Detection of flow instability in natural circulation boiling water reactors using instantaneous spectral entropy measure

Natural Circulation Boiling Water Reactors (NCBWR) are vulnerable to thermal-hydraulic instabilities in coolant flow under certain operating conditions. Considering the strong coupling between neutronics and thermal hydraulics, monitoring of such instabilities is crucial for safe and reliable operation of NCBWRs. However, wider frequency range, significant non-stationarity and time-varying nature of flow oscillations makes this a challenging task. In this paper, we introduce a novel method using Instantaneous Spectral Entropy (ISE) to detect and evaluate flow instabilities. Key advantage of the method is its ability to clearly differentiate between random noise and oscillations. Comparison between Short-Time Fourier Transform (STFT) and Continuous Wavelet Transform (CWT) for calculating entropy revealed that STFT provides superior results. Validation with operational data from a thermal hydraulic facility confirmed the method's effectiveness in identifying and assessing time-varying flow instabilities.