Presentations

Abstract

Mesoscale models are commonly used to assess wind resources, while low-fidelity engineering wake models are applied for layout optimization. However, cluster wakes from neighboring wind farms, becoming increasingly more relevant, are not well captured by engineering models. Mesoscale models can represent them physically, but their coarse resolution and high computational cost make them unsuitable for accurate layout-level yield estimation and optimization. With coupling of the two modeling solutions, one would be able to utilize the strengths of both: external wakes are simulated with a mesoscale model, and the wind farm of interest is modeled with an engineering model. A coupling of the mesoscale model WRF and the engineering modeling suite FOXES was proposed by Sengers et al. (2024), in which the timeseries of wake-reduced background wind fields from the mesoscale model were used as input to the engineering model. In this work, we aim to validate this multi-model approach by comparing model results with SCADA data from the Amrumbank West wind farm in the German Bight. The years 2019-2021 were simulated, predating the commissioning of the neighboring Kaskasi wind farm. In this scenario, the distance to the (upstream) wind farm Nordsee Ost is less than 4 km.  To demonstrate the proposed coupling, three simulation strategies are carried out: 1. FOXES: WRF without turbines provides the background flow. Full wind farm cluster simulated with FOXES with common wake models (a.o. Jensen, TurbOPark). 2. WRF: Mesoscale model simulates all turbines via the wind farm parametrization; FOXES is not used. 3. WRF-FOXES: Multi-model tool chain in which WRF simulates the external wind farms and FOXES simulates the wind farm of interest.  The simulated external wakes of the upstream wind farm were evaluated against the SCADA data along the southern-most row of the Amrumbank West wind farm. Relative production patterns of the evaluated turbines reveal a good representation of the external wakes produced by WRF, outperforming commonly used wake models in FOXES. This contradicts some recent studies which proclaim a consistent overestimation of the deficit by WRF in the near-wake of a wind farm and agrees with studies that state a good match between model and measurement in the far-wake. When focusing on the pattern-of-production of the Amrumbank West wind farm, WRF shows relatively poor performance due to its coarse resolution. Here, wake models are more accurate in representing internal wake effects, showing a generally good match to the SCADA data.  Using the multi-model tool chain, results are highly sensitive to the wake-model configurations, with many configurations overestimating cumulative wake losses when a wind farm is subjected to upstream wakes. The mismatch mainly arises from how wakes from the wake models are superimposed onto the WRF field containing external wakes. We aim to provide concrete, technically grounded recommendations for best practices in wind-farm wake modeling and yield assessment under the influence of cluster wakes. Sengers, B., Vollmer, L., & Dörenkämper, M. (2024). Multi-model approach for wind resource assessment. In Journal of Physics: Conference Series (Vol. 2767, No. 9, p. 092024). IOP Publishing.