| To fulfill the requirement of wind farm building, a high-resolution non-linear numerical model NLMSFD (Non-linear Mixed Spectral Finite Difference Model) of Canada is used to assess wind resource of Shandong Rongcheng area. Experiment as below: In the experiment area, one-year average monthly wind speed observation data of different height in two wind towers are selected to drive the model which simulated wind speed distribution of the experiment area, obtained the wind energy distribution in a pilot area under 100m×100m resolution and assessed the applicability of NLMSFD model in Shandong Rongcheng.Conclusion:(1) With driving the model by one-year average monthly wind speed observation data of different height in two wind towers, the spatial distribution information of experiment area in four seasons of one year are obtained. Through comparing the simulated values with observation values, the results show that the simulation of NLMSFD can reflect spatial and temporal variation of the average wind speed basically. But the relative error of simulated and observation values will change in the different height and season. From the seasonal perspective, the model simulation results are best in winter, better in spring and autumn, worst in summer; from the point of view of height, most of the relative error is less than 10% except for the height of 10m. Moreover, the distribution of simulated wind speed in each height is related to the variation of terrain. In the area with high altitude and big variation of slope, the gradient of wind speed will be bigger. On the contrary, in the area with low altitude and little variation of slope, the gradient of wind speed will also be littler.(2) The accuracy of simulation is also affected by using observation data in different locations or different height to drive model. For example, using observation data of the tower which are away from the coast to drive model, the results are better than the coastal observation data. At the same time, the results which are driven by the observation data higher than 10m are better than using 10m data.(3) Through test of WAsP and NLMSFD model's comparison in tower A, it was found the wind speed which WAsP simulated was greater than NLMSFD model's results, and at 10m height, simulated speed of WAsP is closer to the observation; but in 70m height, NLMSFD simulated and observed values are close. (4) Simulated result will be affected by prevailing wind direction of tower A & tower B through some individual cases analyzes. When the prevailing wind direction of the speed affect tower A & tower B speed, the result of simulation is good, too. When the prevailing wind direction of the speed can not affect tower A & tower B speed, the result of simulation is not good, either. At the same time, when the prevailing wind is constant, different wind speed will also affect the result of simulation. When the wind speed is higher, the result of simulation will be better.(5) Different roughness will affect the result of simulation in January are also discussed. The result which was driven by average wind speed and roughness of windy & cloudy day is different from the result which was driven by average monthly wind speed and roughness. The result of simulation which was driven by the observation date of windy & cloudy day of tower B is better than that driven by tower A.These results indicate that NLMSFD model has some reference and application value in wind resource assessment and wind farm site in Shandong Rongcheng. But because this is the first attempt in China, other aspects of wind energy assessment should be studied to further discuss the value of NLMSFD model. |
PDF Full Text Request