Engineering news
As wind power booms, so have maintenance costs for offshore turbines. A Crown Estate report predicted annual maintenance costs of £2bn for a total offshore turbine fleet of 5,500 by 2025, and experts estimate operations and maintenance account for a third of overall costs for wind farms.
A team of engineers from Brunel University hope the industry could tackle the massive costs with predictive maintenance, where operators fix or replace parts before they break or reach the end of service. To create a commercial predictive maintenance process, the researchers are developing a digital platform called WindTwin, which acts like a control panel for wind farm managers.
The platform will feed data from sound sensors on turbines’ gearboxes, generators and other parts into 3D virtual models, or digital twins, of the turbines. The software will highlight pieces that need fixing and when the work needs completion, allowing companies to scrap scheduled maintenance and replace or repair parts before they do damage and cause downtime.
WindTwin will combine mathematical models describing the physics of a turbine’s operation with the sensor data, said engineer Miltiadis Kourmpetis from the Brunel Innovation Centre. He claimed it could have huge potential benefits for the industry, cutting upkeep costs by up to 30% and downtime losses by 70%.
Several companies and researchers have previously tried to use microphones for monitoring and maintenance, said Stephan Barth, managing director of research centre ForWind. The technology is a “holy grail” for the industry, he added, which would allow operators to massively reduce expenditure.
However, for predictive maintenance techniques to have the desired impact, he said manufacturers must be willing to share technical information for accurate data and mathematical models of turbine operation.
“If a company or an institute gets all the information from the manufacturers, then you have the chance to do that,” he said to Professional Engineering. “The usual problem is nobody except the manufacturers has this kind of information. So you’re listening to a big black box, and hearing all sorts of stuff, and then you have to work out what’s going on. That’s extremely challenging.”
Successful modelling with the full manufacturers’ information could have even bigger savings than new efficient parts, said Barth. He called on manufacturers to pool information, allowing researchers to create models which could benefit society as a whole.
“That is the hope,” he added. “If the industry would agree on that, sharing the data, the impact of reduction in the cost of energy would be huge.”
Funded by the government’s Innovate UK, the Brunel team is working with experts from companies and consultants Agility3, ESI and TWI on the £1.4m, 30-month project. They hope to sell the WindTwin platform worldwide.
Content published by Professional Engineering does not necessarily represent the views of the Institution of Mechanical Engineers.