Wind farm power production reduction by the wake effect in Harmonie.

Abstract

This study investigates how the loss in power production of offshore wind farms, caused by the wake effect, depends on the stability of the atmospheric boundary layer. A windfarm parameterization is used in HARMONIE-AROME, an operational numerical weather prediction model. Three experiments are performed — a control run without turbines, a default wind farm run with turbines installed until January 1 2019 and a future run that consist of a possible scenario for 2050 — in the German Bight for a selection of hours in2019, where wind direction ranges between 78.75-101.25°, wind speed from 8 to 16 m s-1 and the atmospheric boundary layer is considered unstable when RIB, the Bulk Richardson number, < 0, neutral when 0 ≤ RIB ≤ 0.2 or stable when RIB > 0.2. First the performance of the wind farm parameterization is evaluated with measurement of wind speed at FINO1, a research platform 45 km north of Borkum. We find that the default run performs better than the control run, as for all stabilities the bias and mean absolute error at 91 m and 34 m height are reduced. Then, the dependency of the wind farm wake on atmospheric stability is examined. We conclude that our method to define the wake area performs good and the wake area is larger in stable atmospheres where turbulent mixing is suppressed. Finally, the dependency of the power production loss on atmospheric stability is investigated. The median relative production loss for all wind farms under investigation combined is 31.5% for stable hours, 25.5% for neutral hours and 17% for unstable hours.