Table of contents

Short ∼3 ms pulses of 80 keV deuterium neutrals are injected at three different tangency radii into the national spherical torus experiment. The confinement is studied as a function of tangency radius, plasma current (between 0.4 and 1.0 MA), and toroidal field (between 2.5 and 5.0 kG). The jump in neutron emission during the pulse is used to infer prompt losses of beam ions. In the absence of MHD, the neutron data show the expected dependences on beam angle and plasma current; the average jump in the neutron signal is 88±39% of the expected jump. The decay of the neutron and neutral particle signals following the blip are compared to the expected classical deceleration to detect losses on a 10 ms timescale. The temporal evolution of these signals are consistent with Coulomb scattering rates, implying an effective beam-ion confinement time ≳100 ms. The confinement is insensitive to the toroidal field despite large values of ρ∇B/B (≲0.25), so any effects of non-conservation of the adiabatic invariant μ are smaller than the experimental error.

The Helias ignition experiment is an upgraded version of the Wendelstein 7-X experiment. The magnetic configuration is a four-period Helias configuration (major radius 18 m, plasma radius 2.0 m, B = 4.5 T), which presents a more compact option than the five-period configuration. Much effort has been focused on two versions of the four-period configuration. One option is the power reactor HSR4/18 providing at least 3 GW of fusion power and the second is the ignition experiment HSR 4/18i aiming at a minimum of fusion power and the demonstration of self-sustaining burn. The design criteria of the ignition experiment HSR 4/18i are the following: The experiment should demonstrate a safe and reliable route to ignition; self-sustained burn without external heating; steady-state operation during several hundred seconds; reliability of the technical components and tritium breeding in a test blanket. The paper discusses the technical issues of the coil system and describes the vacuum vessel and the shielding blanket. The power balance will be modelled with a transport code and the ignition conditions will be investigated using current scaling laws of energy confinement in stellarators. The plasma parameters of the ignition experiment are: peak density 2–3×10²⁰ m⁻³, peak temperature 11–15 keV, average beta 3.6% and fusion power 1500–1700 MW.

Ion heating experiments have been carried out in the large helical device using ECH (82.5, 84.0, 168 GHz, ⩽1 MW), ICRF (38.5 MHz, ⩽2.7 MW) and NBI (H° beam: 160 keV, ⩽9 MW). The central ion temperature has been observed from the Doppler broadening of Ti XXI (2.61 Å) and Ar XVII (3.95 Å) x-ray lines, which are measured using a newly installed crystal spectrometer with a charge-coupled device. Recently, in ECH discharges, on-axis heating became possible. As a result, a high T_e(0) of 6–10 keV and a high ion temperature of 2.2 keV were obtained at n_e = 0.6×10¹³ cm⁻³. A clear increment of T_i was also observed with the enhancement of the electron–ion energy flow when the ECH pulse was added to the NBI discharge. These results demonstrate the feasibility towards ECH ignition. A clear T_i increment was observed also in ICRF discharges at low density ranges of (0.4–0.6)×10¹³ cm⁻³ with appearance of a new operational range of T_i(0) = 2.8 keV > T_e(0) = 1.9 keV. In low power ICRF heating (1 MW), the fraction of bulk ion heating is estimated to be 60% of the total ICRF input power, which means P_i>P_e. Higher T_i(0), up to 3.5 keV, was obtained for a combined heating of NBI (<4 MW) and ICRF (1 MW) at density ranges of (0.5–1.5)×10¹³ cm⁻³. The highest T_i(0) of 5 keV was recorded in Ne NBI discharges at n_e<1×10¹³ cm⁻³ with the achievement of T_i(0)>T_e(0), whereas the T_i(0) remained at relatively low values of 2 keV in H₂ and He NBI discharges due to less P_i. The main reasons for the high T_i achievement in the Ne discharges are: (1) 30% increment of deposition power, (2) increase in P_i/n_i (five times, P_i/n_i≫P_e/n_e, P_i<P_e) and (3) increase in τ_ei (three times). The obtained T_i(0) data can be plotted by a smooth function of P_i/n_i. This result strongly suggests that the ion temperature increases even in the H₂ discharge if the P_i can be raised up.

