Operational planning of magnetic confinement fusion power plants using a MIP unit-commitment model

Once magnetic confinement fusion power plants are commercialized, they have to compete with other energy supply technologies. Energy system planning models can give insights into the operation and interaction of such supply technologies. Due to their specific operational behavior and power requirements, fusion power plants need dedicated consideration in energy system modeling. In this work, we present an approach for multi-state techno-economic modeling of fusion power plants to enable the estimation of their optimal dispatch. Models of tokamak and stellarator type plants have been developed in a mixed-integer programming framework for unit commitment, which optimizes the operation of a complex of power plants by minimizing their variable costs. Multiple reactor states - hot, dwell, production-start and production - and their corresponding losses are defined. The operation is first validated in a simplified single-node model. Next, the conclusions are transferred to a European power system model for the year 2050 of much higher spatial and structural complexity. The impact of various design pathways on operational strategy is examined. The results show that both reactor types would operate in a load-following mode. By using assumed 20 GW_el of installed fusion power, 20 % of CO₂ emissions can be saved in the European electricity market compared to a scenario without fusion. Significant seasonal differences in the fusion power plant operation are observed. The tokamak reactor operates on average 4,043 cycles per year (2 h pulse length), while the stellarator reactor operates on average 474 cycles per year.