CCR Incident (Feb 2013) – Double Cell Failure, Human Factors – Inquest Report

CCR FatalityAn instructor was teaching two students an advanced trimix CCR class. Unfortunately due to two of the cells being current limited and the voting logic voting out the ‘good cell’, and the instructor being a victim to a number of powerful human factor biases, they had an oxygen toxicity seizure at depth. One of this buddies took the diver to the surface but never recovered.

This incident happened in February 2013 and was widely reported in the (social) media at the time, especially as it quickly became known that the diver was using cells which were 40 months old (cells 1 & 3) and 17 months old (cell 2) and a significant amount of negative criticism raised, primarily how could someone so experienced undertake such a ‘stupid’ activity, using cells well past their use-by date. Immediately afterwards, AP issued a warning about using out-of-date cells.

Unfortunately a number of technical and human factors came to play and these were identified by the Australian Coroner in this excellent report (it is a shame that UK Coroner’s reports are not made public like this as there are many lessons to be learned contained within it).

Confirmation bias and anchoring are extremely powerful biases which influences your decision making without even realising it.

An example I use in my Crew Resource Management training is to show this video. If you want to see this bias in action, watch the video first, follow the instructions to the letter, then come back. Do not read further until you have watched the video.  Now click on the video link! I now want you to remember how many plant pots were in the scene, 2 or 3?

I have not done this in a remote context before so not sure how it will work out in terms of the responses from readers, but the answer is none. You may have been influenced by my placing the answer in the question thinking there were plant pots there, but not sure, or you may have thought it was 2 or 3. When I show this clip in my training and coaching classes I use the power of those present to influence the decision making process. In a one-to-one session, I influence the coachee because I ask a question; I wouldn’t ask a question with an incorrect response in the posed question would I? In the classroom, I speak to the most senior trainee in the room before the class starts and ask him to respond positively to my question when I ask how many plant pots are in the video clip with ‘3’.  I play the and ask the questions posed in the video. Most miss the obvious feature present (selective attention subject). I then ask the senior trainee how many plant pots are in the scene and he responds with ‘3’. Pretty much without failure, everyone in the class or a one-to-one session responds the same, they follow the lead.  This is despite me already asking them whether they would question something that wasn’t right that their boss said. What they perceive they will do, and what they will actually do, are normally different things unless a certain questioning mindset is in place.

Now looking to diving, I know of an incident where an instructor was teaching a trimix class with 3 students and as the instructor had finished his 6m decompression stops, the instructor ‘asked’ students how much deco was left as their own decompression obligation was clear.  The students still had time remaining on their timers but decided that as the instructor had signalled that the deco was clear, they should ascend rather than staying there to finish the decompression.  One of the students had to be treated for DCS as a consequence. The instructor was unaware of the power of influence they had and no longer asks a question with the answer in it. (There is also a level of trust involved such that the student is going to speak up if something goes wrong).

The coroner’s report made reference to ‘anchoring‘ which is

…the tendency to over-rely on one piece of information, especially information that is our first piece of information on the relevant subject, to the exclusion of other information that is inconsistent with the information we rely on.  For example, it is well known that first impressions can be hard to shift.

Furthermore, once a plan has started to be executed and it isn’t going to plan, to stop that execution takes considerable effort and we end up being the subject to another bias called ‘sunk costs’ where we don’t want to throw away what we’ve already gained even if we think it is going wrong. “Just a little bit more, it will be ok…”

Alarm blindness. There are numerous examples of (non-diving) operators cancelling alarms without really understanding what is going on behind the scenes. The first, second, may third are ‘read and understood’ but if the environment hasn’t changed, then the likelihood for the operator to cancel the alarm immediately it goes off for the fourth time without checking to understand what is going on is very high.  When you have too many false positives, you end up with a ‘cry wolf’ scenario and then ignore the alarm (or don’t trust it) – this can have fatal consequences.  At Eurotek 2012 I was involved in the discussion about checklists following RF3.0. During that discussion one audience member said that they didn’t like the electronic checklist on the Sentinel because it stopped him diving. I asked whether it was the automated checklist he didn’t like, or the fact that the unit kept on failing and therefore the checklist was doing its job…there wasn’t a reply.  This article provides more on this subject.

External Pressures.  Although not explicitly covered in the report because it would be hard to verify, there may have been a perceived pressure to use the out of date cells to maintain an income stream from instructing, especially as the cells had not ‘failed’ in the past. Finding reliable cells once Teledyne went out of business was hard and many people had to make the hard decision to either not dive, or dive cells which were past their ‘use by’ date (or change rebreather to use another cell).  Simple to make that decision whilst sitting at your computer, not so easy when the cells haven’t ‘failed’ in the past and you need / want to get that diving ‘fix’.  Divers need to recognise these external pressures and control them – having an external decision making process can certainly help, maybe a regular buddy who understands what is going on and is unafraid to challenge you. A question like ‘if this all goes wrong on the dive because of X, how would it look to others (wife/kids)?’ can certainly help individuals.

Just Culture and Reporting. It is very easy to make snap judgements about why poor decisions are made without fully understanding the context or background, even when so called ‘obvious’ mistakes are made.  As we can see from this inquest report, detail is important. Unfortunately inquests only happen when fatalities occur and fortunately these are rare. When you submit an incident report to something like DISMS think about how someone completely unconnected could recreate the incident and learn from it.

Diving has a level of risk which we can control and mitigate to a certain extent, as long as we are informed of the factors out there. Reports like this are great because they go into significant detail identifying a number of non-technical factors at play. If you know a CCR diver, send them a copy of the report. It doesn’t take long to read, but it will certainly help improve their knowledge and may save their life. If you like this blog, sign up, you’ll get notified when I publish the next one.

3 responses to “CCR Incident (Feb 2013) – Double Cell Failure, Human Factors – Inquest Report

  1. Pingback: Gareth Lock Offers Thoughtful Examination of Phil Gray Fatality | IntoThePlanet·

  2. Pingback: CCR incident Feb 2013 | Rebreather Clean·

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s