Power Dependence of Electron-Cyclotron Current Drive for Low- and High-Field Absorption in Tokamaks

Abstract

A new relativistic, bounce-averaged Fokker-Planck-quasilinear code shows that for experimentally relevant rf power /7 r f(W/cm 3 )/[« e (10 13 cm _3 )] 2^0 .5, electron-cyclotron current-drive efficiency (and absorptivity) is power dependent; at given plasma radius the current-drive efficiency increases above linear values for absorption of the rf on the low-field side of the cyclotron resonance layer, but decreases and even passes through zero for absorption on the high-field side. Thus, electron-cyclotron current drive by high-field launch will generally not be suitable.PACS numbers: 52.50.Gj, 52.55.Fa A tokamak reactor is a much more attractive concept as a fusion power device if it is run at steady state. This can be accomplished with an efficient source of toroidal current, other than inherently pulsed inductive drive. Electron-cyclotron (EC) waves are particularly suitable for this purpose because they readily penetrate into high-density, high-temperature tokamak plasma, and because the damping process is easily controlled in configuration and velocity space. A limitation is that the frequency of the wave source must be sufficiently high to avoid plasma cutoff. At elevated density this presents technological difficulties. However, the required frequency is reduced for waves absorbed on the inboard, high-field side of the resonance layer, compared to waves absorbed on the outboard, low-field side. Thus, in Ref.1, high-field absorption has been proposed for the heating of the conceptual compact ignition tokamak (CIT) at elevated density (n e -10 15 cm -3 ). Moreover, past work 2 has shown that EC current-drive efficiency with absorption outside the cyclotron layer is degraded in practical situations at temperatures greater than 20 keV by overlap of neighboring cyclotron interactions, which drives current in opposite directions. Unfortunately, as evidenced by the results presented here, the inside launch configuration is not suitable for high-power rf current generation. In this Letter we study the EC current drive as a function of applied power, for cases of absorption on the high-or the low-field side of the resonance layer. We find that current drive by absorption on the highfield side is problematical because of a reduction of current-drive efficiency as a function of increasing applied power, but for absorption on the low-field side the efficiency increases at high power.The relationship between the region of velocity space affected by the waves and the spatial location of the absorption elucidates the dominant physical effects. At a given spatial point in the tokamak, the incoming wave resonates with electron-cyclotron motion satisfying v\\ = (co -l(o ce /y)/k\\, where v\\ is the electron velocity parallel to the magnetic field, co and k\\ are wave frequency and parallel wave number, / is the cyclotron har-monic, co ce is the cyclotron frequency which varies inversely with major radius R, and y is the relativistic mass factor. This resonance condition corresponds to an ellipse in (v\\,v±) space (v=p/m e...

Electron-cyclotron-current-drive efficiency in DEMO plasmas

“…The combination of linear absorption and adjoint determination of the CD efficiency offers a well-benchmarked, reliable tool for the investigation of the CD efficiency. The broad absorption profiles typical of high-CD injection geometries result generally in relatively low power densities, ruling out quasi-linear effects [48] even for an injected power of several hundred MW (except possibly for the very centre of the plasma column, which is however usually not affected by the peak EC deposition in our simulations). An evaluation of synchrotron losses on the CD efficiency [49] is still to be addressed.…”

Benchmarking of codes for electron cyclotron heating and electron cyclotron current drive under ITER conditions

“…The threshold for quasilinear effects was found by Harvey [28] to be when 2q e /n 19 2 > 1, where q e is the power density in MW /m 3 and n 19 is the density in units of 10 19 m "3 . For this standard density case, this criterion is only 4 "10 #4 for the benchmark calculation at a power level of 1 MW; even at 20 MW this criterion is well satisfied.…”

On recent results in the modelling of neoclassical-tearing-mode stabilization via electron cyclotron current drive and their impact on the design of the upper EC launcher for ITER

“…The "standard" analysis is usually based on the solution of either the equations of standard geometrical optics or of extended-ray / paraxial-beam-tracing equations as far as the propagation is concerned, while absorption and current drive calculations are based on linear or quasilinear modelling [31]. In ITER, quasilinear distortions of the electron distribution function are usually found to be very small (consistently with the smallness of the Harvey parameter [32] for standard injection scenarios). Adjoint modelling of current drive including momentum-conserving corrections [33,34], as implemented by now in several codes, also matches well Fokker-Planck results [35].…”