Abstract
Operating large offshore wind farms decreases wind speeds in them and their downwind surroundings. This is known as wind farm wake, and it can significantly impact annual energy production, especially in areas with high installed capacity density. To assess this in Danish waters, we use a mesoscale model to capture atmospheric conditions and the effects of wind farms on them, particularly the wind. We simulate the flow to estimate wind resources for the North Sea, South Baltic Sea and the Kattegat using three scenarios: no wind farms, existing farms as of November 2021, and projected deployment in 2030. The 2030 projections were made in the first quarter of 2024. We estimate reductions in wind speed, capacity factors, load hours, and recovery distances for wind speed and changes in other climate conditions.
The main conclusions of this report can be summarised as follows:
- Simulations show that in large wind farms for a 2030 scenario, wind speeds in the southern North Sea may drop by 4 m s−1 (corresponding to a reduction in 15 M
- When focusing on the Danish EEZ (Exclusive Economic Zone), existing wind farms and those envisioned for 2030 produce modest reductions in hub-height wind speed and capacity factors compared to other regions in the southern North Sea, owing to strong winds and relatively low installed capacity densities.
- Simulations show that the recovery distances, i.e., the distance required for the wind speed to return to its background value, for offshore large wind farms vary between 20 km to 80 km. This distance depends on the installed capacity density, the size of the wind farm, and the background wind speed. In the Danish EEZ, recovery distances are 20 km to 30 km.
- The modelling results indicate a greater impact on wind speed in large wind farms when using one parameterization rather than the other; however, the recovery distance to the background capacity factor is similar in both simulations
- The results in this report carry considerable uncertainty, as many aspects of the modelling outcomes remain unvalidated. Comprehensive data needed to validate the simulations is absent, primarily due to the scarcity and availability of extensive observations, and large wind farms similar to those envisioned for the 2030 scenario have yet to be established.
- Simulations show decreases in 2-m temperature (up to 0.2 °C), increases in boundary layer height (up to 120 m), and cloud fractions (up to 7 %). While mean annual changes are not statistically significant at the 95 % level, they may be significant in specific seasons or stability conditions.
- Wind turbines in Denmark would have produced about 20 TW h if driven by the winds simulated by the WRF model and a wind farm parameterization. This aligns reasonably well with the 16 TW h reported by the DEA for 2019–2021, which includes many other losses.