Analysis of List of Rebreather Accidents

Which one of the three (I think) variants of the Inspo was this on? The pre-backlit or backlit Classic or the Vision?

I've asked before that you are clearer about which units you are referring to. As a Vision user, I want to know what affects my unit, not what affects the Rebreather diving equivalent of a 70s Ford Cortina. That's not to say that I don't see the point of what you are doing - just that you should target your comments more closely. Potentially, that would help you reach just the people using pre-backlit Classics that you want to reach.

Cheers,

To be blunt, all the technical faults are irrelevant if the bailout is carried, wouldn't you agree? Everything that you have described should result in an aborted dive and an inconvenience, not a death. If electronics fail, people know what to do.

It is on our web site, and has been for some years.
Diver mole had something even closer to what you describe on his web site for years earlier.

Did not stop the deaths though.

The fix is to meet EN61508 safety standards in eCCRs, of which accident investigation is one little part, and for divers to have zero tolerance of kit that does not do that. It means sharing safety data, and openness, so all companies can benefit from lessons learned elsewhere.

I fundamentally disagree with this stance.

Let's qualify exactly what is meant by the term "technical failures"

It is utterly ridiculous to expect users to carry and deploy bailout gas to mitigate a whole RANGE of potential technical failures (created by bad design),...

some of which are not even notified to the users. You can't mitigate that which you are unaware of

Thanks. I did try to be as clear as I could, but just in case: my comments were specific to pre-backlit Inspirations. These were manufactured prior to September 2000, but some are still in use.

After 2000 there were various other Inspiration Classic fixes, one in later 2003/early 2004: the battery housing. That is worth replacing: it is implicated in at least one fatal accident, and is top of the plausible causes list on another. There are also some software fixes. The best way for users to get these is to send the unit back to APD for service. As the cost for a non-backlit unit will be high, divers should consider the alternatives before doing that. Brent's system looks the most promising, and is what I am doing with an Inspo here, but I have not got it yet or tested it yet.

The Vision controllers are a much better design than the Classic controllers, but still do not meet SIL 1 safety levels for a system in continuous use, as it is uses an unverified controller running unverified code compiled using an unverified compiler. There is also a need for a much wider FMECA: this has not been done because it would have caught a few things, such as those below.

The main safety weakness with Vision, is it should switch itself on when PPO2 falls, and always when the PPO2 is less than 0.19. We have published this algorithm. Methods using depth or wet contacts are not foolproof.

Second safety hazard with the Vision, is ensuring it always goes straight into dive mode when reset underwater.

These are safety hazards because in many accidents the diver has passed out in circumstances that indicate hypoxia, even though there is sufficient gas in the tanks. The above two items would address this risk directly.

Given the capability of electronics to hang, would recommend strongly that eCCR users fit a second HUD, e.g. the Uri HUD, not related at all to the primary controller. This would be after getting all upgrades to the controller in the case of Inspos.

Alex

And that is why I am suggesting to Alex TO PUT THESE ISSUES UP ON HERE, SO WE CAN ALL SEE THEM!!!!!

I work as a designer in the construction industry (which kills more people every year than any other industry). If I was made aware of a risk that had the very real potential to kill someone and did not take steps to design that risk out I'd end up in front of a judge. If I knowingly allowed a flawed design to be incorporated into the finished scheme I'd at the very least be sued and worst be prosecuted.

Bailout is not there to account for design flaws. A known flawed system shouldn't be in the hands of the end user. Any designer that includes a critical system with a known severe failure mode with a reasonable probability of happening needs his bumps reading (and checking his PI insurance is up to date).

Your main thrust is that we should expect almost everything to work and that anything which doesn't should be rectified by the manufacturer - and that details of when it doesn't should be published for all to see.

Well, I'd like you to be right. However my thrust is that when you are in the water, the fact that the PO2 monitoring shouldn't stop because you tried to reboot a handset is neither here nor there. You need to know that this could happen, and what to do if it does. If you have bailout and can use it then you have a last line of defence which in all probability will work.

I take the pragmatic approach. If it can go wrong, I need to know about it and what to do to rectify things. That, IMHO, is a much more practical and successful approach than expecting the unreasonable, such as a manufacturer retro-fitting ten year old equipment which usually works except for in very specific circumstances.

