Large Eddy Simulations investigating the effects of atmospheric stability on wind turbine loads and wakes in complex terrain

Deployment of wind turbines has been continuously increasing over the past years and decades. Currently, priority is set to the exploitation of offshore environments, but to reach climate goals, it will be essential to take into account unused onshore potentials. This presents unique challenges due to surface heterogeneity, including elevation, buildings and plant canopies, which can considerably alter the wind field. The influence of complex terrain also varies depending on atmospheric stability. These combined effects create a variable wind field with locally high wind shear and strong turbulence, which must be considered during site assessments due to their potential impact on turbine loads and wake behavior.

In our study we employed PALM to examine how complex terrain and atmospheric stability influence turbine performance, longevity, and wake behavior. We used geospatial data from the test site WINSENT in southern Germany, located on a plateau with a forested escarpment upwind of the main wind direction. For the simulation we focused on three variable elements: 1) plant canopies, 2) additional terrain upwind and 3) atmospheric stability. With a combination of these we created twelve different scenarios. For each of them we examined load distribution over the rotor blades, power output and wake recovery of a single NREL 5 MW turbine.

Additional upstream terrain can substantially modify the incoming flow by generating internal boundary layers and enhanced turbulence, with effects that strongly depend on stability. Under neutral and unstable conditions, added terrain accelerates wake recovery, while also reducing mean power output. In stable conditions, plant canopies play a critical role by increasing turbulence and boundary-layer depth, leading to more even load distributions across the rotor and significantly faster wake recovery. While mean power differences are smaller in stable stratification, load variability and wake deficits are strongly affected.