Modal processing of the Local Power Range Monitors signals in BWR NPP

In a boiling water reactor (BWR), the power distribution can be varied both axially and radially by control rod manipulation and by void content variations in the moderator. As an example of accidental situations in nuclear power plants (NPP), we consider the power oscillations that can occur in a BWR NPP under certain circumstances. An interesting problem is to study and characterize these instabilities analyzing the neutronic power signals obtained from the Local Power Range Monitors (LPRMs) installed in the reactor core. Several techniques exist to detect and classify the possible oscillations in a BWR. The power modal decomposition method and the signal component analysis methodologies are reviewed. The performance of these techniques is compared analyzing Records 9 and 10 of cycle 14 of the OECD Ringhals 1 Stability Benchmark.

► Analysis of the LPRMs signals using modal decomposition, PCA and ICA methodologies.
► The projections from ICA and modal decomposition are quite similar.
► PCA and ICA can be performed almost online and does not need any extra information.
► A modal decomposition analysis is needed to give physical sense to the projections.
► ICA is more robust and reliable than PCA.