Rebreather incident investigation checklist

York · 9th October 2007, 10:09

Hi all,
in response to Alex's sad-but-true comment

Quote: (Originally Posted by AD_ward9)

Every time I look over these reports, I am struck by how inadequate many accident investigations are. The problem of coroners looking at the end result, instead of the cause, is a major problem.

http://www.rebreatherworld.com/rebre...tml#post141845
I would like to encourage a discussion to compile a checklist for investigators asked to look into rebreather incidents. This may not be of immediate use for us, but the long-term effect could be more complete and consistent analysis, resulting in better design criteria and training requirements, ultimately serving to reduce incident numbers. I have no doubt that, if we manage to generate a good "guide for the rebreather accident investigator", law enforcement personell will be happy to utilise this. Details might be to specific too be asked from someone not knowledgeable in the subject, but maybe we can point out who could answer these questions.

I would like to start the discussion with a few thoughts - certainly incoherent, and I hope you guys jump in:

What information should the recovery team record while still at depth? Removing the unit/victim has the potential to alter conditions significantly, and if the person is dead anyway, a few minutes delay should not be of concern, so there should be time to record a few facts (i.e. pictures). Position of the mouthpiece (in mouth, out of mouth, open, closed, bail-out position). Pressures in the tanks, position of the tank valves. Electronics on or off. HUD blinking? how? PO2 displayed on unit? Should the recovery team close the mouthpiece if it isn't already to preserve the loop? The recovery team needs to know these things before setting out.

What considerations are there regarding the victim? In case of a deceased, how quickly does a post mortem need be carried out before valuable data gets lost? What conditions are to be looked for?

What consideration are there regarding the rebreather? Unitflooded/dry
I remember that in one specific case there was an attempt to analyse the scrubber after it has been sitting in storage for weeks, after it had been completely flooded - IMHO completely pointless. So how fast does an analysis need to be made? in the example of the scrubber, what can be done and who could do it?

In general, I often hear that the manufacturer was asked to analyse the system for faults. Now I believe that most manufacturers are ethical, but it still sounds like a terrible conflict of interest to me. Who could be asked to assist in analysing the system for faults? The competition?

just initial thoughts, you guys are much more experts. I believe we need to keep this simple, but yet inclusive. The idea is that using this checklist, the conditions as they were encountered at the time of the investigation are recorded and can be re-created in court at a later date - the fundamental meaning of "forensics".
At appears to me that, at the moment, any investigator is pretty much doing his own thing, making comparison very difficult.

In anticipation of a good discussion
Joerg

AD_ward9 · 9th October 2007, 11:53

May I suggest, you need a process rather than a checklist.

The main bones of the process should be:

1. At dive scene, photograph everything. Record tank pressures, handset readings, exact configuration including hoses, cylinders and weighting. Do not press any buttons, do not replace any regulators behind bungies or change anything at all: just record it as it is. However, a photograph of the dive profile is most helpful: it is lost too often. The Buddy needs to ring the next of kin: the damage caused by contacting the next of kin later, is grim. Tell it straight, be kind, keep it short.

2. Get a copy of all Police Witness Statements, and all dive computer dumps from the buddies and those recovering the diver. Get information on when the regulators and rebreather was last serviced. Find out if the diver has any known health issues (e.g. is deaf, state of fitness), their certifications and experience. Speak to the diver's regular buddies to find out what sort of diver it was who was involved in the accident.

3. The equipment should go to an independent lab for examination to check all breathing gases, WOB and for equipment faults. NEDU in the States do an excellent job of checking the gas etc, and HSE do a similar thing in the UK. This is the First Examination: two are needed, see below. Measure how much water is emptied out of the unit. Do a pos and neg pressure check of the unit. Test the ADV function (get a lissajou). Photograph the serial numbers of the O2 sensors, obtain the last calibration data from the handsets using a cell emulator, and then test the cells. Test the handset's functionality using a Cell Emulator. Photograph the batteries using a macro lens and the battery contacts in the unit. Do an EN250 lissajou test on the regulators, but remember empty cylinders take in water and will corrode the regulator. Weigh the scrubber, then test its breathing resistance. Let it dry, and measure the weight again. Compare resistance with a reference scrubber that has been packed properly but soaked in water. Peel off the handset case (if soft), or open it if not, and photograph it using a macro lens and inspect for leakage.

4. All known facts should then be reviewed by a panel of equipment and dive experts, and extra data obtained to resolve ambiguities of fact. If you are a lone investigator, bring others in to challenge your assumptions and check the information you gather to ensure it cannot be interpreted differently.

