Wednesday, 26 January 2011
In order to effectively integrate wind energy to the electric power grid systems, it is necessary to accurately predict the power that will be produced at the wind farms, and the 0 - 6 hour forecast is considered to be the most crucial. Since 2009, NCAR-RAL has been working with Xcel Energy in implementing a wind power forecast system for its US wind farms, employing the high resolution (Δx = 3.3 km) NCAR/ATEC WRF-ARW based RTFDDA. In this paper, we explore several approaches to improve the WRF RTFDDA data assimilation algorithm to take full advantage of the variety of observations in and near wind farms to improve the wind farm wind forecast, especially on the 0-6 hour time ranges. This includes 1) adjusting the model and actual terrain difference in the assimilated data in order to reduce the wind speed error and 2) utilizing the model wind direction for assimilating wind turbine nacelle anemometer wind speed measurements. We are also examining the impact of special datasets at a Northern Colorado wind farm from an intensive observational field project. These datasets include a wind profiler radar, eight 10m met tower, two 60m met tower, and a Windcube Lidar installed at and around the selected wind farm and wind speed measurements at wind turbine hub-height. The results from a number of case studies will be reported at the meeting.
- Indicates paper has been withdrawn from meeting
- Indicates an Award Winner