The incident pit and redundancy

Owen Petchey · January 20th, 2007, 12:29 AM

This post is motivated by posts (pasted below) in the thread 'what do you regard as wrong'.

My understanding is that the incident pit is the phenomenon that failures (equpiment failure, loss of kit, separation, etc) tend to increase the likelihood and negative consequence of subsequent failure. Failures make failures more likely and their consequences more serious. This combination of increased likelihood of failure and increased consequence of a failure creates increased risk.

I assume that the increased likelihood of a subsequent failure occurs due to increased stress and task loading caused by the previous failure(s). And that the increased negative consequence of a subsequent failure can result from loss of redundancy caused by the previous failure(s).

For example, separation may increase stress, increase task loading, impair good judgement, and make a second failure more likely. It would also result in zero redundancy if one's only source of redundancy is the team. This reduction in redundancy increases the potential negative consequence of a subsequent failure.

As Mark writes, many of the fatalities in the BS-AC incident reports may be attributed to this phenomenon. It is something we should guard against.

If my representation of the incident pit is correct, what should we do?

My impression (gained from BS-AC and GUE training) is that it suggests that we should be concerned primarily with making an appropriate response to failures. We should aim to respond in a way that gets us out of the pit, or at least stops us descending further into it. Training that includes simulated failures and encourages reasoned and relaxed responses to failures could be one approach to this aim.

I'm less certain of what the incident pit phenomenon means we should do regarding redundancy (i.e., the number of failures we cater for). My understanding of Mark's post is that he thinks it means we should increase our level of redundancy. He might, however, have been only suggesting we cater for separation plus one other failure (which I agree is prudent if its known that a dive involves significant likelihood of separation).

I'm interested in knowing what people think the incident pit phenomenon suggests about the level of redundancy we carry.

Owen

(rjack)
Most of these come down to having 2+ major failuers on 1 dive. [snip]

(Mark Chase)
Look at the BSAC incident report on fatalities. Many fatalities started out with seporation and then a second incident occured causing the fatality.
This is repeated over and over again in the reports.
They call it the incident pit.
ATB
Mark Chase

rjack · January 20th, 2007, 12:53 AM

There comes a point when planning for multiple major failures that you just give up. Lost backgas AND lost deco gas is an example. The buddy requirements become so extreme, you should just take up golf.

And buddy seperation is a major failure IMHO, although it frequently gets short shift in many educational systems (can't speak for BSAC). It certainly did in my early NAUI/PADI classes.

Otherwise... yes, halting the escalating levels of stress is step one to making rationale decisions with the resources remaining at hand. Adding additional redundancy (beyond 1 major failure or a couple of minor ones) is less important to me than making good, clearheaded, decisions with what's typically available.

Richard

lamont · January 20th, 2007, 04:57 AM

The DIR divers that I know who do big dives all carry bags and reels so that if they get separated they can extract themselves. Typically everyone has a bag, a reel, safety spool, a backup mask, backup cutting tools and a deco plan which can handle loss of a single deco gas. That makes any separated diver pretty self-sufficient. This has been tested through real-life sepearation and the separated diver has operated fine independently.

So, what does the argument center around here, is it just the backup depth gauge / timer?

The peril of going down this road too far is that you start to take failures which are not likely to be part of a cascade due to stress (like complete failure of an isolator resulting in total gas loss) and combine that with the separation scenario and wind up switching to independent twins because of the possibility of a massively remote failure case occuring at the same time as team separation. The correct answer there is to make team separation as infrequent as possible (if team separation is 1 in 100 and complete isolator failures are 1 in 100,000 then the combined risk on any dive is 1 in 10,000,000 which is getting into the struck-by-lightning its-just-your-time-to-go category).

So, there's a limitation to the incident pit analysis where different failures are more or less likely to be affected by stress. Messing up shooting a bag is more likely to occur while stressed. Having an isolator valve blow or even a depth gauge crap out is not going to be more likely while stressed (although the impact while already stressed of having a depth gauge go will be magnified). And I don't think that you can take a separated team member and do 'failure mode analysis' on their situation in the same way that you would a solo diver -- already you've hopefully got only a 1 in 100 chance of being separated or less, and if the failure is not stress-related then it should be at least two orders of magnitude less likely to occur on the same dive as separation as it would on a normal dive with a solo diver. I'm not generally going to be worried about those happening unless they're already fairly common (1 in 100 dives or so).

