Disruption Event Characterization and Forecasting Research and First Real-time Application on KSTAR

Disruption prediction and avoidance is critical for ITER and reactor-scale tokamaks to maintain steady plasma operation and to avoid damage to device components. Physics-based disruption event characterization and forecasting (DECAF) research determines the relation of events leading to disruption and aims to provide event onset forecasts with high accuracy and early warning for disruption avoidance. Real-time application of DECAF was recently made on the KSTAR superconducting tokamak. Experiments focused on locking MHD instabilities produced 40 plasmas with nearly equal disrupted / non-disrupted cases that are forecast with 100% accuracy. These real-time forecasts triggered controlled plasma shutdown and disruption mitigation. The warnings were issued well before the expected plasma disruption time and early warning guidance given for ITER disruption mitigation. Offline analysis has access to data from several tokamaks (e.g. KSTAR, MAST, NSTX) to best understand, validate, and extrapolate models. Recent code improvements allow fully automated analysis of up to entire device databases. Such initial analysis shows very high true positive success rates over 99%. *This research is supported by the U.S. DOE under grants DE-SC0020415, DE-SC0018623, and DE-SC0021311.