5. The generic rebreather FMECA (one on the DL site, for example), should be used to assign a probability for every possible cause, given available information. From this identify what information is needed to confirm or reject a specific cause, and have the equipment tested again, by a different expert from that involved in the First Examination, to get that information where appropriate. The investigator often does not need the autopsy report to do this, where the rest of the data is clear, but if it is not obtained state that clearly what assumptions you make on the diver's health.

6. Formal modelling of every plausible cause should be carried out, to eliminate those not consistent with the facts. This involves running a computer model. For example, hypoxia and hyerpoxia can usually be distinguished very easily this way, and the computer can trace the hypoxia back to a specific time range on the dive profile where the problem started. Most other events can be traced in a similar manner.

7. Review the formal modelling and FMECA results, again against all available evidence, and construct an objective timeline of the accident. Where data is unknown, that should be stated. If there are several plausible clauses, there may be several different timelines.

Question every conclusion to ensure there is not an alternative.

8. Manufacturers and relatives should receive a copy of the full report, with all information and data, regardless of who retains the investigator. The manufacturer should examine the events and consider if there is any change to the equipment, or any new feature that is within ALARP, that could have prevented the accident, then implement it, advising users accordingly (recall, or implement at next service).

Alex

depth · 9th October 2007, 12:00

Quote: (Originally Posted by AD_ward9)

May I suggest, you need a process rather than a checklist.

The main bones of the process should be:

1. At dive scene, photograph everything. Record tank pressures, handset readings, exact configuration. Do not press any buttons, do not replace any regulators behind bungies or change anything at all: just record it as it is. However, a photograph of the dive profile is most helpful: it is lost too often.

3. Get a copy of all Police Witness Statements, and all dive computer dumps from the buddies and those recovering the diver. Get information on when the regulators and rebreather was last serviced. Find out if the diver has any known health issues (e.g. is deaf, state of fitness), their certifications and experience.

4. The equipment should go to an independent lab for examination to check all breathing gases, WOB and for equipment faults. NEDU in the States do an excellent job of checking the gas etc, and HSE do a similar thing in the UK. This is the First Examination: two are needed, see below. Measure how much water is emptied out of the unit. Do a pos and neg pressure check of the unit. Test the ADV function (get a lissajou). Photograph the serial numbers of the O2 sensors, obtain the last calibration data from the handsets using a cell emulator, and then test the cells. Test the handset's functionality using a Cell Emulator. Photograph the batteries using a macro lens and the battery contacts in the unit. Do an EN250 lissajou test on the regulators, but remember emtpy cylinders take in water and will corrode the regulator. Weigh the scrubber, then test its breathing resistance. Let it dry, and measure the weight again. Compare resistance with a reference scrubber that has been packed properly but soaked in water.

5. All known facts should then be reviewed by a panel of equipment and dive experts, and extra data obtained to resolve ambiguities of fact. If you are a lone investigator, bring others in to challenge your assumptions and check the information you gather to ensure it cannot be interpreted differently.

6. The generic rebreather FMECA (one on the DL site, for example), should be used to assign a probability for every possible cause, given available information. From this identify what information is needed to confirm or reject a specific cause, and have the equipment tested again, by a different expert from that involved in the First Examination, to get that information where appropriate. The investigator often does not need the autopsy report to do this, where the rest of the data is clear, but if it is not obtained state that clearly what assumptions you make on the diver's health.

7. Formal modelling of every plausible cause should be carried out, to eliminate those not consistent with the facts. This involves running a computer model. For example, hypoxia and hyerpoxia can usually be distinguished very easily this way, and the computer can trace the hypoxia back to a specific time range on the dive profile where the problem started. Most other events can be traced in a similar manner.

8. Review the formal modelling and FMECA results, again against all available evidence, and construct an objective timeline of the accident. Where data is unknown, that should be stated. If there are several plausible clauses, there may be several different timelines.

Question every conclusion to ensure there is not an alternative.

9. Manufacturers and relatives should receive a copy of the full report, with all information and data, regardless of who retains the investigator. The manufacturer should examine the events and consider if there is any change to the equipment, or any new feature that is within ALARP, that could have prevented the accident, then implement it, advising users accordingly (recall, or implement at next service).

Alex

Good points, but what happened to #2 ??? ;-)

AD_ward9 · 9th October 2007, 12:04

Quote: (Originally Posted by depth)

Good points, but what happened to #2 ??? ;-)

Sorry, in the edit, got merged with 3, and you snapped it before I finished my edit. I must use preview more often...

Alex

***caveseeker7*** · 9th October 2007, 12:40

DAN called in a meeting a while back that was centered around Rebreather fazalities.
DAN Rebreather Fatality Investigation Meeting: Tuesday, November 7, 2006
Below a draft by Mike Ward "DIVING REBREATHER ACCIDENT CHAIN OF CUSTODY"
Thought it has been around, I've had it since Dec 06.

