Session GO06: MFE: KSTAR Tokamak

Overview of KSTAR Experimental Research towards Future Fusion Reactors

KSTAR has been focused on exploring the key physics and engineering issues towards future fusion reactors by demonstrating the long pulse operation of high beta steady-state discharge[1]. Advanced scenarios are developed targeting steady-state operation and significant progress has been made in high beta scenarios with β_N of 3. Abundant fast ions exist in the new operation scenario called FIRE and they play an important role in confinement enhancement FIRE mode may not mean that fast ions fully determine all properties of confinement in the complex fusion plasma system [2]. The optimization of 3D magnetic field techniques including adaptive control and machine learning made long-pulse operation and high performance discharge of ELM suppressed discharge possible [3]. Symmetric multiple shattered pellet injections and real-time DECAF[4] are being conducted in order to mitigate and avoid disruption accompanying high performance for long-pulse ITER-like scenarios in KSTAR.  Lastly, the research plan in near term will be addressed with divertor tungsten upgrades with active cooling and KSTAR is steadily progressing the upgrade of extensive NBI and helicon current drive heating and transition to full metallic wall.   

Long Pulse High βN Scenario in KSTAR

The high β_N scenario with a relatively high value of li ~ 1 is one of the promising candidates for the ITER steady-state scenario and future power reactors. Stationary high β_N ~ 3 operation has been extended to ~15 sec longer than multiple current relaxation time (> 5τ_R). No obvious sign of performance degradation has been observed at β_N = 2.8. However, the β_N oscillation often causes further evolution of the current profile at higher target β_N > 3, which eventually triggers resistive n = 2 MHD modes, resulting in substantial confinement degradation. The β_N oscillation are correlated largely with interaction between the fishbone-like n=1 modes and long ELM-free period. The confinement strongly depends on the presence of small amplitude of n = 2 MHD modes, while off-axis Electron Cyclotron (EC) heating and current drive (H/CD) helps to avoid the performance-limiting n = 2 modes. Predict-First approach with the IPS-FASTRAN integrated modeling will be presented to optimize high β_N scenarios with the planned KSTAR H/CD upgrade.   

Integrated process for enhancing the normalized beta during n = 1 RMP-driven ELM-crash-suppression phase in KSTAR

The integrated process is designed to recover and enhance the plasma confinement during the ELM-crash-suppression phase in KSTAR by incorporating recent achievements in the RMP technique. The RMP onset timing, RMP spectrum, and strength are optimized to maximize the normalized beta (β_N), a metric for the plasma confinement in this study, utilizing real-time machine learning classifier, edge-localized RMP (ERMP) spectrum, and adaptive feedback RMP ELM controller, respectively. The integrated process under n = 1 ERMP achieves β_N up to ~2.65 (H_89L ~ 1.99), surpassing the previous KSTAR β_N record of ~2.4. The RMP onset at the L-H transition timing can reduce the external heating power required to achieve the same β_N compared to the conventional RMP onset during the H-mode. The adaptive controller, crucial for maximizing β_N while maintaining the suppression phase, restores the suppression phase by 38-62% quicker and recovers β_N ~ 2.6 by ~51% faster under the ERMP spectrum than the conventional RMP (CRMP) after the onset of the ELM-crash-mitigation phase accompanied by a significant drop in β_N during the adaptive control.   

Highlights of KSTAR-PPPL Collaborations on 3D Fields, Disruptions, Control, and Impurity Powder Dropper

PPPL has several ongoing collaborations on KSTAR in control of plasma transport and ELMs with 3D fields, disruption forecasting, and use of the Impurity Powder Dropper (IPD) as an actuator for real-time wall conditioning.  Applied 3D fields were optimized to successfully suppress ELMs to achieve KSTAR’s record-high β_N = 2.6 and record long ELM suppression of 45 s duration by integrating adaptive ELM control. Validation between TM1 simulations and KSTAR experiments reveal key characteristics of density pump-out by pedestal-foot island formation induced by RMP.  Coupling of the nonlinear 3D MHD JOREK code with the PENTRC neoclassical toroidal viscosity code has consistently captured the experimentally observed bifurcation and density pump-out with increasing RMP fields. Implementation of real-time measurements of MHD modes, plasma rotation, and ECE has permitted developing disruption forecasting and feedback algorithms to help avoid disruptions.  Dynamic control of the IPD has been implemented for real-time wall conditioning to improve KSTAR’s long pulse capability with its new tungsten divertor with experiments due in late 2023. Future experiments are planned to compare glow discharge boronization and real-time wall conditioning with the IPD considering ITER plans for a full tungsten wall and divertor. The latest results to help KSTAR achieve its long pulse goals and plans for future collaborations will be presented.    

