Tank explosion kills one - Cozumel

[mstevens]

Question (and I don't know): Was this a Scuba Club tank (part of their regular rentals) or was it a private (inidivudal's) tank?

In Cozumel it's practically guaranteed that it was the facility's tank. Practically nobody ever takes privately-owned tanks to the island and I don't know anyone who lives on the island who owns a tank except dive operator owners, including my friends who earn their living as divemasters or instructors. There aren't really any dive shops in the north-of-border sense where you go in and buy dive toys you don't need, get your tank filled, and sign up for the underwater pumpkin carving contest.

 

[cb5150]

Yes, a tank exploded. I was about 25 yards away when it exploded. I saw the condition of the tank room, but did not go inside. I did not see the actual tank that blew, but was later told that it was an older tank although I cannot confirm this info. I saw Juan on the ground, but could do nothing to help his condition other than to say a prayer that he was not suffering. There were employees on the scene asking us to stay back because they didn't know if the scene inside the tank room was safe.

The authorities shut down the compressor room for an investigation and tanks had to be shipped in from another source. The tank room is currently under renovation and I am watching the work from my balcony as I type this reply.

 

[Ken Kurtis]

In Cozumel it's practically guaranteed that it was the facility's tank.

That's also my assumption, but that's different than knowing that assumption is correct. I was also wondering if it could have been a staff member or local's tank that they were filling as a favor or off the books.

- Ken

 

[kombiguy]

Regarding burst discs: They designed to SAFELY vent the air out of the tank. In fact, the whole idea of a burst disc is that it gives way BEFORE the tank explodes and in fact PREVENTS the tank from exploding. So the question here would be why that apparently didn't happen.
- Ken

Yes, except that burst disks are rated for higher pressures than the cylinder itself. It will only rupture when the pressure in the cylinder has exceeded its normal working pressure. A 3000 psi cylinder will have a burst disk rated at 3500 psi or higher, so a bad tank can explode before the burst disk ruptures.

 

[nadwidny]

Yes, except that burst disks are rated for higher pressures than the cylinder itself. It will only rupture when the pressure in the cylinder has exceeded its normal working pressure. A 3000 psi cylinder will have a burst disk rated at 3500 psi or higher, so a bad tank can explode before the burst disk ruptures.

A bad tank can explode at any pressure. I've been present when one let go at an estimated 1000 psi.

 

[ibj40]

There aren't really any dive shops in the north-of-border sense where you go in and buy dive toys you don't need, get your tank filled, and sign up for the underwater pumpkin carving contest.

Back in the day, and I have to admit that I haven't been on the island in over five years, there were at least two dive equipment outlets in town, actually across the street from each other, near the original location of Prima. One was Michelle's, and I can't recall the other one, but both sold tanks.

 

[Missy Diver]

The tank did not explode while it was being filled. It had been filled and removed from the fill chamber. It exploded minutes later. Nothing that anyone could have controlled

 

[DjDiverDan]

The tank did not explode while it was being filled. It had been filled and removed from the fill chamber. It exploded minutes later. Nothing that anyone could have controlled

Except perhaps by more careful visual inspection of the tank and regular hydro testing. It's hard for me to believe that a tank which was capable of exploding didn't have visible corrosion and microfractures from metal fatigue which could have been spotted in a close visual inspection, or that it wouldn't show noticeable (unacceptable) deformation if it had been hydro-tested recently. Tanks do not simply explode at random. They explode as a result of metal failure under pressure - corrosion, microfractures due to metal fatigue, a weak spot caused by abuse, or a flaw in the manufacturing (like a void in the cylinder wall) which would have been caught (should have been caught?) by Xray testing before it left the plant.

 

[Ken Kurtis]

Yes, except that burst disks are rated for higher pressures than the cylinder itself. It will only rupture when the pressure in the cylinder has exceeded its normal working pressure. A 3000 psi cylinder will have a burst disk rated at 3500 psi or higher, so a bad tank can explode before the burst disk ruptures.

I think you misunderstand the general concept. While you're correct that a burst disc is set to blow at a SLIGHTLY higher pressure than the working pressure of the tank, the working pressure of a tank is not also the pressure at which it will rupture/fail. Tanks are hydroed at 5/3 their stamped/working pressure. An aluminum tank would be hydrostatically tested at 5000psi. So the order of march would be working pressure, burst disc pressure, hydro test pressure, tank failure pressure.

The whole idea behind a burst disc is that it blows BEFORE the tank explodes, not the other way around. But in some areas, burst discs may not be replaced regulaly and have corroded shut, and they've been deliberately disabled, or discs of significantrly higher pressure are instaklkled to allow for higher over-fills without triggering the burst disc.

No one knows the history of this particular tank. It may have been repeatedly over-filled which can induce metal fatigue. It may have been out of hydro. It may have had undetected or known-but-ignored hairline fractures. But simply put, tanks that are properly inspected and meet certain standards will not rupture at pressures less than that of a burst disc.

- Ken

 

[ZombieZombie]

Very sad. I don't remember Juan but I dived at Scuba Club just in April. It seemed a reasonably squared away operation. I am relatively new to diving and unsure of what is common practice around scuba tanks, but as a mechanical engineer with years of experience as maintenance manager where I was responsible for pressurized tanks of all sizes and materials- this seems like a subject where I should take more interest. Can anyone point me to any resources where guidelines or standards for re-inspection of scuba tanks is outlined?




---------- Post added August 31st, 2014 at 03:32 PM ----------

Except perhaps by more careful visual inspection of the tank and regular hydro testing. It's hard for me to believe that a tank which was capable of exploding didn't have visible corrosion and microfractures from metal fatigue which could have been spotted in a close visual inspection, or that it wouldn't show noticeable (unacceptable) deformation if it had been hydro-tested recently. Tanks do not simply explode at random. They explode as a result of metal failure under pressure - corrosion, microfractures due to metal fatigue, a weak spot caused by abuse, or a flaw in the manufacturing (like a void in the cylinder wall) which would have been caught (should have been caught?) by Xray testing before it left the plant.

A visual inspection will only be effective in revealing major issues like pitting corrosion. Fatigue issues will almost never be visible to the naked eye.
The point of hydro-testing is to bring any significant defects or cracks to failure in a controlled environment. (Explaining extra detail for benefit of anyone who isn't familiar with pressure testing). Over-pressurizing a scuba tank to 5000 psi g should induce failure of a tank that might fail in normal operation soon. Liquid doesn't release energy the same way as a gas so a failed cylinder will not explode- at least not as dangerously as a gas filled cylinder would.

The basic premise behind hydro testing on a defined frequency is that a tank that passes will not have any defects large enough to progress to catastrophic failure before the next hydro-test in normal use. Presumably the industry has plenty of data to support this.

Factory testing does not guarantee there are zero defects. That is simply impossible. Every gas tank ever made contains defects. Factory testing ensures that defects are below a size and number where they will grow under fatigue and cause failure during normal operations. However, tanks get treated pretty roughly after leaving the factory- being dropped, hit, or exposed to salt water can turn innocuous micro-defects into larger ones beyond the critical size for crack propagation in a 3000 to 500 psi g cycle once a day.

Periodic pressure testing (I would guess annual or biannual inspection would be the norm for the scuba industry?) is intended to cover for these in service issues by forcing failure early.