Fusion alpha power control optimisation using RAPTOR

Abstract

The Spherical Tokamak for Energy Production (STEP) is a prototype fusion power plant planned to be operational in the 2040s. STEP’s interaction with the national grid and its responsiveness to demand hinges largely on the effective control of the energy released during the fusion burn phase, i.e., the alpha power, which acts as the dominant form of heating in STEP. This research explores novel trajectory optimisation and control methods to regulate the alpha power, while maintaining other global and local plasma parameters such as internal inductance and safety factor. The RApid Plasma Transport simulatOR (RAPTOR) code is used to self-consistently solve four coupled, 1D PDE equations for poloidal flux diffusion, electron thermal transport, ion thermal transport and electron particle transport. Since STEP has limited space for a central solenoid, the flattop stationary state must be sustained with 100% non-inductive sources (predominately bootstrap current and electron cyclotron heating and current drive). To this end we also introduce a new Dirichlet boundary condition (BC) for the flux diffusion equation which allows the plasma boundary flux to be used as an actuator rather than plasma current (Neumann BC). The RAPTOR code open loop optimisation framework is used to solve this multi-objective, constrained, non-linear, finite-time optimal control problem.