Predictive Multichannel Flux-Driven Modelling to Optimize ICRH Tungsten Control in JET

The evolution of the JET high performance hybrid scenario, including central accumulation of the tungsten (W) impurity, is reproduced with predictive multichannel integrated modelling over multiple confinement times using first-principle based models. Eight transport channels (Ti, Te, nD, nBe, nNi, nW, Vtor, j) are modelled predictively with self-consistent predictions for sources, radiation, and magnetic equilibrium, yielding a predictive system with multiple nonlinearities which can reproduce observed radiative temperature collapse after several confinement times. The mechanism responsible for W accumulation is inward neoclassical convection driven by the main ion density gradients and enhanced by poloidal asymmetries due to centrifugal acceleration. The slow timescale of bulk density evolution sets the timescale for central W accumulation. Prediction of this phenomenon requires a turbulent transport model capable of accurately predicting particle and momentum transport (QuaLiKiz) and a neoclassical transport model including the effects of poloidal asymmetries (NEO) coupled to an integrated plasma simulator (JINTRAC). The modelling capability is applied to optimize the available actuators to prevent W accumulation, and to extrapolate in power and pulse length. Central NBI heating is preferred for high performance, but comes at the price of central deposition of particles and torque which pose the risk of W accumulation. Several benefits of ICRH to mitigate W accumulation are examined: The primary mechanism for ICRH to control W in JET are via its impact on the bulk profiles and turbulent diffusion, which are insensitive to details of the ICRH scheme. High power density near the axis is found to be best to maximize the beneficial effects of ICRH against W, but changing the minority species or its concentration does not significantly change the W behaviour. With attention to the location of the ICRH resonance and MHD stability, high performance hybrid scenario discharges of 5 s at maximum power should be possible in the coming campaign, and a controlled and steady fusion performance in the subsequent JET DT campaign. This work demonstrates the integration of multiple first-principle models into a powerful multichannel predictive tool for the core plasma, able to guide JET scenario development to its objectives of higher performance and longer pulses.