Waterford Garda dead - County Wexford, Ireland

[helodriver87]

Do you have scientific proof for that?

This thread has officially jumped the shark.

 

[KenGordon]

The inquiry stated "O2 spike" as an issue, and the reason behind it won't be discovered by a dick-waving competition on an internet forum. If people want to put their knowledge to use, draw up a list of possible causes for an "O2 spike" and try to assign rough probabilities to each one. Better, find some actual "O2 spike" incidents so that those probabilities are based on real data. Even better still, establish a reporting and analysis process (such as FMECA) where incidents of that type are reported and their causes and consequences analysed. That way, you will build up a database of information that will allow you to say which failure modes are likely to lead to an "O2 spike" and whether or not it is a serious issue that requires further attention, or if it is addressed by current training methods.

That's thinking like an engineer: collect data; analyse; act - and people on this forum are probably well-placed to do that data collection and analysis. That's what they do when an aeroplane crashes. That's why air travel used to be risky and now is very safe.

Many of the people building rebreathers are engineers and they go through some process to get CE certification.

some of those engineers get called in to look at deaths. That has been happening for twenty years for recreational CCR.

As a result the courses cover the failure modes. They are basically quite simple devices. The issues come dow to too much or too little O2.

Since the basic component of a rebreather is the sensor that measures the ppo2 and those age the device is inherently unreliable. That means that the users have to expect failures and behave appropriately.

On top of that you have the normal failed value, run out of gas etc etc. None of those things are a surprise.

What IS a surprise is that people die following simple to remedy failures. Sometimes they have ignored warnings for tens of minutes before becoming unconscious.

We know this happens because there are formal processes in place to examine deaths. The coroner looks for an expert and there are not too many. So those experts get to learn what is really happening. Since at least one runs the biggest manufacturers of rebreathers there is we can expect that knowledge to influence design of both the hardware and the courses.

Yet still people die.

Two examples that left me thinking WTF were two members of the same dive club dying a year apart in near identical circumstances due to having modified their counter lung dump valves in the same way. This led to exhausting the dil, having to make up volume with o2 and toxing. So diver A and B knowingly made the same strange mod, A died and the B carried on with the same mod until he died.

You can’t really explain why that made sense.

Everybody makes mistakes. The training is there to help.

Hands up everyone who has never made a mistake that might have killed them except someone else caught it.

I am not blaming the bloke that died, everyone makes mistakes. Maybe there was a fault with the kit too, who knows. But everyone should be planning on mistakes and failures.

I will go and find my checklist now...

 

[Brad_Horn]

People have been dying on CCR for more than twenty years. The various ways are well explored and training and processes take that into account. If someone dies because they have a freeflow on a twinset but fail to shutdown do you blame the regulator?

Are they really that well explored or covered in training? What do you figure is the highest percentage driver of rebreather fatalities based on causation?
Open Safety Equipment Ltd

I dont believe Deep Life sells regulators.

No but Open Safety do:
Apollo A320 1st stages being the highest quality that OSEL could source which pass the BAM.de Oxygen Service testing
ALVBOV covers the second stage ATM albeit being mostly Apeks internals and fitted on a RB loop
DeepLife did the testing see https://www.opensafetyglobal.com/Safety_files/DV_DL_ALVBOV_Breathing_Params_A3_100318.pdf

No idea why Shearwater left out other signaling. Control systems that came before it like the Hammerhead would vibrate the the HUD if critical messages were on the controller. Not blaming Shearwater, their products are some of the best out there but I wonder how many deaths would have been avoided had the controllers signaled outside of display messages.

How much does the patent licence cost! US8028696B2 - Display integrated vibrating alarm - Google Patents

Wouldn't experienced diver would switch straight to Bailout bottle in fear of oxygen toxicity at depth instead of wasting seconds trying to deal with the oxygen spike?

If he had a BOV, fitted sure. At least if he could switch it rapidly and easily enough to OC. Any accident analysis would have documented this and it would have been an output.

You should verify your sources before you post. A quick search re that 'list' would have shown just how discredited that list, the author and that company is, across all diving forums.

