D.J. Sweeney “Nuclear Weapons Latency”, Ph.D. Dissertation, Nuclear Engineering, Texas A&M University, College Station, TX (2014).
A novel nuclear weapons proliferation assessment method has been developed to determine a state’s Nuclear Weapons Latency, the expected time to be taken by a non-nuclear weapons state to develop a conventionally deliverable nuclear weapon given the state’s position on a path toward or away from a nuclear weapon and accounting for the state’s motivations and intentions. Potential proliferation time is taken as a representation of the latent proliferation capacity of a non-nuclear weapons state. An assessment of proliferation time is critical to crafting an effective policy response within a useful time frame. Current proliferation assessments either neglect proliferation time or are static case-specific assessments frequently built on restricted information and opaque assumptions. The Nuclear Weapons Latency computational tool has been developed to determine a state’s Nuclear Weapons Latency and embodies a stochastic Petri net proliferation simulation. The tool makes only three simple assumptions: a decision to proliferate has been made, the proliferation pathway network is known, and the associated pathway activity times are estimable. Beyond the quantification of a state’s latency, the tool provides a transparent, efficient, adaptable, and highly repeatable platform which allows for extensive sensitivity analysis to better inform the nonproliferation discussion and policy decisions. Functionality of the tool was verified and inherent sensitivities determined through historical analysis with the U.S. case of proliferation in the Manhattan Project. Network and operational parameters were found that drove expected Latencies high while others increased the Latency distribution variance. Further confidence was built with historical analyses of the Pakistani and South African cases of proliferation. These verifications were done in lieu of experimental validation which is impossible for future event simulations like the Latency tool. Analysis revealed that while A.Q. Khan altered the Pakistani proliferation pathway, his impact on proliferation time may have been minimal. A Multi-Attribute Utility Analysis (MAUA) function was implemented for proliferation pathway selection. This function might increase the accuracy of the most-likely Latency estimate in certain cases. However, use of MAUA for adversary modeling also significantly increased the number of assumptions necessary. A Latency investigation of South Korean nuclear fuel cycle facility development, a current nonproliferation policy concern, demonstrates how Nuclear Weapons Latency can help characterize the proliferation risk of different policy options for decision makers. Analysis showed that development of any one of pyroprocessing, PUREX, or especially commercial uranium enrichment technologies reduces South Korean Latency. This risk characterization ability through policy option sensitivity enables the Latency tool to help fill a void of useful proliferation risk information provided by technical assessments to policy makers identified by the 2013 National Academies study Improving the Assessment of Proliferation Risk of Nuclear Fuel Cycles.