The last couple of posts have been on assessing whether a risk is ALARP. The ALARP definition in AS 2885 includes the concept of “grossly disproportionate”. Which means what?
The UK HSE website is useful again, particularly this page, which includes this graph:
It shows a “proportionality factor” (PF) increasing from 1 to 10 as the risk level changes from “Broadly Acceptable” (BA) to Intolerable. Of course this is not gospel, just a guideline, but it’s quite helpful. It gives you a number to put into the Maximum Justifiable Spend (MJS) formula in the previous post.
Translating the risk scale into AS 2885 terms, “Broadly Acceptable” means a Low or Negligible risk rank in the AS 2885 matrix (tolerable risk, hence PF not applicable or zero). “Intolerable” means High or Extreme (unacceptable risk, hence PF is infinite). Of course under AS 2885 the whole ALARP exercise is necessary only for Intermediate risks, so all this is fits together nicely.
If your risk is Intermediate but at the lower end of the Intermediate band then you could use a PF value of perhaps as low as 1. But if it’s at the upper end of the Intermediate range then a PF of 10 is more appropriate. Having said that, I’ve found that I tend to use a high value for risks that have a catastrophic severity, regardless of where the risk falls in the Intermediate band, just out of conservatism.
At one stage I dabbled in F-N curves to help me estimate PF. F-N curves plot the frequency of occurrence against the number of fatalities and can include curve for risk acceptance criteria. But with hindsight it seems unnecessarily complicated for estimating a parameter that is only used to support a judgement. If anyone really wants details then let me know.
Next will be some background on estimating frequency.