Small Edge Localized Modes Research Articles

ELM suppression by resonant magnetic perturbation in high-performance, stationary plasmas

The method of resonant magnetic perturbation (RMP) has been shown to suppress edge-localized modes (ELMs) in the high-performance, stationary (or ‘hybrid’) scenario in the DIII-D tokamak. Calculations of stability to peeling–ballooning (P–B) modes are shown to be consistent with the observed suppression of type-I ELMs, while the ELM suppression, dependence on edge safety factor and density pump-out are similar for hybrids and standard H-mode discharges. However, other small ELMs can appear when the edge safety factor is outside the resonance window or when the H-mode pedestal is perturbed, which are not related to P–B stability. The role of the edge bootstrap current in determining stochastic heat transport during RMP is discussed.

Pedestal stability comparison and ITER pedestal prediction

The pressure at the top of the edge transport barrier (or ‘pedestal height’) strongly impacts fusion performance, while large edge localized modes (ELMs), driven by the free energy in the pedestal region, can constrain material lifetimes. Accurately predicting the pedestal height and ELM behavior in ITER is an essential element of prediction and optimization of fusion performance. Investigation of intermediate wavelength MHD modes (or ‘peeling–ballooning’ modes) has led to an improved understanding of important constraints on the pedestal height and the mechanism for ELMs. The combination of high-resolution pedestal diagnostics, including substantial recent improvements, and a suite of highly efficient stability codes, has made edge stability analysis routine on several major tokamaks, contributing both to understanding, and to experimental planning and performance optimization. Here we present extensive comparisons of observations to predicted edge stability boundaries on several tokamaks, both for the standard (Type I) ELM regime, and for small ELM and ELM-free regimes. We further discuss a new predictive model for the pedestal height and width (EPED1), developed by self-consistently combining a simple width model with peeling–ballooning stability calculations. This model is tested against experimental measurements, and used in initial predictions of the pedestal height for ITER.

MHD stability analysis of small ELM regimes in JET

We have analysed the edge stability of JET discharges with small edge localized modes (ELMs) using the high resolution Thomson scattering system for accurate edge profiles in the equilibrium reconstruction. For the reference plasmas with large Type I ELMs we confirm the results from earlier analyses that the edge stability is limited by intermediate-n peeling–ballooning modes with a relatively large radial extent. The double null configuration needed to replace Type I ELMs by smaller Type II ELMs considerably increases the stability against these modes while the stability against n = ∞ ballooning modes is not affected. When this is combined with high collisionality (which is the other requirement for Type II ELMs), we find that the plasma cannot reach the Type I ELM triggering peeling–ballooning mode stability boundary before it is destabilized by high-n ballooning modes resulting in more benign ELMs. The ELM mitigation by magnetic perturbation causes the edge stability to be limited by pure peeling modes with a narrow radial extent. This explains the smaller ELM size and also why the ELMs are not fully suppressed. The transition from Type I ELMs to Type III ELMs by increasing the edge radiation fully stabilizes the edge plasma against ideal MHD modes. Therefore, the Type III ELMs are due to be triggered by some other mechanism than an ideal MHD instability.

Use of the absolute phase in frequency modulated continuous wave plasma reflectometry

In frequency modulated continuous wave reflectometry, used for density profile measurement in fusion plasmas, it is usual to measure the beat frequency between the launched wave and the reflected wave, and from this to calculate the position of the reflecting layer in the plasma. The absolute phase of the beat signal is usually neglected. The reason is that the phase shift between sweeps is usually comparable with or more than 2pi, leading to an ambiguity that is impossible to resolve. However, recent observations on the MAST tokamak have shown that, under quiet plasma conditions (this term has to be defined), the phase shift between sweeps is small compared with 2pi and the phase ambiguity can be readily resolved. The reflectometer signal is then being analyzed as an interferometer signal would normally be, and there is a substantial improvement in spatial resolution. The method is illustrated by application to small edge localized mode precursor and allows what is believed to be the first quantitative measurement of the displacement of the plasma boundary by such a precursor mode. The errors in both the absolute phase measurement and the more conventional frequency measurement are also estimated.