York · 9th October 2007, 12:43

Quote: (Originally Posted by AD_ward9)

May I suggest, you need a process rather than a checklist.

The main bones of the process should be:

1. At dive scene, photograph everything. Record tank pressures, handset readings, exact configuration including hoses, cylinders and weighting. Do not press any buttons, do not replace any regulators behind bungies or change anything at all: just record it as it is. However, a photograph of the dive profile is most helpful: it is lost too often. The Buddy needs to ring the next of kin: the damage caused by contacting the next of kin later, is grim. Tell it straight, be kind, keep it short.

2. Get a copy of all Police Witness Statements, and all dive computer dumps from the buddies and those recovering the diver. Get information on when the regulators and rebreather was last serviced. Find out if the diver has any known health issues (e.g. is deaf, state of fitness), their certifications and experience. Speak to the diver's regular buddies to find out what sort of diver it was who was involved in the accident.

3. The equipment should go to an independent lab for examination to check all breathing gases, WOB and for equipment faults. NEDU in the States do an excellent job of checking the gas etc, and HSE do a similar thing in the UK. This is the First Examination: two are needed, see below. Measure how much water is emptied out of the unit. Do a pos and neg pressure check of the unit. Test the ADV function (get a lissajou). Photograph the serial numbers of the O2 sensors, obtain the last calibration data from the handsets using a cell emulator, and then test the cells. Test the handset's functionality using a Cell Emulator. Photograph the batteries using a macro lens and the battery contacts in the unit. Do an EN250 lissajou test on the regulators, but remember empty cylinders take in water and will corrode the regulator. Weigh the scrubber, then test its breathing resistance. Let it dry, and measure the weight again. Compare resistance with a reference scrubber that has been packed properly but soaked in water. Peel off the handset case (if soft), or open it if not, and photograph it using a macro lens and inspect for leakage.

4. All known facts should then be reviewed by a panel of equipment and dive experts, and extra data obtained to resolve ambiguities of fact. If you are a lone investigator, bring others in to challenge your assumptions and check the information you gather to ensure it cannot be interpreted differently.

5. The generic rebreather FMECA (one on the DL site, for example), should be used to assign a probability for every possible cause, given available information. From this identify what information is needed to confirm or reject a specific cause, and have the equipment tested again, by a different expert from that involved in the First Examination, to get that information where appropriate. The investigator often does not need the autopsy report to do this, where the rest of the data is clear, but if it is not obtained state that clearly what assumptions you make on the diver's health.

6. Formal modelling of every plausible cause should be carried out, to eliminate those not consistent with the facts. This involves running a computer model. For example, hypoxia and hyerpoxia can usually be distinguished very easily this way, and the computer can trace the hypoxia back to a specific time range on the dive profile where the problem started. Most other events can be traced in a similar manner.

7. Review the formal modelling and FMECA results, again against all available evidence, and construct an objective timeline of the accident. Where data is unknown, that should be stated. If there are several plausible clauses, there may be several different timelines.

Question every conclusion to ensure there is not an alternative.

8. Manufacturers and relatives should receive a copy of the full report, with all information and data, regardless of who retains the investigator. The manufacturer should examine the events and consider if there is any change to the equipment, or any new feature that is within ALARP, that could have prevented the accident, then implement it, advising users accordingly (recall, or implement at next service).

Alex

Excellent start, will try to compile a check-off sheet.
More discussion: Most things mentioned focus on the unit. As mentioned earlier: after photographing, should the mouthpiece be closed prior to retrievintg the body from depth, or left open. Should valves be closed or be left open. How about modified units/homebuilds?

What is to be looked for in the victim in an autopsy that may be specific to rebreather diving?

No1: "Everything" is a mighty task, and we all know how little info we can get from DEMA pictures although the people visiting try to photograph everything. Specifics are given, and the list should probably grow in the process. How about photographing the victims hands also?

Re. No.2: What kind of behavioral past would be of interest - specific to rebreathers?

Re. No3: What's the window of opportunity? If the equipment sits wet in storage for 6 months then there may be no need for further investigation. Simply asap? 24 hours? 12 hours? 6? Now?
Should the face of the sensors be photographed asap to ascertain moisture buildup before sending off for examination? When measuring how much water is emptied out of the unit, would it be important to note where the water is coming from? Flow-direction test of the check-valves?

4: where can "experts" (I love this term) be found?

Joerg

***caveseeker7*** · 9th October 2007, 12:59

Quote: (Originally Posted by York)

4: where can "experts" (I love this term) be found?

United States NAVY Experimental Diving Unit, Panama City, Florida.
You find the adress on page three.

