The Imperfections of Accident Analysis

Article by Fiona Macleod CEng FIChemE

Four key ideas that help us understand the real world of work.

TEN years ago, I had the good fortune to hear Erik Hollnagel speak at a conference. I wondered then why his ideas were not better known in the process safety community.

He articulated a problem that I had been struggling with while managing a high hazard site. Despite a total commitment to process safety at all levels of the organisation, we still had too many near-misses and incidents. Each investigation threw out a raft of new recommendations to prevent recurrence, to the point where maintenance of current good practice risked being eroded by new initiatives.

In our first paper together, The Imperfections of Accident Analysis  (Loss Prevention Bulletin 270, December 2019) we look at  four key ideas that help us understand the real world of work.

1. WAI-WAD – Work as Imagined vs Work as Done

It is practically impossible to provide guidelines or instructions that are detailed enough to be followed ‘mechanically.’ In fact, “work to rule” is often used as a form of industrial action; employees precisely follow all written regulations to the letter in order to cause a slowdown. In real work, people face a variety of difficulties, complexities, dilemmas and tradeoffs and are called on to achieve multiple, often conflicting, goals. How work is actually done, how everyday performance is adjusted to match the conditions and why things go well is a prerequisite for understanding what has or could go wrong.

2. ETTO – Efficiency Thoroughness Trade Off

We make regular trade-offs between the resources we spend on preparing to do something (thoroughness) and the resources we spend on executing it (efficiency). While no activity will succeed without a minimum of both thorough preparation and efficient execution, the ETTO fallacy is that people are required to be both efficient and thorough at the same time – or rather to be thorough when with hindsight it was wrong to be efficient.

3. M&M – Methods and Models

Accident analysis is often based on simple linear causality – one thing causes another. But modern process plants are complex, interlocking systems, designed and run by people – socio-technical rather than pure technical systems. 

4. WYLFIWYF – What you look for, is what you find

The assumptions about the possible causes of an incident (what-you-look-for) will, to a large extent, determine what lessons are learned (what-you-find).

Hindsight gives you 20:20 vision, but looking forward there are multiple branches, interactions and possible futures. In the real world of work – where we lack infinite time and resource – how do we prioritise?

It turns out that human brains are particularly good at this: working with incomplete information, seeing patterns, sorting by difference, juggling limited resources, imagining possible futures.

It is this human ability to adapt, to adjust, to absorb everyday variability based on experience that allows work to get done.

Reinforcing what goes well, understanding why, walking around the workplace, observing and listening to people, damping variation and building resilient teams is one way that high performing organisations minimise adverse events, protect the environment and keep people safe.


Read the full paper, The Imperfections of Accident Analysis, free of charge, here

Article by Fiona Macleod CEng FIChemE

Managing Director of Billions Europe and Volunteer Chair of the Loss Prevention Bulletin Editorial Panel

Recent Editions

Catch up on the latest news, views and jobs from The Chemical Engineer. Below are the four latest issues. View a wider selection of the archive from within the Magazine section of this site.