Unsupervised Causal Deep Learning-Based Anomaly Detection in Nuclear Power Plant Applications

##plugins.themes.bootstrap3.article.main##

##plugins.themes.bootstrap3.article.sidebar##

Published Oct 26, 2023
Abhinav Saxena Helena Goldfarb Jeffrey Clark

Abstract

Nuclear power generation will be key to meeting carbon free energy transition goals. However, nuclear power must provide agility and flexibility to fluctuating power demands when other sources of carbon-free energy like solar and wind may not be as flexible. This has led to small modular reactor (SMR) development where nuclear power will be generated from a distributed fleet of smaller reactors where units can be brought online or offline as needed. Due to its high operational and maintenance (O&M) costs as it is, a distributed fleet will put additional cost burden if remote monitoring and crew sharing is not enabled. This requires prognostics and health management (PHM) capabilities such as early warning, diagnostics, and prognostics to enable predictive maintenance with high accuracy. Typically, monitoring solutions are developed on component and subsystem levels targeting specific failure modes. However, it is argued that a systemwide monitoring, in addition to specific targeted analytics, would be of key importance. This paper presents a deep-causal unsupervised anomaly detector that has been successfully applied in various aerospace and renewable energy applications. In this paper we share our experience applying this method on a nuclear power plant (NPP) application. Specifically, we share how we dealt with practical challenges of data quality, ground truth labeling, performance evaluation and field validation in an unknown-unknown setting where prior knowledge of failures and failure modes were not available to begin with.

How to Cite

Saxena, A., Goldfarb, H., & Clark, J. (2023). Unsupervised Causal Deep Learning-Based Anomaly Detection in Nuclear Power Plant Applications. Annual Conference of the PHM Society, 15(1). https://doi.org/10.36001/phmconf.2023.v15i1.3570
Abstract 364 | PDF Downloads 246

##plugins.themes.bootstrap3.article.details##

Keywords

Deep learning, Causal Learning, Anomaly detection, unsupervised, nuclear plant, PHM, digital twin, maintenance, O&M cost, SMR, Small Modular Reactor

References
Alex Tank, I. C. (2018). Neural granger causality for nonlinear time series. Retrieved from arXiv preprint: arXiv:1802.05842
Coble JB, G. L. (2013). Approaches to Quantify Uncertainty in Online Sensor Calibration Monitoring. ANS Winter Meeting and Technology Expo.
Coble, J., Ramuhalli, P., Bond, L. J., Hines, J., & Upadhyaya, B. (2015). A review of prognostics and health management applications in nuclear power plants. Interational Journal of Prognostics and Health Management, 6, 16.
Feng Xue, W. Y. (2020). Deep anomaly detection for industrial systems: a case study. Annual Conference of the PHM Society. PHM Society.
GE Hitachi. (2017). BWRX-300. Retrieved from https://nuclear.gepower.com/build-a-plant/products/nuclear-power-plants-overview/bwrx-300
Hines, J. W., Coble, J., & Bailey, B. K. (2010). A Novel Method for Monitoring Single Variable Systems for Fault Detection, Diagnostics and Prognostics. International Journal of Performability Engineering, 6(5), 477.
Huang, H., & Kasiviswanathan, S. P. (2015). Streaming anomaly detection using randomized matrix sketching. VLDB Endowment.
Huang, H., Yan, W., Wang, T., & Xue, F. (2018). Imbalanced Time Series Classification with Nonlinear Causal Learning. Woodstock ’18:ACM Symposium on Neural Gaze Detection (p. 9). New York: ACM.
Huang, H., Yoo, S., Yu, D., & Qin, H. (2015). Diverse Power Iteration Embeddings: Theory and Practice. IEEE Transactions on Knowledge and Data Engineering.
INPO. (2021). Institute of Nuclear Power Operations. Retrieved from Nuclear Costs in Context: INPO.info
Kaiming He, X. Z. (2016). Deep residual learning for image recognition. IEEE conference on computer vision and pattern recognition (pp. 770-778). IEEE.
Moleda, M., Momot, A., & Mrozek, D. (2020). Predictive maintenance of boiler feed water pumps using SCADA data. Sensors, 20(2), 571.
NEI. (2021). Nuclear Costs in Context. Retrieved from Nuclear Energy Insititute: https://www.nei.org/resources/reports-briefs/nuclear-costs-in-context
Ramuhalli, P., Walker, C., Agarwal, V., & Lybeck, N. (2021). Nuclear Power Prognostic Model Assessment for Component Health Monitoring. 12th ANS NPIC-HMIT, (pp. 976-986).
Yadav, V., Agarwal, V., Gribok, A. V., Hays, R. D., Pluth, A. J., Ritter, C. S., . . . Iyengar, R. (2021). The State of Technology of Application of Digital Twins. TLR/RES-DE-REB-2021-01 .
Section
Industry Experience Papers

Most read articles by the same author(s)

1 2 3 > >>