Good folks down there, when they have the time they do have a keen
intrest in technical/rebreather diving issues.

AD_ward9 · 9th October 2007, 13:06

Quote: (Originally Posted by caveseeker7)

DAN called in a meeting a while back that was centered around Rebreather fazalities.
DAN Rebreather Fatality Investigation Meeting: Tuesday, November 7, 2006
Below a draft by Mike Ward "DIVING REBREATHER ACCIDENT CHAIN OF CUSTODY"
Thought it has been around, I've had it since Dec 06.

The DAN chain of custody list is an excellent implementation of Process Step 1.

Alex

AD_ward9 · 9th October 2007, 13:14

Quote: (Originally Posted by York)

Re. No3: What's the window of opportunity? If the equipment sits wet in storage for 6 months then there may be no need for further investigation. Simply asap? 24 hours? 12 hours? 6? Now?

It needs to proceed through the stages listed as fast as possible. Getting from Step 1 to Step 3 is critical, and should not exceed 3 weeks. After that the corrosion really starts destroying things. If it cannot get to Step 3 fast, then the water should be drained via the mouthpiece (amount measured, pH measured), and flushed through with clean water. Once it has been drained or washed, the corrosion is arrested, but it is important not to disturb anything else when doing this, particularly the scrubber.

Quote: (Originally Posted by York)

Should the face of the sensors be photographed asap to ascertain moisture buildup before sending off for examination? When measuring how much water is emptied out of the unit, would it be important to note where the water is coming from? Flow-direction test of the check-valves?

Units recovered are usually flooded. Water emptied out comes from the mouthpiece generally, unless someone inexperienced opens the scrubber canister first, with the mouthpiece shut, in which case it floods out all over the place.

A flow check of the valves is part of the equipment fault check: this simply proceeds around the unit to check every item. On the WOB check, the attachment of the CLs to the harness is important: one accident was caused by a CL floating above the diver.

On your autopsy question: very little information comes from autopsies in my personal experience. They usually confirm the diver drowned, and cause a lot of trauma to the family to get at that obvious conclusion. Heart attack or stroke indicates one should focus on CO2 and WOB issues, but that data can come from examining the unit. Bottom line is autopsy information is usually a useless smoke screen as regards identifying the real cause. Witness statements, the dive profile, FMECA check and formal modelling are infinitely more valuable. There are exceptions of course, such as a diver who had a leg bitten off by a shark and then later drowned: an autopsy is useful if the circumstances are unusual.

On your question of behaviour: knowing that the diver was a very careful guy who got things serviced, or had no training, or had a thousand dives on the unit, tells you a lot. How a trimix diver responds to problems is generally very different to how a AOW diver responds. A good example is a public safety diver, trying a dry suit out for the first time, was strangled by it because he did not trim the neck seal. Knowing it was a new dry suit, tried out for the first time, was very helpful in that case (and so was the autopsy report, in that instance).

Alex

g3stanton · 9th October 2007, 15:50

One reason better data is not collected is reflected in our diver rescue training. As you may recall York, we were told body recovery was not to include a camera as this would require defense of the data in court, something not found desirable. Our research found police investigators often reject data from underwater as inaccurately collected. Rescuers deliver a body to the surface for the investigation to then begin. Those rescuing the body from depth are seldom trained crime scene investigators. What you have described is a possible crime scene.

I support the development of a Protocol for the investigation of underwater fatalities. While I understand the brainstorming first step seen here, perhaps we should recall the FDLE format of Principle, Policy and Procedure. Why are we collecting this data, what is the purpose (mitigation, causation), what is the legal basis of the data collection, deliniating the requirements for implimenting the protocol (courts, police standard, ME requirements) and what steps are required, establish the generic proficiency-based guidelines to acheive the Protocol.

To move from a "snatch & grab" mindset to a recovery protocol is quite an undertaking.