Fast imaging of edge localized mode structure and dynamics in DIII-D

Fast-framing images of CIII and Dα emission in the low-field-side plasma boundary of the DIII-D tokamak [J. L. Luxon, Nucl. Fusion 42, 614 (2002)] show that edge localized modes (ELMs) rapidly eject multiple field-aligned filaments from the plasma edge. The toroidal and poloidal mode numbers of these filaments depend on normalized plasma density, with measured ELM toroidal mode numbers ranging from ⩽10 to 20 in low-density plasmas and 15 to 35 in high-density plasmas. In high-density plasmas with moderate collisionality νped*=0.50, ELMs originate at the low-field-side midplane region and the ion parallel velocity in the scrape-off layer is faster for ELMs with larger Dα divertor emission, suggesting that large ELMs eject higher-temperature ions from deeper within the plasma compared to small ELMs. In low-density plasmas with collisionality νped*=0.25, the midplane and divertor ELM signals appear simultaneously, indicating that ELM behavior depends on collisionality. At all νped*, ELMs drive parallel fluxes to the divertor; in addition, ELMs drive cross-field propagation of filaments, which results in plasma-wall interactions that are poloidally localized within 15cm of the midplane. Using the wall interactions as signatures of the filaments in the scrape-off layer, the measured poloidal width of the filament ranges from 1to5cm.

Comparison of deuterium pellet injection from different locations on the DIII-D tokamak

Deuterium pellets have been injected into plasmas in the DIII-D tokamak from the inner wall, top, and outer midplane port locations to investigate fuelling efficiency, mass deposition and interaction with edge localized modes (ELMs). Pellets injected from the outer midplane port (low field side (LFS)) show a large discrepancy in the mass deposition profile and fuelling efficiency from conventional pellet ablation theory, while the penetration depth compares favourably with theory. The mass deposition from pellets injected from inner wall and top locations is deeper than expected from ablation theory. The profile measurements indicate that pellet mass is deposited inside the measured penetration radius, thus verifying that a drift of the pellet ablatant is occurring in the major radius direction during the toroidal symmetrization process. The scaling of the measured drift magnitude in DIII-D is found to depend strongly on the pellet size and plasma pedestal temperature. Extrapolation to a burning plasma configuration on ITER is favourable for inner wall pellet fuel deposition depth well beyond the separatrix. Pellets injected into H-mode plasmas from all locations trigger ELMs with much larger ELM events induced by the outside midplane injected pellets. This suggests that the LFS is more sensitive to ELM triggering and may be the preferred location to inject very small pellets to trigger frequent small ELMs and thus minimize ELM induced damage to the divertor material surfaces.

Pellet fuelling and control of burning plasmas in ITER

Pellet injection from the inner wall is planned for use in ITER as the primary core fuelling system since gas fuelling is expected to be highly inefficient in burning plasmas. Tests of the inner wall guide tube have shown that 5 mm pellets with up to 300 m s−1 speeds can survive intact and provide the necessary core fuelling rate. Modelling and extrapolation of the inner wall pellet injection experiments from present day's smaller tokamaks leads to the prediction that this method will provide efficient core fuelling beyond the pedestal region. Using pellets for triggering of frequent small edge localized modes is an attractive additional benefit that the pellet injection system can provide. A description of the ITER pellet injection system's capabilities for fuelling and ELM triggering is presented and performance expectations and fusion power control aspects are discussed.

Roles of Toroidal Rotation at the Plasma Edge, Toroidal Field Ripple and Configuration on ELMs in the JT-60U Tokamak

Recent experimental studies are presented regarding ELMy H-mode plasma having Type-I Edge Localized Modes (ELMs) in the JT-60U. Toroidal rotation scan experiments in inward-shifted, small-volume plasma show that the ELM energy loss decreases as the toroidal rotation at the plasma edge increases in the direction counter to the plasma current. Specifically for a large volume plasma having a toroidal field ripple of ∼2% at the plasma edge, small Type-I ELMs are observed whose ELM energy loss normalized by the pedestal stored energy is smaller than that of an acceptable ELM size in the ITER. However, the pedestal pressure tends to decrease when plasma volume increases. No remarkable effect of reduced toridal field ripple due to the installation of Ferritic Steel Tile (FSTs) inside the vacuum vessel on the JT-60U on ELMs for large volume plasma is seen in the ELM energy loss normalized by the pedestal stored energy. These new findings suggest that the toroidal field ripple itself may not directly affect the normalized ELM energy loss, and that toroidal rotation at the plasma edge as well as plasma configuration might play important roles in the prediction of ELM size in future devices.

