One-dimensional Simulations of Nonlinear Burn Control in ITER

V. Graber, E. Schuster

33rd Symposium on Fusion Technology (SOFT)

Dublin, Ireland, September 22-27, 2024

Abstract

Actively controlling the plasma temperature and density in future reactor-grade tokamaks will be a formidable challenge due to the multi-dimensional, coupled, and nonlinear characteristics of the burning-plasma dynamics. In ITER, the actuator systems that will be useful for temperature-density control (also known as burn control) include neutral beam injection, ion and electron cyclotron heating, pellet injection, and gas puffing. In this work, a nonlinear, model-based, burn-control algorithm is proposed. A model-based approach is attractive because it directly incorporates the complex plasma dynamics into the burn-control algorithm. Using Lyapunov techniques, the burn-control algorithm is synthesized from a control-oriented plasma model that is zero-dimensional (0-D). The reduced dimensionality of this plasma model renders the burn-control design more feasible. Nonetheless, this 0-D plasma model still captures the nonlinear response of the plasma energy and density to deuterium-tritium (D-T) fusion, radiation, auxiliary heating, external fueling, and other phenomena. More importantly, the control objective is indeed 0-D since the primary goal of a burn controller is usually to regulate volume-averaged properties of the plasma (e.g., overall fusion power Pf or fusion gain Q). This makes 0-D models appropriate for control synthesis. However, the assessment of the effect of the proposed 0-D burn-control algorithms on the spatio-temporal dynamics of the plasma is a critical step of the control-design process. This work presents a one-dimensional (1-D) plasma model that is used in closed-loop simulations. The presented simulation study demonstrates how the spatial profiles of the D-T fuel density, the fusion-born alpha-particle density, the ion energy density, and the electron energy density evolve temporally under 0-D burn control. By testing 0-D controllers in 1-D simulations, the effectiveness of 0-D burn-control techniques can be investigated before implementation in actual tokamaks. Through investigations of this kind, control issues can be identified and addressed, enabling iterative improvement of the burn-control designs.

*Supported by the US DOE under DE-SC0010661.