The power performance characteristic of a wind turbine is defined by its power curve and resulting estimate in annual energy production. It is a key attribute for validating the performance of newly installed wind turbines, and the power curve is monitored throughout the wind turbine life cycle. This report explains power performance, upgrading, and conventional measurement methods. Wind farm stakeholders are keen on understanding the power performance of their wind turbines. The manufacturers’ monitoring software allows the power curve to be tracked in real time. Deviations from normal operation and underperformance can quickly be identified. However, these power curves cannot be trusted for evaluating upgrades and particular changes made to the wind turbine. The power curve is highly sensitive to small deviations in wind speed output of the nacelle anemometer. Upgrades which impact the air flow at the nacelle will introduce significant wind speed bias in such power curves. Thus owners are lacking appropriate tools for evaluating past events’ impact on power performance. The focus of this work lies on testing an alternative approach to measuring change in power performance using only historical wind turbine log data. Side-by-Side Testing as explained by Axel Albers was tested on real wind turbine data. This method uses no wind measurements, but instead simulates the wind speed using a deducted power relation to a neighbouring wind turbine and an assumed power curve behaviour. This allows any change in power output to be tracked onto the power curve, relying only on power output measurements. A full power performance analysis was performed by constructing Side-by-Side Testing in Microsoft Excel exclusively for this MSc thesis work. It was applied on a wind farm whose recent blade upgrade had never before been analysed. Two neighbouring identical 2.3 MW wind turbines were considered for the analysis, one which in May 2013 installed blade add-ons featuring serrated trailing edges to the blades. The analysis was executed completely off site. The power performance analysis was completed, producing meaningful results with known uncertainty levels. The test results indicate an improvement of power performance throughout the power curve, corresponding to an increase of 0.53% in annual energy production, at ±1.35% uncertainty. The analysis needs further work and validation, as the power curve shows signs of artefacts. The complex wind farm settings increase the uncertainty levels. The method could likely be tested in flat terrain or offshore with lower uncertainty of results, targeting below ±0.5% uncertainty in annual energy production.
| Identifer | oai:union.ndltd.org:UPSALLA1/oai:DiVA.org:kth-177878 |
| Date | January 2015 |
| Creators | Carlberg, Marcus |
| Publisher | KTH, Kraft- och värmeteknologi |
| Source Sets | DiVA Archive at Upsalla University |
| Language | English |
| Detected Language | English |
| Type | Student thesis, info:eu-repo/semantics/bachelorThesis, text |
| Format | application/pdf |
| Rights | info:eu-repo/semantics/openAccess |
Page generated in 0.0015 seconds