Evaluation of a roughness length parametrization accounting for wind–wave alignment in a coupled atmosphere–wave model

Abstract

The importance of wind energy as an alternative energy source has increased over the latest years with more focus on offshore winds. A good estimation of the offshore winds is thus of major importance for this industry. Up to now the effect of the wind–wave (mis)alignment has not yet been taken into account in coupled atmosphere–wave models to study the vertical wind profile and power production estimations of offshore wind farms. In this study the roughness length parametrization of Drennan et al. in 2003, and its extension addressing the wind–wave (mis)alignment proposed by Porchetta et al. in 2019, are investigated in the Coupled Ocean–Atmosphere–Wave–Sediment Transport (COAWST) model. This study shows that the yearly mean wind estimation at hub height (100 m) is improved by the roughness length parametrization of Porchetta et al. compared to Drennan. This is mainly due to the increased roughness of the former parametrization compare to the latter, even in aligned wind–wave conditions. This difference in roughness is caused by the dataset used to obtain the constants, deep‐water conditions versus mixed offshore conditions. Moreover, the roughness length parametrization of Porchetta et al. performs better in two of three alignment categories. Furthermore, similar model performances are obtained if we exclude the wind directions from the wind shadow zone of the measurement mast or the wind directions from the recently built Alpha Ventus wind farm, which is in close vicinity of the measurement mast. Investigating different wind conditions shows that the new roughness length parametrization of Porchetta et al. performs best for both offshore and onshore winds. Additionally, we show that the coupled model estimations of the vertical wind are only slightly affected by significant wave height estimations. Similar model performances for different accuracies of significant wave height estimations are presented. One exception is the perpendicular alignment category where the new roughness length of Porchetta et al. outperforms the roughness length of Drennan when investigating the wind estimations related to significant wave heights with a higher accuracy. The roughness length parametrization of Porchetta et al. reduced the power production overestimation of the coupled model from 5.7 to 2.8%. We also show that the standalone atmospheric model including the roughness length of Charnock in 1955 has a degraded performance compared to the coupled model including the roughness length parametrization of Porchetta et al. for yearly average wind profiles.

Impact of ocean waves on offshore wind farm power production

“…The wave model, with two nested domains (Fig. 2), has the same setup as reported in [21]. [33], while the white-capping and the linear growth of waves due to wind are represented by the model of [34].…”

“…In order to be able to capture the largest differences between a standalone atmospheric model and a coupled atmosphere-wave model, we select two weeks of the year of 2010 where we expect the waves to have the largest impact and thus result in the largest differences between the two models. This is made possible by the simulation results of the full year of 2010 from [21] (Appendix B). The first selected week is from July 8 th until July 15 th .…”

“…The numerical simulations for each of the weeks are ran continuously and while the atmospheric model is initialized by the ERA5 reanalysis data set, the wave model is initialized by a precursor simulation of 7 days prior to the actual start of the simulation. This is done in order to obtain an initial wave field [21]. For the simulation results, the first 12 hours are used as spin-up and thus excluded from further analysis, according to [14] and [3].…”

“…As shown by the aforementioned studies [18,19,20], aligned wind-wave directions result in higher velocities and thus in higher power productions compared to no waves. The opposite is probably true for opposed wind-wave directions, since they have a higher roughness length causing the wind speed and thus the wind turbine power production to decrease [21].…”

“…Figure B.13: Domain and week averaged results from the work of[21] including significant wave height and difference in roughness length between a coupled model including the roughness length parameterization of[22] and a stand-alone atmospheric model including the roughness length of[40].…”