January-February 2019 NPJ

Performance Monitoring of Digital I&C Systems By H.M. Hashemian and Josh Cole, Analysis and Measurement Services Corporation. H.M. Hashemian H.M. Hashemian is President and Chief Executive Officer of Analysis and Measurement Services Corporation (AMS); a nuclear engineering consulting firm headquartered in Knoxville, Tennessee, and operating in the United States, Europe, and Asia. His technical and operational vision and leadership have enabled AMS to play a key role in ensuring the safe and cost- efficient operation of virtually every U.S. nuclear power plant, as well as many in Europe and Asia. A globally recognized expert in peaceful applications of nuclear energy for electricity generation and medical diagnostics and treatment, Dr. Hashemian lectures frequently around the world on nuclear power plant instrumentation and control areas. 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 is a Fellow of the American Nuclear Society (ANS), a Fellow of the Institute of Electrical and Electronics Engineers (IEEE), a Fellow of ISA, as well as a member of the European Nuclear Society (ENS). He serves as an adjunct professor of nuclear engineering at the University of Tennessee. Abstract It is customary in daily life to smell food such as meat, fruit, and dairy to determine if they are fresh, ripe, old, or rotten. The same can be done to assess the condition of process equipment by measuring (smelling) its electromagnetic emissions. For example, the instrumentation and control (I&C) systems of nuclear power plants are made of an assembly of analog and/or digital components which together emit a particular electromagnetic signature that could relate to the condition of the assembly. This claim is being substantiated through a hands-on research and development (R&D) project at AMS with a grant from the U.S. Department of Energy (DOE). More specifically, the authors are conducting an experimental project to develop empirical correlations between the condition of representative nuclear power plant I&C equipment and its e l e c t r o m a g n e t i c signature. These correlations can then be used to determine the condition of installed equipment by measurement of electromagnetic fields at or near the equipment. Introduction Vibration analysis and Motor Current Signature Analysis (MCSA) are examples of mature technologies that the nuclear industry uses for predictive maintenance and incipient failure detection of motors and other rotating equipment. However, such tools are not currently available for electrical and I&C (analog and digital) equipment. This gap will be filled through the R&D project described in this paper. The R&D will involve representative electrical and I&C equipment fromnuclear power plants such as power supplies and input/output (I/O) modules to demonstrate the feasibility of electromagnetic measurements near the equipment for tracking the degradation or determination of the equipment. Once the feasibility of this idea is established for representative nuclear plant equipment, the scope of the R&D will be expanded to cover a comprehensive list of vulnerable electrical and I&C equipment and subsequently develop a prototype system for automated equipment condition monitoring through analysis of emission spectrums. The product of the project will be referred to as the “System for Electromagnetic Condition Assessment (SECA)”. This system will be designed and built to provide early warning of impending failures in electrical and I&C equipment. Existing condition monitoring features incorporated into digital I&C systems typically provide an alert after a failure has occurred such as the loss of output of a power supply or a failed communication card. For systems with a singlepointofvulnerability,thisindication is provided too late to take corrective action and prevent a potential impact on plant operation. Even for systems that are fault tolerant, the failure(s) could still impact system availability and reliability. Therefore, SECA can fill a significant void in development of a predictive maintenance program for electrical and I&C equipment in nuclear power plants. SECA will include advanced signal processing algorithms and software packages to automatically identify changes from normal or baseline behavior and empirical correlations to relate the changes in emission spectrums to root causes of equipment degradation. In doing so, big data analytic algorithms will be used to scan through electromagnetic emissions from power supplies, analog and digital printed circuit boards, microprocessors, network switches and other components to identify changes resulting from failing internal components. For instance, a failing capacitor in a DC power supply can lead to a change in the spectrum of the output voltage at the harmonics of the fundamental switching frequency. A system with the capability to passively monitor the output voltage spectrum of the power supply or the energy radiating from the device could identify these changes such that the device itself could be repaired or replaced prior to complete failure. 42 NuclearPlantJournal.com Nuclear Plant Journal, January-February 2019