PROF. AVIV ROSEN
Dean, Faculty of Aerospace Engineering,
Technion – Israel Institute of Technology, Haifa
Research on wind energy was performed in the Faculty of Aerospace Engineering of the Technion during the last seven years, and supported by the Israel Ministry of Energy and Infrastructure. The motivation for this research was based on the needs of wind turbine owners, and to a lesser extent, on the needs of wind turbine manufacturers.
The main factor that determines the success of a wind turbine is its economic achievements. These economic achievements depend on the turbine performance, namely the turbine energy production during its life-time. The turbine performance is a function of the turbine’s technical characteristics, the wind characteristics at a specific site and the coupling between both. The existing engineering tools for predicting the performance of a turbine, located at a certain site, are not accurate enough in many cases. Since many sites are marginal when it comes to economic considerations, the incapability of an accurate prediction of the turbine performance hinders the possibility of a well-founded economic decision. Thus the first part of the research was motivated by the need for a more accurate tool to predict turbine performance.
A main factor in the difficulty of an accurate prediction of a turbine performance. is that wind speed and wind direction are changing continuously. The turbine is a dynamic system that has a "dynamic" output
(power) as a response to the systems dynamic input (wind speed and wind direction). The dynamic nature has an important influence on the turbine performance. Nevertheless, most of the existing methods of calculating wind turbine performance, do not take into account the dynamic characteristics of the wind or the turbine. To cope with this problem. methods that account for the dynamic characteristics of the turbine and the wind (including the coupling between both) were developed at the Technion. These methods differ between themselves by their level of sophistication and the computing effort that they require. These methods were validated by comparing their results with actual measurements. These measurements included a wind turbine in Israel and results of a wind farm in Sweden. It was shown that these methods offer a significantly better accuracy in the prediction of the performance. In the case of the Vestas V25 turbine in Beit-Yatir (Israel), the error while using the regular method was 13.1% (overproduction), but the error was reduced to 2.8% by using the new method. The error was reduced from 14.2% to 5.5 for the wind farm in Alsvik in Sweden, where three Danwin D 23 turbines are located. During the research it was shown that the accuracy can be further increased by considering the influence of yaw errors.
It is not only the total energy product that is important. The quality of the output power. namely the intensity of power fluctuations, is also important. The new model offers methods of calculating and predicting the quality of the output power (predicting the distribution of the intensity of power fluctuations). This capability of the model was also validated by comparison with field measurements of turbines’ output.
Any failure of a turbine component hurts its economic success, since it means lower energy production and repair expenses. The loads that act on the various components cause failures. Thus it is very important to be able to predict correctly the loads that will act on the turbine, located at a certain site. Knowledge of these loads is essential to ensure the high reliability of the turbine.
The second part of the research included the development of an advanced general model to calculate the loads that act on the various components of a wind turbine. These components include: rotor, shafts, gearbox, generator, tower, etc. This model is based on a modern multibody formulation. It is modular and allows an easy replacement of any module of the entire system. The blades are modeled as elastic rods, where the detailed structural and inertia properties are taken into account. The connections between the blades and the hub allow the modeling of elastic joints. Elastic effects are also included in the modeling of the shafts, tower, etc. The modeling is general and allows large rotations and various kinds of nonlineraties. The equations of motion are derived based on e energy principles. The method of introducing the constraints between the various components of the system is based on using Lagrange multipliers.