Optimization-oriented modelling of neutral beam injection for EU pulsed DEMO

Abstract

Neutral beam injection (NBI) is one of the auxiliary power systems considered for the EU DEMO pulsed plasma ('DEMO1'). In this paper, we discuss the characteristics of optimized NBI in terms of the DEMO1 requirements, relating physics and engineering contexts in a novel parameter range compared to current NBI systems and in a larger plasma volume than ITER. Different injection options are investigated to account for various concepts discussed in the literature. The investigation is carried out using a wide range of sensitivity studies by means of the METIS 0.5D transport code and ASCOT Monte Carlo simulations of injected neutral-beam particles. This investigation has generated a series of recommendations for the NBI design, contributing to the system optimization process. We show how tangential injection aimed at the plasma core, with an energy greater than or similar to 800 keV, is recommended to maintain high fusion power performance and fulfill the requirements for bulk heating. Compliance with engineeringconstraints on the NBI design, e.g., the least interference with breeding blanket modules or compatibility with solutions for the beamline components, imposes some restrictions when discussing the beam geometry. The minimum density at which an NBI can be safely operated without harmful shine-through losses is investigated for different injection energies and compared to the ITER case. The NBI operational window for DEMO is shown to be significantly extended to transient, low-density phases, also highlighting the importance of NBI systems designed for modular energy and power output. NBI can therefore sustain the plasma during a considerable portion of the transient phases, i.e., the current ramp-up and ramp-down phases. The final decision on the DEMO heating mix will be made on the basis of system operability and performance: the optimized NBI is shown to be a suitable and effective option for the EU DEMO inductive scenario.