Jet Pump Noise Analysis for BWRs

The use of noise analysis for detection of BWR component malfunction is a powerful tool in determining abnormal operation conditions, during the life of a nuclear power plant. Since the 80s, several nuclear reactors have reported problems related with jet pumps and recirculation loops. The NRC, in the IE Bulletin 80-07, recommended performing periodic monitoring to individual pressure drop jet pumps, to prevent structural failure. In this work, noise analysis methods are used for detection of jet pumps abnormal operation conditions in a BWR. Power signals obtained from the backup process computer of a BWR are analyzed with a home-developed software, called NOISE, for noise diagnostic of power signals. The computer program takes individual signals from the tabular report of the process computer. The normalized power spectral density (NPSD) is then obtained, using a Prime Factor Algorithm to calculate the Fast Fourier Transform. The NPSD of the jet pumps pressure drop, of Unit 2 of the Laguna Verde Nuclear Power Plant, showed a noticeable change in jet pump 6 during 2003, considering the period from the startup test to operation during 2003. This abnormal condition was due to that the jet pump throat was partially blocked. The noise analysis methodology is shown to be a useful tool for malfunction detection, and could be applied to create a data bank for monitoring the dynamic behavior of BWR jet pumps.