Characterization of small, Type V edge-localized modes in the National Spherical Torus Experiment

There has been a substantial international research effort in the fusion community to identify tokamak operating regimes with either small or no periodic bursts of particles and power from the edge plasma, known as edge-localized modes (ELMs). While several candidate regimes have been presented in the literature, very little has been published on the characteristics of the small ELMs themselves. One such small ELM regime, also known as the Type V ELM regime, was recently identified in the National Spherical Torus Experiment [M. Ono, S. M. Kaye, Y.-K. M. Peng et al., Nucl. Fusion 40, 557 (2000)]. In this paper, the spatial and temporal structure of the Type V ELMs is presented, as measured by several different diagnostics. The composite picture of the Type V ELM is of an instability with one or two filaments that rotate toroidally at ∼5–10km∕s, in the direction opposite to the plasma current and neutral beam injection. The toroidal extent of Type V ELMs is typically ∼5m, whereas the cross-field (radial) extent is typically ∼10cm (3cm), yielding a portrait of an electromagnetic, ribbon-like perturbation aligned with the total magnetic field. The filaments comprising the Type V ELM appear to be destabilized near the top of the H-mode pedestal and drift radially outward as they rotate toroidally. After the filaments come in contact with the open field lines, the divertor plasma perturbations are qualitatively similar to other ELM types, albeit with only one or two filaments in the Type V ELM versus more filaments for Type I or Type III ELMs. Preliminary stability calculations eliminate pressure driven modes as the underlying instability for Type V ELMs, but more work is required to determine if current driven modes are responsible for destabilization.

Characterization of coherent magnetic fluctuations in JFT-2M high recycling steady high-confinement mode plasmas

The most important feature of the “high recycling steady” (HRS) high-confinement mode (H-mode) regime in the JFT-2M tokamak [H. Ninomiya et al., Fusion Science and Technology (2002), Vol. 42, p. 7] is the stationary pedestal condition in the absence of the large edge localized modes (ELMs). Accompanying the HRS H-mode transition, the coherent magnetic fluctuations in the frequency range of the order of 10–100kHz with significant variation are seen on the magnetic probes at the vessel wall. Above all, two types of edge magnetohydrodynamics (MHD) activities, which have associated toroidal mode number of n=1 and n∼7, respectively, seem to be more important for the HRS H-mode plasmas. To investigate their interaction, bispectral analysis is applied for the magnetic probe data, exhibiting the phase-coupled oscillations between two types of edge MHD activities. Parameter similarities and differences with small ELMs and stationary ELM-free H-mode regimes in other devices are also discussed.

Small ELM regimes with good confinement on JET and comparison to those on ASDEX Upgrade, Alcator C-mod and JT-60U

Since it is uncertain if ITER operation is compatible with type-I edge localized modes (ELMs), the study of alternative ELM regimes is an urgent issue. This paper reports on experiments on JET aiming to find scenarios with small ELMs and good confinement, such as the type-II ELMs in ASDEX Upgrade, the enhanced D-alpha H-mode in Alcator C-mod or the grassy ELMs in JT-60U. The study includes shape variations, especially the closeness to a double-null configuration, variations of q95, density and beta poloidal. H-mode pedestals without type-I ELMs have been observed only at the lowest currents (≤1.2 MA), showing similarities to the observations in the devices mentioned above. These are discussed in detail on the basis of edge fluctuation analysis. For higher currents, only the mixed type-I/II scenario is observed. Although the increased inter-ELM transport reduces the type-I ELM frequency, a single type-I ELM is not significantly reduced in size. Obviously, these results do question the accessibility of such small ELM scenarios on ITER, except perhaps the high beta-poloidal scenario at higher q95, which could not be tested at higher currents at JET due to limitations in heating power.

Observation of a high performance operating regime with small edge-localized modes in the National Spherical Torus Experiment

We report the observation of a high performance scenario in the National Spherical Torus Experiment with very small edge-localized modes (ELMs). These ELMs, individually, have no measurable impact on the stored energy and are observed by several diagnostics. The small ELMs have clear differences as compared with the ELM types reported in the literature, and this operating mode has distinct features compared with other high performance tokamak scenarios with little or no ELMs. The ELM is termed as ‘type V’, and it has a short-lived n = 1 magnetic precursor oscillation rotating counter to the plasma current and a distinct signature on the soft x-ray system. If we could extrapolate it, this scenario would provide an attractive operating regime for next step fusion experiments.

