Wind Energy Resource Assessment for Cook Islands With Accurate Estimation of Weibull Parameters Using Frequentist and Bayesian Methods

• Published in: IEEE access 2022, Vol.10, p.25935-25953, Article 25935
• Main Authors: Singh, Krishneel A., Khan, M. G. M., Ahmed, Mohammed Rafiuddin
• Format: Article
• Language: English
• Subjects: Assessments; Bayesian analysis; Diurnal variations; Earth; Economic analysis; energy resources; Energy sources; Error analysis; Estimation; Goodness of fit; Ground level; Islands; Least squares method; Mathematical analysis; Parameters; Payback periods; Quartiles; Renewable energy sources; Statistical tests; Turbines; Turbulence intensity; Weibull distribution; Wind energy; Wind power; Wind power generation; Wind shear; Wind speed; Wind turbines
• ISSN: 2169-3536; 2169-3536
• DOI: 10.1109/ACCESS.2022.3156933

Wind energy resource assessments at two islands in the Cook Islands are carried out in the present work. The wind data were collected for one year from sites on Mauke and Rarotonga Islands in the Cook Islands and the daily, monthly and seasonal average wind speeds, the diurnal variations of the wind shear coefficient, average temperature and turbulence intensity were estimated. Eleven frequentist methods and a Bayesian technique were used to determine the Weibull parameters and the wind power density (WPD) for each site. The best method was determined using the goodness of fit test and error measures. The average wind speeds were 4.65 m/s and 3.86 m/s at 34 m above ground level for the sites on Mauke and Rarotonga respectively. Based on the goodness of fit tests and error measures, the Least Squares Method performed best for estimating the Weibull parameters at the Mauke site, while for the Rarotonga site, the median and quartiles method performed the best. For both the sites, the Bayesian method, which is being used for the first time for wind resource assessments, ranked second of the twelve methods, indicating good potential for this method. The annual energy production (AEP) was also determined which was calculated to be 2192.34 MWh from a total of ten Vergnet 275 kW turbines at the two sites. Finally, an economic analysis, carried out for the two sites, indicated a payback period of 7.72 years.