9th October 2007, 10:09	#1 (permalink)
York probubbly not Current Rebreather/s: Megalodon Other Rebreather/s: Join Date: Mar 2005 Location: Bristol, UK Posts: 118	Rebreather incident investigation checklist Hi all, in response to Alex's sad-but-true comment Quote: (Originally Posted by AD_ward9) Every time I look over these reports, I am struck by how inadequate many accident investigations are. The problem of coroners looking at the end result, instead of the cause, is a major problem. http://www.rebreatherworld.com/rebre...tml#post141845 I would like to encourage a discussion to compile a checklist for investigators asked to look into rebreather incidents. This may not be of immediate use for us, but the long-term effect could be more complete and consistent analysis, resulting in better design criteria and training requirements, ultimately serving to reduce incident numbers. I have no doubt that, if we manage to generate a good "guide for the rebreather accident investigator", law enforcement personell will be happy to utilise this. Details might be to specific too be asked from someone not knowledgeable in the subject, but maybe we can point out who could answer these questions. I would like to start the discussion with a few thoughts - certainly incoherent, and I hope you guys jump in: What information should the recovery team record while still at depth? Removing the unit/victim has the potential to alter conditions significantly, and if the person is dead anyway, a few minutes delay should not be of concern, so there should be time to record a few facts (i.e. pictures). Position of the mouthpiece (in mouth, out of mouth, open, closed, bail-out position). Pressures in the tanks, position of the tank valves. Electronics on or off. HUD blinking? how? PO2 displayed on unit? Should the recovery team close the mouthpiece if it isn't already to preserve the loop? The recovery team needs to know these things before setting out. What considerations are there regarding the victim? In case of a deceased, how quickly does a post mortem need be carried out before valuable data gets lost? What conditions are to be looked for? What consideration are there regarding the rebreather? Unitflooded/dry I remember that in one specific case there was an attempt to analyse the scrubber after it has been sitting in storage for weeks, after it had been completely flooded - IMHO completely pointless. So how fast does an analysis need to be made? in the example of the scrubber, what can be done and who could do it? In general, I often hear that the manufacturer was asked to analyse the system for faults. Now I believe that most manufacturers are ethical, but it still sounds like a terrible conflict of interest to me. Who could be asked to assist in analysing the system for faults? The competition? just initial thoughts, you guys are much more experts. I believe we need to keep this simple, but yet inclusive. The idea is that using this checklist, the conditions as they were encountered at the time of the investigation are recorded and can be re-created in court at a later date - the fundamental meaning of "forensics". At appears to me that, at the moment, any investigator is pretty much doing his own thing, making comparison very difficult. In anticipation of a good discussion Joerg
(Offline)

9th October 2007, 11:53	#2 (permalink)
AD_ward9 So much more to learn Current Rebreather/s: Other CCR Other Rebreather/s: Other CCR Join Date: Jun 2005 Location: Scotland Posts: 1,732	Re: Rebreather incident investigation checklist May I suggest, you need a process rather than a checklist. The main bones of the process should be: 1. At dive scene, photograph everything. Record tank pressures, handset readings, exact configuration including hoses, cylinders and weighting. Do not press any buttons, do not replace any regulators behind bungies or change anything at all: just record it as it is. However, a photograph of the dive profile is most helpful: it is lost too often. The Buddy needs to ring the next of kin: the damage caused by contacting the next of kin later, is grim. Tell it straight, be kind, keep it short. 2. Get a copy of all Police Witness Statements, and all dive computer dumps from the buddies and those recovering the diver. Get information on when the regulators and rebreather was last serviced. Find out if the diver has any known health issues (e.g. is deaf, state of fitness), their certifications and experience. Speak to the diver's regular buddies to find out what sort of diver it was who was involved in the accident. 3. The equipment should go to an independent lab for examination to check all breathing gases, WOB and for equipment faults. NEDU in the States do an excellent job of checking the gas etc, and HSE do a similar thing in the UK. This is the First Examination: two are needed, see below. Measure how much water is emptied out of the unit. Do a pos and neg pressure check of the unit. Test the ADV function (get a lissajou). Photograph the serial numbers of the O2 sensors, obtain the last calibration data from the handsets using a cell emulator, and then test the cells. Test the handset's functionality using a Cell Emulator. Photograph the batteries using a macro lens and the battery contacts in the unit. Do an EN250 lissajou test on the regulators, but remember empty cylinders take in water and will corrode the regulator. Weigh the scrubber, then test its breathing resistance. Let it dry, and measure the weight again. Compare resistance with a reference scrubber that has been packed properly but soaked in water. Peel off the handset case (if soft), or open it if not, and photograph it using a macro lens and inspect for leakage. 4. All known facts should then be reviewed by a panel of equipment and dive experts, and extra data obtained to resolve ambiguities of fact. If you are a lone investigator, bring others in to challenge your assumptions and check the information you gather to ensure it cannot be interpreted differently. 5. The generic rebreather FMECA (one on the DL site, for example), should be used to assign a probability for every possible cause, given available information. From this identify what information is needed to confirm or reject a specific cause, and have the equipment tested again, by a different expert from that involved in the First Examination, to get that information where appropriate. The investigator often does not need the autopsy report to do this, where the rest of the data is clear, but if it is not obtained state that clearly what assumptions you make on the diver's health. 6. Formal modelling of every plausible cause should be carried out, to eliminate those not consistent with the facts. This involves running a computer model. For example, hypoxia and hyerpoxia can usually be distinguished very easily this way, and the computer can trace the hypoxia back to a specific time range on the dive profile where the problem started. Most other events can be traced in a similar manner. 7. Review the formal modelling and FMECA results, again against all available evidence, and construct an objective timeline of the accident. Where data is unknown, that should be stated. If there are several plausible clauses, there may be several different timelines. Question every conclusion to ensure there is not an alternative. 8. Manufacturers and relatives should receive a copy of the full report, with all information and data, regardless of who retains the investigator. The manufacturer should examine the events and consider if there is any change to the equipment, or any new feature that is within ALARP, that could have prevented the accident, then implement it, advising users accordingly (recall, or implement at next service). Alex Last edited by AD_ward9 : 9th October 2007 at 12:06.
(Offline)