Edge localized modes (ELMs) are commonly observed in high-energy confinement tokamak plasmas and are thought to be caused by magnetohydrodynamic instabilities driven by the steep pressure gradient and the current in the plasma edge region. Our data show that the divertor magnetic balance, i.e. the degree to which the plasma topology resembles a single-null or a double-null, strongly determines where ELM pulses driven by ballooning instabilities at the plasma edge are distributed to surrounding vacuum vessel surfaces. These data also support the conclusions drawn from the stability analysis that ELMs are generated almost entirely on the outboard side of the main plasma.

A collaboration with a Japanese institute in the field of plasma–wall interaction and dusty plasma has been formed in order to understand the physical properties of edge plasma. Results of the theoretical study on dusty plasma and the experimental study on GAMMA10 plasma are presented in this paper. Part A deals with the results obtained from the theoretical investigation of the properties and excitation of low-frequency electrostatic dust modes, e.g. the dust-acoustic (DA) and dust-lower-hybrid (DLH) waves, using the fluid models. In this study, dust grain charge is considered as a dynamic variable in streaming magnetized dusty plasmas with a background of neutral atoms. Dust charge fluctuation, collisional and streaming effects on DA and DLH modes are discussed. Part B deals with the results of the plasma control experiment in a non-axisymmetric magnetic field region of the anchor cell of GAMMA10. The observations, which indicate the comparatively low-temperature plasma formation in the anchor cell, are explained from the viewpoint of enhanced outgassing from the wall due to the interaction of the drifted-out ions. The drifting of ions is thought to be due to the effect of a local non-axisymmetric magnetic field. Experimental results on the control of the wall–plasma interaction by covering the flux tube of a non-axisymmetric magnetic field region by conducting plates are given. Possible influences of the asymmetric magnetic field and conducting plates on the GAMMA10 plasma parameters are discussed.

JET carbon screening experiments were performed using methane gas injection. L-Mode experiments scanned parameters influencing the JET scrape-off-layer (SOL) and/or intrinsic impurity level. Scaling relations are derived to describe methane injected into L-Mode plasmas from the JET horizontal mid-plane. L-Mode screening was 3–20 times better for plasmas connected to the divertor than for similar limited plasmas. The screening was worse for methane injection from the mid-plane and best for injection from the divertor. The screening was 1.5–2 times worse for H-Mode than L-Mode. Both ELM-averaged and inter-ELM H-Mode screening was documented. The screening results were used to understand the intrinsic impurity levels. Z_eff reduced at higher densities partly due to better carbon screening at the higher density, and partly due to decreased carbon influxes. Diverted L-Mode intrinsic carbon levels arose from both main chamber and divertor sources, while H-mode carbon primarily originated from the divertor. DIVIMP and EDGE2D were used to model the observed screening. The modelling indicated that carbon removal to the divertor required lower temperatures for Coulomb collisions to couple the impurity ions to the SOL deuterium flows. The carbon removal occurred primarily in the outer SOL regions.

The values of Q = (fusion power)/(auxiliary heating power) predicted for ITER by three different methods are compared. The first method utilizes an empirical confinement-time scaling and prescribed radial profiles of transport coefficients; the second approach extrapolates from specially designed ITER similarity experiments, and the third approach is based on partly theory-based transport models. The energy confinement time given by the ITERH-98(y, 2) scaling for an inductive scenario with a plasma current of 15 MA and a plasma density 15% below the Greenwald density is 3.7 s with one estimated technical standard deviation of ±14%. This translates, in the first approach, for levels of helium removal, and impurity concentration, that, albeit rather stringent, are expected to be attainable, into an interval for Q of [6–15] at the auxiliary heating power, P_aux = 40 MW, and [6–30] at the minimum heating power satisfying a good confinement ELMy H-mode. All theoretical transport-model calculations have been performed for the plasma core only, whereas the pedestal temperatures were taken as estimated from empirical scalings. Predictions of similarity experiments from JET and of theory-based transport models that we have considered—Weiland, MMM, and IFS/PPPL—overlap with the prediction using the empirical confinement-time scaling within its estimated margin of uncertainty.

