, so P₂=P₁(T₂/T₁). Let's use 540〫Rankine as T₁, 541〫R as T₂, 3000 psi as P₁, and let's solve for P₂. P₂=3000 (541/540) = 3005.5555 The physics book is correct! (Although it's sad that it's using units of Fahrenheit and psi.) I think the rest of your post is also correct, and I agree that neither tanks in a trunk nor their burst disks will explode from the midday heat if they aren't drastically over-filled or somehow flawed.

Sad eh? Cheer up, you keep your bars and we'll keep our psi. LMAO

 

[ubp_sanity]

On the weekend I got certified, one of the instructors leading us had a burst disc go. It was Canada Day, so middle of summer.

I don't know all the details, but I do know it failed while he was driving. So, there were 2 things that needed replacing, the burst disc, and his underwear.

 

[Emparo821]

i dunno if any of you watch mythbusters, but they tried to make some of these tanks explode once. under NO normal conditions could they get it to explode. so the risk of explosion is all but none-existent.

 

[616fun]

i dunno if any of you watch mythbusters, but they tried to make some of these tanks explode once. under NO normal conditions could they get it to explode. so the risk of explosion is all but none-existent.

The reality is that accidents are rare, but when they happen it's catastrophic. To say all but non existent puts us in a mindset of being less cautious.

**edit - Didn't realize this was specific to overheated cylinders in cars (am readin via my cell phone). I'd agree that a tank actually releasing from the heat of a car is rare. **

 

[Ghanguss]

One thing to keep in mind for tanks / burst disks is fatigue, not just the maximum pressure exerted on a cylinder.

Consider a cylinder with a defect that is undetectable by regular inspection. Metal impurity, incorrrect alloy, incorrect hardness ect. in a part or all of the cylinder. Over time the filling / emptying of the cylinder can fatigue the cylinder / disk and lead to failure.
I prefer to keep my cylinders where the temperature variance is minimal, so the chances of fatigue is also minimised.

 

[Drewski]

i dunno if any of you watch mythbusters, but they tried to make some of these tanks explode once. under NO normal conditions could they get it to explode. so the risk of explosion is all but none-existent.

REALLY? Because the guys on Myth Busters said so? C'mon, now.

 

[charlesml3]

REALLY? Because the guys on Myth Busters said so? C'mon, now.

Well no, they didn't just "say so." They tried. Repeatedly. The video evidence is easily obtainable.

They shot the tanks with a rifle from all kinds of angles. Sure, they could punch a hole through the tank and it would jet around inside the shipping container but there was nothing that resembled an explosion.

-Charles

 

[sled600]

Sound Like a problem for Mythbusters!!!!