Innovative Approaches for Rotor Blade Monitoring
12/02/2024 12:36
EXECUTIVE SUMMARY:
Monitoring the rotor blades of wind turbines poses several challenges and has not yet been implemented as standard in the turbines. A radar sensor installed on the tower could monitor all three blades with little effort.
Complex inspection of highly stressed components
Rotor blades of wind turbines are among the most highly stressed technical components of all [1]. Damage to rotor blades is therefore not uncommon and is detected by regular manual inspections or inspections using drones. Repairs are carried out as required on the basis of these inspections.
The manual inspection of many thousands of wind turbines (currently around 29,000 in Germany [2]) is very time-consuming. The resources required could be significantly reduced through the use of automated condition monitoring.
However, automatic detection of damage to rotor blades is not yet a technical standard. It is therefore extremely desirable to further develop the existing technology approaches for monitoring. This could reduce the effort required for manual inspections and repairs and avoid total failures. This would reduce the levelised cost of energy and lower CO2 emissions during the service life of a wind turbine.
Challenges for monitoring
Monitoring the rotor blades of wind turbines poses several challenges. Firstly, the power supply and data transmission of sensors that rotate with the rotor is not trivial. A simple cable between the stationary nacelle and the rotating blade would be destroyed immediately by the rotation.
In addition, it is very challenging to monitor the extremely long and irregularly shaped blades not only in a small area, but over as large an area as possible.
The new sensor system
Against this background, cp.max has developed and tested a new type of monitoring system for rotor blades together with an interdisciplinary research group [3]. At its core, this system consists of a single radar sensor that can check a large part of all three rotor blades.
The sensor can be easily mounted and dismounted as it is installed on the tower of the wind turbine [Figure 1]. This means that no components need to be attached to/in the rotating rotor. The tubular steel towers of the 3 test turbines were not damaged as the radar sensors were fixed to the tower walls with strong magnets. Only a few millimetres of drilling through the tower wall was necessary to supply the sensor with power and transmit the data.
Installing the sensor on the tower therefore offers the following decisive advantages for this application:
- Only one sensor needs to be installed to monitor all 3 rotor blades.
- The sensor on the tower is easily accessible using rope access technology.
- Only a small hole needs to be drilled through the tower. The sensor adheres to the steel wall with magnets.
- No components need to be installed on the rotating rotor and connected by cable.
(In this research project, the radar sensor was permanently installed on the tower in the main wind direction to simplify matters. This meant that the sensor could not follow the rotor when the wind was blowing from other directions).
Identification of the rotor blades
A prerequisite for being able to check the condition of the rotor blades is, of course, the assignment of the measured values to the correct blade. It must therefore be known whether a recognised defect is located on blade 1, 2 or 3.
This task was initially solved by attaching symbols to the blade surface [Figure 1]. These markings were reliably detected by the sensor system and recognised by means of a neural network. They were used to assign the measured values to the respective rotor blade. Later in the project, the blades could also be differentiated solely on the basis of the individual measured values, the radar fingerprint of the blades, so to speak [Figure 2].
Up to this point, the project objectives were fully achieved. The radar sensor was installed on the tower of a wind turbine and was able to independently record radar programmes of the passing rotor blades and assign them to the respective blade. The data was then transmitted via the Internet.
Problems with damage detection
The neural network for processing the measured values was able to differentiate between the three rotor blades. However, it was not possible to detect any changes in the structure of the rotor blades during the project. However, this does not necessarily indicate a defect in the measurement system, as no damage was found on the blades during an additional manual inspection.
To test the measuring system, an absorber film measuring approximately 1 m² was therefore glued to a rotor blade. After a short test on a laboratory scale, it was assumed that this change in the blade surface would be clearly recognised by the neural network.
Unfortunately, this was not the case in the field test. One possible cause could be unfavourably selected properties of the absorber foil or environmental parameters.
Conclusion
Although not all project objectives were achieved, a great deal of experience was gained in the radar-based monitoring of rotor blades. Monitoring the rotor using sensors attached to the tower appears to be promising and cost-effective.
The detection of damage to rotor blades should be investigated further in the field. As the functionality has already been proven in the laboratory [4], cp.max also assumes good chances of success in the wind farm.
Measurement data from this project for further research [6].
[1] Hau, Erich: Windkraftanlagen; 4. Auflage; Springer-Verlag; Berlin, Heidelberg; 2008
[3] Goethe University Frankfurt, University Siegen, IMST GmbH, BOREAS Energie GmbH, cp.max Rotortechnik GmbH & Co. KG
[4] Arnold, P., Moll, J., Mälzer, M., Krozer, V., Pozdniakov, D., Salman, R., ... & Nuber, A. (2018). Radar‐based structural health monitoring of wind turbine blades: The case of damage localization. Wind Energy, 21(8), 676-680.
[5] Sercan ALIPEK, Moritz MAELZER, Sebastian BECK, Christian KEXEL, Jochen MOLL, Viktor KROZER, Jürgen KASSNER, Thomas HEINECKE, Jonas ROSE (2024). Potential and Limitations of Anomaly Detection via Tower-Radar Monitoring of Wind Turbine Blades in Regular Operation, EWSHM 2024
https://www.ndt.net/article/ewshm2024/papers/266_manuscript.pdf
[6] https://ewshm2024.com/frontend/index.php?page_id=385
Pictures:
[7] https://www.researchgate.net/publication/376535200_RADAR-BASED_STRUCTURAL_MONITORING_OF_WIND_TURBINES_BLADES_FIELD_RESULTS_FROM_TWO_OPERATIONAL_WIND_TURBINES
