I would like to make you understand the notion of ‘safe’ related to small probability values (“After Japan,” Sept. 15).
Does small probability not happen? The Fukushima accident had a probability of (Unit 1 nuclear accident/year probability) 10-6 x (Unit 2 nuclear accident/year probability) 10-6 x (Unit 3 nuclear accident/year probability) 10-6 x (Unit 4 nuclear accident/year probability) 10-6. That means 10-24 nuclear accidents/year.
Is this a very small probability to happen? Yes. But it happened.
The main reason that happened is all operating authorizing are based on safety analysis for each reactor unit, not for all reactor units. It is missing a safety analyses for common events for all reactors units.
Some common events such as earthquakes require a greater than design value; tsunamis are not analyzed as common events for all reactor units, but only for each reactor unit only.
The ‘safe’ notion you understood before is different if you consider the above.
Constantin Robitu