Observation of wide and high pedestal formation triggered by coherent edge mode activities in KSTAR hybrid scenario discharges

Hybrid scenarios, one of the candidate scenarios for the International Thermonuclear Experimental Reactor (ITER), exhibit distinctive features such as a central  above unity and an extended region of low or zero magnetic shear in the core. These characteristics reduce neoclassical tearing modes induced by sawtooth activities and enable higher beta operation. Extensive efforts have been made to comprehend the underlying mechanisms responsible for the improved performance observed in hybrid scenarios on present-day tokamaks.

In recent experiments in the Korea Superconducting Tokamak Advanced Research (KSTAR), a transition-like performance enhancement was observed in hybrid scenario discharges. This enhancement was associated with the formation of a wide and high pedestal. A comparative analysis discovered that coherent edge mode (CEM) activities were responsible for changing the pedestal. CEM, a coherent edge-localized mode, occurs during the pedestal recovery phase and is stabilized after the subsequent edge-localized mode (ELM) crash. Measurements showed that particle and heat transport tended to increase during CEM activities, providing a consistent explanation for the observed pedestal broadening. Furthermore, dedicated experiments changing the amount of external gas fueling revealed a correlation between increased external fueling and intensified CEM activities, indicating an available control knob for accessing high performance.   

The Investigation of Fast Ion Effects in the Internal Transport Barrier Operation in KSTAR

Fast ions generated from the fusion reaction will be used as a self-heating source for sustaining the future burning plasmas. Therefore, it is essential to understand the fast ion effects on fusion plasmas. In recent, it has been known that fast ions can suppress turbulence through various mechanisms including an increase in the pressure gradient [1], dilution effects [2], and changes in the zonal shearing rate [3]. Various physical mechanisms of fast ion effects for turbulence suppression during the internal transport barrier (ITB) operation were investigated in KSTAR plasmas. We observed that a gyrokinetic simulation predicted a significant reduction of thermal energy fluxes when fast ions were included in the KSTAR ITB plasma [4] heated by neutral beam injection with a high fast ion fraction and a peaked fast ion density profile. It turned out that the dominant physical mechanism for turbulence suppression by fast ions in this KSTAR ITB plasma was dilution effects including both effects of reduced main ion density fraction and change of main ion density gradient. Furthermore, it was found that the sole effect of the inverted main ion density gradient was sufficient to suppress the turbulence enough for ITB sustainment. The significance of this finding lies in its potential to explore novel operational modes with improved confinement through effective turbulence control in the future. By manipulating the main ion density gradient through various methods such as pellet injection, impurity puffing, or the addition of fast ions, the ability to regulate turbulence can be achieved. Additionally, fast ion effects on KSTAR ITB plasmas with the different experimental conditions [5,6] will be discussed.   

Gyrokinetic simulations of KSTAR core perturbations and transport with energetic particles

The energetic particle (EP) effect on tokamak confinement is a topic with growing attention for high-performance tokamak operation [1]. The EPs are produced by auxiliary heating or fusion reaction. Recent gyrokinetic simulation studies [2,3] have shown that a comprehensive understanding of various interactions among EP, AE, and microturbulence is needed to properly predict confinement. We aim to achieve an understanding of the impact of EP-AE-turbulence interactions on tokamak core confinement by gyrokinetic simulations of a target KSTAR L-mode discharge #21695 using GTC [4] and GKW [5] codes. In the target discharge [6], control of AE activities has been achieved by ECCD which has changed the q-profile. We have identified unstable BAEs in the AE-active phase, with 3/1 kink-like mode at qmin. We have found linearly the most unstable ETG for turbulence in both AE-active and mitigated phases, but it appears to be ineffective on transport compared to the ion-scale modes, TEM for the AE-active phase and ITG for the mitigated phase. Controlled simulations with AE-only and turbulence-only and a comparison with the reference case will be presented, which reveal the impact of AE-turbulence interaction on the confinement.

