Nuclear Archaeology
Integrated Approach
As of today, two approaches to reconstructing past fissile material production exist: First, concepts for measurements in shut-down facilities and of radioactive waste have been suggested to gain information about the past. However, there are cases where using only measurements is insufficient for reconstructing the production history.
Second, documentation of past nuclear facility operations can be used as input to simulation tools to quantify the produced fissile materials. However, information from such records will have uncertainties, it might have been altered, or it may be missing altogether.
We propose to expand this toolbox by combining both approaches for an integrated assessment. Taking data from various sources into account allows for consistency-checking. Furthermore, using Bayesian inference and numerical methods, more reliable estimates of operational parameters can potentially be obtained.
Archaeology with Reactors
Information about the past operation of reactors can be obtained by examining changes in their structural elements. Components such as fuel tubes are subjected to irradiation with neutrons, leading to activation of the material. Measuring their isotopic composition enables drawing conclusions about the history of neutron flux, and in turn about the reactor operation producing this flux.
We want to obtain a detailed picture of a reactor's operation by analyzing multiple isotopes in parallel. A special focus lies on unstable isotopes, which carry information about, e.g., reactor power or shutdown periods.
Along with the possibility for more detailed information comes an increased difficulty of the analysis process. Since a direct inversion of the processes in a reactor is impossible, we must rely on repeated forward simulations. We can then apply numerical methods to develop an optimization algorithm, matching the input parameters to the measured ratios.
Archaeology with Radioactive Waste
Information about the past operation of reactors can be obtained by measuring the isotopic composition of the produced radioactive waste. This history is characterized by a group of parameters, such as fuel burnup and time after irradiation. Reconstructing these parameters is an inverse problem.
We solve this problem with a Bayesian approach: first, we create initial parameter estimates based on, e.g., public information. Then, we combine these estimates with reactor simulations and isotopic ratio measurements in the Bayesian framework to obtain estimates of the total plutonium production of a given reactor.
Our reconstruction results can be used to verify baseline declarations and possibly detect undeclared production activities that could jeopardize disarmament efforts.
This research is funded by a FREIGEIST-Fellowship of the VolkswagenStiftung and the Deutsche Stiftung Friedensforschung.