Risk Analysis

[skip]

Considering blueprint for survival, accident analysis, and the recent suggested additions to things that kill, i wonder if a risk analysis is possible in which the risk factors are assessed along the way. Like the difference between tables and computers is a square profile versus on the fly profiling, risk assessment may also be possible "on the fly."

Risk may come in packets of behavior. One packet is loading the gear, another is transporting, another is filling tanks, another is unloading gear (all rather boring). Then we get to the more interesting packets: setting up and checking the gear. I know we all make risk assessments at this stage/packet. A minor leak from the first stage - too small to be a worry so dive anyway? We do buddy checks, S-drills, run guidlelines, etc., and each behavioral packet has it's own associated risk.

We've often heard of the "cascade of events" that results in an incident. A packet analysis may be a way of deciding the relative contribution of each event in the cascade and of finding the tipping point. I know we already do something like what I'm trying to define: buddy checks at this point in time, S-drills, air-sharing, at other points in time. But perhaps it needs to be extended to the dive itself. The dive has a beginning, middle, end, naturally, but it also has a lower and upper middle(?) Moments or "packets" during which risk is present. At different times, different packets, are associated with different levels of risk.

Gas management is often based on losing all gas at max penetration (the most critical "packet" is max penetration), but gas loss is a risk, varying risk, at any point in the dive, and not just based on penetration/distance/time (lose deco gas half-way through deco for example). Gas management "rules" may be better designed with a packet analysis of risk. Reading the thread over on TDS on 1/2+200 vs 1/3's may be an argument over packet analysis versus overall dive planning.

Packet analysis may also influence other guidelines (no pun intended). It's pretty obvious that some divers run lines, while others assess the risk of silt out as too improbable to bother with a guideline today. Still others do visual jumps (one "packet" may occur at the jump decision point).

In some very general way, a "packet analysis" view means assessing risk at each significant point during the dive. It may be nothing more than that, but an attitude of dividing the dive up into significant packets in which the diver stops, breathes, thinks, assess risk, and then continues or calls it....well maybe some benefit?

If this sounds rather vague that's because it is! and BTW, I just made up packet analysis, or maybe read the term somewhere in computer programming, or just heard it used for something else once. It's not a "real" thing as far as I know.

-skip

[FW]

Skip, you need to go diving! You are going to hurt your brain thinking too much between dives

[cavediver256]

I gotta agree with Forrest, Skip. I am heading down in just a few hours....come on down and lets go diving !!!!

[skip]

I was just waiting for the sun to come up a little higher.... i'm off to guy-james this very minute!

-skip

[cavediver256]

I feel better just knowing you are going to let the cool waters clear your mind..... Be safe Skip.

TJ

[crawford]

Off topic but please let us know how Guy James looks, even if it's aweful.

[skip]

I had fun in guy-james today. Visibility the first 400 feet was bad, zero to one foot. I just stayed on the line. Then it cleared to 20 feet and was very nice. I turned at 1/6's, sidemount 95's, about 1,000 feet in. lots of fish. Line following again coming out, I just closed my eyes and stayed on the line til the cavern zone.

it did clear my mind too, and now I'm more convinced than ever that there is something to risk packets, or whatever I called them.

-skip

[aainslie]

OK, back ON topic...

I really like your way of thinking, Skip. That probably doesn't surprise you given that thread!

But actually, that thread really got me thinking and changed some of the way that I think about gas management. I think the "packets" are considering the worst events individually, and in particularly thinking about what causes cascades ,i.e. how packets affect each other.

Some examples: 1) I really started thinking in that thread, what are the WORST places to have gas failures? They are, at the furthest point of penetration, and at each stage changeover. That seriously bears thinking about. 2) A scooter failure at the furthest point. 3) HOW you intend handling a silt-out. One of the most valuable things I learned on an incredible two days with Steve Bogaerts (THERE is a guy that should write a book on cave diving, it'd be brilliant) is when he taught me to learn to swim just as fast lights out as I do with lights... otherwise guess what, a lights out or siltout will cascade to a gas issue, as you swim so slowly that you run through your reserves. I have NEVER heard anotehr instructor mention this critical issue. Doing a lights out drill with Bogaerts is very different to anything you did on your dad's cave class.

