Low-frequency wind speed variations and their impact on wind farm performance

Standalone large-eddy simulations (LES) of wind farms often assume steady inflow conditions, overlooking low-frequency, large-scale wind variabilities. As a result, they capture only higher-frequency boundary-layer turbulence and miss wind speed variations on timescales longer than several minutes. We demonstrate that the low-frequency wind speed variations significantly impact wind farm performance. We address this by incorporating dynamic wind speed changes in wind farm LES using a non-inertial moving reference frame within the concurrent precursor framework, allowing the simulation domain to follow mass-conserving, time-varying wind speed signals with minimal changes to standard setups. Simulations show that such variability can strongly influence wake flows and wind farm power output, with wake losses decreasing during flow acceleration and increasing during deceleration. These effects result from changes in wind speed on timescales significantly longer than the wake advection time between successive downstream turbines, which traditional LES do not resolve. Accounting for this variability will be important for improving wind farm performance predictions and for developing control strategies that are robust to changing inflow conditions.