Wind Energy Resource Assessment for Suva, Fiji and Design of a 30 kW Wind Turbine

Abstract

Abstract Fiji needs to invest in renewable energy sources to meet its energy needs to reduce the country’s dependence on imported fossil fuels. For investing in wind energy projects, a detailed assessment of wind energy resource is required. In this work, wind speeds were measured at 34 m and 20 m above ground level at a site in Suva for three years and the daily, monthly, yearly and seasonal averages were estimated. Average turbulence intensities at the two heights were also estimated. The Weibull parameters, average wind speed and the wind power density were estimated by using eleven frequentist methods and a Bayesian technique. These twelve methods were compared against each other for their performance using five goodness of fit test and error measures. The best method was found to be the empirical method of Lysen (EML) which gave a mean wind speed of 5.04 m/s and a wind power density (WPD) of 147.79 W/m2. A horizontal axis wind turbine of 30 kW capacity was designed and optimized using Harp_Opt software which works on a multi-objective genetic algorithm. The blade sections (airfoils) were designed using an in-house multi-objective genetic algorithm code by mathematically parametrized 7th order Bezier curve coupled with XFOIL software. The lift and drag coefficients were interpolated using AirfoilPrep to get the data in the required format as needed by the Harp_Opt GUI. The Weibull parameters from the statistical analysis of the measured data were used to optimize the performance characteristics of the wind turbine. The output power curve shows a cut-in speed of about 2 m/s and a rated wind speed of 10 m/s. The AEP was optimized from around 47.3 MWhr/year to 48.3 MWhr/year after 50th iteration of Harp_Opt.