9th October 2007, 12:00	#3 (permalink)
depth Custom Title Allowed! Current Rebreather/s: Megalodon Other Rebreather/s: Join Date: Aug 2006 Location: Denmark Posts: 159	Re: Rebreather incident investigation checklist Quote: (Originally Posted by AD_ward9) May I suggest, you need a process rather than a checklist. The main bones of the process should be: 1. At dive scene, photograph everything. Record tank pressures, handset readings, exact configuration. Do not press any buttons, do not replace any regulators behind bungies or change anything at all: just record it as it is. However, a photograph of the dive profile is most helpful: it is lost too often. 3. Get a copy of all Police Witness Statements, and all dive computer dumps from the buddies and those recovering the diver. Get information on when the regulators and rebreather was last serviced. Find out if the diver has any known health issues (e.g. is deaf, state of fitness), their certifications and experience. 4. The equipment should go to an independent lab for examination to check all breathing gases, WOB and for equipment faults. NEDU in the States do an excellent job of checking the gas etc, and HSE do a similar thing in the UK. This is the First Examination: two are needed, see below. Measure how much water is emptied out of the unit. Do a pos and neg pressure check of the unit. Test the ADV function (get a lissajou). Photograph the serial numbers of the O2 sensors, obtain the last calibration data from the handsets using a cell emulator, and then test the cells. Test the handset's functionality using a Cell Emulator. Photograph the batteries using a macro lens and the battery contacts in the unit. Do an EN250 lissajou test on the regulators, but remember emtpy cylinders take in water and will corrode the regulator. Weigh the scrubber, then test its breathing resistance. Let it dry, and measure the weight again. Compare resistance with a reference scrubber that has been packed properly but soaked in water. 5. All known facts should then be reviewed by a panel of equipment and dive experts, and extra data obtained to resolve ambiguities of fact. If you are a lone investigator, bring others in to challenge your assumptions and check the information you gather to ensure it cannot be interpreted differently. 6. The generic rebreather FMECA (one on the DL site, for example), should be used to assign a probability for every possible cause, given available information. From this identify what information is needed to confirm or reject a specific cause, and have the equipment tested again, by a different expert from that involved in the First Examination, to get that information where appropriate. The investigator often does not need the autopsy report to do this, where the rest of the data is clear, but if it is not obtained state that clearly what assumptions you make on the diver's health. 7. Formal modelling of every plausible cause should be carried out, to eliminate those not consistent with the facts. This involves running a computer model. For example, hypoxia and hyerpoxia can usually be distinguished very easily this way, and the computer can trace the hypoxia back to a specific time range on the dive profile where the problem started. Most other events can be traced in a similar manner. 8. Review the formal modelling and FMECA results, again against all available evidence, and construct an objective timeline of the accident. Where data is unknown, that should be stated. If there are several plausible clauses, there may be several different timelines. Question every conclusion to ensure there is not an alternative. 9. Manufacturers and relatives should receive a copy of the full report, with all information and data, regardless of who retains the investigator. The manufacturer should examine the events and consider if there is any change to the equipment, or any new feature that is within ALARP, that could have prevented the accident, then implement it, advising users accordingly (recall, or implement at next service). Alex Good points, but what happened to #2 ??? ;-) __________________ I am not qualified to make my own meg batteries!!! I am not qualified to make my own meg batteries!!! I am not qualified to make my own meg batteries!!! I am not qualified to make my own meg batteries!!! I am not.............
(Offline)

9th October 2007, 12:04	#4 (permalink)
AD_ward9 So much more to learn Current Rebreather/s: Other CCR Other Rebreather/s: Other CCR Join Date: Jun 2005 Location: Scotland Posts: 1,732	Re: Rebreather incident investigation checklist Quote: (Originally Posted by depth) Good points, but what happened to #2 ??? ;-) Sorry, in the edit, got merged with 3, and you snapped it before I finished my edit. I must use preview more often... Alex Last edited by AD_ward9 : 9th October 2007 at 12:08.
(Offline)

