Speaker
Description
The presence of water in a spent nuclear fuel (SNF) analogue has been determined by measuring the prompt 2.223 MeV $\gamma$-ray produced by neutron capture in hydrogen (or deuterium production). These measurements have been made using a 1” diameter Stilbene scintillator, an EJ-309 liquid scintillator (both from Scionix) and a GR5021 reverse electrode germanium detector (REGe) from Mirion. The detectors from Scionix have been used for initial proof of concept measurements before the REGe detector was set up. The basis of this work focuses on the UK’s large inventory of spent advanced gas cooled reactor (AGR) fuel, the majority of which is currently in wet storage. Water ingress into spent fuel assemblies due to fuel cladding failures presents a risk of instability for long term dry storage. Water present and the subsequent radiolysis may lead to, and latterly accelerate, corrosion of the cladding, resulting in an additional radiological hazard. There is also a limited risk of conditions suitable for criticality arising within the storage vessel. Early, non-destructive detection of water to take proactive, preventative action is necessary for appropriate measures to be taken. Prompt gamma activation analysis (PGAA) is a promising method for the measurement of water in SNF due to the highly penetrative nature of ~2 MeV -rays. PGAA also presents an opportunity to detect water in larger fuel masses which may have been wetted. These include fuel containing materials such as those at Chernobyl and Fukushima where the (n, $\gamma$) reaction may be induced by the intrinsic radioactivity of these materials containing large quantities of fission products. Simulations have been performed using the simulation toolkit Geant4 and compared to other work which utilised MCNP6 for validation of the Geant4 physics in use. Geant4 simulations comprised two main focus areas: detection of water in a AGR fuel storage element and selection of a suitable analogue for spent nuclear fuel. For the former modelling work, an AGR fuel assembly was constructed in Geant4. Different configurations of flooding the annular space within the fuel elements, space between the fuel pins and both empty volumes were simulated. In these simulations, limits of detection were similar to that observed in the previously mentioned MCNP6 model. For experimental work, a fuel assembly composed of SNF is not feasible for use in the laboratory. Subsequently, an analogue for SNF is required to build mock fuel assemblies. Calculations informed the selection of potential fuel analogues which were then defined within Geant4 and exposed to a simulated 252Cf source. The neutron attenuation was measured and materials comparable to SNF for AGR pellet dimensions were zinc oxide, molybdenum oxide, tin oxide, and 316L stainless steel. In this work, experimental validation of the results obtained in the previously discussed simulations has been carried out. Targets composed of zinc oxide and 316L stainless steel were placed in front of a neutron source and the neutron attenuation was measured for comparison to the simulations. These targets were produced by pressing powders of each material resulting in rectangular solid masses with a thickness of 14.5 mm, equivalent to an AGR fuel pellet. Results using the Stilbene and EJ-309 detectors for targets with thickness equivalent to the diameter of AGR fuel pellets were agreeable with the results observed in the simulations. For measurement of the 2.223 MeV $\gamma$-ray, stearic acid was added as a water analogue due to it possessing a hydrogen density similar to water. Stearic acid was added to the metal powders at 5% and 10% concentrations by volume and pressed to form targets with the same dimensions as the “water”-free targets. Using the GR5021 REGe, a clear variation in the 2.223 MeV peak intensity was observed between the 5% and 10% volume concentrations of water in the metal analogues. A similar intensity of 2.223 MeV $\gamma$-rays from SNF containing water is seen in the simulations when only neutrons from spontaneous fission are considered. This suggests the materials identified as analogues for SNF from calculations and simulations are suitable for this application.
This work has shown the viability of PGAA by means of the 2.223 MeV $\gamma$-ray emitted from deuterium production for the detection of water in a spent nuclear fuel analogue. Further work is to be done into the spatial resolution of the germanium detector as well as the optimisation of fuel analogues to more closely represent the behaviour of SNF. Spatial resolution will be determined using activated copper foils spread around the lab space so as to represent a distributed source. Understanding the spatial resolution of the detector will inform work concerning the localisation of water pockets. Optimisation of the spent fuel analogue will be carried out using Geant4 now the model has been experimentally validated.
