Posters

Abstract

To meet the expansion targets for wind energy defined by the German government, a large number of wind farms must be developed within a short period of time. The planning of a wind farm at a new site is fundamentally based on estimating the expected energy yield and selecting suitable wind turbines. At present, yield estimations are associated with considerable uncertainties and involve high time and cost expenditures, particularly due to the currently required one-year wind measurement campaigns. The objective of this project is therefore to enable more accurate yield assessments within a significantly shorter timeframe and at substantially reduced costs by improving the entire process chain. To achieve this objective, new methods are being developed to provide the wind energy sector with an improved wind data basis, including a new anemos wind atlas with a spatial resolution of 1 km. The 1 km anemos wind atlas for Germany constitutes a comprehensive database containing long-term time series of all relevant atmospheric parameters. Following simulations with the Weather Research and Forecasting (WRF) model, a key task is the optimization of the results through the correction of systematic deviations and extensive validation using a large number of measurement datasets. This optimization process enhances the quality of the wind atlas by improving time series, statistical characteristics, frequency distributions, as well as diurnal and annual wind patterns. Methods were developed to improve the wind data base and enable shorter measurement periods. Seasonal biases in the 1 km anemos wind atlas for Germany were analyzed and corrected using long-term reference data. The accuracy of the diurnal wind speed profile was further enhanced through a correction approach, which is particularly important for assessing time-dependent losses and revenue calculations. Regression analyses were used to identify dependencies between diurnal and annual cycles, and systematic errors were reduced by applying site-specific correction functions. Using the optimization method developed by anemos, the new 1 km wind atlas based on ERA5 reanalysis data achieves a high level of accuracy, further enhanced by corrections of the vertical profile as well as diurnal and annual cycles. An accurate representation of the diurnal cycle is particularly crucial for time-dependent loss and revenue calculations and was validated in the project using more than 100 measurement sites. An optimization approach is presented that applies post-processing techniques to WRF simulations using a large set of high-quality wind measurements. This approach substantially reduces wind speed bias, improves correlation, and leads to significant accuracy gains at most sites. Sub-grid information and measurement data are combined through multiple regression analysis to derive wind direction-dependent scaling factors, which are applied to the wind speed time series across the entire dataset.