Use this URL to cite or link to this record in EThOS:
Title: Wind turbine condition monitoring based on SCADA data
Author: Wang, Yue
Awarding Body: University of Strathclyde
Current Institution: University of Strathclyde
Date of Award: 2014
Availability of Full Text:
Access from EThOS:
Access from Institution:
Wind energy has an increasingly essential role in meeting electrical power demand and achieving environmental sustainability. The excellent offshore wind resource and the need to reduce carbon emissions from electricity generation are driving policy to increase offshore wind generation capacity in UK waters. Access and maintenance offshore can be difficult and will be more costly than onshore and availability correspondingly lower and as a result there is a growing interest in wind turbine condition monitoring allowing condition based, rather than responsive or scheduled, maintenance. Existing wind turbine condition monitoring methods, such as vibration analysis and oil debris detection, require expensive sensors. The additional costs can be substantial considering the number of turbines typically deployed in offshore wind farms and in addition, costly expertise is generally required to interpret the results. In contrast, the potential to extend the Supervisory Control and Data Acq uisition (SCADA) data based analysis approach is considerable and could add real value to the condition monitoring with little or no cost to the wind farm operator. This thesis focuses on wind turbine condition monitoring that utilises exclusively data from SCADA systems. The aim is to detect incipient wind turbine operational faults or failures before they evolve to catastrophic failures, so that preventative maintenance or corrective action can be scheduled in time, hence reducing downtime and potentially preventing wider damage. Useful component condition indicators are derived by comparing incoming operational SCADA data with the results for relevant variables, like component temperature that reflect component condition, derived from relevant models trained on SCADA data from a healthy wind turbine. Incipient failures are identified through anomalous behaviour in the variables of interest manifest in the SCADA data. This approach is first applied to individual wind turbines, but then extended to include other wind turbines operating under similar conditions to derive component condition indicators through inter-machine comparison. This is demonstrated to facilitate significant savings in computational effort and model complexity compared to the repetitive development of individual turbine models. In addition, a real time wind turbine power curve is implemented based on SCADA data, and compared with a reference power curve to identify anomalous behaviour, through minor changes in the power curve, in a timely manner.
Supervisor: Not available Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID:  DOI: Not available