And yet oddly enough all these years later the DeepLife list remains the gold standard for anyone to use as a reference when looking at rebreather fatalities!
Every time someone has mentioned to me that their accident list is inaccurate DeepLife add fatalities to it. And as per Ken Gordon’s post below it looks like they missed another two.

When this depth of design verification supports your product - which isn’t available from anyone else - it does upset the apple cart doesn’t it Deep Life Design Team: Selected Design Validation Reports for DL & Open Safety Equipment Ltd's Rebreathers
Noting that not a single one of those tests has ever been questioned, by anyone!

I’m in the market for another BOV, but I want to buy one CE certified with a known published OC and CC WOB. Whom can I buy from?

You will find that using the Engineers who designed both the unit and the electronics in a Coroners inquest involving one of their units would lead to a legal conflict of interest and their findings would be open to criticism of bias.

I However, rebreather diving is such a niche activity, that when an accident happens, authorities have no other choice but to ask an instructor or even the manufacturer. There is no real independent examination similar to aviation.

How much does that independent aviation examination cost?
How much does an accident analysis on a rebreather cost? There has been a few done.
I’d argue that the latter is considerably cheaper… and does it really need to be independent? Transparent and openly published would certainly be lightyears ahead of the past 23 years! From the 600 odd rebreather fatalities you’d probably get some pretty good trend analysis and some real clear trends of things that need to be engineered out.

The difference would seem to be the SIL requirement based on the risk assessment What Is IEC 61508? How To Determine Safety Integrity Level (SIL) Basics? | Perforce Software
Sure some airplane crashes only kill the pilot; but even a cheap plane probably averages 10x that of an expensive rebreather and for some, the J35 comes to mind the $$$ are certainly a factor. But also typically rebreather fatalities usually only kill the user, or at worst 2 divers in maybe 3 or 4 instances total; whereas the average for planes is likely quite a few higher.

The rebreather initially passed CE testing before sale (ok this opens another can of worms). So let's assume the rebreather is in acceptable working order before it is bought. Then something "breaks" and the diver cannot react in time and dies.

Did it? Certainly plenty of rebreathers on the market that were sold by some manufacturers well before CE was ever achieved.

Why does something on a rebreather need to ‘break’ for a diver to get outside an individual rebreathers known safe underwater life support envelope?

Is it a manufacturer issue? Since the rebreather passed CE, the manufacturers are "safe"

Are they? I don’t believe this has ever actually been stress tested.

But at least some testing has been published for this rebreather though to enable some pre-dive planning:
https://www.bluelabeldiving.com/wp-content/uploads/2014/06/JJ-CCR-test_report.pdf
https://jj-ccr.com/wp-content/uploads/2018/06/QQ-14-01561-JJ-CCR-June-2014-V1.pdf
I’m sure the diver had additional scrubber duration testing available to him in order to plan a 70m dive. As using the results from the 100m profile test would seem a little inefficient on bottom time with only 15min available.

What does seem clear is that either the high PO2 warnings were missed or the decision was made to stay on the loop despite them. At the end of the day, that's the proximal cause regardless of the precipitating event, to borrow a phrase.

Was it actually Hyperoxia?
How have you ruled out it not being Hypoxia from two water blocked cells or cells that have failed high…. Both of which might work perfectly fine on the surface!

Again as there is a plethora of testing from the different manufacturers covering empirical testing for how their individual units handle specific faults this always makes interesting reading. And I’m glad as someone has mentioned in this thread that this is covered in modern training courses.
One example https://www.opensafetyglobal.com/Safety_files/DV_O2_sensor_fusion_120625R.pdf

 

[Brad_Horn]

Better find some actual "O2 spike" incidents so that those probabilities are based on real data. Even better still, establish a reporting and analysis process (such as FMECA) where incidents of that type are reported and their causes and consequences analysed. That way, you will build up a database of information that will allow you to say which failure modes are likely to lead to an "O2 spike" and whether or not it is a serious issue that requires further attention, or if it is addressed by current training methods.

Funny this keeps getting raised and everyone has to keep going to the same source: Deep Life Design Team: databases and analysis of rebreather accident data

On FMECA for rebreathers I’m aware of two, happy to get educated about more…
Deep Life Design Team: Design Submission for Open Revolution
Rebreather Products from ISC | Innserspace Systems Corp.

