The first conclusion that could be drawn is that the capability of improving acoustic performance of wind turbines by profile optimization depends dramatically on the geometrical and aerodynamic constraints applied during the optimization process and on the definition of reference case for optimization. The fact that noise from untreated and clean blades are quite similar suggest that the clean case is a better selection for reference case that the rough one. The tolerances during manufacturing process could modify the acoustic behaviour of the profiles and, therefore, they have to be taken into account during the optimization process to obtain a feasible geometry and reduce the sensibility to manufacturing quality. The design of new profiles that could control the properties of the boundary layer is a promising line of investigation to improve the new generation of wind turbines blades. The results obtained in this project have confirmed that optimization processes are applicable to improve the acoustic performance of modern airfoils. The wind tunnel measurement campaign has proven the validity of theoretical work and the confirmation on real wind turbines have showed a smaller noise reduction than expected from wind tunnel tests. The noise reduction obtained during Sirocco campaign is of the order of 0.6 dB for the reference case. This value should be used as an indication, because the geometrical constraints imposed during the optimization phase could be relaxed if a new blade was designed from scratch. More investigation is needed to fully understand the mechanisms of noise in wind turbines and in this context the collaboration between the industry and other institutions should play an important role in future noise reduction. The support of EU, national committees and other authorities is indispensable to obtain the degree of maturity in this technology that the community is demanding. In this context, the main success of Sirocco project has been its capability of creating a network of companies and institutions that work together across Europe with the common objective of reducing noise emitted by wind turbines. The application of profiles designed following the procedures developed during this project is questionable for current blades, where the cost of replacing the manufacturing tools would be high compared to the economical benefits derived from a reduction in noise. However the use of this method for new blades is under evaluation inside GAMESA, as although very promising, it involves some risks. More research is required to reach the state of technological maturity that will allow the implementation of new blades designed for low noise levels and high aerodynamic performance. Nevertheless, within GAMESA, the outcomes of the project are considered a technological breakthrough, and the project itself is regarded as a major success, not only for the knowledge gained through it, but also in terms of establishing a cross-European network of research and development in the field of wind turbine technology.
An experimental method has been demonstrated with which noise sources on a wind turbine can be localized and quantified. The technique employs a planar microphone array on the ground, which has a diameter of about 8 m and contains about 150 microphones. Using conventional beam-forming, the noise sources in the rotor plane are localized. In this way e.g. noise from the nacelle can be distinguished from noise from the blades. Moreover, it can be assessed on which part of the revolution most of the noise is produced. By using a trigger signal which monitors the RPM of the rotor, an alternative processing method can be applied which localizes and quantifies the (dedopplerized) noise sources on the rotating blades. In this way noise from the individual blades can be distinguished, for any segment of the revolution. In addition, the technique yields an improved signal-to-noise ratio.
The following outlines the major findings of the final SIROCCO measurement campaign, conducted on the GE 2.3MW prototype wind turbine in Wieringermeer, as well as the business potential of the noise reducing blade design and trailing edge serrations. The 2007 measurement campaign showed that the relative apparent sound power level reduction achieved with the SIROCCO blade was 0.5 dB while the SIROCCO serration design achieved a reduction of 3 dB. These values are an average of all measurement points gathered for the clean blade state (state 2). The noise reduction of the SIROCCO blade is low and typically within the uncertainty level of an IEC 61400-11 test. This low reduction could be caused by: - slight defects introduced during the manufacturing of the SIROCCO blade - the physical differences between the full-scale and wind-tunnel flow fields - the presence of other unexpected dominant noise mechanisms besides the trailing edge noise. The limited noise reduction measured on the SIROCCO blade does not justify the full-scale production of this blade in its current form. The competitive noise advantage is too low to offset the required investment into new manufacturing hardware. However, the airfoil designed as the basis for the SIROCCO blade could be implemented in future low noise blade designs. Furthermore, it is likely that the performance of low noise airfoils could be greatly increased using the airfoil design tools developed and validated during the SIROCCO program. To achieve this it would be necessary to relax the airfoil design constraints imposed by the turbine that was used for testing. Trailing edge serrations achieved the targeted noise reduction, and provide the basis for a possible field implementation where suitable.
An existing numerical optimization environment was extended to enable the constrained aeroacoustic optimization of subsonic airfoil sections. The coupling of the aerodynamic calculation method and the noise prediction scheme was improved based on detailed boundary-layer experiments conducted during the present project. The improvements significantly increased the accuracy and consistency of the airfoil trailing-edge noise prediction as demonstrated by comparison to acoustic measurement results. One field of application of the method is the design of airfoils for wind-turbine blades with reduced noise emission. The design of silent, high performance airfoil sections can be offered to the wind-turbine industry. The aerodynamic and aeroacoustic properties of the airfoils are verified in the institute's Laminar Wind Tunnel.
The NLR-measurements have been used to validate the aero-acoustic wind turbine code SILANT. This code was developed in 1996 by a Dutch consortium that consisted of Stork Product Engineering (SPE), the Dutch Aerospace Laboratory (NLR) and TNO. The SILANT code calculates the sound power level of the wind turbine blades and sums it to the overall wind turbine sound power level. The input for the code consists mainly of geometrical and aerodynamic data, operational conditions and external conditions. Basically SILANT calculates the noise level as follows: "The wind turbine blades are divided in a number of segments (usually 10 to 20); "For every blade segment two noise sources are calculated: - Trailing edge noise: According to the model of Brooks, Pope and Marcolini - Inflow noise: According to the model of Amiet and Lowson The noise sources are ('acoustically') summed over the segments in order to obtain the total blade and turbine sound power level. On basis of the NLR measurements and by comparing with results from a 2D noise prediction code from NLR many improvements could be made to the original code. The improvements eventually led to a code which delivered results which are very close to the measurement results.