Optimising energetic particle transport in 3D fields in the ITER tokamak

The high-gain ITER baseline tokamak plasma scenario depends upon reliable H-mode operation with edge-localised modes (ELM) suppressed by the application of resonant magnetic perturbations (RMP) using an ELM-control coil (ECC) system. However, these perturbations can lead to significant fast-ion transport and resulting power loads on plasma-facing components (PFC). This is of particular concern to the high-power discharges in ITER, which are extremely challenging to predictively study using conventional experimental devices and the physics codes currently used to study them.

In this work, software was created with which novel, high-performance computing components were assembled to enable, for the first time, routine, high-fidelity simulation of realistic fast-ion transport due to 3D fields in ITER. The primary fast-ion component, LOCUST, was verified and validated. The assembled workflow was then deployed to determine methods of operating the ITER ECC system where plasma heating efficiency, PFC power loads and ELM suppression are optimised simultaneously.

The response of fast-ion confinement to ECC operating parameters, such as coil current amplitude and phase, was discovered to correlate with ELM suppression. With this knowledge, the optimal method for operating the ECC system was determined, over a range of plasmas and applied RMP mode spectra, to increase total heating efficiency by 1.7-3.2% points in the baseline scenario. Even in the worst case, PFC power loads were found to be tolerable and rotation of the applied RMP to reduce power loads by up to 0.44MWm^-2 (64%). In the optimal setting however, rotation may not be required, as minimum power loads and global losses were found to correlate. Methods for experimentally verifying these findings, using low-power plasmas and the diagnostic neutral beam, were also studied. Lowering the ECC coil current was found to be a low-risk approach to predicting fast-ion confinement in the ITER baseline scenario.