P63 Creating wind power production data sets with WRF: The WIND toolkit project

Draxl, Caroline, Dennis Elliott, Scott George, Wesley Jones, Jim McCaa, John Michalakes, and Kirsten Orwig, National Renewable Energy Laboratory

Site-specific wind resource characteristics can have major impacts on wind turbine design and durability, plant siting and layout, and electric power system operation. The existing electric grid infrastructure poses significant limitations on wind power expansion. The WIND toolkit project is thus aimed at producing a state-of-the-art national wind resource dataset with WRF covering the contiguous United States for the years 2007-2011 for use in a variety of wind integration analyses, and wind power planning. WRF has been run with a 2 km horizontal grid spacing, and a vertical grid spacing every 20 m up to 160 m with model output every 5 min. The data will be freely available online via a data extraction tool for 110,000 onshore and offshore sites representing existing and potential wind facilities.

We will discuss the challenges associated with creating and storing many terabytes of multi-year wind resource output data. Parallel asynchronous I/O (pNetCDF combined with WRF Quilt-I/O) were used to keep pace with continuous generation of output data resulting from very high spatial and temporal resolutions for a large geographical area (whole US).

We will further discuss selections of WRF settings to optimize our output for the purpose of wind turbine arrays and include lessons learned from past projects. The validation approaches will include wind power variability and wind power ramping behavior from actual wind plants, as well as wind speed validation against tall tower measurements and remote sensing devices.