What events qualify for inclusion in this database?
An event meeting any one (or more) of these criteria is included in this database:
The 1979 accident at Three Mile Island Unit 2 (Pennsylvania, USA) does not meet any of the qualifying criteria for listing here (see criteria above).
On 28 March 1979 operator error combined with deficiencies in monitoring equipment at the TMI-2 commercial nuclear reactor caused melting of reactor core material and buildup of gases within the containment structure. The containment structure performed as designed, containing core material and gases. Controlled gas releases were conducted over the next few days.
No acute radiation injuries occurred, and the evidence indicates no chronic radiation injuries resulted. Certainly no individual chronic injuries can be attributed to the TMI accident; even claims of statistically observable changes in possible chronic effects (such as cancer) are not consistent with the evidence. Maximum possible dose to any individuals near the plant was about 0.02 to 0.07 rem in the case of a few hundred individuals; this is well below natural background radiation levels. The findings of exhaustive epidemiological studies of the population near TMI are illustrated by Talbott et al., 2003, who found "no consistent evidence that radioactivity released during the nuclear accident has had a significant impact on the overall mortality."
During the TMI accident total radioactive releases were between 0.043 and 0.14 megacuries, almost exclusively noble gases (i.e. biologically inert). Total releases of radioactive iodine, a biologically active isotope, were about 15 curies (i.e. 0.000015 megacuries).
To sum up, the TMI accident resulted in significant property damage for the plant operator, but no injuries resulted.
Why isn't the Windscale accident included in this list?
The 1957 accident at Windscale Pile 1 (Cumbria, UK) does not meet any of the criteria for inclusion (see criteria above).
On 10 October 1957 a fire broke out in the core of the Windscale Pile 1 plutonium production reactor (Cumbria, United Kingdom), resulting in the release of radioactive core material. The fire was extinguished the following day.
No acute radiation injuries resulted (or would be, based on the size of the release). The radiation release was about 0.02 megacuries, mostly biologically active radioiodine. The British government took action to stop consumption of contaminated milk; these actions were apparently sufficient to prevent any demonstrable chronic radiation injury.
Why is the K-386 submarine accident no longer included in this list?
In 1976 an accident reportly occurred on the Soviet nuclear submarine K-386. Little information has been reported beyond an accident listing by the Bellona Foundation, citing this among accidents "causing radiation discharge" and citing a "main condenser breakage (2 persons injured to death)". Radiation injury is not specifically cited here or elsewhere as the cause of death. (Additionally, the designation K-386 has not been identified with a particular class of Soviet nuclear submarines in Bellona listings or elsewhere.)
Given that the condenser in naval nuclear propulsion systems is in the secondary water system, it appears more likely that the accident involved a leak of steam causing fatal burns or trauma injuries. Only the primary water system enters the reactor core and is significantly radioactive. Lacking further confirmation, this accident cannot be concluded to have involved radiation exposures sufficient to have caused injury.