Interplay between ballooning and peeling modes in simulations of the time evolution of edge localized modes

The time evolution of edge localized modes (ELMs) in the Joint European Torus tokamak [P. H. Rebut et al., Nucl. Fusion 25, 1011 (1985)] is investigated using the JETTO predictive modeling code [M. Erba et al., Plasma Phys. Controlled Fusion 39, 261 (1997)]. It is found that both pressure-driven ballooning and current-driven peeling modes can play a role in triggering the ELM crashes. In the simulations carried out, each large ELM consists of a sequence of quasicontinuous small ELM crashes. Each sequence of ELM crashes is separated from the next sequence by a relatively longer ELM-free period. The initial crash in each ELM sequence can be triggered either by a pressure-driven ballooning mode or by a current-driven peeling mode, while the subsequent crashes within that sequence are triggered by current-driven peeling modes, which are made more unstable by the reduction in the pressure gradient resulting from the initial crash. The HELENA and MISHKA ideal magnetohydrodynamic stability codes [A. B. Mikhailovskii et al., Plasma Phys. Rep. 23, 713 (1997)] are used to validate the stability criteria used in the JETTO simulations. This stability analysis includes infinite-n ideal ballooning, finite-n ballooning, and low-n kink∕peeling modes.

Reduction of divertor heat load in JET ELMy H-modes using impurity seeding techniques

The main objective of this paper is investigation of methods for reduction of divertor heat loads in order to increase the lifetime of divertor tiles in future fusion reactors. Special emphasis is given to studies of reduction of transient heat loads due to edge localized modes (ELMs). Two methods are compared: argon seeded type-I ELMy H-modes and nitrogen seeded type-III ELMy H-modes. In both scenarios, the impurity seeding leads to a reduction in the pedestal energy and hence a reduction in the energy released by the ELM. This consequentially reduces the power load to the divertor targets. At high radiative power fractions in type-III ELMy H-modes, part of that released ELM energy (small ELMs, below 20 kJ) is dissipated by radiation in the scrape off layer (SOL). Modelling of the ELM mitigation supports the experimental findings. This ELM mitigation by radiative dissipation is not effective for larger ELMs. In between ELMs, the plasma is detached and radiates strongly from the X-point region. During an ELM, the nitrogen in the X-point and divertor region becomes ionized into more weakly radiating higher charge states and the plasma re-attaches for large ELMs. At JET, argon radiates predominantly in the main plasma and not so much in the cold divertor region. Hence, the effect of radiative dissipation of ELM heat fluxes by argon is very low due to the limited argon density in the divertor region. Nevertheless, both scenarios might be compatible with an integrated ITER scenario, with respect to acceptable divertor lifetime and acceptable confinement.

Edge localized mode physics and operational aspects in tokamaks

Recent progress in experimental and theoretical studies of edge localized mode (ELM) physics is reviewed for the reactor relevant plasma regimes, namely the high confinement regimes, that is, H-modes and advanced scenarios.Theoretical approaches to ELM physics, from a linear ideal magnetohydrodynamic (MHD) stability analysis to non-linear transport models with ELMs are discussed with respect to experimental observations, in particular the fast collapse of pedestal pressure profiles, magnetic measurements and scrape-off layer transport during ELMs.High confinement regimes with different types of ELMs are addressed in this paper in the context of development of operational scenarios for ITER. The key parameters that have been identified at present to reduce the energy losses in Type I ELMs are operation at high density, high edge magnetic shear and high triangularity. However, according to the present experimental scaling for the energy losses in Type I ELMs, the extrapolation of such regimes for ITER leads to unacceptably large heat loads on the divertor target plates exceeding the material limits. High confinement H-mode scenarios at high triangularity and high density with small ELMs (Type II), mixed regimes (Type II and Type I) and combined advanced regimes at high βp are discussed for present-day tokamaks. The optimum combination of high confinement and small MHD activity at the edge in Type II ELM scenarios is of interest to ITER. However, to date, these regimes have been achieved in a rather narrow operational window and far from ITER parameters in terms of collisionality, edge safety factor and βp.The compatibility of the alternative internal transport barrier (ITB) scenario with edge pedestal formation and ELMs is also addressed. Edge physics issues related to the possible combination of small benign ELMs (Type III, Type II ELMs, quiescent double barrier) and high performance ITBs are discussed for present-day experiments (JET, JT-60U, DIII-D) in terms of their relevance for ITER. Successful plasma edge control, at high triangularity (∼0.5) and high density (∼0.7nGR), in ITB scenarios in JET is reported.Active control of ELMs by edge current, pellet injection, impurities and external magnetic perturbations creating an ergodic zone localized at the separatrix are discussed for present-day experiments and from the perspective of future reactors.

