Improving the accuracy of wind speed statistical analysis and wind energy utilization in the Ningxia Autonomous Region, China

• Published in: Applied energy 2022-08, Vol.320, p.119256, Article 119256
• Main Authors: Dong, Zuo, Wang, Xianjia, Zhu, Runzhou, Dong, Xuan, Ai, Xueshan
• Format: Article
• Language: English
• Subjects: Probability distribution function; Wind energy utilization efficiency; Wind power; Wind speed statistical analysis
• ISSN: 0306-2619; 1872-9118
• DOI: 10.1016/j.apenergy.2022.119256

•Statistical analysis of hourly wind speed series.•Finding out the optimal distribution to characterize the hourly wind speed series.•Guiding wind farm operation with the hourly expected wind power. Accurate wind speed characterization is important for wind farms to efficiently utilize wind energy which plays a significant role in reducing the dependence on fossil fuels and promoting global carbon peak and carbon neutrality. Since hourly wind speed density exhibit a variety of shapes, their characteristic features change substantially within a day, influenced by meteorological factors, therefore, the hourly wind speed analysis is considered to improve the accuracy of wind speed characterization. In this paper, 4 commonly used wind speed probability distributions (Weibull, Rayleigh, Gamma, Lognormal distribution) are applied for hourly wind speed analysis, and the best fitting distribution of each hourly wind speed series is identified by the goodness-of-fit test. Then the hourly wind power distribution of wind farms is determined by the corresponding optimal hourly wind speed distribution and the relation between wind power and wind speed. Finally, the hourly expected wind power calculated by the hourly wind power distribution is used to guide the hourly operation of the wind farm. The measured wind speed and wind power data of Dashuikeng wind farm in the Ningxia Autonomous Region of China is taken as an example to illustrate the results, 1) the hourly wind speed analysis performs a more accurate characterization of wind speed than annual wind speed series analysis. 2) The range of the average value of hourly wind speed series is 5.7 m/s–6.9 m/s, and the average wind speed, overall, during the daytime is smaller than that during the nighttime. 3) Gamma distribution shows a better performance than the other three distributions in characterizing the hourly wind speed series. 4) Wind energy resources are not fully consumed during the daytime, especially the period of 11:00–18:00. The proposed method improves the characterization of wind speed and guidance of the hourly operation for the wind farm, which has an important significance in raising the efficiency of wind energy utilization.