Old steels denied fills due to store "policy"

[kelemvor]

Good, well done

Here's a couple of Doc for information.

Diving cylinders – the danger of internal corrosion Bear in mind this is UK centric where Vis is only required every 2.5 years (lets not open that can of worms) But it does highlight how little moisture is need to create a substantial loss of wall thickness in a short period of time. Corrosion can happen 100x faster in a compress gas environment

Cylinder Corrosion and Prevention Techniques Is a slightly longer info document - still worth a read

One from Undercurrent Incident, months post inspection

The Truth About Cracks and Scuba Tanks (I Think...) Be very careful not to jump to conclusions on this one. Read a couple of times, the devil is in the detail on this one

Sure, but water in a tank isn't going to corrode a 51 year old tank any faster than it will corrode a 6 year old tank.

 

[2airishuman]

Science keeps getting quoted:

The US requires Annual Vis, 5yr Hydro
The UK requires 2.5yr Vis, 5yr Hydro
Australia & New Zealand require Annual Vis, Annual Hydro

All based on the same evidence (science) all with different interpretations - Who's right?

The first seamless high pressure steel cylinders were manufactured in Germany in 1890. The original application was the shipment of liquid CO2. The cylinders were imported into the USA until 1902, when the first cylinders were produced here.

There were a number of accidents, and in 1911, the goverment mandated that new cylinders be subject to a hydro test before use.

In 1914, a periodic requalification was required, including a hydro test.

The rules have been tinkered with since then very slightly. For example, the requirement for a hydro test is now once every ten years for most steel high pressure cylinders. There are a number of exceptions that require more frequent tests, and SCUBA service is one of them. Another notable example is carbon dioxide service, which can cause internal corrosion. Tests are required every five years.

As for the science, well, it's widely acknowledged that there are better ways to test cylinders now than hydro+visual, and that the intervals are probably overly conservative. But actual field experience over the 105 years with millions of cylinders and fills that have to be well into the billions shows that the interval and the testing approach, though perhaps not ideal, are enough to provide a fantastic safety record.

As for the UK, Austraila, and NZ, who knows. They have their own history. Their test protocols are safe but excessive, just as the USA test protocol is safe but excessive. All I can say for sure is that the new UN cylinders are probably going to be accepted there, too, and they have a 10-year hydro interval.

 

[2airishuman]

I think you'll find (but I've no supporting evidence of their policies) that these entities would retire their cylinders after a relatively short time (15yrs?) mainly because of the frequent use they get and wear and tear.

The reality in the USA has been that SCBA equipment is subject to its own regulatory structure. You can't mix brands of SCBAs and cylinders and remain in regulatory compliance. The SCBA is certified by NIOSH and MSA as a unit and any replacement parts or accessories have to come from the same manufacturer. Historically there has been a new generation of SCBAs about every 10-15 years that offers significant safety improvements for firefighting. These have been accompanied by changes in cylinder technology.

In general the early SCBAs (1960s) used steel 2216 PSI cylinders, then there was a transition to aluminum 2216 PSI cylinders, then to composite 3000 PSI cylinders, then to composite 4500 PSI cylinders. There are a few 6000 PSI composite cylinders out there but they haven't really caught on.

At the same time the SCBAs themselves had significant improvements. The early ones used a vacuum-demand approach with a second stage on the belt. The face mask was a atmospheric pressure, and drawing a breath opened a demand valve connected to it with a corrugated hose much like a dual-hose scuba regulator. Later systems used a lightweight second stage on the mask itself and used a pressure-demand valve that kept the mask pressurized slightly at all times, reducing the risk of inhaling toxic gases in the event of a leaking seal between the mask and the skin. Subsequent systems added electronics, integrated radios, heads-up displays, and integrated means of signalling low air to teammates and tracking people as they moved in and out of a burning structure to reduce the incidence of firefighters being left behind in a chaotic fire ground.

As individual departments adopted each of these new generations of SCBAs, they obtained new cylinders with them, and in many cases new compressors to handle the higher pressure.

The cylinders and SCBAs are generally surplussed out and end up on ebay and similar places, or are shipped to departments (often outside the USA) that cannot afford the extremely high price of new SCBAs.

In most cases the existing cylinders are perfectly good. I have some steel SCBA cylinders (from the 1960s) that I use with a newer SCBA (1990s) when working with toxic marine finishes. Since I am looking after my own safety and not that of employees, I am not obligated to follow the NIOSH rules about mixing and matching.

The newer composite cylinders have a life limited to 15 years because that's what the original manufacturer wanted. The available evidence suggests that they can last far longer than that (there was a Navy study), and there are now competing composite cylinders that have a longer life. The expired 15-year old composite cylinders are -- you guessed it -- ending up at departments outside the USA, and in the hands of paintball enthusiasts.

In any case, wear and tear on the cylinders themselves has nothing to do with why departments retire them.

 

[Diving Dubai]

Sure, but water in a tank isn't going to corrode a 51 year old tank any faster than it will corrode a 6 year old tank.

Did I say that?

 

[2airishuman]

12% is probably not even statistically significant to be honest. Like you mentioned, it's all design parameters.

