In 2009 we have reached a total of 13,000 years of use. We should have got 13,000 * 5 / 100,000 = 0.65 accidents (or much less because recent reactors are more secure)
Number of accidents with fuel melting that I have found :
- one in the US. (1979 TMI-2 Three Mile Island )
- one in England. (1957 Windscale )
- one in Ukraine. (1986 RMBK Tchernobyl )
- one in Russia. (1977 ABM200 Beloyarsh)
- one in Japan. ( 2011 BWR Mark 1 Fujushima)
- two in France.(1969 1980 Centre de St Laurent du Cher)
7/.65= 10 times the probability given. The 13,000 years include also the most secure nuclear installations.
That is why I believe we have got to find a solution that works in the worst case: the loss of everything but a reactor about to explode in its confinement : no staff, no electricity, no cooling (no water, no air),even the gravity might not be completely usable because the reactor might lose its horizontality.
If we decide to achieve such a security target then we might find that nuclear corporations are a good investment. Oil price would rise accordingly.
They still need to work a little, to involve a lot of people in the research.