The ferromagnetic and resistive wall effects on the beta limit in a high aspect ratio tokamak are investigated. It is shown that the beta limit is reduced to 90% of that without ferromagnetic effect for a high aspect ratio tokamak, when the relative permeability of the ferromagnetic wall is 2. In these analyses, parabolic profiles for both plasma current and pressure are employed and the radius and thickness of the resistive wall are r_w = 1.43a and d = 0.07a (a is plasma minor radius), respectively. The stability window with respect to the external kink modes is shown to be reduced by a high permeability effect in the case of a uniform current tokamak plasma, which is different from the case where only the finite resistivity effect is considered. The effect of toroidal plasma flow is also investigated, and it is shown that the toroidal background flow velocity of 0.3v_pa, v_pa is poloidal Alfvén velocity, is sufficient for the resistive wall to have the stability effect of an ideal wall. The ferromagnetic effect of the wall destabilizes both resistive wall and ideal kink modes.

Plasma beta in magnetized target fusion systems is sometimes much greater than 1, and the plasma may be in direct contact with the imploding liner. Plasma processes are strongly dominated by inter-particle collisions. Under such conditions, the plasma microturbulence, behaviour of α-particles, and plasma equilibria are very different from conventional fusion systems. This paper contains the most comprehensive analysis of the corresponding phenomena to date. Two-dimensional numerical simulations of plasma convection in the targets of a diffuse pinch type demonstrate an onset of convection in this configuration.

Two examples of non-perturbative models of intermittency in drift-wave (DW) turbulence are presented. The first is a calculation of the probability distribution function (PDF) of ion heat flux due to structures in ion temperature gradient turbulence. The instanton calculus predicts the PDF to be a stretched exponential. The second is a derivation of a bi-variate Burgers equation for the evolution of the DW population density in the presence of radially extended streamer flows. The PDF of fluctuation intensity avalanches is determined. The relation of this to turbulence spreading, observed in simulations, is discussed.

H-modes are routinely obtained in the National Spherical Torus Experiment (NSTX) and have become a standard operational scenario. L–H transitions triggered by NBI heating have been obtained over a wide parameter range in I_p, B_t, and _e in either lower-single-null (LSN) or double-null (DN) diverted discharges. Edge localized modes are observed in both configurations but the characteristics differ between DN and LSN, which also have different triangularities (δ). An H-mode duration of 500 ms was obtained in LSN, with a total pulse length of ∼1 s. Preliminary power threshold studies indicate that the L–H threshold is between 600 kW and 1.2 MW, depending on the target parameters. Gas injector fuelling from the centre stack (i.e. the high toroidal field side) has enabled routine H-mode access, and comparisons with low-field side (LFS) fuelled H-mode discharges show that the LFS fuelling delays the L–H transition and alters the pre-transition plasma profiles. Gas puff imaging and reflectometry show that the H-mode edge is usually more quiescent than the L-mode edge. Divertor infrared camera measurements indicate up to 70% of available power flows to the divertor targets in quiescent H-mode discharges.

The physics of internal transport barrier (ITB) formation in JET has been investigated using micro-stability analysis, profile modelling and turbulence simulations. The calculation of linear growth rates shows that magnetic shear plays a crucial role in the formation of the ITB. Shafranov shift, ratio of the ion to electron temperature, and impurity content further improve the stability. This picture is consistent with profile modelling and global fluid simulations of electrostatic drift waves. Turbulence simulations also show that rational q values may play a special role in triggering an ITB. The same physics also explains how double internal barriers can be formed.

High performance scenarios in the HL-2A tokamak are studied by numerical modelling. Through shifting the plasma column outwards, a shaped plasma with significant triangularity is achieved with sufficient room left for the RF antenna. For the out-shifted, shaped plasma, ripple loss of high energy ions during neutral beam injection (NBI) is analysed, and the results show that the ripple loss fraction of NBI power for the shaped plasma is no higher than that for the unshifted circular plasma. The time dependent TRANSP code is used to model realistic reversed magnetic shear (RS) operation in such plasmas. In order to sustain the RS operation towards steady-state, an off-axis current drive with a lower hybrid wave at 2.45 GHz is used to control the current profile. A steady-state RS discharge is formed and sustained until the LH power is turned off; the plasma confinement is enhanced with the development of an internal transport barrier. In the RS discharges with shaped plasma geometry, a double transport barrier is developed. To understand the underlying physics for the current profile control with LHCD in HL-2A, the LH wave deposition in plasmas with RS is analysed.

