Simulated State accounting data for development of
advanced and automated State-level safeguards analysis

In compliance with Comprehensive Safeguards Agreements (INFCIRC/153(corr.))[1] and Additional Protocols (INFCIRC/540)[2], States Parties to the NPT are obligated to report to the IAEA all changes in their nuclear material inventory and movement of nuclear material across boundaries of IAEA recognized material balance areas (MBAs). Based primarily on these State nuclear material accounting reports, the IAEA plans and conducts safeguards verification activities, including on-site inspections, audits, measurements, and deployment of various safeguards equipment, to detect and deter proliferation-related noncompliance. The overarching issue addressed by this project is to ensure that data analysis capabilities are in place to detect irregularities in State accounting reports, thus ensuring their accuracy and completeness—and in the broader context, States’ compliance with safeguards obligations of the NPT. At a primary level, State declarations to the IAEA can be only complete or incomplete, and either correct or incorrect, whether the reason for mismatch is intentional or an inadvertent technical or human error. This project demonstrates how analysis of dynamic correlations in nuclear material movement within the entire fuel cycle of a State (viewed as a single virtual process) can reveal irregularities consistent with and potentially indicative of clandestine proliferation activities. Using this concept of “cadence of operations” analysis, we have modified the Cyclus nuclear fuel cycle simulator to produce State reporting data reflective of individual MBAs and compatible with Code 10, the formal reporting format used between the States and the IAEA. The resultant realistic fuel cycle simulations of a State produce synthetic high-fidelity State declarations, which can then be subjected to various data analytical approaches to test sensitivities to spot different types and magnitudes of disruptions. These could be either benign reporting mistakes or results of deliberate deception. The ability to analyze dynamic correlations in declared nuclear material movement across and within fuel cycles of States under nuclear safeguards enables the detection of mis-declared or undeclared activities, which could indicate clandestine proliferation.