Skip to main content

Condition Monitoring of Wind Turbine Drive-Trains via Non-Contact Acoustic Sensors

Article Category

Article available in the folowing languages:

Listening for faults and flaws in wind turbine drivetrains

An EU-funded project unveiled the first non-invasive system that uses acoustic emissions to detect failures in the drivetrain of onshore wind turbines. The groundbreaking low-cost condition monitoring system can save up to 50 % of annual drivetrain maintenance costs.

Industrial Technologies
Energy

The drivetrain is the powerhouse of a wind turbine, containing the gearbox and the generator – components needed to convert the rotation of the blades (torque) into electricity. Although vibration is normal in most rotating machinery, excessive vibration levels can be detrimental to component health and are usually an early indicator of machine failure. The EU-funded CMDrive project developed a groundbreaking wind turbine condition monitoring system that can dramatically reduce the amount of downtime required for preventive maintenance. “ACMWind is the first non-intrusive condition monitoring system on the market that uses an acoustic sensor to detect faults in wind turbine drivetrain components at an early stage,” notes project coordinator Natalia Fernandes. The compact reliable system can warn wind turbine operators of potential damage before it becomes severe and causes prolonged downtime.

Sound-sensitive system

ACMWind comprises a microphone that listens and memorises noise patterns generated by the machinery. This acoustic sensor can detect sound being radiated from the different drivetrain components, including the main bearing, gearbox and generator. Vibrations from all the rotating components within the nacelle are sensed through the sound generated in the surrounding air. Through a signal processing tool, the system can differentiate the noise stemming from the faulty piece of machinery. The software automatically recognises the structural health signature of each monitored component by analysing the frequency and time dependence of the acoustic responses. Operators can also make use of the quantitative history of the structural integrity of the machine components to estimate the time left until the component fails. Field tests conducted in two countries successfully demonstrated ACMWind’s ability to identify several failure modes in the frequency spectrum. In particular, it detected the vibration frequency signatures of four accelerometers installed in the main bearing, gearbox and generator of wind turbines from just one microphone installed in the nacelle.

Low-cost, non-intrusive, easy to install

Wind farm owners are increasingly investing in solutions to prevent failure instead of retaining turbine operation until damage occurs. Typical solutions for fault detection of wind turbine machinery include vibrational analysis, acoustic emission or oil debris analysis systems. Fernandes elaborates: “The condition monitoring systems currently available on the market are intrusive and thus present difficulties during the installation process. Such tampering also affects the initial warranty of the drivetrain components.” The condition monitoring systems are often calibrated according to the manufacturer’s machinery, and hence cannot adapt to the specific requirements of each operator’s wind turbines. Furthermore, they often trigger false alarms or only warn operators when the component is very close to catastrophic failure. “ACMWind is a disruptive innovation aiming to revolutionise how condition monitoring of wind turbines is carried out, shifting it from an intrusive to a contactless process,” notes Fernandes. The cost of the product is well below the cost of commercial systems due to the microphone used that is inexpensive compared with the widely used accelerometers on the market. The new system complies with the standards set by the European Agency for Safety and Health at Work defined to ensure safer operation and maintenance and increase the detection probability for wind turbine defects. Its installation is also rather simple and straightforward as it requires fewer components.

Keywords

CMDrive, wind turbine, drivetrain, condition monitoring, ACMWind, non-intrusive, vibration, acoustic sensor

Discover other articles in the same domain of application