17,000 PSI is nuts. Why aren't we routinely filling to 10,000 PSI if that's common? A safety factor of 750% is pointless. The obvious answer is that 750% of design is an outlier and they routinely fail at much lower pressures when tested to their catastrophic failure point. I'd be interested in seeing the test parameters involved. New cylinder? Old cylinder? How many load cycles to what pressure? Is there a failure mode that's at a lower pressure that they chose to ignore and let it cato? Bunch of variables could put those numbers all over the place so if you've got the source I'd love to see how they did it. I'd imagine there's a reason the tank(s) in question are rated for 2250 PSI, and I would posit that it's not because they could reasonably guarantee a MBS of 17,000 with any routine level of confidence.

I wasn't trying to disprove your point, I agree, I just think markm was using a bit too much hyperbole. We do similar testing with fall protection and rigging equipment. Intended use plays a role in rating all of that stuff, and different industries use different types of ratings. MBS vs WLL, etc. Physics can be a bitch, but it's cool as frig.

I came across some navy test reports from the 1970s where they pressurized LP72s to the point of rupture. Typically it would occur at a fill pressure of around 5000-6000 PSI, in most cases slightly higher than the design strength. The test pressure of 3AA cylinders is supposed to be 70% of the yield strength of the steel, as I recall, so with a 3750 test pressure the steel should yield at 5357 PSI.

There's no way any of them could withstand 17,000 PSI.

 

[2airishuman]

Wait a minute, you are telling me that there are 5 times as many 6351 cylinders made then Ford Mustangs. Last year there was a big publicity about the 10-millionth mustang built. I see a lot of Mustangs. Owned a few as well. But my sightings of 6351 cylinders are rather rare. Where are all these millions of 6351s hiding at? They sure don't come through any dive shop or boat around me. A couple show up, but they are few and far between. Not even that many on Craig's list, and that is loaded with Mustangs.

Something about your numbers isn't adding up.

Most of them weren't dive cylinders.

They were in widespread use for SCBAs, medical oxygen, and CO2 cylinders for the beverage industry.

 

[Wookie]

I came across some navy test reports from the 1970s where they pressurized LP72s to the point of rupture. Typically it would occur at a fill pressure of around 5000-6000 PSI, in most cases slightly higher than the design strength. The test pressure of 3AA cylinders is supposed to be 70% of the yield strength of the steel, as I recall, so with a 3750 test pressure the steel should yield at 5357 PSI.

There's no way any of them could withstand 17,000 PSI.

That came from me, and it was testing to failure Hoop Wrapped cylinders....

 

[2airishuman]

To reiterate the point, bullet point number one is scuba specific, and is a subset of each of the other two points. There is a tendency for people to read that second point’s lower rate for ALL Luxfer 6351 cylinders, and want to apply it directly back to scuba.
Interesting that in the context of these three statements, it appears that non-scuba, non-Luxfer cylinders have a much higher rate of SLC than Luxfer non-scuba. (I am assuming all manufacturers produced scuba in similar proportion to total cylinder output.)

The various methods used to detect SLC in the field are notoriously unreliable and produce both false positives and false negatives.

There may be perceived to be more SLC in scuba cylinders because they are typically inspected for SLC every year rather than every five years as is the case for other industries. Therefore, there are five times as many opportunities for a false positive.

Another few words on what I will agree is an extremely low incidence of catastrophic failure of 6351.
First, that low rate is poor solace for the fill station operator that experiences it first hand.
Second, that failure rate is highly skewed - it does not exist in a vacuum where there were zero proactive measures to remove a cylinder prior to failure. They occur in spite of those measures. Even if they can be traced back to a missed opportunity due to substandard testing, that testing is part of the world in which we live, and must remain included in any risk assessment.

I agree, and would add that the incidence of actual cases of SLC was predicted to increase over time, and evidence suggests that it actually is. These cylinders are less safe today than they were when the problems were first identified.

I will still fill a 6351 cylinder that has an annual VIP and annual eddy current test, but would not lose much sleep over it if tomorrow the DOT banned them outright.

Once in a while I get an old 6351 cylinder in a lot of used dive gear that I buy for other reasons.

I can either spend $90 on this cylinder for a hydro, eddy current test, and new valve, or sell it to a scrapyard for $20. I choose the scrapyard. There is so little residual value in the used cylinder, I don't see any reason to take a risk.

 

[Diving Dubai]

The various methods used to detect SLC in the field are notoriously unreliable and produce both false positives and false negatives

But they shouldn't be. It frustrates the heck out of me that any Eddy Current (EFD) "defects" have to be confirmed by Vis.

This is clearly because the equipment, procedure and perhaps training aren't fit for purpose.

Aluminum is the most receptive material for Eddy Current inspection, that because the defect signal differ so widely. We call it phase separation. Simply put different defect signals appear at different clock positions around a circle. A Natural defect (crack) will have a very tight signal, whereas a mechanical defect will almost have 2 lines because it's so "wide"

EFD can also spot sub surface defects and corrosion too. Why the visual plus system was accepted is beyond me. It's junk

 

[2airishuman]

Why the visual plus system was accepted is beyond me. It's junk

On that, we can agree.