Mark Chase · January 20th, 2007, 09:53 AM

[quote]

Quote: (Originally Posted by lamont)

The DIR divers that I know who do big dives all carry bags and reels so that if they get separated they can extract themselves. Typically everyone has a bag, a reel, safety spool, a backup mask, backup cutting tools and a deco plan which can handle loss of a single deco gas. That makes any separated diver pretty self-sufficient. This has been tested through real-life sepearation and the separated diver has operated fine independently.

So, what does the argument center around here, is it just the backup depth gauge / timer?

No, the argument centres arround the DIR divers I have met who consider the above (prudent in my view) precautions unnessary. Worse they apear to consider them NOT DIR.

I am enormously pleased that DIR divers have come forward on these threads pointing out that this is not true and i hope it will change the attitude of some of the more recent additions to DIR.

ATB

Mark Chase

Ahmed Adly, Marlin Inn DC · January 20th, 2007, 10:22 AM

Quote: (Originally Posted by lamont)

(if team separation is 1 in 100 and complete isolator failures are 1 in 100,000 then the combined risk on any dive is 1 in 10,000,000 which is getting into the struck-by-lightning its-just-your-time-to-go category).

I think I will use this in my class this week.

Ahmed Adly, Marlin Inn DC · January 20th, 2007, 10:25 AM

Hi Owen,

I was reading your post and actually scared that it might be some magical incantaion that could turn me into a frog or something.

Well put.

jerry.mobbs · January 20th, 2007, 10:30 AM

Quote: (Originally Posted by Owen Petchey)

My impression (gained from BS-AC and GUE training) is that it suggests that we should be concerned primarily with making an appropriate response to failures. We should aim to respond in a way that gets us out of the pit, or at least stops us descending further into it. Training that includes simulated failures and encourages reasoned and relaxed responses to failures could be one approach to this aim.

[/i]

Owen

DIR is more about not even getting into the pit in the first place. Its a subtle but VERY important difference.

The fundamentals such as mindset, balanced rig, teamwork, correct gas choice and standardisation mean you should not even have to look over the edge of the pit, let alone fall into it.

It is always easy to construct scenarios where DIR in theory does not work, but they often range from the sublime to the ridiculous.

Jerry

*JGrogan* · January 20th, 2007, 10:34 AM

Quote: (Originally Posted by lamont)

The DIR divers that I know who do big dives all carry bags and reels so that if they get separated they can extract themselves. Typically everyone has a bag, a reel, safety spool, a backup mask, backup cutting tools and a deco plan which can handle loss of a single deco gas. That makes any separated diver pretty self-sufficient. This has been tested through real-life sepearation and the separated diver has operated fine independently.

This is standard operating procedure for any of the experienced DIR divers in the UK.

There is a lot of BS going around on what is DIR and what is not. The most important thing to remember is to use your brain and think things through!

John.

Martin Burnard · January 20th, 2007, 02:26 PM

Whether diving twins or single (less than 20m no deco I use singles), both I and my buddy (or team when three of us) each carry the same basic kit whatever.

This basic kit is
2 bottom timers/ depth gauges
2 spools
2 dsmb.
1 knife
1 pair of shears
1 spare mask
1 compass
1 set of relevant tables
1 set of wet notes.
1 main torch
1 back up torch.

When diving with the peeps from this board they have been likewise equipped.

So has that just been coincidence, or is it typical of the divers on this board whether DIR or not?

On seperation we run the 1 minute search, then surface. Each diver is equipped for solo ascent. So minimising incident pit slope.

In summary "When it goes wrong, go up", the incident pit slope is down hill all the way, so get your ass up.

Redundacy is the surface as last resort on shallow dives no deco, and the shut downs etc for deco.

tim · January 20th, 2007, 02:41 PM

Quote: (Originally Posted by Martin Burnard)

So has that just been coincidence, or is it typical of the divers on this board whether DIR or not?

That matches my default kit carried for any dive. I'll normally also carry a spare double ender.

