DEVELOPING DELIBERATE REACTIVITY INSERTION SOURCE TERMS FOR
REGULATORY APPLICATIONS

The Nuclear Regulatory Commission (NRC) is developing alternative physical protection requirements for new nuclear power reactors as part of its effort to "risk-inform" its licensing and oversight processes to address the enhanced safety and security characteristics that some assert the next generation of reactors will exhibit. These alternatives include a "limited scope" rule to provide a mechanism for exempting new reactor licensees from certain requirements that apply to the operating fleet, such as maintaining a minimum number of armed responders capable of preventing a design basis threat (DBT) adversary from causing radiological sabotage of the reactor(s). In parallel, the Part 53 rule under development would allow any new reactor to be entirely exempt from the requirement to protect against the DBT. To qualify for the exemptions in either case, applicants would have to provide analyses to demonstrate that postulated security-initiated events would not lead to exceeding certain dose limits to members of the public. NRC has not yet finalized the process for determining specific events to be analyzed. However, for certain reactor applications, it may be necessary to show that even adversaries with full access to the reactor could not "break" it by any available means prior to being neutralized by a local law enforcement response. Depending on reactor and facility design, sabotage attacks that cause rapid reactivity insertions may be plausible and could pose challenges to meeting dose limits. However, assessing the consequences of such events will be difficult because they cannot be experimentally validated today in real-world integral tests. Fortunately, an experimental database of transient tests resulting in core damage and radionuclide release exists from early programs such as BORAX and SPERT at the National Reactor Testing Station and the nuclear rocket and ramjet tests at the Nevada Test Site. However, in most of those tests the cores were unirradiated before the transients, which would not generally be the case for a sabotage attack on a power reactor. This paper will provide some observations about how the historical data may be able to inform sabotage analyses for regulatory applications.