🚀 go-pugleaf

Murray Love wrote:
> 
> "D.A.Kopf" <dak@dakx.com> wrote in message
> news:393D6C16.AB65AC41@dakx.com...
> > Donald Borowski wrote:
> > >
> <snip>
> 
> > However, it seems reasonable to me that individual risk-taking
> > should not imperil the lives of others. Criminal law seems to
> > agree, by allowing punishment for involuntary manslaughter.
> > If I fell off a roof onto someone, well, I'm bound to say I'd
> > feel responsible for their injury. If a nuke that I designed had
> > an accident that caused injuries, I'd feel uncomfortable arguing
> > that I'm not reponsible becaused I used accepted engineering standards.
> 
> This is a quandary of any industrialized society.  Substitute "chemical
> plant", "gas flares", "car", "toaster", or "space shuttle" for "nuke", and
> you'll see what I mean.  Pretty much all engineered items carry a risk, and
> that's OK if society decides they're worth it.

Although individuals can decide whether to use cars, toasters, or (here in
North Carolina) collapsing walkways, while these other risks are forced upon
them. It thus seems to me their acceptance is properly a political, not an
engineering, decision.
> 
> By the way, the formula used to evaluate risk is:
> 
> Risk = Probability x Consequences

I understand expectation values pretty well, but differentiate between known
probability distribution functions ("One in a million will win the lottery")
and calculated pdf's ("Chances of failure are one in a million"). The former
allows an objective decision to be made, but the latter involves subjectivity
until such time as sufficient real-world statistics are accumulated. I
personally use the time-honored 5% estimate in these cases - one in a hundred,
one in a million, one in a billion, they are all one in twenty as far as I'm
concerned (If I recall correctly, it was the twentieth shuttle mission that
blew up).

Other people can of course choose a different a priori threshold probability,
but if this is done, then for all lower-probability events, relative risk
becomes the same as relative consequences.

> 
> An airplane flight, for instance, is low-risk, even though the consequences
> of an accident are drastic--because the probability of an accident is very
> low.  Roofing (to amuse wmbjk) is relatively high-risk, because both
> probability and consequences are significant.
> 
> Now, by this measure, nuclear energy is low-risk (if we're talking about
> LWR's and CANDU's).  Anti-nuclear advocates get a long way by stating the
> _consequences_ of a nuclear accident, while ignoring the probability of one.
> One Canadian group quoted verbatim from an AECB report about the possible
> consequences of an accident during transportation of MOX fuel in Canada
> (AECB is our nuclear regulatory body).  Unfortunately, the anti-nukes
> ignored all the qualifiers; for instance, the containers (made of black-box
> material) would have to be completely vaporized in a truck explosion,
> releasing all the plutonium into the air, within a short distance of an
> urban centre, with a strong wind blowing directly towards the centre, during
> a temperature inversion (to keep the smoke plume close to the ground),
> assuming that no evacuation measures were taken, and that everyone inhaled
> enough plutonium to cause cancer within 10-40 years.  They also neglected to
> quote the report's estimate of the probability of such an accident:  I
> believe it was 1x10^-9, or one in a billion, though how they got THAT
> number.... I can dig up the reference if anyone's interested.
> 
> Whew.