资源科学 ›› 2017, Vol. 39 ›› Issue (1): 125-135.doi: 10.18402/resci.2017.01.13

• 能源 • 上一篇    下一篇

GlobeLand30地表覆盖产品应用于精细化风能资源评估

张双益1, 2, 3, 胡非1   

  1. 1. 中国科学院大气物理研究所大气边界层物理与大气化学国家重点实验室,北京 100029;
    2. 中国科学院大学,北京 100049;
    3. 中国三峡新能源有限公司,北京 100053
  • 收稿日期:2016-08-15 修回日期:2016-10-24 出版日期:2017-01-20 发布日期:2017-01-20
  • 通讯作者: 胡非,E-mail:hufei@mail.iap.ac.cn
  • 作者简介:张双益,男,湖北老河口人,博士生 ,主要从事风能资源评价研究。E-mail:zhshy_002@163.com
  • 基金资助:
    国家自然科学基金(11472272)

Application of GlobeLand30 land cover product in refined wind energy resource assessment

ZHANG Shuangyi1, 2, 3, HU Fei1   

  1. 1. State Key Laboratory of Atmospheric Boundary Layer Physics and Atmospheric Chemistry,Institute of Atmospheric Physics,Chinese Academy of Sciences,Beijing 100029,China;
    2. University of Chinese Academy of Sciences,Beijing 100049,China;
    3. China Three Gorges New Energy Company,Beijing 100053,China
  • Received:2016-08-15 Revised:2016-10-24 Online:2017-01-20 Published:2017-01-20

摘要: 高山风电场具有多方面复杂特性,因此有必要开展精细化风能资源评估。本文创新性地引入了中国自主制作的全球首套30m分辨率地表覆盖产品GlobeLand30,制作高精度地表粗糙度资料,并应用于精细化风能资源评估。该方法在云南大团山风电场案例进行了尝试实施,使用计算流体力学(CFD)模式Meteodyn WT获得了30m分辨率的平均风速和平均风功率密度图谱。采用风电场内测风塔的实测数据对模拟结果进行了检验,观测期6个月平均风速的模拟误差在1%以内,平均风功率密度的模拟误差在4%以内。模拟与实测的风速频率分布和风向玫瑰均保持了较好的一致性。采用模拟数据计算了两种典型风电机组的发电量,观测期的总发电量误差低于1%。利用欧洲空间局地表覆盖产品开展了对照实验,其模拟结果的误差水平显著高于上述结果。本文研究成果基本满足风电场预测长期发电量水平的业务需求,具备应用与推广价值。

关键词: GlobeLand30地表覆盖产品, Meteodyn WT模式, 地表粗糙度, 精细化风能资源评估

Abstract: Due to the complex characteristics of high mountain wind farms,it is necessary to carry out refined wind energy resource assessment. Here we innovatively introduce China’s self-developed and world-first 30 m resolution land cover product:GlobeLand30. The product was used to produce high precision land surface roughness data applied to refined wind energy resource assessment. In a case study of the Datuanshan Mountain wind farm in Yunnan,China,30 m resolution mean wind speed and wind power density distribution maps were obtained by Com-putational Fluid Dynamics(CFD)modeling Meteodyn WT. The simulation results were verified by observed data from meteorological masts inside the wind farm. For the entire six months’ observation period,the average wind speed simulation error was less than 1%,and the average wind power density simulation error was less than 4%. The simulated and observed wind speed frequency distribution and wind direction were found to be consistent. Simulated data was used to calculate two typical wind turbines’ energy production. Total energy production in the whole observation period had a simulation error within 1%. The European Space Agency land cover product was used in the control experiment,and its simulation error was generally greater than our model. These results basically meet the demand of predicting long-term power generation levels for wind farms and have the value of application and promotion.

Key words: GlobeLand30 land cover product, land surface roughness, Meteodyn WT model, refined wind energy resource assessment