How Much Gas is Enough - Running on Empty (Ainslie)

[aainslie]

Hey Ken,

I have mixed feelings about buddies during exploration. Frankly there are not a lot of people willing to go into some of these caves with me. Next, you barely see each other during the most critical parts of the dive as there are extended sections of sidemount only, narrow passage. But, I must say, it would have hugely helped after the emergency started to have someone calming me down.

[RN]

I'm with you on that, Andrew. There are some passages that I just won't go into unless I'm solo. Trying to dive these passages in a team only increases the risk.

[LiteHedded]

great article man. glad you made it out ok

[argyris]

Great article Andrew, its very nice to share the knowledge after such an event like this. Slow movement, and common sense are key elemnts in survival psychology.

[Michael Stroeck]

In no flow caves thirds is definitely not enough and I do not understand why it is an accepted and taught safety margin, I think you can never go wrong with having too much air, but you can very easily go very wrong with having exactly enough.

I agree and am not comfortable diving to thirds in most cases.

As to why it's taught, that probably comes down to an over-reliance on accident analysis and statistics. Diving to thirds hasn't killed enough people to consider it completely unsafe and never will. It's just very unlikely. Consider the following scenarios:

1) No flow; dive planned to thirds; one catastrophic gas loss in a two person team at planned max penetration. Even if nobody panics and they share gas, both divers are in deep trouble. The stress and logistics of gas-sharing will probably not allow for a safe exit.

That's the worst case-scenario, however. Additionally, as far as I'm aware, the very few times that actually happened the divers miraculously made it out anyway. After all, there are some mitigating factors: jump and primary reels that might have taken a lot of gas to place; the fact that you might have been surveying, sight-seeing, etc. on the way in; deco bottles somewhere close to the exit, etc...

2) Same as above, but in a three person team. Unless one of them panics or another gas loss occurs, they will be fine.

3) Same as 1), but at around 80% or less of planned max penetration. Again, they will probably get out.

It's common sense that the vast majority of cases of complete gas loss while diving thirds will resemble scenarios 2) and 3). Therefore, it's reasonably unlikely that diving to thirds will kill a lot of divers, which is why it's still taught...

[DA Aquamaster]

I agree and am not comfortable diving to thirds in most cases.

As to why it's taught, that probably comes down to an over-reliance on accident analysis and statistics. Diving to thirds hasn't killed enough people to consider it completely unsafe and never will. It's just very unlikely. Consider the following scenarios:

1) No flow; dive planned to thirds; one catastrophic gas loss in a two person team at planned max penetration. Even if nobody panics and they share gas, both divers are in deep trouble. The stress and logistics of gas-sharing will probably not allow for a safe exit.

That's the worst case-scenario, however. Additionally, as far as I'm aware, the very few times that actually happened the divers miraculously made it out anyway. After all, there are some mitigating factors: jump and primary reels that might have taken a lot of gas to place; the fact that you might have been surveying, sight-seeing, etc. on the way in; deco bottles somewhere close to the exit, etc...

2) Same as above, but in a three person team. Unless one of them panics or another gas loss occurs, they will be fine.

3) Same as 1), but at around 80% or less of planned max penetration. Again, they will probably get out.

It's common sense that the vast majority of cases of complete gas loss while diving thirds will resemble scenarios 2) and 3). Therefore, it's reasonably unlikely that diving to thirds will kill a lot of divers, which is why it's still taught...

I agree. Statistically, a total gas failure is extremely rare and a total gas loss failure at max penetration would be an extremely rare subset of an already rare event.

In practice, if the failure occurs a 2-3 hundred psi prior to reaching thirds, it gets all of the benefits of more conservative planning with none of the intent, so in the vast majority of cases where a total gas loss failure has occurred, the dive was not actually done to "thirds" regardless of the intended gas plan.

Also, as you point out, how divers conduct the dive potentially adds safety. Some instructors teach and some divers use a "sight see on the way in" approach as even if you also sight see on the way out, the longer penetration time for a given distance builds in a fudge factor not otherwise present. And this is on top of natural delays due to placing a primary line, jump/gap reels etc that are not pulled on the way out in an emergency.

Additionally, 3 person teams further mitigate the gas loss/reserve issue with thirds that would be present on a two person team, and some instructors suggest additional reserve gas for two person teams.

Finally, rounding down thirds adds some unintended reserve. For example, having 3800 psi in double 95s on the surface just before the dive still results in 1200 psi "thirds" and the 200 psi left over equates to 14 cu ft of additional reserve not factored into the "thirds" calculation. Similarly 3500 psi and 1100 psi thirds results in the same 14 cu ft pad, but results in a slightly greater reserve as a percentage of pentration gas than with 3800 psi.

I think it is roughly analogous to diving with the old US Navy tables. When I started diving they were the only game in town and it was well known that 1) the normative group was not very compatible with recreational divers, 1) the hit rate on square profiles for deco diving and repetitive diving reflected operatioan neccesity and/or on board recompression capability. In light of that, everyone built in fudge factors or operational practices that individually all added small safety margins and cumulatively served to ensure the overall hit rate was acceptably low.

In practice, "thirds" statistically works the same way with a very small incidence of events where divers ran out of gas in circumstances where slightly more conservative gas planning would have made a difference - assuming divers are prudent with it, don't cut corners and recognize when it is clearly not appropriate.

[Michael Stroeck]

Finally, rounding down thirds adds some unintended reserve. For example, having 3800 psi in double 95s on the surface just before the dive still results in 1200 psi "thirds" and the 200 psi left over equates to 14 cu ft of additional reserve not factored into the "thirds" calculation. Similarly 3500 psi and 1100 psi thirds results in the same 14 cu ft pad, but results in a slightly greater reserve as a percentage of pentration gas than with 3800 psi.

There's an important point to be made about that. Many divers I know will round down numbers that are not evenly divisible by three. So if they have 220 bar in their tanks, they'll do 210/3=70 bar for penetration. HOWEVER, when they have 210 bar or 240 bar in their tanks, they'll just use that number and divide by three WITHOUT deducting a safety factor afterwards. That makes no sense and I think it's incredibly dumb. You always need to deduct a fudge factor of at least 10 bar, IMHO.

[DA Aquamaster]

There's an important point to be made about that. Many divers I know will round down numbers that are not evenly divisible by three. So if they have 220 bar in their tanks, they'll do 210/3=70 bar for penetration. HOWEVER, when they have 210 bar or 240 bar in their tanks, they'll just use that number and divide by three WITHOUT deducting a safety factor afterwards. That makes no sense and I think it's incredibly dumb. You always need to deduct a fudge factor of at least 10 bar, IMHO.

I agree, and on those dives where I really start with 3600 psi (or 3900 psi if the tank monkey is feeling really generous) I tend to turn the dive a couple hunered psi earlier than normal to reflect the loss of the "left over"gas in the orignal thirds calculation.

[ntheath]

Is AA's article available online or elsewhere (email, etc.) yet?

[aainslie]

Give me one more month, and let me clear it with NACD. I'm in High Springs starting tomorrow so I'll clear itwith Larry and Deb.