[SystemSafety] Historical Questions

[Matthew Squair]

>> When was the first regulation that required explicitly that *risk assessment* should be conducted?


Not strictly regulation (sorry). But I’m aware of the physicist George Compton’s narrative of how the risk of a runaway fusion reaction in sea water or the air was considered in the Manhattan project. He quoted 3 chances in a million as being considered ‘acceptable'. The actual declassified report LA-602 is determinedly safety factor oriented and not at all a probabilistic risk assessment. So perhaps a rhetorical device of George's. Good news though, you’d need a bomb of about a 1000 square meters volume (e.g real big) and even then Compton scattering would prevent a full runaway. There are also the classic operations research problems from WWII which introduced probability of survival/success as outcomes to be quantitatively assessed. Not strictly safety or regulations but very similar problems.


> On 9 Mar 2017, at 6:20 pm, Peter Bernard Ladkin <ladkin at causalis.com> wrote:
> 
> On 2017-03-09 07:18 , Drew Rae wrote:
>> When was the first accident report that found explicitly that failure to conduct a *risk assessment*
>> was a cause of the accident? 
> 
> I would say that that is, in general, mistaken causal reasoning.
> 
> When an operator fails to perform a necessary action, that is a reified non-event in WBA.  For
> example, failure to brake was a cause of the Berajondo derailment.
> 
> But how can there be a causal connection, according to the Counterfactual Test, from failure to
> perform an analysis to an accident event?
> 
> Try it. The CT statement: In the closest possible world to ours, had a risk analysis been performed,
> the accident would not have happened.
> 
> Generally not passed. Had a risk analysis been performed, *and been acted upon*, and *had all
> subsequent actions been conformant with the risk analysis*, then *maybe* the accident would not have
> happened. But, after all, a risk is a risk and it is consistent with any non-zero risk and
> concomitant action that the risk is realised, that is, the accident happened.
> 
>> When was the first regulation that required explicitly that *risk assessment* should be conducted?
> 
> Very interesting question!
> 
> PBL
> 
> Prof. i.R. Peter Bernard Ladkin, Bielefeld, Germany
> MoreInCommon
> Je suis Charlie
> Tel+msg +49 (0)521 880 7319  www.rvs-bi.de
> 
> 
> 
> 
> 
> _______________________________________________
> The System Safety Mailing List
> systemsafety at TechFak.Uni-Bielefeld.DE