P1.34
Factors affecting wind power production in mountainous terrain
Mark Zagar, Vestas Wind Systems, A/S, Randers, Denmark; and A. Hald Collins, L. Gulstad, S. Hahn, Y. Hristov, C. H. Hu, T. Maric, S. Holm Mogensen, and G. S. Oxley
Interesting locations from a wind power perspective are commonly located in a hilly or even mountainous terrain, with scales of terrain features far beyond a typical resolution of nowadays' operational numerical weather prediction models. High model resolution is also required for estimating the local climatology for the atmospheric parameters affecting power production and wind turbine fatigue load. Turbulence is the main driver of wind turbine loads, and after the wind speed the most influential variable. The applied numerical tools are required to provide a quantitative information about the four dimensional structure of turbulence. We are presenting preliminary results of a coupled mesoscale-CFD model suite applied on a complex site with extreme sub-mesoscale wind variability and associated turbulence.
In order to resolve flow features in downslope accelerated winds we run the WRF model in it's LES mode, i.e. resolving the largest boundary layer eddies. Computations at these resolutions are too expensive for operational use. Hence the required model resolution to provide useful information about wind exceeding specified thresholds in the gusts called pulsations is investigated. These quasi periodic large amplitude wind speed fluctuations present a particular challenge for wind power estimation and forecasting. The speed throughout the periods can be within the design specifications and only the energy production will fluctuate, or the speed can occasionally exceed the design specifications and the wind turbines will shut down thus causing production loss. In the worst case the sudden wind change will inflict damage to the wind turbines. We are presenting the newest findings regarding microscale dynamics of such downslope winds and site-specific methods for diagnosing their occurrence from the larger scale models.
Poster Session 1, Monday Poster Session
Monday, 30 August 2010, 5:30 PM-7:00 PM, Alpine Ballroom B
Previous paper Next paper