Investigating the effect of thermal stratification on wind farm blockage using LES

A wind farm acts as an obstruction, influencing incoming air flow as it approaches the turbine-array. Unfortunately, this blockage effect prevents the first row of a wind farm from acting as a perfect reference because the incoming flow is impacted by what follows downstream. Understanding how and to what extent blockage affects power production is gaining interest in the wind energy field. However, this effect remains relatively unaccounted for compared to other well-established physical phenomenon impacting wind farm performance.

 

Most wind farm blockage studies only consider neutral boundary layer conditions for simplicity, despite the fact that the atmospheric boundary layer is almost never truly neutral. In our study, we use large eddy simulations to investigate how thermal stratification of the atmospheric boundary layer influences blockage. While a large amount of mixing is characteristic of an unstable boundary layer, a stably stratified boundary layer hinders mixing and is expected to exacerbate the blockage effect. To address wind farm blockage in a stable boundary layer, we simulate and compare the performance of a large-scale wind farm to that of a solitary infinite row of turbines, while systematically varying the streamwise turbine spacing and the surface cooling rate.

We observed, in agreement with previous studies, that the first row of a wind farm produces less power than a solitary row when turbines are placed closely downstream and this blockage effect increases as the streamwise spacing decreases. Additionally, we observed that the relative performance of the first row is dependent on thermal stratification. As the surface cooling rate increases and the atmosphere becomes more stable, the relative power output of the first row of the wind farm decreases. The influence that turbine placement and thermal stratification have on wind farm performance through flow blockage should be considered in wind farm planning.