Bivariate empirical mode decomposition applied to the estimation of out-of-phase oscillations in BWR

Abstract

In this paper a new method based on the bivariate empirical mode decomposition to estimate the phase of regional (out-of-phase) or global (in-phase) modes associated with instabilities in boiling water reactors (BWR), is explored. The proposed method allows decomposing the analyzed signal (constructed from two different Local Power Range Monitors, LPRMs) in different levels or intrinsic mode functions (IMF). The estimation of the phase between these LPRM signals can be achieved by tracking the modes associated to the instability of the BWR and obtaining the cross-correlation function of their corresponding IMF. This phase determines possible out-of-phase oscillations, which play an important role in the BWR instability. The method is relatively simple to implement and it does not represent a high computational complexity. The methodology was tested with simulated signals and validated with two events reported in the Forsmark and Ringhals stability benchmarks. The results of the cases studied show that the proposed method clearly contributes on the fact to detect possible cases of out-of-phase oscillations.