Detection of Incipient Signal or Process Faults in a Co-Generation Plant Using the Plant ECM System
A 161 megawatt co-generation plant operated by the US Steel Corporation is highly instrumented, providing real-time monitoring of turbine, extraction, generator, condenser and bearing functions, including pressure, valve position, temperature, amperage, voltage and displacement. An equipment condition monitoring (SmartSignal eCMTM) system based on an empirical profile of normal operational conditions and utilizing pattern recognition techniques was used to monitor 60 signals on this plant at one-minute intervals. A study of plant data from a two month period shows that the Plant eCMTM fault detection and identification technology generates a high fidelity estimated value for each of the signals and enables early detection of significant, but subtle, variations in pressures, temperatures, flows, and energy level. Alerts annunciated by the system focus attention on groups of signals that can be early indications of equipment failure or process imbalance.
Unable to display preview. Download preview PDF.
- 1.S. Wegerich, R. Singer, J. Herzog, A. Wilks, “Challenges Facing Equipment Condition Monitoring Systems,” Proceedings, International Maintenance and Reliability Conference (MARCON 2001), Gatlinburg, TN, USA (6–9 May 2001 )Google Scholar
- 2.K. C. Gross, R. M. Singer, S. Wegerich, J. P. Herzog, R. Van Alstine, and F. Bockhorst, “Application of a model-based fault detection system to nuclear power plant signals,” Proc. Intelligent Power System Application To Power Systems (ISAP 97), Seoul, Korea (6–10 July, 1997 ) 66–70Google Scholar