An Estimation of Extreme Wind Speeds on Jeju Island Using Measure-Correlate-Predict Method

Abstract

This study shows how wind turbine classes were determined for the following possible installation sites on Jeju Island, South Korea: Jeonhul, an islet region; Handong, a coastal region; and Cheongsoo, a mountainous region. These different region types were selected to compare the effects of their topographical conditions on extreme wind speeds. Long-term wind speeds at 60 m above ground level were estimated using the Measure-Correlate-Predict technique based on wind data for the last ten years taken from 10- to 14-meter high anemometers at meteorological observatories. Gumbel and Weibull distribution functions were used to estimate extreme wind speeds at various topographical sites. Then appropriate wind turbine classes were determined for the sites on the basis of IEC 61400–1 ed.3. The results showed that extreme wind speeds at the Jeonhul site were higher than those at the other sites. The Gumbel distribution function was determined to be more suitable than the Weibull distribution function for predicting extreme wind speeds. Wind turbine class I was therefore recommended to be installed at the Jeonhul site because of its comparatively high extreme wind speed.