These types of issues bear a LOT of thinking about. On the one or two issues I've had it really helped to think, what's the best thing I can do? Working through "packets" in your surface time REALLY helps.

An example... I once did a 7000 ft traverse. Along the way, I got temporarily waylaid, and basically swam a silly 800 to 1000 ft circuit before realizing what I'd done. I recomputed gas reserves... tight, but I decided to continue. This was about 3000 feet in. Then, at the 5000 ft point, a stage totally died on me. I switched it on, and all it did was stream bubbles. I swam., trying to fix it... already Steve's training was kicking in, I did NOT stop and try to fix it. I couldn't - I later found an O ring had split on the inner part of the DIN fitting. Then - a big help in gas issues - I started feathering the stage to see how much I got out of it. It was incredible - it was pissing gas, but I still got it to last about 15 minutes before it died. That was 15 minutes more than I would have had otherwise.

So... two issues. The first I handled poorly, I should have turned. The second I'm prouder of... I simply started feathering that stage, and it gave me a lot more gas than if I ahd just treated it as dead. Oh... and another thing, I had a double marking system whereby I marked both the way to the far side and to the entrance side of the traverse, and that allowed me to recover easily when I did the unintended circuit.

Bottom line - I like the idea of "packets", and of trying to work out how to deal with each one. The more we work this out when dry, the more we stand a chance of surviving when the stuff hits the fan underwater.

[Mart]

he taught me to learn to swim just as fast lights out as I do with lights... otherwise guess what, a lights out or siltout will cascade to a gas issue, as you swim so slowly that you run through your reserves. I have NEVER heard anotehr instructor mention this critical issue.

Don't get me wrong, I heard from several people Steve is a very good instructor. But this example is total nonsense. I have all my students think about this, draw their conclusions, and try to swim as fast with lights out as they do with lights. And all instructors I know do the same. Maybe there are some that do not cover this issue, there are always those that miss the point.

[caver]

Considering blueprint for survival, accident analysis, and the recent suggested additions to things that kill, i wonder if a risk analysis is possible in which the risk factors are assessed along the way. Like the difference between tables and computers is a square profile versus on the fly profiling, risk assessment may also be possible "on the fly."

Risk may come in packets of behavior. One packet is loading the gear, another is transporting, another is filling tanks, another is unloading gear (all rather boring). Then we get to the more interesting packets: setting up and checking the gear. I know we all make risk assessments at this stage/packet. A minor leak from the first stage - too small to be a worry so dive anyway? We do buddy checks, S-drills, run guidlelines, etc., and each behavioral packet has it's own associated risk.

We've often heard of the "cascade of events" that results in an incident. A packet analysis may be a way of deciding the relative contribution of each event in the cascade and of finding the tipping point. I know we already do something like what I'm trying to define: buddy checks at this point in time, S-drills, air-sharing, at other points in time. But perhaps it needs to be extended to the dive itself. The dive has a beginning, middle, end, naturally, but it also has a lower and upper middle(?) Moments or "packets" during which risk is present. At different times, different packets, are associated with different levels of risk.

Gas management is often based on losing all gas at max penetration (the most critical "packet" is max penetration), but gas loss is a risk, varying risk, at any point in the dive, and not just based on penetration/distance/time (lose deco gas half-way through deco for example). Gas management "rules" may be better designed with a packet analysis of risk. Reading the thread over on TDS on 1/2+200 vs 1/3's may be an argument over packet analysis versus overall dive planning.

Packet analysis may also influence other guidelines (no pun intended). It's pretty obvious that some divers run lines, while others assess the risk of silt out as too improbable to bother with a guideline today. Still others do visual jumps (one "packet" may occur at the jump decision point).

In some very general way, a "packet analysis" view means assessing risk at each significant point during the dive. It may be nothing more than that, but an attitude of dividing the dive up into significant packets in which the diver stops, breathes, thinks, assess risk, and then continues or calls it....well maybe some benefit?

If this sounds rather vague that's because it is! and BTW, I just made up packet analysis, or maybe read the term somewhere in computer programming, or just heard it used for something else once. It's not a "real" thing as far as I know.

-skip

You may want to watch the show "Numbers"...It is on CBS, Friday nights.