"Usually it takes a very long time before new recommendations are accepted by certifying bodies," says Jessica Holierhoek of ECN Wind Energy. "So we were pleasantly surprised when Germanischer Lloyd incorporated part of our results into guidelines published this year."
Of course it helps that Germanischer Lloyd (GL) was one of ECN's partners in ‘PROTEST’ (PROcedures for TESTing and measuring wind energy systems, a project financed by the European Commission). However, at least as important is the number of unexpected defects and the losses they cause, especially offshore. There are too many cases of bearings and gearboxes which had to be replaced soon after their first working day. For instance, the British wind farm Kentish Flats (in the Thames Estuary) was plagued by numerous gearbox failures, as was the Offshore Wind farm Egmond aan Zee (OWEZ) near the Dutch coast - within two years of becoming operational. The owners of these farms think or hope that the problems are now solved.
"But designing a new wind turbine remains too much of a gamble," says Holierhoek, who coordinates PROTEST. "A major cause is the use of far too simple models to represent certain components. Like a gearbox which only has to deal with a perfect torque on its input shaft, or a blade pitch control mechanism which handles nothing but smooth and steady loads. The reality isn't like that at all. And the differences are known, at least in part."
Ideally, the stiff bedplate of the nacelle ensures that the main shaft and the gearbox remain properly aligned, reducing the forces on the latter to just the desired torque. In reality, severe and variable loads cause a certain amount of misalignment.
A good start
Standards for the certification of wind turbines are partly based on the analysis of a number of carefully defined design load cases. PROTEST has generated recommendations for several new definitions: for the calculation of resonances in the drive train consisting of the rotor, the gearbox, the generator and the nacelle; for dealing with misalignment occurring under load and for a better analysis of the consequences of a (partial) loss of the connection with the network which receives the turbine's power. These recommendations have been incorporated in GL's new Guideline for the Certification of Wind Turbines.
Holierhoek:"It's a good start, but not much more than that. Calculations must be validated by prototype measurement campaigns. As it is impossible to validate what you can't calculate, the mathematical models and simulations one uses determine what can be measured in a meaningful way. And the models aren't good enough."
The Kentish Flats wind farm, in the Thames Estuary.
In five years, all thirty turbines here have had their gearboxes replaced twice.
Six steps approach
Another stumbling block is communication. Wind turbines pose unique demands, which their designers find difficult to convey to suppliers of components such as bearings – not in the least because both parties want to keep commercially sensitive information to themselves. That doesn't cause much harm when discussing the mast or the rotor blades, because new masts and blades will always be much like their predecessors, allowing the use of standard methods for analysing their reliability. But there are many quite different ways to control blade pitch, to yaw the nacelle and to design a drive train.
It is even possible to do away with the gearbox and connect the rotor directly to a specially designed generator.
"A detailed, rigid list of standards can't deal with such a variety of concepts," says Holierhoek. "That's why we have devised a 'six step approach' which allows the companies involved to arrive at a custom set of models and measurements, just right for their particular new design."
The six steps approach also makes clear what information has to be made available. But whether the certifying bodies will be as quick to accept this method remains to be seen. It is a new way of doing things, which puts more responsibility on the shoulders of the designers.
Modular models
Holierhoek: "We would like to introduce mathematical models with a modular structure, allowing a component supplier to tell his story by presenting a model which adequately describes his product, a model that would fit directly into effective simulations of complete systems."
This modular approach would make the results of the six steps approach better comparable, when the method is applied to various turbine designs. But that is still in the future, a possible sequel to PROTEST. And even if the procedure for designing wind turbines ever becomes as good as it can be, there will be many problems left to be solved. Predictions based on simulations do not match reality well enough by a long shot, where the forces acting on a wind turbine are concerned. Experience teaches manufacturers what usually works and what tends to fail, but such lessons are rarely shared with others.
Transparency
"In the aviation industry it's normal to be very open about such matters," says Holierhoek. "When Airbus hits a snag, Boeing doesn't have to suffer from exactly the same trouble. But turbine makers regard the causes of failures as commercially sensitive information, which means that many wheels need to be invented more than once."
That costs money and above all, it wastes more time than the industry can afford to lose. Offshore wind farms are rapidly gaining in size, while the turbines aren't anywhere near reliable enough to make a profit without being subsidized.
Contact
Jessica Holierhoek
ECN Wind Energy / Integrated Wind Turbine Design
Tel.: +31 (0)22 456 42 78
Email: Jessica Holierhoek
Text: Steven Bolt
Info
PROTEST, Final report
Suggestions for improved standards
Recommended Practices for Measuring in Situ the "Loads" on Drive Train, Pitch System and Yaw System
Website PROTEST
This ECN Newsletter article may be published without permission provided reference is made to the source: www.ecn.nl/nl/nieuws/newsletter-en/
