Year
2000
Abstract
The September 1999 nuclear criticality accident at the JCO, Co. plant in Tokaimura, Japan has graphically demonstrated that serious radiological impacts to the public can result from criticality accidents. This observation has implications both for nuclear safety and for security. One major concern is that terrorists possessing relatively small quantities of special nuclear material (SNM) could induce a deliberate criticality event in a homemade apparatus which would remain undetected for days or even weeks, while causing serious radiation injury to many individuals within a large area. The Tokaimura accident was caused when 16.1 kilograms of 18.8%-enriched uranium in nitrate solution was placed in a vessel designed to hold no more than 2.4 kg. Because the vessel was cooled by a water jacket (which also acted as a neutron reflector), heat-induced disassembly did not occur and the resulting chain reaction continued for twenty hours before it was halted by active intervention to drain the jacket. Thousands of individuals were exposed to neutron and gamma radiation during the event and hundreds of millions of dollars in damages were incurred. The quantity of uranium that triggered the accident (3.03 kg U-235 contained in uranium enriched to 18.8%) is currently considered to be Category III according to the Convention on the Physical Protection of Nuclear Material (or SNM of \"low strategic significance\" according to the U.S. Nuclear Regulatory Commission) - which is the SNM category requiring the least stringent physical protection. Moreover, by designing a critical system appropriately, a malevolent actor could initiate and sustain a chain reaction with a considerably smaller amount of U-235 contained in uranium of the same or lower enrichment - even commercial low-enriched uranium could be used. Given the potential severity of such an event, it is prudent to reconsider both safeguards and physical protection requirements for nuclear materials currently considered to be of \"low strategic significance,\" but which could potentially be used for construction of criticality devices for radiological terrorism. In this talk, we provide estimates of the worst-case consequences of this threat and the minimum credible amounts of nuclear material that could be used for this purpose, and recommend further action.