Facing tomorrow's challenges for wind energy

The opening of the first German Offshore Wind farm alpha ventus fired the wind industry's imagination. Since the conditions are rough and the reachability is limited out at the sea, wind turbines that were designed for these dynamic loads can assure high availability and therefore profitability.

Fatigue is one of the key design drivers for the design of rotor blades of offshore wind turbines. For larger rotor blades, the swept area increases with the square of the rotor diameter. However, the weight of the blade increases to the third power. Therefore, in-depth knowledge about the failure mechanisms of materials is necessary.

Stefan Wessels, Project engineer at the competence center rotor blades of Fraunhofer IWES presents the results of a detailed fatigue analysis of rotor blades considering non-linear Goodman analysis combined with finite element analysis. “Considering the material properties more in depth and using their capacities, lead to a weight reduction without a reliability reduction”, states Wessels.

Not only the material, also the assembly of rotor blades is a critical point that needs profound investigation to as-sure blade design optimization, especially for larger blade lengths. Since the composite parts of a rotor blade are bonded together, the adhesives play a key role in the structure. Stefan Wessels presents a beam scaled test for the evaluation of the structural adhesive bond between the web and spar cap of a MW scale wind turbine blade. In contrast to coupon size testing for the mechanical characterization of the bonding paste, this method takes into account the shape of the bonding line, material thickness variability and the stochastically distributed voids, and above all a more realistic stress distribution in the bond line.

In 2009 about 24 GW of installed wind power capacity was integrated into the German power supply system. The system operator requires accurate and reliable forecasts of the electricity generated by wind turbines for the next hours to days ahead. Fraunhofer IWES works on the development of advanced shortest-term predictions in order to improve accuracy of wind power forecasts, and models for estimations of the forecast uncertainty. “Improved ac-curacy of wind power forecasts in combination with estimations of the forecast uncertainty does not only lead to higher system security but also to attractive cost savings concerning the allocation of balancing power”, declares Jan Dobschinski, researcher for signal analysis, model de-velopment and prediction systems at Fraunhofer IWES.

One approach to enable high penetration of wind power is for wind farms or wind farm groups to be operated as far as possible as conventional power plant. “In the German “Renewable Model Region Harz” (RegMod-Harz) wind farms and other renewable energy producers, controllable consumers and energy storage devices will be coupled to a large virtual power plant (VPP)”, reports Dr. Kurt Rohrig, head of division Energy Economy and Grid Operation. Beside scheduled energy supply, the VPP will provide ancillary services like frequency and voltage control to ensure the reliability of the power system.

IWES researcher Stefan Faulstich shows that electrical subassemblies are a significant contributor to wind turbine unreliability and the resulting consequences for maintenance procedures. By evaluating empirical data the “reli-ability based maintenance strategy” may identify weak points, find fault propagations and predict remaining life time to optimise maintenance. Therefore, detailed documentation of all maintenance measures and a purposeful structured database are necessary.

A shortest-term wind power prediction with integration of offsite wind speed measurements was presented by Dr. Bernhard Lange, Head of information and prediction systems at Fraunhofer IWES. This forecast of wind power is needed for balancing the feed in of wind power in the electrical grid. Therefore a large effort is done to improve the quality of the forecast and their availability.

In the present study the use of offsite measurements for forecasting is investigated for two aims: To improve the forecast quality, especially at high power classes and to guarantee a high quality forecast even if other input parameters of the forecast system, like actual power measurements, are missing. This forecast system is currently installed at three of four German TSOs.