24 Dec Tolerability of Risk
The ‘Tolerability of Risk’ (TOR) concept is one of the earliest and most misunderstood ways of demonstrating that risks are As Low As Reasonably Practicable (ALARP).
The TOR approach was developed by the UK nuclear industry to address the annual risk of fatality for individuals. If the probability is less than one in a million (i.e. 10-6) it is thought to be broadly acceptable but if it is higher than 1 in 1,000 (1 in 10,000 for members of the public) it is considered as being intolerable and must be reduced whatever the cost. It is important to note that the approach:
- Has no basis in UK law but provides a useful yardstick when considering high-risk activities
- Is for individual risk – it is meaningless to apply the approach for ‘collective’ or ‘societal’ risk since larger systems are inevitably more risky
- Applies to the most exposed person, not the average risk to passengers or workers
- Applies to fatalities only – injuries or weighted injuries are not included.
People are attracted to the technique because it defines intolerable and broadly acceptable levels of risk, and then try to apply it to collective risk. For example, I have seen the thresholds being used to set safety targets by multiplying them by the number of workers/passengers on a system. This is fundamentally wrong (creating a very ‘soft’ target for fatalities and a very misleading one if weighted injuries are included). The TOR approach should simply be used to ensure that individual passengers or staff are not exposed to intolerable levels of risk. This is, however, extremely unlikely since the intolerable level for workers is the sort of risk that trawler men and deep sea divers are exposed to. To illustrate the point, the risk of fatality for commuters (the most exposed passengers) is approximately 3.7 x 10-6 per year in the UK, which is in the bottom half of the Tolerable region.