Peter

Thanks a very interesting and well written article. I have always been a sceptic of those who claim they have the best safety records or systems - you're only as good as your last or next major incident. Covering up LTIs and manipulating AFR/IFRs is very tempting when your reputation is on the line. Measuring performance and aligning it with current practices is fraught with difficulties. The tendency is to 'measure what is easy to measure as opposed to what should be measured'.

Part of the problem is the Heinrich/Bird triangle.

Assumes a directly linear relationship from the top to the bottom of the triangle, but……

Some minor incidents are not precursors of a fatality

BST (2011) analysed incidents at six multinationals and estimate only 20% of less serious injuries are precursors of Serious Injuries and Fatalities (SIFs).

Common bias towards dealing unsafe behaviour rather than unsafe conditions.

Heinrich – “Selection of remedies is based on practical cause analysis that stops at the selection of the first proximate and most easily prevented cause” (usually human error). Heinrich 88% down to unsafe behaviour, DuPont upped this to 96%.

BST - Selection of measures to address unsafe behaviour may result in greater reduction in minor injuries than reduction in SIFs.

Recordable injury rates in North America and Europe have steadily declined in last decade and longer, but fatality rates have levelled off or reduced by less.

Ditto - tend to put a veil over the risks that could result in low probability, high consequence events or the management of occupational health risks.

BST -Causes and correlates of SIFs are often different to those of less serious injuries.

and almost all accidents are multi-causal……

Study of 92,000 accidents in Pennsylvania in 1953 (with similar findings in 1960 follow up)

Over 90% of both fatal AND non-fatal accidents linked to both unsafe behaviour AND unsafe conditions.

I was exposed to Dupont some 20 years ago and made my mind up then it was a myth created by PR.
Ditto BP say about 10 years ago.
All big companies are big on words but less good at actions, simply because middle managers tell their bosses what they want to hear.

Regarding "Triangles" under various names, it was clear to anyone with a smattering of statistics knowledge that they were dodgy. Actually their authors pretty much said so.
That did not stop people, who could not be bothered to think about accident causation, hijacking them.

"Proper" behavioral safety is very important in accident prevention IMHO.
But again we have managers who see it as a "blame the workers" get out of jail card.

Interesting short video on Heinrich's Principle (& triangle) by Prof Tim Marsh.

I worked for ExxonMobil here and in Europe during the 90's. They used Dupont STOP at first and then moved to the Loss Prevention System. Both added real value but needed careful use.

However, these were not the only tools they used. They also used Taproot for root cause analysis and had a well developed safety management system ( the OIMS mentioned earlier by Jay).

Can't talk for BP or for what Exxon are like now but in my time Exxon were very serious and very pro-active on the safety front. I have been to a lot of places since and not seen anything as good.

The fall of Dupont from safety leader to also ran would make an interesting book.

I am another old oil man (BP mostly and some time with Exxon). I remember the flirtations with OIMS in Exxon and ISRS (the Bird based system) in BP from the mid 80's to mid 90's. I think I probably went to at least one of Mervs, or his colleagues, presentations as we looked over the fence to see if there was anything to learn from Dupont. (and yes there was back then)

The big difference that I saw happening during that time was the dumbing down and reduction of technical management and more reliance on systems and audits for what we now call governance.

In my opinion, for both the OIMS and ISRS systems, the focus on activity at the coal face was much less signficant than engineering and management design. I still remember the 'people, equipment, materials, environment' base for those systems. However, as 'managers' and 'finance' got their hands on them and the technical input waned then they focused more and more on the 'easy to get' achievements.

It is oh so easy to confuse activity with accomplishment and outcomes and in the wrong hands these sytems inevitably lead to focus on the minor and mundane at the expense of looking at the overall picture.

Human error amongst senior staff through not understanding/following/updating or properly modifying their own procedures and protocols is just as common as those errors lower down the order. Put that with the fact that we know that poor management equals poor OSH in the workplace and it follows that any system can only be as good as those using and/or managing it.

These systems can and do bring success and improvement. But they all take on-going hard work and being open to change at all levels in the organisation. And that is the hardest thing of all to achieve and then maintain over a long period.

No system is the golden nugget of guaranteed success. That is the fundamental error that can be made when using these systems. Many of the large petrochem companies have found that out to their cost in the 21st century.

I would generally agree that one major accident in itself is not a true reflection of an organisation's health and safety capability, at least until I read the investigation report (all 400 pages!) into BP's Deep Water Horizon (Gulf of Mexico) oil spill disaster.

The report identified systemic failures at every level within the organisation, multiple latent and active failures on the day. Having had an identical near miss some months before it could be strongly argued the Gulf spill was an event waiting to happen.

Quote IanBell.
"yet when fundamental issues about plant design and operation are pointed out, solutions are ignored or swept under the carpet because of difficulty of implementing solutions i.e. cost."

Exactly the same issue as 25 years ago. Makes you think doesn't it?

I think Bird's Triangle and the nonsense of behavioural safety etc - hold the hand rail, wear your hat etc are much too simplistic, when considering major accidents.

But to question such h&s dogma such as these theories has been akin to the Spanish Inquisition and heresy in the sometimes small minded world of occupational h&s.