Highly Time-Resolved Production Data Analyzed Using Engineering Wind-Farm Models

Wind farm design and operation require a good understanding of the complex intrafarm wind flows, with the level of sophistication depending on the concrete objectives. Often speed and thorough validation are essential and one prefers to focus on the average flow field using analytical, engineering wake models [6, 5] or highly simplified computational fluid dynamics calculations [1]. However, when standard engineering wind-farm wake models are employed for e.g. the maximization of wind-farm power through collaborative control schemes [8, 3], the results from experiments do not fully match the theoretical prediction. In fact, intra-farm wind flows are very dynamic with wake meandering, wake interaction and the entrainment of kinetic energy from the atmospheric layer above the wind-farm boundary layer. It is desirable take these dynamical effects into account, while still working with descriptions simple enough to be used efficiently in an optimization calculation. As a step in this direction, we here present a methodolgy where high-frequency intra-farm power generation data from the offshore wind farm Nysted is analyzed by comparison to several variations of engineering wake models, which differ in the description of single wakes as well as multiple wake interactions.