Correlation between ion/electron distribution functions and device performance, i.e. potential structure, density profile and neutron production rate, in spherical inertial electrostatic confinement plasmas is studied by solving the Poisson equation for various deuteron and electron distribution functions. For several combinations of the ion and electron convergences, dependence of the total neutron production rate on discharged current is discussed. It is shown that when electrons have high convergence and energetic component compared with ions, the neutron production rate can increase in proportion to more than a power of the discharged current, even if the neutron production is sustained mainly by the fusion reactions between the beam (deuteron) and background (deuterium) gases.

Steady-state and transient power deposition profiles have been measured in the JET MIIGB divertor using improved diagnostic techniques involving the use of fast infrared (IR), thermocouples and Langmuir probe arrays. In unfuelled type I ELMy H-modes a very narrow power profile is observed at the outer target, which we associate with the ion channel. Systematic parameter scans have been carried out and our analysis shows that the average power width scaling is consistent with a classical dependence of perpendicular transport in the SOL. Using the fast IR capability, factors such as rise time, broadening, variability and in/out asymmetry have been studied and lead to the conclusion that, within the extrapolation errors, type I ELMs in ITER are likely to have energies comparable to the ablation limits of the divertor material. JET disruptions are very different from type I ELMs in that only a small fraction of the thermal energy reaches the divertor and what does arrive is distributed uniformly over the divertor area. This is very different from the current ITER assumption that puts most of the energy from the thermal quench onto the divertor strike points.

The expected approval for the construction of ITER will have great impact on the way magnetic confinement research will be pursued. Many national laboratories will have to, at least partially, refocus their programmes, as they have to procure various subsystems for ITER. This means a gradual move from physics to more technological R and D and construction of hardware. This will be reflected in the publication output of the fusion community. Nuclear Fusion will anticipate this change by publishing special issues related to the more technological IAEA Technical Meetings and to other workshops on plasma engineering issues such as heating systems, diagnostic systems, feedback control aspects and subsequent test results. The greater emphasis on technology will also be reflected in the increase of Nuclear Fusion articles related to the technology contributions to the IAEA Fusion Energy Conferences. However, our Board of Editors advises that the chosen technology subjects should be related to the more traditional, physics, subjects of Nuclear Fusion.

This new line is reflected in the choice of special issues for the end of 2003/beginning of 2004:

This issue: Aspects of steady state operation (3rd IAEA TM Arles/Greifswald, May 2002);

Electron Cyclotron Waves in Fusion Plasmas (EC-12 and SMP-2002 together);

Overview Papers from the FEC-2002.

Later next year another three special issues of Nuclear Fusion are planned:

The ITER Physics Data Base Revised (this is an initiative of the IPTA-group);

IFSA-2003: results of ICF research presented in Monterey on 7-12 September 2003;

Workshop on Stochasticity in Fusion Edge Plasmas (SEP), 6-8 October 2003, Jülich.

This current issue of Nuclear Fusion contains a number of regular articles as well as a group of 11 articles related to Steady State Operation of Magnetic Fusion Devices from the 3rd IAEA Technical Meeting (Arles/Greifswald, 2-7 May 2002)

These 11 articles are related only to a subset of all the presentations given. Most of the presentations were too specialized or otherwise not suitable to be reworked into an article of interest for the general readership of Nuclear Fusion. The selection of articles addresses the following issues:

experimental results of the long pulse operation of Tore Supra and DIII-D;

simulations of long pulse scenarios for Tore Supra operation;

state-of-the-art feedback control of JET plasmas;

technology aspects of long pulse NB-heating and cryogenics (W7-X);

plans for steady state devices in South Korea (KSTAR) and China (HT-7U).

A short review of diagnostics and techniques for active power exhaust control is given from the point of view of experiences gained at Tore Supra. The physical and technological aspects are presented with the aim of illustrating strengths, difficulties and potential for future long pulse and steady state fusion devices such as W7X and ITER-FEAT. One of the main features of the Tore Supra approach to power exhaust control is the reliance on infrared (IR) thermography to survey nearly 100% of the high flux target zones. This is applied to the largest elements, the toroidal pump limiter and the heating antennae, by a number of IR endoscopes. For high flux target areas inaccessible to direct view, optical fibres in the IR and the near-IR range are used. The large wall panels, which evacuate the rest of the power, are controlled by calorimetry. Bolometry is used to achieve a complete power balance. These measurements can be inserted in control loops acting typically on the injected heating power or the radiation cooling.

