Posters

Abstract

Loss of yield due to noise-related curtailment of wind turbines (WTG) is a regular occurrence in Germany, especially at night, and will occur more and more frequently in the future due to the ongoing expansion of onshore wind energy. For the economic calculation of new wind farms, it is necessary to estimate these losses as accurately as possible in advance during the planning phase and to adjust them continuously in the event of changes to the planning. At Fraunhofer IEE, methods are being developed to calculate the losses due to noise-reduced operation with a high degree of automation: The noise pollution from existing wind turbines is considered by including WTGs from the German “Marktstammdatenregister”. As a next step an automated selection of the relevant immission points based on the type of settlement area and a corresponding sound threshold level is carried out. In particular, these points involve a considerable amount of work with conventional noise calculation methods and represent a challenge for automation. In the event of missing information, such as octave bands or attenuation modes, approximations were chosen or generic models were developed. A calculation model was developed that incorporates the various models and data sets, in which only the input of plant coordinates and types of new wind farms is required to identify available noise quotas and, if necessary, to determine the required throttle modes of the new WTG for each plant. The model was validated using 37 noise reports from the German environmental impact assessment portal. In the comparison with these noise reports from different federal states, it was shown that the results were good overall, but that in some federal states there were major errors and that it makes sense to use a federal state-specific parameterization of the calculation basis. Finally, the yield losses due to the reduced sound power levels are calculated based on reanalysis time series. Overall, such automated loss estimates have higher uncertainties due to the simplification by the underlying data, but can provide a good approximation for an initial estimate of the losses.