Wind turbines are supposed to last about 20 to 25 years. Within that timeframe, repairs are regularly carried out. But how do you know when a certain component will fail? It’s a question that people like Anders Aaen deal with on a daily basis. Aaen is Reliability Modeling Manager and key expert at Siemens Gamesa Renewable Energy. He is also one of the international speakers at Holland Innovative’s Reliability Seminar to be held on October 14, 2021. At the event, he will talk about the challenges related to reliability predictions of wind turbines, among other things
Aaen helps companies to understand the reliability of wind turbines. For example, they need to estimate how many components will break down. Not to mention when they will break down. Aaen explains: “To be able to say something about the reliability of the turbines, we look at the data and the design. It’s not something we can predict exactly. But we can evaluate it from a statistical point of view.”
After all, it’s important for customers to know what to expect. “When we sell the wind turbines, we provide a warranty of two to five years. Customers can also choose to leave the turbine in our hands throughout the design lifetime. We then take care of all repairs throughout its lifetime”, he explains. So one has to estimate in advance how much it will cost. And that’s not so easy. “People know, of course, that repairs have to take place. It’s the same as with a car. You don’t expect it to never need repairing.”
But a wind turbine is not like a car and the costs quickly run into the millions, especially when it comes to offshore wind turbines. Many factors come into play: “You first have to wait for the right weather conditions. And it takes a crane as big as a soccer field.” Reliability and Root Cause Analysis (RCA) are very important here. If there is a problem or if something breaks down, the cause must be figured out. “That’s complicated because there can be many causes. And sometimes something is also just a coincidence. But you can only solve a problem if you know the cause. Based on that, repairs are not only made on site. It is also taken into account in the production of new units.”
Estimating reliability is inextricably linked to this. This indicates the direction in which the cause must be sought. Aaen: “It tells us which things are important and what we need to look out for. We always start from a certain hypothesis. Using statistical analysis we then test the hypothesises.”
Aaen and his team have recently started working with a special calculation method to achieve faster and more accurate calculations. Such a new method is needed because wind turbine manufacturers are facing ever-greater challenges. Aaen gives an example: “The pitch bearing is a component that connects the blades to the turbine. We have to find out how long such a bearing can last. That’s quite a challenge, especially now that the turbines are getting bigger. We can’t sit around waiting for something to break or for the wind turbine to reach the end of its life. So we have to speed up the learning process somehow.”
In the new method, therefore, such a component is exposed to overloads. This simulates, in a short period of time, the load that wind turbine components face over the full design life with normal loads. At the reliability seminar, Aaen will elaborate on this. Finally, he says, “I hope that there will be many people who work in the same field and try to solve the same kind of problems.” A more modern approach to reliability – Reliability 2.0 – is at the heart of this.