Uncertainty Quantification at the 100 Million Degree Edge of the Burning Tokamak Plasma

he United Kingdom Atomic Energy Authority (UKAEA) is one of Europe’s leading nuclear research centres, closely collaborating with both academia and industry to develop a demonstration fusion power plant for construction in the 2030s. CCFE (Culham Centre for Fusion Energy) contributes to a range of key areas in magnetic confinement fusion research. Our work is part of a co-ordinated European programme led by the EUROfusion consortium, focused on providing Europe’s input to the next-step, multi-billion Euro international fusion experiment, ITER, and the demonstration power station that will follow it, known as DEMO. CCFE is at the forefront of fusion energy research including development of power plant technology and operates the Joint European Torus (JET) – the world’s-largest fusion device, holding the world record for generated fusion power and the only tritium compatible device in the world. Key to the success of both ITER and DEMO is the accurate simulation of plasma discharges, and underpinning this, routine quantification of the uncertainties associated with the underlying models and experimental data.

A key element for delivering high fusion gain is the establishment of a high “edge pedestal” (the pedestal is a comparatively narrow layer of plasma that forms the interface between the hot, dense central plasma and the cooler, more rarefied plasma that lies adjacent to the machine first-wall components). The pedestal is characterised by significantly improved energy confinement than the rest of the plasma, leading to a very steep temperature gradient. If its confinement properties can be optimised, then the confinement of the whole plasma will improve, leading to efficient power production. However, not surprisingly, given that the desired temperature gradient is of order 100 million Kelvin over just a few centimetres, the region is very challenging to model (and measure), and much needs to be done to increase the reliability of predictions for both ITER and DEMO plasmas.

The student assigned to this project will contribute significantly to a pan-European, British-led endeavour to deliver an advanced pedestal prediction model. Moreover, the thesis work may, to a larger extent than is usual, be directed by the student’s preferences and interests.

There are opportunities to examine model discrepancy in the application of both experimental data, and simulation codes, or a combination of the two, since the main aim of the pedestal project is to find surrogate models for some six physical processes thought to be important in the tokamak pedestal. These processes, which include linear magnetic field instability, heat transfer and plasma turbulence, are modelled by a range of simulation techniques, ranging from computational fluid dynamics to Monte Carlo particle transport. The student might, for example, collaborate to help identify sources of model discrepancy, and/or to provide advice on how to combine the phenomena into a single surrogate for implementation into a model of the complete tokamak plasma. Alternatively, or additionally, the student might help quantify uncertainty in the advanced model, using the COSSAN platform produced at Liverpool, perhaps contributing to the further development of the COSSAN framework.

The student would in addition to generally useful skills in uncertainty quantification, acquire a background in plasma dynamics, well-suited to a subsequent career developing tokamaks into a practical and economic energy source.