The Korea Institute of Nuclear Nonproliferation and Control (KINAC) is developing a simulation model to estimate nuclear material production. This model is a foundational technology in interpretation and evaluation in preparation for denuclearization verification. Through this model, it is possible to estimate the amount of nuclear material that can be produced based on information on the activities of facilities related to the nuclear fuel cycle in the actual denuclearization verification stage. This model makes it possible to determine whether the declared amount of nuclear material is reliable. In addition, the reliability of the reported information can be confirmed through on-site inspection. However, there is a possibility that proliferation-related activities cannot be detected even through this inspection, and a normal state may be misdiagnosed as carrying out nuclear proliferation-related activities. Therefore, it is unreasonable to specify activities related to nuclear proliferation with only one inspection. Since each inspection method has its diagnosis rate and false diagnosis rate, measures such as repeating the same inspection method or combining different inspection methods are required to detect activities related to nuclear proliferation reliably. Therefore, a model capable of estimating the number of repetitions to obtain a reliable nuclear activity detection probability was developed by using each inspection method’s diagnosis rate and false diagnosis rate as input information through a Bayesian inference method. Through this model, it can be concluded that repetitive inspections increase the probability of detecting nuclear proliferation-related activities. This approach confirmed the possibility of repeatedly breaking away from the high-intensity inspection method that causes political and diplomatic resistance from the target country and substituting it with a more readily acceptable, low-intensity inspection method.

Concerns about North Korea’s 7th nuclear test have been rising recently, and it is a significant threat to the situation around the Korean Peninsula. Amidst these threats, the Korean government also shows a strong will for denuclearizing the Korean Peninsula, referring to the “Audacious Initiative.” For denuclearization negotiations with North Korea, it is essential first to understand North Korea’s nuclear capabilities. However, since access to information is complicated and contains many uncertainties, many studies have been conducted to estimate it. Among them, Von Hippel surveyed to estimate the total amount of uranium ore based on information on uranium mining, which is relatively widely known throughout North Korea’s nuclear fuel cycle, and the amounts of HEU and Pu suggested by many experts. KINAC has conducted a study on a methodology that can narrow the estimation range and improve reliability through the Bayesian Network based on Von Hippel’s research results. However, in this study, the probability distribution is assumed to be the simplest form of uniform distribution, and the estimation formula for the amount of Pu produced compared to the amount of uranium loaded in the core is used as it is, which is an error in Von Hippel’s study. Improvement is needed. This study proposes a more reliable BN model by supplementing this and attempts to estimate the amount of uranium ore that North Korea produces or possesses. Of course, the data used as the basic structure of the model is insufficient, and the estimation formula is straightforward, so it is somewhat unreliable to trust the estimate for uranium ore. However, it is expected to be a suitable methodology that can narrow the scope of North Korea’s nuclear material production estimate or compensate for the uncertainty of the nuclear material production estimation model being developed at KINAC.

Bayesian statistics, which is an approach to analyzing data based on Bayes’ theorem, is currently widely used in all fields. However, it has been applied very limitedly to studies related to nuclear nonproliferation. Therefore, this paper provides a knowledge base and directions for using various Bayesian techniques in nuclear non-proliferation. First, the concepts and advantages of the Bayesian approach are summarized and the basic solving methods of Bayesian inference are explained. The Bayesian approach enables more precise posterior estimation using the prior probability and the likelihood functions. To solve Bayes’ theorem, it is necessary to use the conjugate prior distribution, which is analytically solvable, or to use a numerical approach with computing power. Next, for several Bayesian statistics methods, the purpose of use and the mathematical derivation process are described. Bayesian linear regression analysis aims for obtaining a function that outputs the closest value to data of variables and results. Factor analysis is mainly used to derive a smaller number of unobserved latent variables that can represent observed variables. The logit and probit model are nonlinear regression models for when the outcome is binary. The hierarchical model is to analyze by introducing hyper-parameters in an integrated manner when there are several groups of similar data. The Bayesian approach of these methods is generally based on the numerical solution of the Bayesian inference of the multivariate normal distribution. Finally, the previous researches that each introduced method have been applied to nuclear non-proliferation are investigated, and research topics that can be applied in the future are suggested. Bayesian statistics have been mainly used for precise estimation of the amount, location, and radioactivity spectrum of nuclear materials using detectors. Using Bayesian approach, it will be possible to perform various analyzes. For example, the change of activeness of nuclear program can be estimated by Bayesian inferences on the frequency and scale of nuclear tests. And it can be tried predicting the production of plutonium according to the core configuration and burnup using the Bayesian linear regression. Also, by introducing the Bayesian approach to factor analysis or logit analysis of nuclear development motives or nuclear proliferation probability, it can be expected to improve precision. With the development of computer technology, the use of Bayesian statistics increases rapidly. Based on the theory and applied topics summarized in this paper, it is expected that Bayesian statistics will be more actively used for nuclear non-proliferation in the future.

KINAC is trying to build a comprehensive aerial view of the nuclear material balance to predict North Korea’s weapons-grade nuclear material production capacity. We are creating a visualization model for North Korea’s nuclear facilities as part of these efforts. However, information on North Korea’s nuclear facilities is scarce, and it is not easy to consider additional facilities other than those already known. In addition, in the case of a model that targets only exceptional situations, it is not easy to secure objectivity for model validation, so it is necessary to upgrade to a general-purpose analysis tool that can be applied more generally. The following two examples are proposed as an analysis tool that can be a high degree of analysis. The first case is an Acquisition Path Analysis (APA) utilized to introduce IAEA’s State-Level Approach (SLA). The acquisition path analysis aims to find and evaluate the technically possible pathways to obtain nuclear materials for nuclear weapons or other nuclear explosive development. It can be an acquisition route if it is possible to produce at least 1 Significant Quantity (SQ) of weapongrade nuclear material within five years. The assessment of technologically feasible pathways is based on available information about the country’s past and present nuclear cycle capabilities. The second is the IAEA Physical Model. The IAEA Physical Model was carried out to introduce a comprehensive approach to all information on a country’s nuclear activities. It describes and characterizes the technologies and processes expressed at all levels of the acquisition path, depending on the development objectives. The IAEA Physical Model attempts a multi-tiered acquisition path analysis to identify all known technologies and processes in the nuclear fuel cycle, from raw material production to weapon usable material acquisition. Based on this analysis, the IAEA evaluates the signs of nuclear proliferation in a specific country. Based on the two cases discussed above, we intend to derive the following implications and priorities for extending the existing nuclear cycle model to a more general-purpose for a specific country. First of all, the requirements necessary to evaluate nuclear non-proliferation or verification of denuclearization must be at a level that the international community can recognize. In the stage of actual denuclearization verification, since verification will be conducted through the IAEA, a corresponding level of tools and technology will be required. From this point of view, the following is presented as a prerequisite for adding versatility to the existing physical model: It is necessary to derive all processes related to the nuclear cycle and standardize relevant indicators and data. In order to determine the signs of nuclear activity, detailed information on technologies, materials, by-products, and wastes, which are essential for each process, is required. For denuclearization verification, cumulative information from the past to the time point is required, and a comparative analysis of the operation history information of all facilities and the amount of nuclear material is required. To this end, it is necessary to make it possible to trace the history at every point where it can be determined that nuclear material has been diverted so that missing nuclear material can be found. Based on this, it is expected that it can be possible to evaluate a hypothetical threat state, but it is also expected that it will be easy to verify the model through the evaluation of easily accessible domestic facilities.