Risks in Cave Diving

[deepdiverbob]

I asked the workshop chairman (Gene Melton) to contact Ed, and offer to do it there as well. If you bug Ed too, it would help

That would be my pleasure. Thanks again.

[Gary]

I sorted out a number of errors in the calculations I made but none of them made any difference in the graphs or used calculations. One row was shifted down by one but the numbers were insignificant. I also tried some linear redistributions across the range of risk but they made no real sense (with multiples as high as 372 per 100 psi in the vicinity of 2/3s reserve)

The original guestimate of "10 times per hundred psi" on a 3000psi tank is probably closer to "5.5 times per 100 psi" average (5000^(1/5)) - ranging from as low as 2 times per to as high as 17 times per 100 psi. For cave fills 3600 psi yields 600 psi change 1/3->1/2 in thirds so 5000^(1/6)=4.1 would be more like an average of "4 times per 100 psi" average across the range with the other assumptions the same.

4-5 times risk per 100 psi sounds like a reasonable result for the given assumptions. Varying assumptions might put it in the range of multiplying risk by 3-6 times risk for each 100 psi.

Multiplying risk of death by tens, hundreds or thousands for minimal returns is probably a lot more risk than most people may have considered.

Whatever the death figures are based on I think the cumulative survival chance graph gives good reason safety planning to at least the levels others have in the past and to consider adding extra safety when practical/needed - which is no different than what has been taught and what everyone above has recommended. I don't think anyone got into cave diving thinking about making a couple dozen cave dives and then quitting - and even if they did it wouldn't be worth risk.

[aainslie]

Gary,

This is an interesting idea. But there is a much simpler way to do this.

Let's say you determine the risk of dying from running out of gas when reserving double the gas let's call that "thirds"). To make things simple, let's say that it's one in a million, or .000 001.

Now you have to decide what the risk is of dying from running out of gas if you only reserve half. Well, it's the square root of the previous probability.

Why? Because we'll assume that losing each half is an independent event itself a huge assumption unless you dive sidemount, but let's stick with it). So the risk of losing your main supply AND your reserve is the square of losing just your main. If diving 1/2's, then essentially all you need to do is lose your main and you're screwed.

Hence the 1/2 power.

This is a simple starting point. It doesn't take into account losses at points other than at exactly the furthest point in the system. But at least it offers a start to modelling this. Note that it gives a probability of .001, NOT .5.

Now let's do a "sniff test" in your 1/2 assumption. For this to be true, divers would have to dip into their reserves i.e. if starting iwth 3000, go BELOW 1000) on 50% of their dives. This simply doesn't happen.

So... while I think this is a laudable topic to explore, your numbers just don't add up.

[OFG-1]

Gary,

After reviewing your information, I have one request. Please go and take a statistics class. Your numbers have no meaning, your method has a ton of built in bias based on your own beliefs, and you are manufacturing data points to fit your conclusion. I am not by any means saying that if any person on any given dive uses too much gas going in that they will not drown. They may. But you have no way of telling exactly what happened to any person who has cacked themselves.

When you recover a body, what you see is the aftermath of a drowning. Sometimes the person has gas left, sometimes they don't. Sometimes they are entangled, sometimes not. You find disjointed pieces of information that show you the aftermath of some event or events. Deployed spools, silt, rolled off valves, this type of thing. The original event that overloaded the diver is rarely known. As example, does anyone really know why Berman ran out of gas?

If you read the data that Jeff's group publishes, you will notice that very often any conclusions are left out, or are stated as an opinion of the person doing the recovery. You are taking overall accident data and trying to make it fit one conclusion, that they violated gas management rules.

I personally knew several of the people on Jeff's list. One died doing one of the first tri mix dives on deco. One lady had a hot mix. One missed a jump in Peacock diving solo. One had a broken line and siltout. Several were doing "trust me" dives where they had no business diving there at all. A couple that had plenty of gas left lost it, and probably would have recovered if they kept their wits. None of these, nor many more I suspect, originally had anything to do with gas management, except of course, when they ran out.

You are beating the same horse that you beat with cave conservation, just a different whip. People are going to act responsibly or not. They will manage their gas or not. They will be careful in the caves or not. I do appreciate your concern, but face it, when one of the most experienced divers in the world dies trying to make 1000 feet on SCUBA, you have to reach the conclusion that people are going to do dumb things.

I will share the OFG rule of gas management. Never run out.

[Gary]

So the risk of losing your main supply AND your reserve is the square of losing just your main. If diving 1/2's, then essentially all you need to do is lose your main and you're screwed.

Note that it gives a probability of .001, NOT .5.

Since hardly anyone who has died cave diving has died due to equipment failure looking at it as a chance of "two items failing" vs "one item failing" situation is useless.

But let's look at your numbers anyway:

[Are you saying (whole reserve -> half) or (2/3's -> 1/3)? I'll assume you are going with all -> 1/2 to leave the most air possible.]

So diving to (all reserve) is 1 in a million chance of dying
and diving to half is the SQRT of .000001 is .001
and diving to half of that is 1/4th is .0316
and diving to 1/8 is .17
and diving to 1/16th is about .42

42% chance of dying if you press all the way to 200 psi before turning... (3000 psi used going in and 200 psi going out.)

and 90% if you press all the way to 50 psi ...
you still have a 2.5% chance to make it if you turn with 12 psi remaining...

I don't think so....

[MengTze]

Analysis paralysis comes to mind...............

[aainslie]

Since hardly anyone who has died cave diving has died due to equipment failure looking at it as a chance of "two items failing" vs "one item failing" situation is useless.

But let's look at your numbers anyway:

[Are you saying (whole reserve -> half) or (2/3's -> 1/3)? I'll assume you are going with all -> 1/2 to leave the most air possible.]

So diving to (all reserve) is 1 in a million chance of dying
and diving to half is the SQRT of .000001 is .001
and diving to half of that is 1/4th is .0316
and diving to 1/8 is .17
and diving to 1/16th is about .42

42% chance of dying if you press all the way to 200 psi before turning... (3000 psi used going in and 200 psi going out.)

and 90% if you press all the way to 50 psi ...
you still have a 2.5% chance to make it if you turn with 12 psi remaining...

I don't think so....

Gary, I'm going to go with John on this. Take a stats course, learn how independence assumptions affect probabilities, make it clear that you understand a few basic distributions that might be belpful here like Poisson, Binomial, Lognormal and Normal, then let's chat.