July-August 2019 NPJ

40 NuclearPlantJournal.com Nuclear Plant Journal, July-August 2019 Plant Condition Monitoring By H.M. Hashemian and Greg Morton, Analysis and Measurement Services Corporation. H.M. Hashemian H.M. “Hash” Hashemian is President and Chief Executive Officer of Analysis and Measurement Services Corporation (AMS). He holds three doctorate degrees in engineering including a Ph.D. in nuclear engineering, a Doctor of Engineering degree in electrical engineering, and a Ph.D. in computer engineering. Dr. Hashemian and his company, AMS, specialize in testing the instrumentation and control systems of nuclear power plants in the area of predictive and automated maintenance of critical plant equipment and processes. Dr. Hashemian is the author of three books: Sensor Performance and Reliability (ISA, 2005), Maintenance of Process Instrumentation in Nuclear Power Plants (Springer Verlag, 2006), and Monitoring and Measuring I&C Performance in Nuclear Power Plants (ISA, 2014). In addition, he holds 20 U.S. patents and has written more than 400 papers including journal and magazine articles, book chapters, and reports. Dr. Hashemian is a Fellow of the American Nuclear Society, a Fellow of the Institute of Electrical and Electronics Engineers, and a Fellow of International Society of Automation. Abstract This article presents a review of new tools for automated on-line monitoring of integrity of reactor internals using data from existing neutron detectors and process sensors analyzed with a menu of advanced signal analysis and analytical modelling techniques. This is important not only to support light water reactor sustainability but also to develop the foundation for automated predictive maintenance in the next generation of reactors. Description of Technology The sustainability of existing fleet of nuclear power plants depends on aging management of components that cannot be replaced easily or economically [1]. These components nclude the reactor vessel and its nternals such as he core barrel, hermal shield, fuel ssemblies, and core hroud. For example, he core barrel in pressurized water eactors (PWRs) is fixed on the top and ree on the bottom. Therefore, the core barrel vibrates like a pendulum while the plant is operating. As plants age, it is important to verify that the amplitude and frequency of this vibration is not changing beyond an acceptable limit for safe operation. The accelerometers that exist in the current generation of nuclear power plants for loose parts monitoring and other purposes cannot measure this vibration with any appreciable resolution. This is because accelerometers are better suited for high frequency vibration measurements (>100 Hz) while the vibration of core internal components is typically below 30 Hz. Fortunately, existing neutron detectors that are normally used for neutron flux monitoring can also yield low-frequency vibration data as needed to verify the integrity of reactor internals. The technique to bring this about is referred to as “noise analysis”. It is based on monitoring the natural fluctuations (noise) that exist at the output of neutron detectors and other sensors during plant operation. If the output of the sensors is sampled at a high rate (>1000 Hz), the noise can be extracted from the sensor output, amplified, filtered, and analyzed to yield the vibrational characteristics of the reactor vessel and its internals. Figure 1 illustrates how noise data for online condition monitoring of reactor internals is extracted from the normal output of a plant sensor. Applications In addition to vibration monitoring, existing sensors can be used with the cross-correlation technique to monitor for flow anomalies in the reactor coolant system and identify and locate flow blockages. For example, in PWRs, the ex-core and in-core neutron detector signals can be cross-correlated with core exit thermocouples to yield fluid flow data that may be tracked to identify changes in flow rate or flow path and determine if flow anomalies or flow blockages are developing within the i i t t a s t r f Figure 1. Block Diagram of the Noise Data Acquisition Sequence.