Calculation of electron-cyclotron-current drive (ECCD) with the comprehensive CQL3D Fokker-Planck code for a TCV tokamak shot gives 550 kA of driven toroidal current, in marked disagreement with the 100-kA experimental value. Published ECCD efficiencies calculated with CQL3D in the much larger, higher-confinement DIII-D tokamak are in excellent agreement with experiment. The disagreement is resolved by including in the calculations electrostatic-type radial transport at levels given by global energy confinement in tokamaks. The radial transport of energy and toroidal current are in agreement.

Electron cyclotron current drive (ECCD) is an important prospective tool for tailoring the current profile in nextstep devices. To fill the remaining gaps between ECCD theory and experiment, especially in the efficiency and localization of current drive, a better understanding of the physics of suprathermal electrons appears necessary. In TCV, the fast electron population is diagnosed by a multichordal, spectrometric hard x-ray camera and by a highfield side electron cyclotron emission radiometer. The main modelling tool is the quasilinear Fokker-Planck code CQL3D, which is equipped with a radial particle transport model. Systematic studies of fast electron dynamics have been performed in TCV with modulated or pulsed electron cyclotron power, followed by coherent averaging, in order to identify the roles of collisional relaxation and radial transport in the dynamics of the suprathermal population. A consistent picture is emerging from experiment and modelling, pointing to the crucial role of the radial transport of suprathermal electrons in the physics of ECCD.

With the very high electron cyclotron (EC) wave power density achieved in the TCV tokamak, more than 20 MW m−3, quasilinear modelling predicts an electron cyclotron current drive (ECCD) efficiency well in excess to the experimental value, by up to a factor of 10. Experimentally, radial transport of suprathermal electrons consistent with a diffusion coefficient larger than 1.5 m2 s−1 has been observed. This implies that the radial transport time is of the same order as the electron deflection time, suggesting that the key to resolving the discrepancy is to include radial transport in the kinetic simulations. In this paper we show that with a diffusion coefficient in accordance with the experimental estimation, we can reproduce the observed current drive efficiency in the fully EC current driven plasmas of TCV by solving the Fokker–Planck equation. Experimentally the total wave-driven current is well-known since the current in the Ohmic transformer is set to a constant value. We study the radial profile and the velocity dependence of the radial diffusion coefficient.A specific model is employed for steady-state electron internal transport barriers, produced by off-axis ECCD, with the electrons divided into two groups according to their energy. A small diffusion coefficient is assigned for the low-energy electrons, while at higher energies the diffusion level is chosen such as to obtain the experimental ECCD efficiency. The total current density, which is the sum of the wave-driven part and the bootstrap current, is found to be hollow, supporting the hypothesis that the reversed shear is the cause of the transport barrier.

Improved electron energy confinement in tokamak plasmas, related to internal transport barriers, has been linked to nonmonotonic current density profiles. This is difficult to prove experimentally since usually the current profiles evolve continuously and current injection generally requires significant input power. New experiments are presented, in which the inductive current is used to generate positive and negative current density perturbations in the plasma center, with negligible input power. These results demonstrate unambiguously for the first time that the electron confinement can be modified significantly solely by perturbing the current density profile.

Abstract. Measurements of electron cyclotron emission from the high field side of the TCV tokamak have been made on plasmas heated by second and third harmonic X-mode Electron Cyclotron Heating (ECH) and Electron Cyclotron Current Drive (ECCD). Suprathermal Electron Cyclotron Emission (ECE), up to a factor of 6 in excess of thermal emission, is detected in the presence of second harmonic X-mode (X2) ECCD and of third harmonic X-mode (X3) ECH. The measured ECE spectra are modelled using a bi-Maxwellian describing the bulk and the suprathermal electron populations. Suprathermal temperatures between 10-50keV and densities in the range 1 · 10 17 − 6 · 10 18 m −3 are obtained, and correspond to 3 -15 bulk temperatures and 1% -20% bulk densities. Good agreement between ECE suprathermal temperatures and energetic photon temperatures, measured by a hard X-ray camera, is found. For optically thin X3 Low Field Side (LFS) injection in presence of X2 CO-ECCD, the suprathermal population partly explains the discrepancy between global and first pass absorption measurements.