5th International Conference on Renewable Energy Research and Applications - ICRERA 2016, Birmingham (United Kingdom). 20-23 November 2016
Summary:
This paper presents a method for the possible detection of abnormal performance of a wind turbine based on real observed pairs of wind speed and generated power. Usually the power curve satisfies this objective but its use in real-time for the detection of abnormal performance is not easy because the pairs observed do not follow a single line, but rather cloud of points around an ideal definition of it. This is the reason why the concept of the power curve has been extended in this paper to a normal power generation area for a period of time taken as reference for later comparisons. This area has to cover the most part of the data observed in the period considered as reference of performance. Once a normal power generation area is defined, it can be used to detect possible deviations in future observations. The paper includes some examples of application to four wind turbines where some abnormal deviations of performance were detected. Also, an analysis is included of other measured variables that could feed a complementary view about the results observed in the study carried out.
Keywords: wind turbine performance; power curve model; anomaly detection; wind turbine health assessment.
DOI: 
https://doi.org/10.1109/ICRERA.2016.7884562
Published in ICRERA 2016, pp: 1-6, ISBN: 978-1-5090-3389-8
Publication date: 2017-03-23.
Citation:
M. Carmona, M.A. Sanz-Bobi, Normal power generation area of wind turbines for the detection of abnormal performance, 5th International Conference on Renewable Energy Research and Applications - ICRERA 2016, Birmingham (United Kingdom). 20-23 November 2016. In: ICRERA 2016: Conference proceedings, ISBN: 978-1-5090-3389-8
