A study on ohmic plasma initiation for JA DEMO

Highlights

• Available loop voltage and prefill gas pressure for plasma initiation in JA DEMO are clarified.

• Conducting shell does not necessarily hinder plasma initiation.

• A possible ohmic plasma initiation scenario for JA DEMO is developed.

Abstract

We have investigated ohmic plasma initiation for JA DEMO, a design concept of the tokamak demonstration fusion reactor examined in Japan. A calculation model describing the tokamak plasma initiation phase consisting of the two-dimensional analysis of the coil and eddy currents and the zero-dimensional plasma analysis is developed. It is revealed that the conducting shell does not necessarily hinder plasma initiation, while it provides the stabilizing effect of vertical displacement events. The available loop voltage and prefill gas pressure for initiation in JA DEMO are clarified for developing the initiation scenario. We demonstrate a possible scenario of ohmic plasma initiation, which is consistent with the design specifications of superconducting poloidal field coils.

Introduction

Tokamak plasma initiation is an important phase as it affects the subsequent plasma current ramp-up phase and reactor designs, especially, of the poloidal field (PF) coil system. In large superconducting tokamak devices, plasma initiation is generally difficult due to a low loop electric field and a large eddy current induced in passive conducting structures. Hence, meticulous consideration is essential for developing the plasma initiation scenario for the design activities of the tokamak DEMO reactor.

The plasma initiation phase achieves breakdown, burnthrough, and the formation of initial tokamak equilibrium. The breakdown phase is characterized by the electron avalanche process and is described analogously to the Townsend avalanche theory [1]. Ionizing and elastic collisions with prefilled neutral particles are dominant in the electron scattering process. The breakdown phase transits to the next phase when the collision frequency between electrons and ions becomes equal to that between electrons and neutrals [2]. The subsequent burnthrough phase aims for full ionization of the main gas. To raise the electron temperature and achieve full ionization, the plasma heating power must exceed the power of radiation and other losses. Burnthrough analyses have been performed using the zero-dimensional models for developing the initiation scenario and analyzing experimental data [3], [4], [5], [6], [7].

The low loop electric field due to the use of superconducting PF coils limits the prefill gas pressure for plasma initiation. During the initiation phase, the vertical magnetic field in the plasma region needs to be controlled smoothly connecting from the distribution at the end of the initial magnetization to the initial equilibrium configuration, via the null field configuration. The possible maximum loop voltage and the control of the vertical magnetic field are limited by the allowable voltage and change rate of the current of PF coils. They also vary depending on the vertical field distribution and the location of the breakdown region. The maximum rate of change of the coil current is given by the allowable alternating current loss, and the maximum coil voltage is determined by the supply voltage and the voltage endurance of the superconducting coil. The magnetic field experienced by a central solenoid (CS) coil also has an allowable value to maintain the superconducting state. The eddy current is induced in passive conducting structures such as the vacuum vessel (VV) due to the change of the coil current. The large eddy current complicates breakdown by causing loss of electrons from the plasma region by generating the stray magnetic field and reducing the net loop voltage which contributes to breakdown by generating the opposite loop voltage. It is necessary to develop an operation scenario of PF coils which is consistent with the design specifications and allows plasma initiation, considering the effects of the eddy current. It is required for the design of PF coils to be able to obtain such a coil-current scenario. To determine the supply voltage of PF coils, it is essential to clarify the required coil voltage for plasma initiation.

The enlargement of the tokamak adversely affects plasma initiation. The loop electric field at the center of the breakdown region is inversely proportional to the major radius. The large, thick VV increases the eddy current and the time-constant of the decay of the eddy current. The large vessel volume prolongs the time required for the diffusion of the supplied fuel gas. The large plasma makes burnthrough of low-$Z$ impurities more severe [3]. The effects of plasma and vessel sizes on breakdown and burnthrough are thoroughly discussed in [8]. For instance, the maximum prefill gas pressure for burnthrough is proportional to the square root of the ratio of the plasma volume to vessel volume; that is, the larger vessel is disadvantageous for burnthrough when the plasma size is the same.

Demonstration reactors will be larger than ITER and use superconducting PF coils, according to the current DEMO design activities [9], [10]. The conducting shell can be used to stabilize the high-ellipticity plasma [11]; the efficient induction of the eddy current in the conducting shell can have a negative effect on plasma initiation. Thus, plasma initiation of DEMO will be more severe than that of ITER. It is necessary to clarify the possibility of the plasma initiation scenario consistent with the DEMO reactor design and the design specifications of PF coils required for plasma initiation.

In this paper, we investigate ohmic plasma initiation for the current design of JA DEMO [9]. A calculation model consisting of the two-dimensional analysis of the coil and eddy currents and the zero-dimensional plasma analysis is constructed for developing the plasma initiation scenario. The effect of the conducting shell proposed for JA DEMO on plasma initiation is examined. The parameter region of plasma initiation which is consistent with the engineering design is clarified. Based on the obtained results, we illustrate a possible ohmic initiation scenario for JA DEMO.