Advanced Modelling And Stability Analysis For Nuclear Reactors

This work focuses on the development and improvement of advanced modelling strategies for the study of traditional and innovative nuclear reactor design. The primary aim is the integration of MOR and DA techniques to provide accurate and computationally efficient algorithms suitable for real-time and control applications. As nuclear reactors present unique features compared to conventional systems, the identification of the best performing advanced modelling strategies remains an ongoing challenge. In this sense, MOR techniques offer a promising solution to the trade-off between solution accuracy and computational times, especially in multi-query simulation scenarios. Their joint use with data-driven techniques can help improve the computational accuracy by integrating the model simulation with the experimental data efficiently. The three techniques considered and developed in this study have been tested on the TRIGAMarkIIreactor, using experimental data and a CFD model of the reactor core, focusing on the prediction of the steady-state beyond the observed data (DMD), the improvement of the prediction compared to the experimental data (POD-KF), and the optimisation of the sensor positioning within the core along with the reconstruction of fields for which measurements are not available (GEIM).