Death of the Safety Triangle

01 January 2015

The Baker report on the BP Texas City Oil Refinery explosion in 2005 found there is little relationship between everyday near misses and the circumstances that result in a major process disaster.

The theory that there is a fixed ratio between the small number of serious accidents in an organisation and the far greater number of minor accidents and higher still volume of near misses, the so-called accident triangle, was first put forward by Herbert Heinrich in the 1920s and was refined by later safety academics and is still widely taught today.

But in recent years the triangle has been widely attacked and was even dropped by the HSE from the latest version of its HSG65 management systems guidance.

The numbers used by Heinrich – one major injury for every seven minor incidents and 189 near misses - have been questioned, as has the idea that paying attention to low level incidents in a workplace helps ensure the worst cannot happen.

The Baker report on the BP Texas City Oil Refinery explosion in 2005 found there is little relationship between everyday near misses and the circumstances that result in a major process disaster.

But recent research has suggested it is possible to derive consistent ratios for specific accident types, such as being struck by a workplace vehicle or falls on the level, though they vary enormously by hazard.

On the issue of small incidents being a poor tell-tale for major process failures, perhaps what we need is quality control, refining our focus to relevant minor injuries and near misses. If the overnight scaffolding collapse was during weather that would have stopped work anyway, then, as long as you are confident that control will be applied, there may be no need for further action; if the scaffolding collapsed overnight but could equally have happened during the day when people were working on and under it, there is a problem that needs fixing.

Use of the accident triangle may be no magic bullet to inoculate an organisation against major injury and tailoring the ratios to hazard class and organisation may mean we end up with multiple triangles (some of them barely triangular), but perhaps there is life in the old theory yet.

Biffa's adoption of a systematic approach to reducing incident levels actually aligns with the requirements for the OHSAS 18001 certified health and safety management system it has in place. This too cannot necessarily be used to eliminate all unforeseeable risks, but it has assisted in driving the following positive results:

A 40% reduction in lost time incidents (LTI) over a 12 month period from July 2012 to July 2014 is.

In April – March 2012/13 Biffa recorded the following incident levels:

73 - RIDDOR 95 accidents
135 - Lost Time incidents
1623 - minor injuries
15000 - near misses
30000 – hazards

In April – March 2013/14 it is clear to see from the positive improvements below, that the number of incidents that occurred had decreased, whilst the number of near misses and hazards reported had significantly increased.

64 - RIDDOR '95 accidents (76 - RIDDOR 95)
97 - Lost Time incidents
1,271 - Minor injuries
18000 - near misses
40,000 - hazards

Biffa's Head of Safety, Health & Quality – Matthew Humphreys says:

"OHSAS 18001 continues to be a critical part of our management system foundations and a key part of our continual improvement safety strategy."

A full feature article on accident triangles is available in this month's issue of Health and Safety at Work magazine.