A prototype ion cyclotron range of frequency (ICRF) antenna with RF power of 6 MW has been developed for the long pulse (300 s), high power operation in the Korea superconducting tokamak advanced research (KSTAR) tokamak. Cooling paths in the antenna were carefully designed to remove the dissipated RF power loss. An RF power test has been performed to estimate the standoff capability of the antenna. A high power RF test at a frequency of 30 MHz gives a standoff voltage of 30.5 kVp for 60 s and 23.2 kVp for 300 s (without cooling). During the RF pulse, the peak voltage, forward/reflected powers, temperature of the antenna, and gas pressure are measured. A vacuum feedthrough of 1 MW RF power has been developed, which has two alumina ceramic cylinders and an O-ring seal. For cooling of the ceramic parts, dry air is injected into the ceramic surface through two outer nozzles. Independent cooling water channels are installed to cool the inner conductor of the feedthrough. RF high voltage tests show that stable operation is possible, with a peak voltage of 28.9 kVp for 300 s, without any severe damage.

One of the main goals for the DIII-D research programme is to establish an advanced tokamak (AT) plasma with high bootstrap current fraction that can be sustained in-principle steady-state. Substantial progress has been made in several areas during the last year. The resistive wall mode (RWM) stabilization has been done with spinning plasmas in which the plasma pressure has been extended well above the no-wall beta limit. The 3/2 neoclassical tearing mode (NTM) has been stabilized by electron cyclotron heating (ECH) of the magnetic islands, which drives current to substitute the missing bootstrap current. In these experiments either the plasma was moved or the toroidal field was changed to overlap the ECH resonance with the location of the NTMs. Effective disruption mitigation has been obtained by massive noble gas injection into shots where disruptions were deliberately triggered. The massive gas puff causes a fast and clean current quench with essentially all the plasma energy radiated fairly uniformly to the vessel walls. The run-away electrons that are normally seen accompanying disruptions are suppressed by the large density of electrons still bound on the impurity nuclei. Major elements required to establish integrated, long pulse, AT operations have been achieved in DIII-D: β_T = 4.2%, β_P = 2, f_BS = 65%, and β_NH_89P = 10 for 600 ms (∼4τ_E). The next challenge is to integrate the different elements, which will be the goal for the next five years when additional control will be available. Twelve RWM coils are scheduled to be installed in DIII-D during the summer of 2003. Future plans include upgrading the tokamak pulse length capability and increasing the ECH power, to control the current profile evolution.

The Tore-Supra program is oriented towards the control of multi-megawatt plasmas for very long durations. During its first years of operation, both lower hybrid (LH) waves and ion cyclotron (ICRH) heating have been used, leading to a world record in injected energy of 280 MJ. In these conditions, plasma–wall interactions becomes a critical issue, as the injected power has to be removed continuously, and recycling fluxes must be controlled to maintain the density at its desired value and avoid plasma contamination. On the basis of these results, an enhancement in the capability to handle large input powers on a steady-state basis and to control the particles over long duration has been implemented during the 2000–2001 shut-down. In 2002, experiments will resume with the ultimate goal of operating with steady-state advanced scenarios based on purely non-inductive current drive at moderate density, with both LH current drive and ICRH heating, during hundreds of seconds, leading up to 1 GJ of injected energy.

Scenarios of steady-state, fully non-inductive current in Tore Supra are predicted using a package of simulation codes (CRONOS). The plasma equilibrium and transport are consistently calculated with the deposition of power. The achievement of high injected energy discharges up to 1 GJ is shown. Two main scenarios are considered: a low density regime with 90% non-inductive current driven by lower hybrid waves—lower hybrid current drive (LHCD)—and a high density regime combining LHCD and ion cyclotron resonance heating with a bootstrap current fraction up to 25%. The predictive simulations of existing discharges are also reported.

The stellarator WENDELSTEIN 7-X (W7-X) includes water-cooled plasma facing components (PFCs) to allow steady-state operation and to provide an efficient particle and power exhaust up to 10 MW for a maximum pulse duration of 30 min. Ten divertor units are arranged along the helical edge of the fivefold periodic plasma column. The three-dimensional shape and positioning of the target surfaces are optimized to address physics issues for a wide range of experimental parameters, which influence the topology of the boundary. The three-dimensional target surfaces are reproduced by a series of consecutive plane target elements as a set of parallel water-cooled elements positioned onto the frameworks of target modules. The design and arrangement of target modules and elements are described.