[1] H. Han et al., Nature 609, 269 (2022)

[2] Di Siena et al., J. Plasma Phys. 87, 555870201 (2021)

[3] P. Liu et al., Phys. Rev. Lett. 128, 185001 (2022)

[4] Z. Lin et al., Science 281, 1835 (1998) 

[5] A.G. Peeters et al., Comput. Phys. Comm. 180, 2650 (2009)

[6] J. Kim et al., Nucl. Fusion 62, 026029 (2022)   

High-Accuracy Disruption Event Characterization and Forecasting for database analysis and 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 physical and technical events leading to disruption and can provide event onset forecasts with high accuracy and early warning for disruption avoidance [1]. Real-time application of DECAF on the KSTAR tokamak to over 50 experimental plasmas subject to disruption by locking MHD instabilities, and which produced a nearly equal number of disrupted / non-disrupted cases, were forecast with 100% accuracy. These real-time forecasts triggered controlled plasma shutdown, disruption mitigation, and disruption avoidance actuators. The warnings were issued well before (0.5s – 1.5s) the expected plasma disruption time and early warning guidance given for ITER disruption mitigation. High accuracy exceeding 99% was also found in DECAF analysis of tokamak databases. This fully automated analysis now expands to examine the plasma state as a general dynamical system to best validate DECAF physical events and event chains, thereby allowing reliable extrapolation of models across devices and to future machines.

[1] S.A. Sabbagh, et al., Phys. Plasmas 30 (2023) 032506; https://doi.org/10.1063/5.0133825    

DECAF cross-machine comparison of born-rotating mode locking forecaster developed for real-time implementation on KSTAR

Operation of reactor scale tokamaks with high thermal and magnetic energy density will require low occurrence of plasma disruptions in which the quenching of these energies can compromise critical plasma facing components or vessel integrity. The presence of rotating MHD modal instabilities can deteriorate the plasma performance in tokamaks and their locking to the wall leading to plasma disruptions. Prediction, observation, and avoidance of these modes is therefore essential in the operation of tokamaks. A mode locking forecaster based on a torque balance model has been developed across devices of ranging aspect ratio and error fields. Analysis of the KSTAR, NSTX, MAST-U, and DIII-D databases has been conducted and compared. Based on the success of database analysis, real-time forecasting and identification modules were written and installed on the KSTAR superconducting tokamak. Over 50 dedicated plasma, of different scenarios, were run experimentally to test this system (real-time DECAF [1]). The results show a 100% success rate in identifying true positives in this collection of nearly equal disrupted / non-disrupted plasmas. Real-time results are shown, with comparisons to the analogous offline analysis, and lessons learned in this process are discussed. [1] S.A. Sabbagh, et al., Phys. Plasmas 30, 032506 (2023); https://doi.org/10.1063/5.0133825   

Improved confinement by non-axisymmetric magnetic field in KSTAR

Non-axisymmetric (3D) magnetic field induces a variety of plasma responses impacting on the pedestal stability, toroidal plasma rotation, and MHD activities. Thanks to flexibility of the 3D magnetic field coils, KSTAR tokamak has demonstrated 3D field capabilities for control of transport and stability of thermal and supratheramal plasmas in the core, edge, and SOL regions. In this presentation, we report the observation of improved confinement in the magnetic braking experiment on KSTAR. The improved confinement is achieved with reduced toroidal plasma rotation by non-axisymmetric magnetic field induced toroidal rotation braking along with significant reduction of edge localized modes (ELMs). Modifications in multi channel transport raise fast ion slowing-down time and improve neutral beam deposition, leading to improved fast ion confinement. We show that modifications of radial electric field and E×B shear flow by magnetic braking provoke an enhanced pedestal to sustain thermal confinement against degradation in the typical 3D field experiment.   

Consequences of charge exchange in plasmas and the intrinsic rotation of KSTAR

Ion gyro-center shift induced by the charge exchange reaction between ion and neutral atom is analyzed to explain the electric field formations and turbulence suppression in various plasmas including Er of tokamak edge, H-mode transition, reversed motion of arc discharge, E-fields of equatorial electro-jet (EEJ), black aurora, and Bohm diffusions. [1][2] And new intrinsic plasma rotation analysis is developed based on the unbalanced ion-electron momentum exchange induced by the ion-neutral charge exchanges. The origin of intrinsic rotation has been researched for decades because the plasma rotation controls the MHD instabilities. However, no quantitative agreement is available thus far. Very strong counter-current intrinsic rotation is induced on KSTAR when the ion-neutral collision frequency is comparable to the electron-ion collision frequency, especially at the edge of tokamak. The comparison of experimental observations on KSTAR and VEST is well agreed with the analysis based on this charge exchange effect. The magnitude and time evolution of the intrinsic rotations agrees for the cases with different plasma parameters. It is also found that the neutral density is not sensitive to the intrinsic rotation and the impurity concentration is an important parameter for the intrinsic rotation.

