Subscribe to our newsletter to receive the latest news and events from TWI:

Subscribe >
Skip to content

Structural Health monitoring for wind turbine blades

TWI has completed research into enhanced condition monitoring systems for wind turbine blades that will decrease maintenance costs and increase installation reliability, allowing the development of more remote offshore wind farms and thus reducing pressure for onshore installations. This project will enable the UK manufacturing industry to compete with overseas suppliers in the production of advanced designs of offshore wind turbines.


Reducing costs through remote monitoring

Operation and maintenance costs constitute a sizeable share of the total annual costs of a wind turbine. For a new machine, these costs might easily have an average share over the lifetime of the turbine of more than 30% of the total cost per kWh produced. Therefore, maintenance costs are increasingly attracting the attention of manufacturers, who are seeking to develop new maintenance strategies.

For offshore turbines, costs of access are disproportionately high, increasing the incentive for remote monitoring. In addition, blade and structure failures seem to correspond to a high percentage of the failures that have occurred during the last six years across the world market.

Developing acoustic emission systems and methodologies

The acoustic emission (AE) technique for structural health monitoring (SHM) is an active area of research primarily due to the requirement for complex signal processing to extract and identify the signals of interest from a noisy background.  Additional complexity arises from the fact that the AE signals are strongly non-stationary (frequency varies as a function of time as in a chirp) which leads to errors if classical signal processing techniques are employed (such as Fourier analysis). 

Blade test programme

The main purpose of this study was to investigate the feasibility of in-service monitoring of the structural health of blades using AE. The experiment was developed at the National Renewable Energy Centre (Narec) 100m Blade Test Facility. A 45.7 metre long blade, in which a crack had been initiated, was subject to fatigue loads at different load values using compact resonant masses (CRMs) to excite the blade, accurately simulating in-service load conditions over a six-week period, during which AE monitoring was performed with four sensors.

Over the course of the experiment over 9000 datasets were recorded. The challenge was to separate the signals caused by an AE event due to crack propagation, from those caused by the vibration and ‘coherent’ noise generated by the movement and friction of the blade.

An AE event was defined as one where all four sensors were hit within a time span ranging from 100 to 550 microseconds. Eliminating those datasets where only 1-3 of the sensors were triggered, left 277 datasets with four signals crossing the threshold at least once. These were selected for additional analysis to confirm if they referred to an AE event.

As AE signals propagate within the material at a specific speed, this value can be used to determine the defect localisation co-ordinates.  To determine the location of the event, the relative differences in time of arrival were calculated using the sensor that detected the first threshold crossing (45dBAE) as a reference point. The localisation process was then performed using trilateration method, finally obtaining the source localisation for four sensors being triggered by the AE. Clustered event locations can be confidently interpreted as representing crack propagation.

For more information, please email


Testing a 45.7m long turbine blade.
Testing a 45.7m long turbine blade.
Signals reaching 1, 2, 3 and 4 sensors
Signals reaching 1, 2, 3 and 4 sensors
AE signals reaching the four-sensor array at different times
AE signals reaching the four-sensor array at different times
AE event locations
AE event locations

The study confirmed that we can successfully extract and classify AE signals, demonstrating that AE has been developed to the point where we are confident that it can yield vital information that will meet our objective to reduce O&M costs with a viable remote condition monitoring system.

For further information, please visit the project website: