Jim Green - 1998
There have been at least three fatal research reactor accidents:
Almost certainly there have been other fatal accidents. Indeed the Australian Academy of Science, in written evidence to the Senate Nuclear Reactor Inquiry (Senate Economics References Committee), says there have been five fatal research reactor accidents yielding a total of seven corpses.
The Research Reactor Review (RRR, p.13) referred to a 1980 report by the US Oak Ridge National Laboratory (ORNL) which listed nine serious accidents involving prototype power reactors or experimental reactors, and a further three involving multipurpose research reactors (Bertini et al., 1980). The multipurpose research reactor accidents were two accidents involving Canadian research reactors (discussed below), and a fuel element melting at the ORR reactor at the ORNL in 1963. None of these three accidents resulted in immediate deaths but the longer-term effects are disputed. The Review (p.13) went on to say that there has been only one further report of a research reactor accident since 1963 - the above-mentioned fatal accident in Argentina in 1983.
ANSTO (RRR Submission) claims that: "Excluding experimental and prototype reactors, only four accidents designated as serious have occurred in multi-purpose research reactors. No off-site consequences were identified in any of the four cases."
This discussion is somewhat complicated by semantics. ANSTO claims that there has never been a fatal research reactor accident (part of its broader, preposterous argument that no research reactor has ever adversely impacted its community). ANSTO's argument rests on a sharp distinction between research reactors and test/training/prototype reactors. But ANSTO (1993) is itself on record defining research reactors as encompassing a range of facilities from subcritical assemblies to prototype power reactors.
The argument was put to the RRR that accidents are more common in multipurpose research reactors than in power reactors because there are more frequent start-ups, shut-downs, fuel and rig movements, and more opportunities for human error. These arguments were drawn from industry literature and so could not easily be refuted. On the other hand there is no dispute that, in general, accidents involving power reactors pose a far greater risk to the general public because of the far greater volumes of fissile material used to fuel power reactors. (RRR, 1993, p.228.)
Two accidents in Canadian reactors warrant mention. The National Research X-perimental (NRX) reactor was built in 1947. It was essentially a pilot factory for the production of plutonium, which was supplied to the US until 1963. NRX was involved in an accident in 1952. A power excursion destroyed the core of the reactor, causing some fuel melting. The core of the reactor was buried as waste. Hundreds of US and Canadian servicemen were ordered to participate in the clean-up. (Edwards, n.d.) According to ANSTO (1993B, pp.3.16-3.17), the accident led to a significant release of radioactivity, but there were no reported injuries. Given the acknowledgment that there was a significant release of radioactivity, ANSTO's claim that "no off-site consequences were identified" should by no means be taken as meaning that there were no off-site consequences.
The Canadian National Research Universal (NRU) reactor first went critical in late 1957. In 1958 there was a fire in the reactor which badly contaminated the inside of the reactor building with some release of radioactivity outside the building. (ANSTO, 1993B, pp.3.16-3.17.) Several fuel rods overheated and ruptured, one catching fire. The ventilation system was jammed in the open position, thus allowing the spread of radioactivity down-wind from the reactor site. The burning fuel rod was extinguished by a relay team of scientists and technicians running past the maintenance pit and dumping buckets of wet sand on it. Over 600 men were involved in the clean-up. Atomic Energy of Canada Ltd. (AECL) claims that very few men were exposed to radiation doses exceeding the then permissible levels. It also claims that no adverse health effects were caused by the exposures received. The methodology for this second conclusion appears to have been the ostrich technique: no follow-up studies were carried out, the men involved in the clean-up were told to observe strict secrecy about the operation, claims that adverse health effects were linked to the clean-up were vigorously denied, and AECL has refused to supply information that would assist in the location of men involved in the clean-up and thus facilitate follow-up studies. (Edwards, n.d.)
ANSTO, 1993, Submission to the Research Reactor Review, Attachment A, Working Paper 3, p.3-2.
ANSTO, 1993B, Submission to the Research Reactor Review, Attachment A: "Research Reactors: Local and International Experience", Working Paper 3: "International Experience with Research Reactors".
Bertini, H.W., et al., 1980, "Descriptions of selected accidents that have occurred at nuclear reactor facilities" Springfield: NTIS.
Edwards, Gordon, (Canadian Coalition for Nuclear Responsibility), n.d., Reactor Accidents at Chalk River: The Human Fallout.
Research Reactor Review, 1993, Future Reaction: Report of the Research Reactor Review, Sydney : Wentworth Press.
Vallentine, Jo, 1992, Let the Facts Speak: An Indictment of the Nuclear Industry, Perth : Office of Jo Vallentine, Senator for the Greens (WA).