Re-orienting human reliability in safety critical systems: from ALARP to AHARP

Major hazard industries are increasingly expected to demonstrate what action has been taken to mitigate the risk of human error both by the choices made during the design of facilities, and in the way they organise and manage operations. In some countries, “safety cases” are expected to include a demonstration of how the risks of human error have been reduced to a level that can be shown to be ‘as low as reasonably practical’ (ALARP). Increasingly, companies are imposing their own internal requirements not only to produce safety demonstrations, but for them to include a demonstration of how the risk of human error has been mitigated both by engineering design and by operational controls.

The need to provide a human error ALARP demonstration inevitably focuses attention on the ways people can make mistakes or otherwise represent risk.  This an inherently negative view of the role of people in systems.  And it is a view that is out of step with the real role and value that the people in systems bring not only to safety and environmental control, but to production.  People are probably the most resilient and robust element of any socio-technical system, capable of immense feats of adaptation, working round problems and finding creative solutions to novel, unforeseen and unexpected difficulties.  Rather than seeing the human as a liability or a risk to safety and integrity, it is the very ability of people to cope and adapt to the unexpected that is so often relied on to maintain safety. 

Organisations that are serious about running their operations in such a way that no one gets injured, everyone goes home safely and there are no spills, can make a major step towards those aspirations by seriously challenging their own expectations of human performance.

The presentation will argue that process safety can be improved by a re-orientation away from focusing on human performance as a risk (i.e. from focusing on reducing the risk of human error to ALARP) and towards viewing human performance as an inherent requirement for safety and productivity. Rather than concentrating on what has been done to reduce the risk of human error to ALARP, effort should focus on what has been done to ensure human reliability will be ‘as high as reasonably practical’ (AHARP). The focus should move away from focusing on how people can defeat controls and towards assuring controls will actually deliver the standards of human performance that are expected and intended, and that controls that rely on human performance will be as effective and as reliable as can reasonably be achieved.