Tim

January 20th, 2007, 12:29 AM	#1 (permalink)
Owen Petchey(Offline) New Member Join Date: Nov 2005 Location: Sheffield Posts: 40	The incident pit and redundancy This post is motivated by posts (pasted below) in the thread 'what do you regard as wrong'. My understanding is that the incident pit is the phenomenon that failures (equpiment failure, loss of kit, separation, etc) tend to increase the likelihood and negative consequence of subsequent failure. Failures make failures more likely and their consequences more serious. This combination of increased likelihood of failure and increased consequence of a failure creates increased risk. I assume that the increased likelihood of a subsequent failure occurs due to increased stress and task loading caused by the previous failure(s). And that the increased negative consequence of a subsequent failure can result from loss of redundancy caused by the previous failure(s). For example, separation may increase stress, increase task loading, impair good judgement, and make a second failure more likely. It would also result in zero redundancy if one's only source of redundancy is the team. This reduction in redundancy increases the potential negative consequence of a subsequent failure. As Mark writes, many of the fatalities in the BS-AC incident reports may be attributed to this phenomenon. It is something we should guard against. If my representation of the incident pit is correct, what should we do? My impression (gained from BS-AC and GUE training) is that it suggests that we should be concerned primarily with making an appropriate response to failures. We should aim to respond in a way that gets us out of the pit, or at least stops us descending further into it. Training that includes simulated failures and encourages reasoned and relaxed responses to failures could be one approach to this aim. I'm less certain of what the incident pit phenomenon means we should do regarding redundancy (i.e., the number of failures we cater for). My understanding of Mark's post is that he thinks it means we should increase our level of redundancy. He might, however, have been only suggesting we cater for separation plus one other failure (which I agree is prudent if its known that a dive involves significant likelihood of separation). I'm interested in knowing what people think the incident pit phenomenon suggests about the level of redundancy we carry. Owen (rjack) Most of these come down to having 2+ major failuers on 1 dive. [snip] (Mark Chase) Look at the BSAC incident report on fatalities. Many fatalities started out with seporation and then a second incident occured causing the fatality. This is repeated over and over again in the reports. They call it the incident pit. ATB Mark Chase

January 20th, 2007, 09:53 AM	#4 (permalink)
Mark Chase(Offline) New Member Join Date: Nov 2005 Location: Kent England Posts: 905	[quote] Quote: (Originally Posted by lamont) The DIR divers that I know who do big dives all carry bags and reels so that if they get separated they can extract themselves. Typically everyone has a bag, a reel, safety spool, a backup mask, backup cutting tools and a deco plan which can handle loss of a single deco gas. That makes any separated diver pretty self-sufficient. This has been tested through real-life sepearation and the separated diver has operated fine independently. So, what does the argument center around here, is it just the backup depth gauge / timer? No, the argument centres arround the DIR divers I have met who consider the above (prudent in my view) precautions unnessary. Worse they apear to consider them NOT DIR. I am enormously pleased that DIR divers have come forward on these threads pointing out that this is not true and i hope it will change the attitude of some of the more recent additions to DIR. ATB Mark Chase __________________ The only DIR Inspiration diver in the village

January 20th, 2007, 10:22 AM	#5 (permalink)
Ahmed Adly, Marlin Inn DC(Offline) New Member Join Date: Feb 2006 Location: hurghada Posts: 849	Quote: (Originally Posted by lamont) (if team separation is 1 in 100 and complete isolator failures are 1 in 100,000 then the combined risk on any dive is 1 in 10,000,000 which is getting into the struck-by-lightning its-just-your-time-to-go category). I think I will use this in my class this week. __________________ It has just dawned on me.... We are all just recycled stardust Ahmed Adly, www.deepvoyage.com

January 20th, 2007, 10:25 AM	#6 (permalink)
Ahmed Adly, Marlin Inn DC(Offline) New Member Join Date: Feb 2006 Location: hurghada Posts: 849	Hi Owen, I was reading your post and actually scared that it might be some magical incantaion that could turn me into a frog or something. Well put. __________________ It has just dawned on me.... We are all just recycled stardust Ahmed Adly, www.deepvoyage.com

January 20th, 2007, 10:30 AM	#7 (permalink)
jerry.mobbs(Offline) New Member Join Date: May 2006 Location: asia Posts: 288	Quote: (Originally Posted by Owen Petchey) My impression (gained from BS-AC and GUE training) is that it suggests that we should be concerned primarily with making an appropriate response to failures. We should aim to respond in a way that gets us out of the pit, or at least stops us descending further into it. Training that includes simulated failures and encourages reasoned and relaxed responses to failures could be one approach to this aim. [/i] Owen DIR is more about not even getting into the pit in the first place. Its a subtle but VERY important difference. The fundamentals such as mindset, balanced rig, teamwork, correct gas choice and standardisation mean you should not even have to look over the edge of the pit, let alone fall into it. It is always easy to construct scenarios where DIR in theory does not work, but they often range from the sublime to the ridiculous. Jerry __________________ -------------------- www.crossfit.com The DIR of Fitness