9th October 2007, 12:43	#6 (permalink)
York probubbly not Current Rebreather/s: Megalodon Other Rebreather/s: Join Date: Mar 2005 Location: Bristol, UK Posts: 118	Re: Rebreather incident investigation checklist Quote: (Originally Posted by AD_ward9) May I suggest, you need a process rather than a checklist. The main bones of the process should be: 1. At dive scene, photograph everything. Record tank pressures, handset readings, exact configuration including hoses, cylinders and weighting. Do not press any buttons, do not replace any regulators behind bungies or change anything at all: just record it as it is. However, a photograph of the dive profile is most helpful: it is lost too often. The Buddy needs to ring the next of kin: the damage caused by contacting the next of kin later, is grim. Tell it straight, be kind, keep it short. 2. Get a copy of all Police Witness Statements, and all dive computer dumps from the buddies and those recovering the diver. Get information on when the regulators and rebreather was last serviced. Find out if the diver has any known health issues (e.g. is deaf, state of fitness), their certifications and experience. Speak to the diver's regular buddies to find out what sort of diver it was who was involved in the accident. 3. The equipment should go to an independent lab for examination to check all breathing gases, WOB and for equipment faults. NEDU in the States do an excellent job of checking the gas etc, and HSE do a similar thing in the UK. This is the First Examination: two are needed, see below. Measure how much water is emptied out of the unit. Do a pos and neg pressure check of the unit. Test the ADV function (get a lissajou). Photograph the serial numbers of the O2 sensors, obtain the last calibration data from the handsets using a cell emulator, and then test the cells. Test the handset's functionality using a Cell Emulator. Photograph the batteries using a macro lens and the battery contacts in the unit. Do an EN250 lissajou test on the regulators, but remember empty cylinders take in water and will corrode the regulator. Weigh the scrubber, then test its breathing resistance. Let it dry, and measure the weight again. Compare resistance with a reference scrubber that has been packed properly but soaked in water. Peel off the handset case (if soft), or open it if not, and photograph it using a macro lens and inspect for leakage. 4. All known facts should then be reviewed by a panel of equipment and dive experts, and extra data obtained to resolve ambiguities of fact. If you are a lone investigator, bring others in to challenge your assumptions and check the information you gather to ensure it cannot be interpreted differently. 5. The generic rebreather FMECA (one on the DL site, for example), should be used to assign a probability for every possible cause, given available information. From this identify what information is needed to confirm or reject a specific cause, and have the equipment tested again, by a different expert from that involved in the First Examination, to get that information where appropriate. The investigator often does not need the autopsy report to do this, where the rest of the data is clear, but if it is not obtained state that clearly what assumptions you make on the diver's health. 6. Formal modelling of every plausible cause should be carried out, to eliminate those not consistent with the facts. This involves running a computer model. For example, hypoxia and hyerpoxia can usually be distinguished very easily this way, and the computer can trace the hypoxia back to a specific time range on the dive profile where the problem started. Most other events can be traced in a similar manner. 7. Review the formal modelling and FMECA results, again against all available evidence, and construct an objective timeline of the accident. Where data is unknown, that should be stated. If there are several plausible clauses, there may be several different timelines. Question every conclusion to ensure there is not an alternative. 8. Manufacturers and relatives should receive a copy of the full report, with all information and data, regardless of who retains the investigator. The manufacturer should examine the events and consider if there is any change to the equipment, or any new feature that is within ALARP, that could have prevented the accident, then implement it, advising users accordingly (recall, or implement at next service). Alex Excellent start, will try to compile a check-off sheet. More discussion: Most things mentioned focus on the unit. As mentioned earlier: after photographing, should the mouthpiece be closed prior to retrievintg the body from depth, or left open. Should valves be closed or be left open. How about modified units/homebuilds? What is to be looked for in the victim in an autopsy that may be specific to rebreather diving? No1: "Everything" is a mighty task, and we all know how little info we can get from DEMA pictures although the people visiting try to photograph everything. Specifics are given, and the list should probably grow in the process. How about photographing the victims hands also? Re. No.2: What kind of behavioral past would be of interest - specific to rebreathers? Re. No3: What's the window of opportunity? If the equipment sits wet in storage for 6 months then there may be no need for further investigation. Simply asap? 24 hours? 12 hours? 6? Now? Should the face of the sensors be photographed asap to ascertain moisture buildup before sending off for examination? When measuring how much water is emptied out of the unit, would it be important to note where the water is coming from? Flow-direction test of the check-valves? 4: where can "experts" (I love this term) be found? Joerg
(Offline)

