DescriptionElectrical power generation via large scale wind turbine farms continues to grow worldwide as a cost-effective renewable energy resource. The proper placement of wind turbines within a given resource area must consider the optimal power capture and the wake interference effects from turbine to turbine. Flow simulations at the scale afforded by HPC systems will enable wind farm developers to account for these effects and reduce the total cost of energy.
Prof. Dimitri Mavriplis and his students at the University of Wyoming have been developing W2A2KE3D code, a high order finite element solver instrumented with VisIt/Libsim for in situ data processing. It has been applied to various aerodynamic applications such as airplanes, helicopters and now wind turbines. This HPC Impact Showcase will highlight how the Cheyenne supercomputer located at National Center for Atmospheric Research / University Corporation for Atmospheric Research (NCAR/UCAR) Computational and Information Systems Lab (CISL) was used to simulate the Lillgrund Wind Plant. This work represents the most highly resolved simulation of a wind farm to date. It includes the detailed unsteady fluid structure interactions occurring within each turbine and has the capability for atmospheric turbulent inflow conditions. These simulations required over 1.1 billion degrees of freedom to perform the simulation on over 32,000 compute cores.