Integrated deterministic and probabilistic safety assessment of a superconducting magnet cryogenic cooling circuit for nuclear fusion applications

The most promising configuration of a nuclear energy fusion system is the tokamak, the largest of which, called ITER, is under construction in Cadarache, France, which uses a complex system of superconducting magnets to generate a field of several tesla (T), aimed at confining the plasma in the toroidal chamber where nuclear fusion reactions occur. For industrial development, the safety of nuclear fusion systems has to be proved and verified by a systematic analysis of operational transients and accidental conditions. Although the final aim of fusion reactors is to reach steady state operation, present-day tokamaks present complex dynamic features, as their operation is based on the transformer principle with a subset of the superconducting magnets operating in a pulsed mode, to inductively generate plasma currents of the order of several MA. We adopt the framework of Integrated Deterministic and Probabilistic Safety Assessment (IDPSA), for identifying the component failures that may cause a Loss-Of-Flow-Accident (LOFA) in the cooling circuit of a superconducting magnet for fusion applications. Post-processing of the simulated scenarios for the identification of the abnormal transients is performed in an unsupervised manner resorting to a spectral clustering approach embedding a Fuzzy-C Means (FCM) that is compared with an Extended Symbolic Aggregate approximation (ESAX) from the literature that also resorts to the FCM for the classification. The proposed approach turns out to be more efficient than ESAX in the identification of clusters and “prototypical states” of abnormal system behavior. Results show that none of the identified scenarios (even those leading to a LOFA) are critical for the ITER central solenoid module integrity, in the mode of operation considered