The Sodium Fast Reactor (SFR), among Generation IV reactor concepts, has promising features such as its feasibility, scalability, and sustainability. However, despite these compelling technical attributes, the SFR design carries some less-desirable features, when considered from a nonproliferation perspective. In case of large-scale adoption, the SFR would inject into the global system an additional amount of special nuclear materials (SNMs), mostly plutonium, that, depending on some design choices such as the adoption of fuel blankets, would increase overall proliferation risks. In addition, the SFR is very often envisioned as, and in some cases relies upon, having its own fuel cycle. Most of the proposed SFR designs today come with a colocated reprocessing facility and a fabrication facility. The introduction of new facilities and the circulation of SNMs in forms and morphologies different from a traditional fuel assembly will also contribute to the increase of the overall proliferation risk.

The worldwide experience available for this technology suggests that its economic feasibility is related to different metrics such as plant operational availability, safety, and nonproliferation performance. To obtain good outcomes concerning this multigoal objective, a metric expressing the nonproliferation performance for the SFR energy system performance needs to be defined and to be related to other performance metrics. This paper describes our efforts toward this multigoal objective. We focus on design metrics for evaluation of the proliferation risks introduced by the SFR as well as the models, tools, and ultimately the design criteria that would be required in order to assess the proliferation risk. We follow the paradigm of nuclear safety and replicate the methods and the approaches used for risk-informing nuclear designs from a safety standpoint. The paper describes the characteristics required by a risk-informed framework for proliferation evaluation, the model used to conduct the assessment of the SFR, and an overview of the features that designers and regulators would need to relate proliferation performance metrics to cost and safety metrics.

A novel nonproliferation metric is proposed and described through a competition model between two actors: the energy system's safeguarder and a potential proliferator interested in acquiring SNMs. The method developed is analogous to that of plant reliability fault tress in that it includes the creation of an event and success tree models that start from basic events and lead to the top metric, measuring the proliferation performance. The method allows measuring the proliferation performance by means of a metric expressing the proliferator's success probability in acquiring SNMs. However, measuring the likelihood of a scenario is not enough if not coupled with the assessment of the consequences associated with its realization. For this, a new probability-consequences (P-C) curve is proposed for use in regulating the nonproliferation scenarios dealing with acquisition of SNMs.

The addition of new design options such as detection devices and of degraded materials as means to reduce the opportunities attracting a proliferator to divert the SNM are used as illustrative applications of the new framework. The examples use a reference design based on preliminarily proposed designs. Then the framework is implemented using security and safety criteria in a fashion consistent with recently proposed U.S. regulations, which are risk informed and performance based.

This proposed framework demonstrates a method for the quantification of proliferation risks and shows how it can be integrated with other design considerations. Ultimately, the framework proposed is a first step toward strengthening the International Atomic Energy Agency's current ability to protect nuclear energy systems from proliferation risks.