First Demonstration of Tokamak Disruption Avoidance by Disruption Event Characterization and Forecasting in the KSTAR Superconducting Fusion Device by Sabbagh Research Group
Limitless energy produced by magnetic fusion has been a long sought goal. This process, which is the energy source of stars, including our Sun, has no carbon emissions and is now being pursued by a worldwide nascent industry. The leading type of magnetic fusion device, the tokamak, has a long-known critical shortcoming - the plasma that creates fusion can “disrupt”, which stops energy production. To make the device viable for energy production, this issue needs to be overcome with high reliability.
Recently, a scientific and technological breakthrough has shown the first demonstration of a multi-element system for prediction, avoidance, and mitigation of plasma disruptions in the $1B USD-class South Korean KSTAR superconducting fusion device (https://www.kfe.re.kr/eng). An international team of scientists and engineers from the Columbia University APAM department (Figs. 1-3), the Korea Institute of Fusion Energy, the Princeton Plasma Physics Laboratory, and General Atomics, led by APAM senior research scientist and adjunct professor Steven A. Sabbagh, produced the first experimental demonstration of a multi-element, high accuracy plasma disruption prediction approach (Disruption Event Characterization and Forecasting, or DECAF [S.A. Sabbagh, et al., Physics of Plasmas 30 (2023) 032506]; https://doi.org/10.1063/5.0133825) with the ability to avoid plasma disruptions in the device thereby allowing an otherwise disrupting plasma to continue operation until it was intentionally shut down after a long pulse duration.

Figure 1: (left to right) APAM students Freddie Sheehan, Matt Tobin, Prof. Steven A. Sabbagh, Dr. Guillermo Bustos-Ramirez, and Dr. Hankyu Lee, part of the Columbia U. APAM Team producing the first disruption avoidance with real-time Disruption Event Characterization and Forecasting (DECAF) on the KSTAR fusion device.

Figure 2: (left-right) APAM students Matt Tobin and Juan Riquezes (first and third from left) with Prof. Steven A. Sabbagh, Dr. Guillermo Bustos-Ramirez, Dr. Hankyu Lee, and colleagues Dr. Shira Morosohk (General Atomics) and Dr. Jun-Gyo Bak (KFE) touring the $1B USD-class international superconducting KSTAR device

Figure 3: (left-right) Dr. Guillermo Bustos-Ramirez, Mr. Matt Tobin, Dr. Hankyu Lee, and Mr. Freddie Sheehan in front of the floor-to-ceiling display wall in the KSTAR control room.
Prof. Sabbagh’s present Columbia research group includes APAM associate research scientist Dr. Veronika Zamkovska, post-doctoral researchers Dr. Guillermo Bustos-Ramirez, Dr. Joseph Jepson, Dr. Hankyu Lee, and students Mr. Juan Riquezes, Mr. Matt Tobin, Mr. Frederick Sheehan, and Mr. Grant Tillinghast. The majority of the group visited KSTAR in December 2024 and January 2025, tirelessly working until early AM hours to produce the first result on January 15th, and reproducing it under different plasma conditions and producing initial optimizations on January 17th (Figure 4).

Figure 4: Time-evolving elongation and normalized beta stability values for three KSTAR plasmas: the shortest is terminated by a disruption (solid line). The second (dashed line) shows disruption avoidance by real-time DECAF, and the third shows an initial optimization of the disruption avoidance.
The experiments showed the versatility of the DECAF system with “DECAF Event feedback”, which includes device hardware actuation by a variety of internal feedback controllers and also couples to and commands external controllers seamlessly, such as a plasma shaping controller produced by Dr. Jayson Barr of General Atomics. All such controllers were used in these initial experiments.
Recently, plasma control management of the $22 billion dollar ITER Project (https://www.iter.org), led by Dr. Peter de Vries, has endorsed the implementation and use of DECAF for real-time disruption prediction and avoidance in ITER, along with associated physics analysis. (https://www.apam.columbia.edu/worlds-largest-fusion-energy-tokamak-device-endorses-sabbagh-research-group-solution-enabling).
This initial demonstration of tokamak disruption avoidance by DECAF on the KSTAR device paves the way for a comprehensive disruption avoidance solution by expanding the present system, providing a critical device control component that can help enable tokamak devices to produce abundant fusion energy for the planet.