Imaging and material identification of nuclear waste with muon scattering tomography

Nuclear waste currently created and placed in interim storage is expected to be fully known and characterised, with records kept and regularly checked, per IAEA regulations. However, there is also historical nuclear waste that was created at a time when these records were not required and presumably...

Full description

Bibliographic Details
Main Author: Frazao, Leonor
Other Authors: Velthuis, Jaap
Published: University of Bristol 2019
Online Access:https://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.767989
Description
Summary:Nuclear waste currently created and placed in interim storage is expected to be fully known and characterised, with records kept and regularly checked, per IAEA regulations. However, there is also historical nuclear waste that was created at a time when these records were not required and presumably the problem of nuclear waste disposal was not such a concern as it is today. Not only the initial materials stored may not be known, but they may also have undergone changes, such as the oxidation of uranium that produces hydrogen gas. This brings a demand for techniques to characterise nuclear waste that both make sure that its records are current, and that historical waste can also be fully characterised and properly taken for final disposal or long term storage. Muon scattering tomography is a technique that can be used for this purpose. It consists of measuring individual cosmic-ray muons before and after they cross the volume of interest, and obtaining the angular distribution of their scatter and related variables. The width of the angular distribution is larger for materials with higher atomic numbers, so it allows for high-Z materials to be found in concrete. Several methods were developed that, when combined, can give a description of the contents of nuclear waste. This starts with an imaging algorithm that can first find lumps of high-Z materials and then detect the edges of these materials with a good precision. The same algorithm can also be used to determine the amount of gas present in the containers. This thesis shows that this algorithm can measure the length of uranium blocks in concrete with a resolution of 3.2±0.6 mm when not using momentum information, for lengths down to 5 mm. A resolution of 0.98 ± 0.03 mm was obtained when including the muon momentum, for lengths down to 2 mm. In a following step, high-Z materials can be identified to verify if they come from nuclear fuel (uranium and plutonium), or if they are other materials such as lead and tungsten. It is shown that the distinction between uranium and lead or tungsten is possible for block sizes down to a cube of 2 cm side, with data taking times up to 70 hours. Some discrimination between uranium and plutonium was also obtained, for 3 cm side cubes and requiring more data, corresponding to 200 hours.