Large Wind Farm Wake Impacts: Estimates from Turbine Data, Meteorological Tower Data, and WRF Model Runs

Wind turbine wakes have an important impact on wind energy production, particularly in large wind farms, where impacts of many wakes have the potential to accumulate. Several studies have looked at such impacts in large wind farms over the ocean, where the flat surface makes it relatively straightforward to quantify the turbine wakes. Over land, the wake impacts are much more difficult to distinguish from the competing impacts of terrain variation, surface roughness changes, and diurnally varying atmospheric boundary layer turbulence, making it difficult to quantify the impacts on downstream portions of large wind farms.

We used wind turbine supervisory control and data acquisition (SCADA) data from several large wind farms in the central United States to estimate the impacts of wind turbine wakes within the farms. As a separate analysis, we performed pre- and post-build comparisons of meteorological tower data using support vector machine (SVM) methods to separate the effect of natural annual variations in wind speed from the wind farm impacts. A third technique involved running the Weather Research and Forecast (WRF) model with a wind turbine parameterization to estimate the aggregate full farm impacts during an overlapping time period of SCADA data and making comparisons between WRF output and SCADA data.

Using the SCADA data alone, wind turbine wake impacts on energy production were roughly half of what was predicted by the standard wake models during the pre-build wind resource assessments for the farms. This outcome is likely at least partially a result of the fact that only the near field wake is consistently detectable in the data. Attempts to measure far-field or aggregate wakes using SCADA data alone do not reveal distinct wake signals. The SVM analysis reveals wind speed deficits that vary from farm to farm, due to varying distances of tower locations relative to the turbines in the farm, but the overall magnitude of deficits suggests wake impacts are larger than what is measured in the SCADA data. WRF runs show distinct wakes that extend tens of kilometers downstream during stable conditions, which also implies larger wake impacts than measured. The WRF output also shows evidence for upstream wind speed deficits, indicating an obstructing effect on the flow that could at least partially reconcile the difference between SCADA and SVM estimated impacts.