2024 – Integrated control for access to and maintenance of Wide-Pedestal QH-mode

Integrated control for access to and maintenance of Wide-Pedestal QH-mode

2024 Research Campaign, Plasma Control

Purpose of Experiment

The goal of this experiment is to demonstrate real-time prediction and avoidance of a particular instability not tolerable by future reactors known as edge-localized modes (ELMs) in the wide-pedestal quiescent H-mode (WPQH) scenario. WPQH is a promising reactor-relevant scenario, as it achieves high confinement without an intrinsic tendency toward ELMy behavior, which stands in contrast with other standard high-confinement scenarios that typically exhibit ELMs at a frequency of 10-100 Hz. However, WPQH can still be susceptible to “breakthrough ELMs,” and this experiment seeks to predict these and preemptively engage an external actuator to steer the plasma’s evolution away from its forecasted ELMy fate.

Experimental Approach

Our approach to achieving this experimental goal is to collect in real-time information from the beam emission spectroscopy (BES) diagnostic, which provides in-depth knowledge of the turbulence properties of the plasma edge. Turbulent transport dominates the plasma edge in WPQH, and we can expect that a change in the turbulence properties are related to the onset of the ELMs. This real-time information is passed to a machine-learning model trained to predict ELMs in WPQH using BES precursor signals. When the ELM prediction rises to an alarming level, the controller registers this to actuate resonant 3D fields to enhance the edge transport and thus steer away from the ELM.