Search

Fusion energy for use in power plants is a continually developing area and many of the related parameters are not yet fixed. The investigation of fusion neutronics and development of computational approaches for assessment is imperative in the road to commercial realisation of fusion power. This research has explored blanket performance, including tritium breeding and the shielding requirements, and assessed radioactive waste, utilising the 3-D Monte Carlo transport code MCNP, and the activation inventory code FISPACT. The performance of some solid and liquid breeder materials has been compared with regards to tritium breeding, energy production and shielding. In the case of novel spherical tokamak concepts, that make use of high temperature superconducting magnets and have no inboard blanket, scoping studies have been performed to investigate the impact of shielding requirements on how small the tokamak can be. Fusion power plants will not produce high level waste, as seen in nuclear fission plants, however the components and structures will become active as a result of interactions with high energy neutrons. A suitable radioactive waste management plan will be required in order to deal with this material appropriately, with an aim to recycle or clear from regulatory control all materials 100 years after shutdown. The study indicates that through suitable material selection and the use of component dismantling the requirement could potentially be satisfied. In terms of computational methods, the neutron flux averaging has been assessed throughout the work and has shown in neutronics estimates to produce some substantial differences. The recently developed unstructured mesh approach to neutronics modelling has been explored and the potential use for more accurate radioactive waste inventory calculations. Although the analysis and comparison shows promising results, it still requires significant development and improvement in the work flow to create a robust neutroni

A neutron collimator has been designed in MCNP5 and tested for feasibility of use in imaging applications. Tungsten, polyethylene, PVC and lead have been compared as collimating materials for neutrons in the MeV energy range; tungsten is predicted to be the most successful material for a restricted volume, giving the highest signal-to-noise ratio and the best resolving power. Experimental data has been used to confirm that tungsten works effectively as a neutron collimator although some discrepancies between real and MCNP5 results were observed. A suspension of tungsten powder in polyethylene has also been tested to address the machining difficulties, mass and cost issues associated with tungsten. This material performs midway between tungsten and polyethylene for a constant volume, and more successfully than tungsten for a constant mass therefore giving this material potential as a collimation material in some scenarios. Further MCNP5 modelling has been performed by varying model parameters and monitoring the collimator functions produced by these changes. These results are conclusive but dependent on the applications of the imaging system.

Sellafield Nuclear Reprocessing Plant in Cumbria contains storage ponds built in the 1950s which was originally intended to hold spent nuclear fuel for reprocessing, and eventual production of weapons grade plutonium. Parts of the spent fuel have corroded; some are buried under a layer of sediment or intertwined with other debris and removal and destruction of this nuclear waste is not a trivial task due to elevated radiation levels. We propose a system in collaboration with the National Nuclear Laboratory (NNL) to characterise the ponds using a system containing three main parts; an ultrasonic SONAR system used to physically map the pond, scintillator based radiation detector (known as RadLineTM) used to map the pond from a radiation point of view, and bespoke software intended to combine the physical and radiation plots of this environment to create an overall 3D source map.

The viability of various materials and blanket designs for use in nuclear fusion reactors can be tested using computer simulations and as parts of the test blanket modules within the International Thermonuclear Experimental Reactor (ITER) facility. The work presented here focuses on blanket model simulations using the Monte Carlo simulation package MCNPX (Computational Physics Division Los Alamos National Laboratory, 2010) and FISPACT (Forrest, 2007) to evaluate the tritium breeding capability of a number of solid and liquid breeding materials. The liquid/molten salt breeders are found to have the higher tritium breeding ratio (TBR) and are to be considered for further analysis of the self sufficiency timing.