High-confinement-mode edge stability of Alcator C-mod plasmas

For steady state high-confinement-mode (H-mode) operation, a relaxation mechanism is required to limit build-up of the edge gradient and impurity content. Alcator C-Mod [Hutchinson et al., Phys. Plasmas 1, 1511 (1994)] sees two such mechanisms—EDA (enhanced D-alpha H mode) and grassy ELMs (edge localized modes), but not large type I ELMs. In EDA the edge relaxation is provided by an edge localized quasicoherent (QC) electromagnetic mode that exists at moderate pedestal temperature T<500 eV, high pedestal density, and high edge safety factor, q95>3.5, and does not limit the buildup of the edge pressure gradient. The q boundary of the operational space of the mode depends on plasma shape, with the q95 limit moving down with increasing plasma triangularity. At high edge pressure gradients and temperatures the mode is replaced by broadband fluctuations ( f<50 kHz) and small irregular ELMs are observed. Ideal MHD (magnetohydrodynamic) stability analysis that includes both pressure and current driven edge modes shows that the discharges where the QC mode is observed are stable. The ELMs are identified as medium n (10<n<50) coupled peeling/ballooning modes. The predicted stability boundary of the modes as a function of pedestal current and pressure gradient is reproduced in experimental observations. The measured dependence of the ELMs’ threshold and amplitude on plasma triangularity is consistent with the results of ideal MHD analysis performed with the linear stability code ELITE [Wilson et al., Phys. Plasmas 9, 1277 (2002)].

High density, high performance high-confinement-mode plasmas in the Joint European Torus (JET)

Recent experiments at the Joint European Torus [Rebut et al., Fusion Eng. Des. 22, 7 (1993)] aim to improve confinement quality in high-confinement-mode (H-mode) plasmas at high densities. Energy confinement time as predicted by the International Thermonuclear Experimental Reactor ITER-H98(y,2) scaling at densities near or in excess of 85% of the Greenwald density limit scaling has been obtained by (i) strong plasma shaping (triangularity 0.35<δ<0.5), or (ii) impurity seeding, or (iii) high-field side pellet injection. Slow peaking of central density without confinement degradation is observed. Loss of sawteeth and core impurity accumulation is prevented by central ion cyclotron resonance heating. In high triangularity and impurity seeded plasmas, reduction of average power loss associated with type I edge localized modes (ELMs) is found which is attributed to the occurrence of additional losses in between ELMs. Broad band magnetic fluctuations are seen which are reminiscent of regimes with small ELMs in other tokamaks. Plasma configurations have been varied to find best combinations of edge pedestal parameters and small ELM losses.

H-mode pedestal characteristics and MHD stability of the edge plasma in Alcator C-Mod

Under most operational conditions of Alcator C-Mod, the dominant type of H-mode is the steady state enhanced Dα (EDA) mode, characterized by good energy confinement, continuously degraded impurity confinement and absence of regular edge localized modes (ELMs). In this regime, a quasicoherent (QC) electromagnetic mode (kθ~5 cm-1, f~100 kHz) is observed, localized in the region of the density pedestal. Experimental evidence suggests that the mode is responsible for enhancement of particle transport. It is shown experimentally that the QC mode can exist in a well defined region in edge temperature-safety factor space, favouring high edge q values (q95⩾3.5) and requiring moderate pedestal temperature Teped ⩽ Tc (Tc~400 eV). As edge temperature and pressure gradient increase, the QC mode is replaced by broadband low frequency fluctuations (f<50 kHz). Small grassy ELMs are observed in these discharges. Analysis of ideal ballooning stability of the C-Mod edge shows that the edge pressure gradient is not limited by infinite n ideal ballooning mode if edge bootstrap current, even reduced by high edge collisionality, is taken into account. Linear stability analysis of coupled ideal peeling/ballooning medium n modes, driven by combination of edge pressure and current gradients, shows that the modes become marginally unstable at C-Mod edge in the range of pressure gradients where the grassy ELMs are observed (∇P⩾1.2×107 Pa Wb-1 rad). This result is consistent with a model of the ELMs as intermediate n peeling/ballooning modes. On the other hand, demonstrated stability of the ideal modes in EDA regime, together with the fact that low Teped is required for its existence, supports theoretical models showing resistive character of the QC mode.

The physics of large and small edge localized modes

The operational boundaries for the occurrence ofvarious types of edge localized mode (ELM) instabilityin high-confinement-mode (H-mode) fusion plasmas are reviewed.The phenomenology of ELMs and the occurrence of differentELM types in parameter space are described.Favourable ELM regimes are observed in various experimentswhich combine small ELM losses withhigh edge pressure and thereby allow for good H-mode confinementeven in situations where core confinement is strongly dependenton edge parameters.Recent progress of ELM models based on ideal or resistive MHDis described and their ability to describethe various ELM regimes discussed.