Uncertainty of toggling in wake steering experiments under diurnal cycle atmospheric conditions: an LES study

Abstract Large Eddy Simulations are used to study wake steering under a 24hr diurnal cycle inflow, which provides realistically varying atmospheric conditions. The impact of the diurnal cycle on wake steering effectiveness and the uncertainty introduced by toggling are investigated. Toggling is frequently used in field experiments on wind farm flow control, involving activating and deactivating the control at a given frequency, and comparing the quantities of interest in the active and inactive periods. This inherently assumes similarity in the flow characteristics between consecutive toggling periods. In LES, identical atmospheric conditions can be simulated with and without wake steering, meaning the baseline performance is precisely known and hence the impact of toggling can be quantified. Wake recovery, power output and effectiveness of wake steering vary significantly during the diurnal cycle, due to the changes in inflow turbulence intensity, veer and shear. Convective periods lead to total power losses, while large power gains are observed for stable periods. The toggling strategy leads to an overprediction of the spread of the total power change distributions across all toggling time periods, leading to occasional false positives and false negatives. Using shorter periods (10-30 min) results in the best agreement with the actual power gains in both spread and mean, as the assumption of similar flow characteristics is more valid than for longer periods, mainly due to the stability changes in the diurnal cycle. However, both the toggling approach and actual power gain predict that wake steering is overall beneficial for this case.