Inscrição na biblioteca: Guest

ISSN Online: 2377-424X

ISBN Print: 978-1-56700-421-2

International Heat Transfer Conference 15
August, 10-15, 2014, Kyoto, Japan

Numerical Assessment and Optimization of Wind Farm on Complex Terrain

Get access (open in a dialog) DOI: 10.1615/IHTC15.rne.008834
pages 7321-7335

Resumo

Wind power generation is one of the utilization forms of renewable energies with the most mature technologies. However, difficulties still exist in extracting wind energy due to its non-uniformly distribution in both space and time, especially over complex terrain. A set of methods for numerical assessment and optimization of wind farm on complex terrain have been developed by the wind energy research group of Institute of Engineering Thermophysics, School of Aerospace, Tsinghua University: (1) Airflow simulation using Computational Fluid Dynamics (CFD) and terrain-following grid system. Computer codes of both Reynolds Averaged Numerical Simulations (RANS) and Large Eddy Simulations (LES) were programmed. (2) To avoid generating unstructured grids for adapting both terrain surface and turbine blades, the Virtual Particle Model to simulate wind turbine wake flow is presented. The model treats the effect of wake flow as convective and diffusive matter and simulates it by tracking scattered particles. It resolves the distortion of wake affected area in non-uniform velocity field. This concept decouples the calculation of original airflow and wake effect, leading to significant reducing of computational time. Moreover, the simulating of particle motion does not depend on grid resolution. (3) Optimization method is developed to automatically determine turbine positions with the target of maximizing total power output. The method is based on the results of CFD calculations, and employs the Virtual Particle Model. It can be used to optimize wind turbine layout on complex terrain. The optimizing process is performed with considerations of multiple incoming wind speeds and directions. (4) Because of the instability of wind resources, forecast is essential for the safety and efficiency of wind power generation system and power grid. A prediction system combing statistical model and physical model is developed. Several configurations are presented for different circumstances of requirements, employing AutoRegressive Moving Average (ARMA) model and CFD method with RANS and LES. In this paper, the concept and algorithms of the models and methods are presented. Numerical and experimental cases are used for validation. Results demonstrate that the present method is effective to evaluate wind resources, to optimize wind turbine layouts, and to predict power output for wind farms on complex terrain.