SCADA dataset of a 2 MW SIEMENS wind turbine drivetrain located at a wind farm on the Baltic Sea coast in northern Poland
Description
This dataset contains Supervisory Control and Data Acquisition (SCADA) measurements from a 2 MW Siemens wind turbine drivetrain located at a wind farm on the Baltic Sea coast in northern Poland. The data were extracted to investigate whether early indicators of a gearbox fault could be detected using data-driven analysis. The monitoring period spans 30 days, from November 1, 2012 (00:00) to November 30, 2012 (23:50). Operational parameters were recorded at 10-minute intervals, resulting in 4,320 time-series samples for each parameter. The dataset includes twelve process parameters describing the turbine’s operational condition, grouped into rotational dynamics, electrical power generation, and thermal conditions. Rotational parameters include wind speed, rotor speed, and generator speed. Electrical parameters include active power, generated power, reactive power, reactive power delivered, generator voltage, and generator current, representing the turbine’s power generation and load conditions. Thermal parameters include gearbox bearing temperature and two generator temperature sensors, indicating the thermal state of key components. During operation, a gearbox bearing failure occurred and was recorded on November 9, 2012 at 13:00 (sample 1232). The dataset therefore contains both normal operational data and data preceding the fault event. In the related study, generator speed and gearbox bearing temperature were used to validate a stationarity-based anomaly detection method. SCADA measurements represent 10-minute averaged values, typical for wind turbine monitoring systems. The dataset contains no missing or corrupted values, making it suitable for research on condition monitoring, anomaly detection, time-series analysis, and predictive maintenance of wind turbines. Related published papers: 1) P.B. Dao, W.J. Staszewski, T. Barszcz, and T. Uhl, “Condition monitoring and fault detection in wind turbines based on cointegration analysis of SCADA data,” Renewable Energy, vol. 116, part B, pp. 107–122, 2018. 2) P.B. Dao, “A CUSUM-based approach for condition monitoring and fault diagnosis of wind turbines,” Energies, vol. 14, no. 11, 3236, 2021. 3) P.B. Dao, “Condition monitoring and fault diagnosis of wind turbines based on structural break detection in SCADA data,” Renewable Energy, vol. 185, pp. 641–654, 2022. 4) P.B. Dao, “On Wilcoxon rank sum test for condition monitoring and fault detection of wind turbines,” Applied Energy, vol. 318, 119209, 2022. 5) P.B. Dao, T. Barszcz, and W.J. Staszewski, “Anomaly detection of wind turbines based on stationarity analysis of SCADA data,” Renewable Energy, vol. 232, 121076, 2024. 6) K. Zolna, P.B. Dao, W.J. Staszewski, and T. Barszcz, “Nonlinear cointegration approach for condition monitoring of wind turbines,” Mathematical Problems in Engineering, vol. 2015, Article ID 978156, 11 pages, 2015.
Files
Steps to reproduce
To reproduce the analysis, load the SCADA dataset recorded from November 1, 2012 to November 30, 2012 at 10-minute intervals. Use the average values of the 12 process parameters. No preprocessing is required, since the dataset contains no missing, corrupted, or unphysical values. For validation of the stationarity-based method, analyze in particular the generator speed and gearbox bearing temperature time series. The gearbox bearing failure occurred on November 9, 2012 at 13:00, corresponding to sample 1232.
Institutions
- AGH University of KrakowLesser Poland, Krakow
Categories
Funders
- National Science CentreKrakowGrant ID: UMO-2023/51/B/ST8/01253, Non-classical Approaches for Condition Monitoring and Fault Detection of Wind Turbines.