Posters

Abstract

Wind turbine wake losses are commonly modeled using the analytical Park1 and Park2 wake models, which are based on the model of N. O. Jensen. The accuracy of these models strongly depends on the wake decay constant (WDC), which controls wake expansion and thus the wake recovery rate. Wake recovery is mainly influenced by site-specific meteorological conditions, particularly turbulence intensity. Selecting an appropriate WDC that reflects local conditions therefore remains a key challenge in wake loss modeling. One approach to address this challenge is to parameterize the WDC as a function of turbulence intensity or surface roughness length. However, this approach is typically applied to microscale or single-turbine applications, as detailed meteorological information - such as turbulence intensity - usually requires on-site measurements. In addition, wake loss modeling software is often designed for microscale applications only. In contrast, when modeling wake losses on a large scale, site-specific meteorological conditions are often neglected. Instead, coarsely resolved data and flat-rate assumptions to estimate wake losses are commonly applied, which reduces the accuracy of wake loss estimates. This study presents an approach that leverages high-resolution spatial and temporal meteorological data, together with detailed wind turbine information on a microscale, to model wake losses on a large scale. To this end, a fully automated model chain was developed that integrates multiple data sources, including onshore wind turbine data from the German Core Energy Market Data Register and time-series-based meteorological datasets. The WDC is dynamically calculated for each wind turbine and time step based on EMD International A/S recommendations, considering prevailing meteorological conditions. Two data sources are used to derive the WDC: (1) turbulence intensities from the Global Atlas of Siting Parameters and (2) effective roughness lengths derived from the CORINE Land Cover dataset. For each data source, wake losses are simulated using the Park2 wake model and calculated at turbine level. For comparison, an additional simulation is performed using a constant WDC of 0.09 for all turbines, as recommended by the Technical University of Denmark for the Park2 model and onshore applications. Resulting wake losses are subsequently cumulated at the federal state level and compared across simulations. The results show that federal states with high turbine densities, such as Brandenburg and Schleswig-Holstein, exhibit higher overall wake losses, while states with lower turbine densities, such as Bavaria and Baden-Württemberg, show lower losses. Wake losses derived from WDC values based on the Global Atlas of Siting Parameters are systematically higher than those obtained using either a constant WDC or WDC values based on roughness length data. These findings demonstrate the strong sensitivity of wake loss estimates to the chosen WDC and indicate that site-specific meteorological data should be considered when calculating wake losses. By transferring the micro-scale modeling of wake losses to the macro-scale level, the proposed framework allows wake losses to be better represented spatially and temporally, which can be used to model wind power feed-in for regional grid studies and contribute to a secure power supply.