University of Strathclyde
Automated fault detection for wind farm condition monitoring
Abstract
dc:description.abstractWind turbines are becoming more established as an economically viable alternative to fossil-fuelled power generation. Recently wind farms consisting of hundreds of units are being built in various locations around the country adding a significant amount of electrical generating capacity. As the size of wind farms continues to increase, business economics dictate the need for effective condition monitoring systems that allow for careful asset management to minimise downtime and maximise availability and profits. Most modern turbines are built with integrated condition monitoring systems that acquire data and store this through Supervisory Control and Data Acquisition (SCADA) Systems. This data quickly becomes unmanageable and brings with it the problems of managing and interpreting it. This thesis considers the development of an automated SCADA data analysis system that aims to interpret the large volumes of data that are generated, with the intention of identifying faults in their early stages before they manifest into more serious catastrophic failures. A number of different analysis techniques for interpreting the SCADA are considered and a methodology of identifying faults in their incipient stages in the gearbox and generator using basic SCADA temperature data is described. Most CM techniques in the research literature focus on one aspect of a wind turbine with regards to identifying faults that may manifest within it. This research also puts forward the development of a multi-agent platform capable of combing multiple data sources and analysis techniques into one system to improve the opportunity of extracting and interpreting interesting information found in the SCADA and present it through a single point of contact for the operator. This provides the possibility of developing a more complete condition monitoring system that can monitor all of the main components of each turbine across a complete wind farm using both, existing and future condition monitoring techniques developed for the interpretation of wind farm data.
Degree
thesis:*- Name dc:type.qualificationname
- phd
- Level dc:type.qualificationlevel
- doctoral-pg
- Grantor dc:publisher.institution
- University of Strathclyde
- Year dc:date.issued
- 2010
Author and committee
dc:creator, dc:contributor.*- Author dc:creator
-
- Zaher, Ammar S.
Rights
- Licence dc:rights.uri
Identifiers
dc:identifier.*- Identifier
- T13107