9th October 2007, 13:06	#8 (permalink)
AD_ward9 So much more to learn Current Rebreather/s: Other CCR Other Rebreather/s: Other CCR Join Date: Jun 2005 Location: Scotland Posts: 1,732	Re: Rebreather incident investigation checklist Quote: (Originally Posted by caveseeker7) DAN called in a meeting a while back that was centered around Rebreather fazalities. DAN Rebreather Fatality Investigation Meeting: Tuesday, November 7, 2006 Below a draft by Mike Ward "DIVING REBREATHER ACCIDENT CHAIN OF CUSTODY" Thought it has been around, I've had it since Dec 06. The DAN chain of custody list is an excellent implementation of Process Step 1. Alex
(Offline)

9th October 2007, 13:14	#9 (permalink)
AD_ward9 So much more to learn Current Rebreather/s: Other CCR Other Rebreather/s: Other CCR Join Date: Jun 2005 Location: Scotland Posts: 1,732	Re: Rebreather incident investigation checklist Quote: (Originally Posted by York) Re. No3: What's the window of opportunity? If the equipment sits wet in storage for 6 months then there may be no need for further investigation. Simply asap? 24 hours? 12 hours? 6? Now? It needs to proceed through the stages listed as fast as possible. Getting from Step 1 to Step 3 is critical, and should not exceed 3 weeks. After that the corrosion really starts destroying things. If it cannot get to Step 3 fast, then the water should be drained via the mouthpiece (amount measured, pH measured), and flushed through with clean water. Once it has been drained or washed, the corrosion is arrested, but it is important not to disturb anything else when doing this, particularly the scrubber. Quote: (Originally Posted by York) Should the face of the sensors be photographed asap to ascertain moisture buildup before sending off for examination? When measuring how much water is emptied out of the unit, would it be important to note where the water is coming from? Flow-direction test of the check-valves? Units recovered are usually flooded. Water emptied out comes from the mouthpiece generally, unless someone inexperienced opens the scrubber canister first, with the mouthpiece shut, in which case it floods out all over the place. A flow check of the valves is part of the equipment fault check: this simply proceeds around the unit to check every item. On the WOB check, the attachment of the CLs to the harness is important: one accident was caused by a CL floating above the diver. On your autopsy question: very little information comes from autopsies in my personal experience. They usually confirm the diver drowned, and cause a lot of trauma to the family to get at that obvious conclusion. Heart attack or stroke indicates one should focus on CO2 and WOB issues, but that data can come from examining the unit. Bottom line is autopsy information is usually a useless smoke screen as regards identifying the real cause. Witness statements, the dive profile, FMECA check and formal modelling are infinitely more valuable. There are exceptions of course, such as a diver who had a leg bitten off by a shark and then later drowned: an autopsy is useful if the circumstances are unusual. On your question of behaviour: knowing that the diver was a very careful guy who got things serviced, or had no training, or had a thousand dives on the unit, tells you a lot. How a trimix diver responds to problems is generally very different to how a AOW diver responds. A good example is a public safety diver, trying a dry suit out for the first time, was strangled by it because he did not trim the neck seal. Knowing it was a new dry suit, tried out for the first time, was very helpful in that case (and so was the autopsy report, in that instance). Alex Last edited by AD_ward9 : 9th October 2007 at 13:40.
(Offline)

9th October 2007, 15:50	#10 (permalink)
g3stanton New Member Current Rebreather/s: Inspiration Classic Megalodon Sport Kiss Classic Kiss Ouroboros Optima Dolphin Ray Other Rebreather/s: Dolphin Ray Join Date: Nov 2006 Location: Wakulla County, Florida Posts: 8	Re: Rebreather incident investigation checklist One reason better data is not collected is reflected in our diver rescue training. As you may recall York, we were told body recovery was not to include a camera as this would require defense of the data in court, something not found desirable. Our research found police investigators often reject data from underwater as inaccurately collected. Rescuers deliver a body to the surface for the investigation to then begin. Those rescuing the body from depth are seldom trained crime scene investigators. What you have described is a possible crime scene. I support the development of a Protocol for the investigation of underwater fatalities. While I understand the brainstorming first step seen here, perhaps we should recall the FDLE format of Principle, Policy and Procedure. Why are we collecting this data, what is the purpose (mitigation, causation), what is the legal basis of the data collection, deliniating the requirements for implimenting the protocol (courts, police standard, ME requirements) and what steps are required, establish the generic proficiency-based guidelines to acheive the Protocol. To move from a "snatch & grab" mindset to a recovery protocol is quite an undertaking.
(Offline)

Thread Tools
Show Printable Version Email this Page
Display Modes
Linear Mode Switch to Hybrid Mode Switch to Threaded Mode