Has Reason’s Cheese gone stale?

Courtesy of James Reason, the Swiss Cheese Model has advanced thinking considerably in the explanation, and in the prevention, of major accidents in the energy industry. Reason forced us to think in more than one dimension and to consider ‘defences in depth’ to avoid undesirable events happening in our industry. These defences include hardware, people, systems and processes.

Reasons ‘Defences in Depth’, 1997


In recent years, industry has adopted many tools, processes and models based on this barrier model aimed at preventing and minimising the risk of Major Accident Hazards being realised. This has without doubt progressed the understanding of the primary causal factors of these events, but we are still, with an unnerving regularity, having unplanned hydrocarbon releases and major accidents. In the UK North Sea, hydrocarbon leaks have reduced in recent years, however, in the period April 2013 to April 2014 there was a 20% increase in leaks (source HSE). This year alone, four workers were killed and several injured, following a fire and explosion on the Pemex Abkatun platform in the Gulf of Mexico, and six workers were killed on the Cidade de Sao Mateus FPSO in Brazil following a pump room explosion.

Certainly, there will have been in-depth analyses carried out by investigators to understand what barriers have failed and resulted in these tragic events. We have a very effective tool to analyse and present these findings through the barrier model, but, it did not prevent the event. Is this because Reason’s model was not applied, not understood, or limited in some other way?

I think there is a limitation of this model, which is this – it is targeted at avoidance of an undesirable event in isolation. To offer a simple analogy of football, if the only strategy was to avoid conceding goals, we wouldn’t be a very successful team. A holistic approach would have a strong defence, midfield and forwards. We would have clear strategies for developing the performance of the team as a whole, identifying barriers to improving overall performance, not just defending. We would have arrangements in place to reflect on performance and identify what areas require further improvement.

By adjusting our mindset to go beyond barriers to avoiding undesirable events happening, we would open up fresh thinking and new ideas, such as identifying barriers to Best Practice. By understanding the barriers to best practice we could strengthen our prevention safeguards and create a larger gap between daily operations and the risk of a major accident, and at the same time, providing significant leeway when organisational drift reduces performance over time.


Moving towards Best Practice

Moving towards Best Practice













Of course, it can be argued that this is just another way of presenting an ‘ALARP’ justification, but there is a difference. In our day to day management of operations, if we think ‘what can I do today that will make a shift towards better practice’, rather than the approach of ‘what do I need to do to meet a minimum standard’ then new opportunities arise. It might be something as simple as focussing on better quality risk assessments for work activities, or investing more effort in integrity management strategies, or looking at opportunities for the team to work more to its strengths.

During the design stage of major developments, how well do we really meet Inherent Safety in Design goals (rather than falling into the reverse ALARP trap)? Testing the design against best practices, and taking into account all the necessary trade-offs, will result in a safer design and a smoother passage through regulatory assessment.

Taking yet another view, Dave Brailsford of Team Sky provides a powerful example of moving towards best practice over a long period of incremental improvements, or as he called it;

the aggregation of marginal gains

To address the question raised in the title of this blog, Reason’s cheese model has been immensely successful in advancing preventive measures, and is perhaps not yet stale! But we need new metaphors to energise our thinking and to move to a different level of performance and we need alternative models to stimulate fresh debate on achieving best practice.







You may also like