Understanding Uncertainties in Nuclear Archaeology

There is a lack of techniques to verify the completeness of baseline declarations of fissile material holdings. Such a capability is important whenever states declare existing fissile material inventories, e.g. as part of joining the Non-Proliferation Treaty and disarmament treaties such as Treaty on the Prohibition of Nuclear Weapons. One of the most promising approaches is nuclear archaeology, i.e data collection and analysis to reconstruct the past fissile material production history. Some of these data are collected from past operating records and declarations while other data are generated using different measurement approaches, e.g. measuring the activation of structural and moderator materials in reactors. For each of these sources, the data are affected by uncertainties. These uncertainties result in an overall uncertainty of the final assessment of material stockpiles.It is thus imperative for nuclear archaeology to be able to identify and understand these uncertainties, and then - to the extent possible - reduce them. This paper provides an overview of the possible sources of uncertainty in nuclear archaeology for plutonium production. Some uncertainties may be directly contained in records (e.g. from past measurements), others will be the result of incomplete information in the records. Uncertainties of measurements taken during inspections must be taken into account. In addition, to deduce past activities from such measurements, calculations and simulations are required that again bear uncertainties, for example from the use of compiled nuclear data sets (e.g. cross section libraries), or computational representations of physical realities. For the paper, a comprehensive list of uncertainties sources will be compiled. For each of these uncertainties, we will qualitatively analyze their impact on the final results and discuss to which extent possibilities to reduce them may exist.