Wind Engineering

The extent to which a wind energy site is affected by a coastal cliff is presented by studying numerically a neutral Atmospheric Boundary Layer (ABL) flow using an RNG k-ε model and different geometries. Initially, the classical flow over a forward-facing step is modelled, followed by the modelling of a neutral ABL over a rough plane in two and three dimensions with various types of ground conditions. Finally, the two and three-dimensional flows over a forward-facing step, representing a coastal cliff, in a neutral ABL were modelled and applied to the Atlantic Wind Test Site (AWTS), site composed of a wind turbine testing facility and a 13 MW wind farm, in the province of Prince Edward Island, Canada. After assessing that the model can predict classical flows and ABL with relatively good accuracy and robustness, the results show that the extent of the effect of a coastal cliff on the flow above the AWTS is limited to a distance of 5h downstream of the (height h = 10 m) cliff. Based on the sitting of the existing wind power infrastructure, it appears that the coastal cliff does not influence the power capacity of the site.

A very detailed 2D-solid finite element model is developed representing the load-carrying box girder of a wind turbine blade. Using typical geometrical values for the girder dimensions and public available material data, the overall cross-sectional behaviour is analysed for a simple compressive line load. The results are compared with result from similar shell models, which typically are used for practical design. Usually, good agreement between the shell models and the detailed 2D-solid model is found for the deflections, strains and stresses in regions with loads from pure bending. However, large differences can exist in regions where the loading is dominated by shear. It is found that geometrical non-linearity starts to become important when deflections are of the same order as the laminate thickness.

This paper describes the simulation model of a controller that enables an active-stall wind turbine to ride through transient faults. The simulated wind turbine is connected to a simple model of a power system. Certain fault scenarios are specified and the turbine shall be able to sustain operation in case of such faults. The design of the controller is described and its performance assessed by simulations. The control strategies are explained and the behaviour of the turbine discussed.

This paper addresses the dynamic modelling of large (MW) capacity Doubly Fed Induction Generators (DFIG) wind turbines using Pscad/Emtdc™ and Matlab/Simulink™. Suitable speed and reactive power controllers are implemented and described. Application studies are conducted to observe the performance of the DFIG during power system disturbances such as three-phase faults developing in different points of the network. These studies are used to determine the DFIG's fault current contribution and power converter rating required for fault ride-through capability and protection requirements. Simulations results are presented and discussed using typical turbine and network data for wind farm facilities.

Two hybrid stand-alone (autonomous) power systems, each with wind and PV generation, were studied as installed at health clinics in semi-desert and mountainous region in Mongolia. Meteorological and system operation parameters, including power output and the consumption of the system, were generally monitored by sophisticated monitoring. However, where wind and solar site information was lacking, justifiable estimates were made. The results show that there is a seasonal complementary relationship between wind and solar irradiation in Tariat Sum. The users understood the necessity of Demand Side Management of isolated wind-PV generation system through technology transfer seminars and actually executed DSM at both sites.

Many tests have been conducted on models in a wind tunnel for an optimum configuration of a Savonius rotor wind turbine. Three types of Savonius rotor have been used (a simple Savonius rotor of mild steel, an overlapped Savonius rotor of mild steel and one overlapped Savonius rotors of aluminium). The effect of different design parameters, namely rotor shape, overlap between rotor blades was studied. The results have corroborated some of the original findings of Savonius.

The wind speed at Inshas, Egypt in 2003 is considered as a case study. A statistical analysis gave the annual average wind speed as 3.80 m/s, in the range of 3.0 to 5.0 m/s for 89% of the time. Comparison among Weibull, Rayleigh, and actual data distributions of wind speed and wind power indicates that the Weibull distribution is the better fit. The mean monthly wind speed had its maximum in December and its minimum in February. The maximum wind energy potential was 63.6 W/m² in December and, the annual mean value was 34.4 W/m².