Limited long-term impacts of wind farms on local meteorology

Abstract:

Long-term weather and climate observatories may be affected by their changing environs, such as the growth of urban areas or changing vegetation. The Dept. of Energy Atmospheric Radiation Measurement Central Facility (ARM CF) near Lamont, Oklahoma, is located in a region optimal for wind energy development (see Figure). We explore if and how this wind development has affected the long-term measurements at the ARM CF.

Mesoscale numerical weather prediction simulations with and without inclusion of the wind plants demonstrate the expected impacts, with the largest impacts occurring during nighttime stable conditions. Enhanced mixing caused by wind turbines redistributes heat in the lower boundary layer, increasing nighttime surface temperature increases on the order of 0.5 °C, accompanied by similar cooling at altitudes in the top half of the turbine rotor disk. Turbulent kinetic energy (TKE) increases at the surface, only during nocturnal stable conditions. Wind speeds decrease in both convective and stable conditions. These “wake impacts” extend downwind, with the wind speed deficits extending ~ 30km downwind and the temperature impacts being constrained to within 5-10km.

However, this clear picture from simulations becomes murkier when we consider observations. Using surface-based temperature, wind speed, and flux measurements from the ARM CF over 17 years, we compare “pre wind plant” and “post wind plant” measurements to quantify the impacts on measurements from wind plants 3-7 km away from the CF. We segregate the data by atmospheric stability, hub-height wind speed, and wind direction. Generally, surface temperatures show a surprising decrease in temperatures in “post wind plant” stable conditions, with one outlier case of +0.2°C. TKE generally increases in stable conditions “post wind plant”, except for one outlier case that shows decreases of TKE at wind speeds when turbines should be operating. Assessments of surface sensible and latent heat fluxes are similarly mixed. As a result, no statistically-significant difference between the pre- and post- wind plant emerges.

The mixed results from the analysis of the ARM CF data suggest that wind plant impacts on local microenvironments are frequently obscured by other phenomena that can overwhelm the subtle wind plant wake effect signal.