    

Progress in Understanding Transport across Core Magnetic Islands in KSTAR

This study presents updated findings from KSTAR experiments, revealing increased turbulence near the X-point of the 2/1 core magnetic island. Through accurate comparisons with KSTAR ECE and ECEI measurements, we analyze the validity of linear and nonlinear MHD simulations and plasma response models (M3D-C1, NIMROD). The results of the plasma response modeling are in good agreement with ECEI measurements for the width of the resulting magnetic island. Nonlinear GTC simulations using the DKE model and incorporating EFIT pressure profiles and M3D-C1 equilibrium with n=1 islands demonstrate the spreading of turbulence across the 2/1 magnetic surface and show the observed fluctuations across the island X-point. The experimental and numerical results are consistent with the previous BES measurements across core and edge RMP islands in DIII-D discharges. BES measurements also show a decrease in fluctuation levels in front of the island O-point compared to the no-RMP case. The observation of localized island transport is consistent with the previously proposed ExB convection mechanism across the island X-point. The results of the simulations give confidence that the understanding of the underlying physics of transport in the presence of 3D fields and magnetic islands can be extrapolated to edge islands, giving us insights into RMP ELM suppression mechanisms.   

Recent Development Status of Virtual-KSTAR

We present Virtual-KSTAR (V-KSTAR), a digital twin study aimed at establishing a unified machine/fusion data framework and simulation workflows. Through the utilization of the game engine technology, V-KSTAR effectively visualizes machine data obtained from CAD (Computer-Aided Design) files, as well as plasma/simulation data derived from simulations. By accumulating simulation technology and validating simulations on KSTAR and ITER experiments, V-KSTAR would play a key role in the design and construction of the Korean DEMO reactor.

Recent advancements in V-KSTAR include the integration of the tokamak/fusion data model from IMAS (Integrated Modeling & Analysis Suite) and the conversion of legacy data into a standardized format. Building upon the data framework, V-KSTAR incorporates an analysis system with specific simulation workflows such as ECH, NBI, and RMP. This not only enables researchers to investigate discharges in more detail in a traditional way, but also to gain novel insights from a 3D perspective. The utilization of computational meshes derived from the CAD data ensures the consistency of the data throughout the analysis workflows.

In this presentation, the current status and recent achievements of V-KSTAR development are presented. The integration of the standard data model into the digital twin and analysis system are shown in detail. Additionally, we will discuss our plans to leverage machine learning or artificial intelligence to accelerate simulations.   

Edge Plasma Density Profile Estimations Using Bayesian Model and Gaussian process in the KSTAR Hydrogen Beam Emission Spectroscopy

We have developed a Bayesian model for inferring edge plasma density profiles using the KSTAR hydrogen beam emission spectroscopy (H-BES) system. The system measures Doppler-shifted D-alpha emission and consists of 16 radial and 4 poloidal channels, each with approximately 1cm spatial resolution. The poloidal channels could provide an increased precision in radial profile measurement through the integration of equilibrium information. The intensities of the Doppler-shifted D-alpha lines are represented as a function of plasma density, guided by a multi-state model that encompasses the interactions of neutral deuterium beam atoms with plasma particles. The Gaussian process prior is employed to model density profiles and the posterior distribution is explored using a Markov Chain Monte Carlo (MCMC) method. Through the use of a Gaussian process and relative intensity calibration, the method can estimate instrument effects and absolute calibration factor. To provide a more accurate calculation of the absolute calibration factor, we have integrated this model with an edge interferometer. Given the growing significance of the 3D field in plasma control, the ability to simultaneously measure both the density profile and density fluctuations at a single toroidal position using a single diagnostic will significantly enhance our understanding of edge plasma physics.   

Remote automatic absolute-polarization calibrations for KSTAR motional Stark effect diagnostic

The absolute-polarization angle calibration procedure that is conventionally performed inside the vacuum vessel manually has been updated in a remote and automatic manner. It is postulated and experimentally confirmed that the polarized light source does not have to be positioned at the exact intersection between the beam trajectory and the line-of-sight. That is, one can position the linearly polarized light source at any point along the line-of-sight as long as the face of the polarizer is perpendicular to the wave propagation vector. This implicates the calibration hardware including the supporting jig can be installed at a position of convenience and cost-effective. The light source module consists of two motorized rotation stages, one long-distance motorized linear stage, and a two-axis inclinometer sensor. The remote operation is realized by a LabVIEW-based graphical user interface program that also functions automation associated with the KSTAR shot cycle. Precise scans of incident polarization angle with the resolution of down to 1 degree are possible and scans of different polarimeter settings such as photoelastic modulator retardation are performed with the same precision which would never be possible without remote automatic procedures.