Integrated Condition Monitoring For Plant-Wide Prognostics
Wallace, C.
Costello, J.
West, G.
Mcarthur, S.
Coghlan, M.
Download PDF

How to Cite

Wallace C., Costello J., West G., Mcarthur S., Coghlan M., 2013, Integrated Condition Monitoring For Plant-Wide Prognostics, Chemical Engineering Transactions, 33, 859-864.
Download PDF

Abstract

The use of Condition Monitoring (CM) in the power industry is well established, with traditional manual analysis increasingly giving way to intelligent, automated analysis. Further, there is a growing interest in the extension of such CM systems to derive prognostic information for asset management. It is common however for CM applications to be retrofitted to existing equipment, to monitor a particular component or subsystem, resulting in several isolated CM systems attached to a larger system from which it is not trivial to extract plant-wide prognostic information. The extension of CM applications to deliver prognostic information is an area of current research in the nuclear industry as operators seek to maximise the operational life of ageing plants. The interdependence of multiple safety critical systems, each of which is monitored individually, is a critical challenge in determining the health of a life-limiting component which may be part of a chain of monitored sub- systems. The generation of new information about plant-wide state, based on subsystem prognostics and monitoring, can help inform operation and maintenance. This paper describes two CM systems used to monitor key systems on the UK fleet of Advanced Gas-Cooled Reactors, at different ends of the coolant cycle in the reactor. The first system monitors vibrations of the gas circulators used to pump coolant into the core, while the second system monitors the thermal power of the fuel channels inside the core. The systems make diagnoses of anomalies in their respective data-sets, however a fault in the gas circulator would directly affect the cooling of the core, potentially affecting the fuel channels further along the coolant loop. The paper discusses how closer integration of the outputs of each system may better inform the prognostic information about the plant, by identifying related anomalies and providing corroborating information to better inform prognostic estimates.
Download PDF