Risk Assessment in Light of Radiaiton

[PLeonat]

Very interesting report that doesn't get lots of coverage!

[HowardE]

I read this and many similar studies and it all sounds so good, less than 0.1% that overall risk of death, such a low probability of the accident in the frst place, it sounds great. How about real life and real data points? How does he low calculated probability of the accident jive with what has already happened? Let's see, back of the envelope says, There are about 430 working reactors and if we assume hey have been running for 40 years that's 1800 rector years ( in reality it is less, but we ll be conservative). And we've had TMI, Chernobyl, and 3 cores melt at Fukushima. That's 5/1800 =2.8E-03 core melts per reactor year REAL occurrence - not just probability! And, of CORE MELT not just an accident! And tha doesnt account or "other" less severe accidents that have happened. How does that jive with all these published theories?

Instead of just posting these reports that make ou feel good, please, someone with expertise address that, please!!!!

[KimChin]

Howard, you may be a bit confused about the whole probabilities concept. I can't claim to be a statistician or risk expert, but when some event has a probability of occurrence of, let's say, 2 per 1000 years, it doesn't mean that it will happen once at 500 years and once more at 1000. The even, for example, could occur twice in the first 100 years, and no time in the next 900 years. It's like the odds in Vegas. For example, if slots have a 56% chance ow winning for the player, it doesn't mean that if one pulls the lever 25 times, 16 times it will be a win for the player. It's possible that there will only be 5 wins, but there may be 18 wins in the next 25 times and possibly 14 wins in the 25 pulls after that. It means that Given a "Large" number of pulls, on the average, 56% will be wins. Hopefully I didn't confuse the issue even more.

[HowardE]

What is it with people always trying to muddy-up the waters by making totally inappropriate parallels? So, what are you saying, Kimchin, that events and accidents that already took place do not and should not change our calculations of the likelihood of these "rare" events? Why can't the industry admit that they simply got it wrong and go back and see that they get it "right" instead of all this cr*p ?

[KimChin]

No, Howarde, that's not what I'm saying at all. I am saying that if en event has a calculated expected frequency on 1 per 1000 year and we started measuring today and the event took place in the first year, it's frequency of occurrence doesn't all of a sudden change to 1 per year, though that event may be used to adjust the calculated, expected frequency.