South Korea’s KSTAR tokamak sustained a plasma at 100 million degrees Celsius for 102 seconds in H-mode during its most recent campaign, more than doubling the facility’s previous record of 48 seconds at the same temperature. The Korea Fusion Energy Institute, led by President Suk Jae Yoo, credited upgrades to the machine’s tungsten divertor and refined magnetic control systems for the leap. The result puts KSTAR closer than any superconducting tokamak has come to the kind of sustained burn that a commercial fusion reactor would require.
Why 102 Seconds at 100 Million Degrees Changes the Fusion Timeline
Fusion devices have hit extreme temperatures before. The harder problem is holding those temperatures long enough for the plasma to do useful work. A reactor designed to feed electricity into a grid would need to maintain a burning plasma not for seconds but for minutes or hours. KSTAR’s jump from 48 seconds to 102 seconds in H-mode, the high-confinement state where energy loss drops and performance rises, represents a concrete step toward that goal rather than an incremental gain.
The advance matters because of what made it possible. During the campaign, researchers applied error-field correction techniques to suppress instabilities that typically force operators to shut down a pulse early. A peer-reviewed study published in Nature Communications detailed how tailoring tokamak error fields can control plasma instabilities and transport. That paper, which grew out of KSTAR operations, was cited in institutional communications surrounding the record-setting run and provides the scientific foundation for the magnetic adjustments that kept the plasma stable well past the one-minute mark.
The same research is also accessible to registered readers through a Nature portal, underscoring that the physics behind the achievement has been scrutinized beyond internal laboratory reports. By grounding the operational strategy in peer-reviewed analysis of error fields and transport, the KSTAR team positioned the 102-second pulse as a test of validated theory rather than an isolated experimental surprise.
Si-Woo Yoon, director of the KSTAR Research Center, framed the result as proof that continuous operation technology is within practical reach. His statement, distributed through the National Research Council of Science and Technology, signals that the Korean program views the 102-second pulse not as a ceiling but as a waypoint. The next target is pushing durations past 200 seconds while maintaining the same ion temperature, a threshold that would test whether the error-field correction methods scale or whether new instability modes emerge at longer timescales.
Error-Field Science and the Tungsten Divertor Behind the Record
Two hardware and physics developments converged to produce the result. First, KSTAR’s carbon divertor was replaced with a tungsten version. Tungsten handles higher heat loads and produces less contamination inside the plasma, which matters when the goal is to keep a 100-million-degree discharge running for minutes at a time. The institutional release from the National Research Council of Science and Technology specifically credited tungsten divertor performance as a factor in the extended pulse.
Second, the team refined how external magnetic coils correct tiny asymmetries in the tokamak’s confining field. Those asymmetries, called error fields, can trigger edge-localized modes and other instabilities that dump energy onto the vessel wall and end a shot prematurely. The published analysis showed that by deliberately shaping the applied error field rather than simply canceling it, operators can reduce instability-driven transport losses. Researchers used this tailored approach during the campaign to hold the plasma in H-mode for the full 102 seconds.
The combination of better wall materials and smarter magnetic control addresses two of the biggest obstacles to long-pulse operation in any tokamak. Wall erosion limits how long a machine can run before maintenance is needed. Instabilities limit how long a single plasma discharge survives. Solving both simultaneously is what allowed KSTAR to more than double its prior mark. It also provides a template for other superconducting tokamaks that aim to reach similar temperatures and pulse lengths without damaging their hardware.
The tungsten divertor carries particular weight because it mirrors design choices being made for next-generation devices. Tungsten’s high melting point and low sputtering yield make it attractive for handling the intense heat and particle flux at the edge of a fusion plasma. Demonstrating that such a component can survive a 102-second, 100-million-degree campaign without reported degradation suggests that long-pulse operation and plant-scale engineering are starting to align. However, the long-term behavior of tungsten under repeated stress cycles remains to be fully characterized.
What the 102-Second Pulse Does Not Yet Prove
The record is real, but several questions remain open. The 102-second achievement has not yet appeared in a dedicated peer-reviewed paper with full shot logs, time traces, and error-field coil current profiles for that specific pulse. The Nature Communications study on error-field tailoring provides the physics rationale, and the institutional announcement provides the headline number, but independent researchers have not yet had access to the detailed diagnostics of the record shot itself.
A second gap involves the hypothesis that error-field corrections can suppress edge-localized modes by 30 percent or more at durations beyond 200 seconds. That claim is plausible based on the published science, but it has not been tested. KSTAR’s next official campaign will be the proving ground. If the team can push past 200 seconds without a significant rise in instability frequency, the error-field method will have cleared a meaningful hurdle on the path to steady-state operation.
A third open question is how the tungsten divertor performs under repeated long pulses. A single 102-second shot demonstrates capability. Hundreds of such shots demonstrate reliability. Fusion power plants will need both. The institutional release described the divertor’s performance in positive terms but did not publish erosion rates or surface condition data from the extended run. Without that information, it is difficult for outside analysts to estimate maintenance intervals or component lifetimes for reactors that might adopt a similar design.
There is also the broader issue of how efficiently the plasma produced energy during the record pulse. The announcement highlighted temperature and duration, not fusion gain or net power balance. For commercial viability, a reactor must not only confine a hot plasma but do so in a way that yields more energy than the system consumes. KSTAR is an experimental device, not a power plant, so net energy production is not its immediate goal. Still, future publications that pair confinement results with performance metrics would help clarify how close the underlying physics is to power-plant conditions.
For anyone tracking the global fusion race, the next development to watch is whether KSTAR publishes full diagnostic data from the February campaign in a peer-reviewed journal and whether the upcoming experimental season targets 200 seconds or longer. Those two milestones would convert a striking announcement into a validated engineering baseline that other tokamak programs, including devices now under construction or upgrade, can compare against their own performance. If the error-field strategy and tungsten divertor continue to hold up under longer pulses, KSTAR’s latest run may be remembered less as a singular record and more as the moment when continuous-operation fusion research entered a new, more practical phase.
More from Morning Overview
*This article was researched with the help of AI, with human editors creating the final content.
