Modeling of Wake Effects on Power Loss, Fatigue Damage and Noise on a Cluster of Wind Turbines

点击翻译按钮获取中文摘要

Wake effects of a wind turbine (WT) are related to a decreasing velocity and an increasing turbulence intensity downstream of the rotating blades of a WT. This typically results in a 10%~20% decline in the total wind farm power production and a 5%~15% increase in the WT blade fatigue load. In addition, WT wakes also influence the sources of aeroacoustic noise generation (particularly, low-frequency noise in the frequency range from about 20-160 Hz and infrasound with frequencies less than about 20 Hz) in a cluster of WTs close to residential areas. This has become an issue of growing importance as it can potentially affect human health. This is commonly referred to as the “wind turbine syndrome”. It is against this background that a research program is proposed which involves the development of a new “aeroelastic actuator line” (AAL) wake model and the subsequent integration of this model into an in-house Computational Fluid Dynamics (CFD) code STREAM developed by the PI's group. As the predictability of AAL is sensitive to the choice of turbulence models, a new turbulence model based on the Partially Resolved Numerical Simulation (PRNS) approach will be proposed. The solution accuracy and computational efficiency will be enhanced by implementing the adaptive mesh refinement (AMR) algorithm in order to properly resolve each WT blade. In order to estimate fatigue damage and lifespan of isolated and clusters of WTs, fatigue models based on M-N and S-N curves (where M, S and N denote the moment, stress and number of load cycles, respectively) will be interfaced with the structural component in AAL. Finally, an acoustic code based on the hydrodynamic/acoustic splitting technique will be developed and coupled to STREAM to predict WT noise.******The novelty of the program is the development of an integrated wind energy tool called WATWind. This unique tool will couple STREAM with the newly proposed AAL wake model, which in turn can be interfaced with (1) an acoustic module for noise prediction, (2) a fatigue module for fatigue damage and total lifetime estimation of a WT blade, and (3) an artificial neural network (ANN) toolbox for wind power forecasting. It is expected that WATWind can assist WT blade manufacturers, such as Siemens Canada, to predict the lifespan of a WT blade, and the Independent Electricity System Operator (IESO) in Ontario to improve the performance of their wind power forecasting models. In addition, WATWind can also be used to assist health researchers and the Ministry of the Environment and Climate Change (MOECC) in answering questions such as what are the physiological effects (anxiety, tinnitus or hearing loss) associated with WT low-frequency noise and infrasound. The PhD and MASc students trained under this proposal will have acquired and contributed to the critical knowledge base that is needed for design and technology development of wind energy as one of the emerging sustainable energy sources.