The refrigeration system for the W7-X superconducting magnet and the divertor cryo-vacuum pumps is presented. In total, five main helium cooling circuits have to be supplied by the refrigerator—four for the magnet including auxiliary equipment like support structure, thermal shield and current leads, and one for the cryo-pumps. For the shields of the latter, an additional LN₂—cooling circuit is required. The lowest operating temperature is 3.3 K. It will be provided by evacuating a sub-cooler bath using a cold or warm compressor. Three of the helium cooling circuits use altogether four identical cold circulators. Apart from the current leads which are supplied with the coolant from a LHe storage tank, the peak reserve power required is equal to 7 kW at 4.5 K entropy equivalent. However, this potential maximal demand occurs continuously for periods of only a few hours at most, and altogether for less than 1% of annual time. The refrigerator thus will be designed for 5 kW continuous power at 4.5 K_equiv. corresponding to 1.5 MW compressor connected rating. The reserve peak power will be covered, if necessary, by using the latent heat and vapour enthalpy of LHe from a storage tank. This supporting LHe stream is added to the phase separator and fed subsequently to the low pressure return stream at the cold end of the cold box. LN₂-pre-cooling equipment of the cold box—which is installed for W7-X cool-down anyway—can also be used to increase refrigeration power. The LHe required for maintaining reserve refrigeration power as well as for running the current leads is generally produced overnight when W7-X is in idle current mode.

Equilibrium reconstruction is the essential diagnostic tool for determining the magnetic field and current density of a tokamak discharge. This method parametrizes the unknown current profile with a suitable set of test functions and determines the coefficients from measurements. For steady state discharges, a continuous equilibrium analysis is required. The purpose of the Tore Supra CIEL project is the study of discharges with duration up to 1000 s. We report on the development of a real-time version of the widely used equilibrium code EFIT.

Long pulse operation has an important impact on the design and utilization of a neutral beam injection system. This paper, first describes briefly the injectors designed for ITER FEAT as they are the first to be designed for long pulse operation under conditions approaching those that will be experienced in future machines. The important consequences of long pulse operation on the injector design will be then discussed, and finally some suggestions will be made for future, continuously operating systems.

Fusion energy is an ultimate and inexhaustible source of energy for mankind and is expected to be obtained in controlled operation within this century. Among various possible candidates for fusion, the tokamak is presently the most qualified one, and since it uses superconducting magnetic coils, it will be adequate for steady-state operation. The HT-7U superconducting tokamak is a part of national project in China on fusion research, scheduled to become available on-line by the end of 2004 (Wan Y.X. and HT-7 & HT-7U Groups 2000 Overview of steady state operation of HT-7 and present status of the HT-7U project Nucl. Fusion40 1057). The control system of the HT-7U is designed as a distributed control system (HT7UDCS), including many subsystems that provide the various functions of supervision, remote control, real-time monitoring, data acquisition and data handling. The major features of the HT-7U tokamak, which make long-pulse (∼1000 s) operation possible are the flexible poloidal field (PF) system, an auxiliary heating system, the current-driving system and a divertor system. In order to realize these features simultaneously, real-time data handling and analysis, along with a significant control capability is required. This paper discusses the design of the HT7UDCS.

In order to simultaneously control the current and pressure profiles in high performance tokamak plasmas with internal transport barriers (ITB), a multi-variable model-based technique has been proposed. New algorithms using a truncated singular value decomposition (TSVD) of a linearized model operator and retaining the distributed nature of the system have been implemented in the JET control system. Their simplest versions have been applied to the control of the current density profile in reversed shear plasmas using three heating and current drive actuators (neutral beam injection, ion cyclotron resonant frequency heating and lower hybrid current drive). Successful control of the safety factor profile has been achieved in the quasi-steady-state, on a timescale of the order of the current redistribution time. How the TSVD algorithm will be used in the forthcoming campaigns for the simultaneous control of the current profile and of the ITB temperature gradient is discussed in some detail, but this has not yet been attempted in the present pioneering experiments.