See also
Closed Circuit Rebreather Safety – Fault Tree Analysis Research Papers

Interestingly enough I understand APD presented DeepLife’s FMECA with the top/bottom of it cut off at the Barrett case so I guess that’s one way of making use of it.

some of those engineers get called in to look at deaths. That has been happening for twenty years for recreational CCR.

And yet you’d struggle to find much on the accident analysis or lessons learnt…. Poseidon published some good ones for the first few on their units https://mkvi.poseidon.com/downloads/Accident_Report_IR-2010-05-31-1.pdf
Not so much nowadays Scientific diving community mourn loss of future talent

But this NEDU presentation probably still sets the standard for what is required of accident analysis of rebreathers https://www.diversalertnetwork.org/files/Tech_Proceedings_Feb2010.pdf Downloading the dive computers just a data logging step. You then still need to actually test the unit to identify why it happened and what might cause the rebreather to give the same outcome as the fatality.

As a result the courses cover the failure modes.

Do they really? For the most comprehensive list of cell failure modes that I’m aware of see https://www.opensafetyglobal.com/Safety_files//Test_Data_for_Sensor_Fusion_Algorithms_RV_130325.pdf

The coroner looks for an expert and there are not too many. So those experts get to learn what is really happening. Since at least one runs the biggest manufacturers of rebreathers there is we can expect that knowledge to influence design of both the hardware and the courses.

Yet despite that expertise there really hasn’t been that much change by Drager in 30 odd years. They still don’t meet CE. WOB’s still high. Scrubber duration really hasn’t improved much in that time. They can use EACs but they are aftermarket fit and the rebreathers not designed around them to optimise the duration. It is as if certain manufacturers don’t want divers using their units really knowing too much about their actual performance.

Two examples that left me thinking WTF were two members of the same dive club dying a year apart in near identical circumstances due to having modified their counter lung dump valves in the same way. This led to exhausting the dil, having to make up volume with o2 and toxing. So diver A and B knowingly made the same strange mod, A died and the B carried on with the same mod until he died.

Yet this is probably the first time this issue has ever been publicly documented. Certainly no accident analysis or lessons learnt published by anyone for others to learn from. If B knew A had died from the modification, would he have carried on with diving it? Have others done the same thing and died from it? Who knows….

 

[helodriver87]

Was it actually Hyperoxia?

Does it matter to the diver when it happens? If an unsafe PO2 is indicated by the majority or all of the cells, the correct action is to get off the loop. We can play devil's advocate all day while hawking our junk (well, for most of us maybe not that last part), but at the end of the day, we know he was presented with unsafe PO2 indications and didn't take corrective action. I'd be interested to hear if the unit was in working order or how it may have failed, but it doesn't change the diver's failure to respond to unsafe indications.

 

[Turtle78]

Are they? I don’t believe this has ever actually been stress tested.

I'm not aware of any manufactures having been held liable for a diver dying using their rebreather, regardless of CE or not. But I'm happy to be corrected, if anyone has more information.

See also: https://www.davidconcannon.com/product-liability

or: Dive Rite Cleared of Responsibility in $25M Trial over Underwater Filmmaker's Death​

At least in Europe, where CE is mandatory, CE should provide some protection for the manufacturer in a liability case.

 

[KenGordon]

Yet this is probably the first time this issue has ever been publicly documented. Certainly no accident analysis or lessons learnt published by anyone for others to learn from. If B knew A had died from the modification, would he have carried on with diving it? Have others done the same thing and died from it? Who knows….

I learned this from a presentation at a conference about accident analysis. Maybe going to conferences or being on good enough relations with anyone else in the industry would mean you would hear it too.

On the other hand only existing to blacken the reputations of every other manufacturer means nobody takes any notice of your list.

 

[John the Pom]

Yet despite that expertise there really hasn’t been that much change by Drager in 30 odd years. **SNIP**

Am I right in thinking that Dräger sells a lot of rebreathers to the military? Defence people are notoriously conservative. There's a famous quote from a senior navy officer after WW2 along the lines of, "Now we can get back to being a proper navy, without that damned radar!" Dräger would have the benefit of incumbency, plus a customer who is very suspicious of anything new - i.e. little incentive to innovate.