UKERC Energy Data Centre: ETI Publications

UKERC Home >> EDC Home >> ETI Publications >> Back to Results >> Condition Monitoring: Final Report on an Holistic Approach to Wind Turbine Monitoring

Condition Monitoring: Final Report on an Holistic Approach to Wind Turbine Monitoring


Citation Futter, D.N., Chevalier, R., Gilbert, D., Muguelanez, E., Whittle, M. and Infield, D. Condition Monitoring: Final Report on an Holistic Approach to Wind Turbine Monitoring, ETI, 2013. https://doi.org/10.5286/UKERC.EDC.000598.
Cite this using DataCite
Author(s) Futter, D.N., Chevalier, R., Gilbert, D., Muguelanez, E., Whittle, M. and Infield, D.
Project partner(s) E.ON Engineering (UK) Ltd, Electricité de France SA (EDF SA), Insensys Limited, University of Strathclyde, Romax Technology Limited, SeeByte Limited
Publisher ETI
DOI https://doi.org/10.5286/UKERC.EDC.000598
Download WIN_WI1004_1.pdf document type
Abstract The Condition Monitoring project was led by Moog Insensys and included Romax, SeeByte, the University of Strathclyde, E.ON and EDF. It looked towards developing an intelligent integrated, predictive, condition monitoring package for wind turbines, which improves reliability, increasing availability by reducing downtime by up to 20% and leading to potential savings of 6,000 per turbine.

The InFLOW project was initiated to develop an holistic, predictive condition monitoring system. This was seen to be distinct from conventional condition monitoring systems (CMS) in that it did not restrict itself to a single technology, but brought together a range of sensing technologies and available turbine data to generate holistic diagnostics and real-time damage modelling to provide prognostic information relating to the life used on various parts of the turbine. It was shown that therewere significant savings to be made by optimising the inspection and maintenance regimes for off-shore turbines, in large part due to the expense of jack-up barges with weather defined access constraints

Conclusions:
  • The application of holistic relational models has been applied for the first time in wind turbines across a wide dataset.This capability shows promise to codify expert knowledge but would require further development and validation.
  • The use of prognostic damage models provides additional information to support inspection and maintenance optimisation. However, there is a need to build more experience with these models.
  • The introduction of SCADA fault algorithms into the system reflects current thinking in wind turbine fleets, and the models produced represent advanced examples of what can be achieved. These real time algorithms can be implemented in the control system or a data historian.
Associated Project(s) ETI-WI1004: Condition Monitoring
Associated Dataset(s) No associated datasets
Associated Publication(s)

Condition Monitoring - Key Findings - One Page Summary