Er...excuse me? We carry bailout so that we can return from a dive in the event that our Rebreather fails for whatever reason. Whatever reason. Design has absolutely zip to do with this - the Rebreather is working or it isn't - the bailout required for a total failure is just the same, whether the loop fell out of a PPort, a microswitch shorted via a cracked handset or a passing shark bit the scrubber in two. Quite so. And that is why I am suggesting to Alex TO PUT THESE ISSUES UP ON HERE, SO WE CAN ALL SEE THEM!!!!!

What follows is not addressed to you Joe, but your point that because some people survive on a death trap it must be OK is one that hit a raw nerve

People get upset here and complain about me when I point things out with specific units,....

People who complain about me highlighting safety issues shoud go and see the little babies born to the divers who simply passed out on his Rebreather when his wife was pregnant, go see the distraught teenagers they left behind, go read and see the consequences of what this is doing.

I agree on your point that divers who do not really need to use eCCRs should not just now, as the risk is too high. I also agree with the need to look at all issues: servicing, training, design, approval.

However, in taking old YBODs to illustrate that the cause is not equipment, you picked about the worst example imaginable: the YBOD has numerous very serious safety design faults. What follows is not addressed to you Joe, but your point that because some people survive on a death trap it must be OK is one that hit a raw nerve. Your timing on this is not good: I received a new fatal accident report this morning on a pre-backlit YBOD, and it reads just like too many others.

This really frustrates me because this attitude that "if a majority survive it must be OK", is killing people. Nothing personal: many divers have the same attidude, and they is why they die. They are smug and smart until they are dead. When they are dead, I tell you now that the manufacture can write to the coroner saying what an idiot the diver was, even if he was one of the best divers around. In fact the manufacturer may not even be bothered to look over his long list of qualifications before doing that. I have seen those letters.

I will try my best to make this absolutely clear:
Inspos before Sept 2000 were designed with breathtaking incompetence: a single non-verified processor controlling the system with no watchdog timer, no working brown out circuit, random code in unused memory, badly designed battery contacts that are prone to bounce, handsets prone to crack and let in water - I could go on. The fatal accident rate on those units is hard to determine because the manufacture hides information on the numbers sold or in use. APV and APD are the only Rebreather manufacturer to do that completely: I ask Peter Readey and he tells me how many Prisms there are, ISC state on every unit they sell how many they have made. Representatives from APD band around impossible numbers, with no obvious sequential serial numbers and their followers challenge any attempt to get that number without coming up with any believable alternative. This system of bad design combined with methods intended, IMHO, to destroy statistical data is killing people.

Based on observations of the numbers on dive boats in the UK in 2000, we believe a diver using non-backlit Inspo in regular use for its 7 year service life has a 30% risk of dying on that equipment. The figure on the latest equipment seems to be arond 10 times lower. Some divers took their time before finding out the truth: there is the odd diver still using an Inspo without backlit handsets, and divers are still dying on them. For those still out there, for goodness sake, upgrade to Vision or Brent's system. Or put a decent value on your life and buy a unit from CCR Ltd, or a KISS with a FFM. At least, send it back for annual service so APD can put some safety upgrades in.

People get upset here and complain about me when I point things out with specific units, and in doing so frankly share this blood guilt. These units are death traps and should have been recalled when the manufacturer received written notice of those faults in 2000. Let is stop this crazyness. It is obscene.

People who complain about me highlighting safety issues shoud go and see the little babies born to the divers who simply passed out on his Rebreather when his wife was pregnant, go see the distraught teenagers they left behind, go read and see the consequences of what this is doing. Divers die in their prime on rebreathers. That means when they have maximum responsibility: when the most number depend on them. Most RB divers die by passing out. Very few O.C. divers do that.

I do not care which company it is that makes a death trap: if I find safety errors in equipment, I will jolly well say so. Those that have done a better job in avoiding these errors, I have praised. However, most eCCRs are designed by people completely ignorant of good safety design practice, and I will do my best to change that. As the fatal accident list is verified and published in increasing detail, it will make clear the social cost of us accepting a lower standard.

The statistics from accident reports, and particular what happens, is all we have to go on to find things that design reviews do not find. Let us mine that data and use it, instead of searching for any excuse for why it might not apply.

Yes, rebreather divers are VERY much more experienced than average O.C. divers. As a result they do engage in solo diving a lot more. Very many Rebreather accidents are in solo dives, but many are not. Solo diving on a rebreather right now does not seem a clever thing to do. Find a buddy who can manage the same dive, and understand the difference between a buddy to help each of you survice and a tourist who happens to be in the same ocean. But let us also eliminate the obvious design faults in equipment that prevent them getting to SIL 1 even (the safety level of a drill guard)!

Alex