Wind energy is currently the most important renewable energy source in Germany. To effectively integrate wind power into the energy system, it is essential to understand both the temporal and spatial characteristics of wind power feed-in. In Germany, the current feed-in tariff system largely compensates for losses, resulting in typically high losses due to factors such as wake and noise. To accurately assess turbine yields, several loss models have been developed. These models are integrated into a Python-based framework that utilizes geodata and meteorological time series to calculate the power feed-in from wind turbines, including losses. These calculations are fully automated and can be performed on various scales, from individual wind farms to large regions (e.g., federal states).

To calculate wake losses, the analytical Park1 and Park2 models are implemented. The noise model, based on DIN ISO 9613-2, estimates power reductions of wind turbines when permissible noise levels, as defined by the German TA Lärm, are exceeded. In addition, a model was developed to estimate losses due to restrictions imposed to protect bats. This model incorporates state regulations that mandate shutdowns based on meteorological conditions and the time of day and season, which influence bat activity.

Existing wind turbines are considered via the market master data register (MaStR). Power and thrust curves are generically calculated based on rotor diameter and nominal power. Wind speed is converted into electricity using the respective wind speed at hub height. Finally, wind turbine yields and losses are calculated for each onshore wind turbine, enabling regional comparisons.

Wake losses were higher in northern regions with advanced wind energy expansion compared to southern regions. Furthermore, wake losses are significantly influenced by the wake expansion parameter k, with higher values resulting in reduced losses.

The noise model was validated using noise reports from the German environmental impact assessment portal, showing mixed overall results. Some wind farms exhibit substantial errors due to the challenges in the automated classification of immission limit values, highlighting the need for future data improvements.

Strong State-specific dependencies for bat curtailment losses arise from statewide regulations and regional meteorological conditions.

The developed framework demonstrates the feasibility of estimating wind energy yields and losses for individual wind farms as well as large regions based on time series data, providing valuable insights from this analysis. It is also capable of simulating future wind energy expansion (including repowering) and forecasting future wind power feed-in while accounting for time series-based losses. Consequently, the framework serves as a remarkably accurate foundation for modeling wind power feed-in for regional grid studies or system analyses, thereby contributing to enhanced security of power supply.