January 20th, 2007, 12:53 AM	#2 (permalink)
rjack(Offline) wet behind the ears Join Date: Nov 2005 Location: Seattle, WA Posts: 863	There comes a point when planning for multiple major failures that you just give up. Lost backgas AND lost deco gas is an example. The buddy requirements become so extreme, you should just take up golf. And buddy seperation is a major failure IMHO, although it frequently gets short shift in many educational systems (can't speak for BSAC). It certainly did in my early NAUI/PADI classes. Otherwise... yes, halting the escalating levels of stress is step one to making rationale decisions with the resources remaining at hand. Adding additional redundancy (beyond 1 major failure or a couple of minor ones) is less important to me than making good, clearheaded, decisions with what's typically available. Richard

January 20th, 2007, 04:57 AM	#3 (permalink)
lamont(Offline) APLP Join Date: Nov 2005 Location: Seattle, WA Posts: 181	The DIR divers that I know who do big dives all carry bags and reels so that if they get separated they can extract themselves. Typically everyone has a bag, a reel, safety spool, a backup mask, backup cutting tools and a deco plan which can handle loss of a single deco gas. That makes any separated diver pretty self-sufficient. This has been tested through real-life sepearation and the separated diver has operated fine independently. So, what does the argument center around here, is it just the backup depth gauge / timer? The peril of going down this road too far is that you start to take failures which are not likely to be part of a cascade due to stress (like complete failure of an isolator resulting in total gas loss) and combine that with the separation scenario and wind up switching to independent twins because of the possibility of a massively remote failure case occuring at the same time as team separation. The correct answer there is to make team separation as infrequent as possible (if team separation is 1 in 100 and complete isolator failures are 1 in 100,000 then the combined risk on any dive is 1 in 10,000,000 which is getting into the struck-by-lightning its-just-your-time-to-go category). So, there's a limitation to the incident pit analysis where different failures are more or less likely to be affected by stress. Messing up shooting a bag is more likely to occur while stressed. Having an isolator valve blow or even a depth gauge crap out is not going to be more likely while stressed (although the impact while already stressed of having a depth gauge go will be magnified). And I don't think that you can take a separated team member and do 'failure mode analysis' on their situation in the same way that you would a solo diver -- already you've hopefully got only a 1 in 100 chance of being separated or less, and if the failure is not stress-related then it should be at least two orders of magnitude less likely to occur on the same dive as separation as it would on a normal dive with a solo diver. I'm not generally going to be worried about those happening unless they're already fairly common (1 in 100 dives or so).

January 20th, 2007, 10:34 AM	#8 (permalink)
JGrogan(Offline) GUE Join Date: May 2006 Location: London Posts: 442	Quote: (Originally Posted by lamont) The DIR divers that I know who do big dives all carry bags and reels so that if they get separated they can extract themselves. Typically everyone has a bag, a reel, safety spool, a backup mask, backup cutting tools and a deco plan which can handle loss of a single deco gas. That makes any separated diver pretty self-sufficient. This has been tested through real-life sepearation and the separated diver has operated fine independently. This is standard operating procedure for any of the experienced DIR divers in the UK. There is a lot of BS going around on what is DIR and what is not. The most important thing to remember is to use your brain and think things through! John.

January 20th, 2007, 02:26 PM	#9 (permalink)
Martin Burnard(Offline) New Member Join Date: Nov 2005 Location: Plymouth, UK Posts: 580	Whether diving twins or single (less than 20m no deco I use singles), both I and my buddy (or team when three of us) each carry the same basic kit whatever. This basic kit is 2 bottom timers/ depth gauges 2 spools 2 dsmb. 1 knife 1 pair of shears 1 spare mask 1 compass 1 set of relevant tables 1 set of wet notes. 1 main torch 1 back up torch. When diving with the peeps from this board they have been likewise equipped. So has that just been coincidence, or is it typical of the divers on this board whether DIR or not? On seperation we run the 1 minute search, then surface. Each diver is equipped for solo ascent. So minimising incident pit slope. In summary "When it goes wrong, go up", the incident pit slope is down hill all the way, so get your ass up. Redundacy is the surface as last resort on shallow dives no deco, and the shut downs etc for deco. __________________ I once enjoyed a sado-masachistic frenzy of total tintintabulation leading to a state of metampsychosis... or did I make that up? www.bananafrogcars.co.uk

January 20th, 2007, 02:41 PM	#10 (permalink)
tim(Offline) Big Grin Syndrome Join Date: Dec 2005 Location: Oxford Posts: 455	Quote: (Originally Posted by Martin Burnard) So has that just been coincidence, or is it typical of the divers on this board whether DIR or not? That matches my default kit carried for any dive. I'll normally also carry a spare double ender. Tim