Vibration, neutron noise and acoustic monitoring in German LWRs

Abstract

Operational experiences in nuclear power plants prove that — in spite of high quality, selected materials and repetitive tests or inspections — defects and failures of structural components cannot be completely excluded during operation. Therefore, on-line monitoring and diagnostic techniques receive an ever increasing interest. In order to guarantee safe plant operation great efforts are spent on further development of prevention measures to avoid malfunctions and system failures. Modern signal analysis methods and state-of-the-art computer technologies can contribute to this aim: more sophisticated, advanced surveillance systems can be developed in an economical way. In Germany monitoring methods for early failure detection and on-line diagnosis at primary circuit components have been developed and are successfully applied for assuring structural integrity in all modern German LWRs. Vibration, noise signature and acoustic monitoring in frequency and/or time domain are performed with multi-sensor-systems distributed over the primary circuit. Advanced signal analysis methods, like correlation-, FFT-, pattern recognition techniques, are applied to calculate relevant features and signatures from stochastical signals. Deviations of actual signatures from reference signatures indicate incipient failure or malfunction developments and are used for on-line diagnoses. A detailed knowledge base about normal and abnormal feature behaviour is needed for correct interpretation. GRS was deeply involved in the basic method development for vibration, neutron noise and acoustic monitoring together with industry (mainly KWU) in the 70ies and has spent great efforts during the subsequent years, to collect operational experiences and to build up a comprehensive signature data bank for interpretation and diagnosis purposes in its laboratories. Recently new computer-based monitoring systems have been developed by GRS and installed in operating plants which allow more “intelligent” data analysis and feature trending on-site as well as an optimal link to the experts and the signature data bank in the GRS diagnosis centre. The progress made by these new systems for vibration and loose parts monitoring in NPPs is briefly reviewed.