Estimating potential wind energy from sparsely located stations in a mountainous coastal region

For the spatial planning of a potential wind power plant (WPP) site, the efficient assessment of the wind power potential is important. However, observations are lacking for most potential sites. Here, a statistical three-step approach is presented to estimate the spatial distribution of potential wind power in a region in which observational stations are sparsely located. First, multiple linear regression is used to fill the gaps in the wind data collected at the stations. Second, spatial interpolation with hybrid Kriging is used to construct the grid-based data based on the station-based, gap-filled data. Third, the wind power potential is estimated according to the theoretical relationship between wind speed and power. As a case study, the spatial distribution of wind power potential is estimated over Jeju Island, where WPPs currently exist. Initially, the selected regression model in this case study showed correlation co-efficients ranging from 0.48 to 0.82. In a second step, the spatially interpolated data with the gap-filled data reasonably captured the expected spatial distribution of wind speed than those with the observed data. Finally, the results showed that the coastal region might be the best region for WPPs rather than the high elevation region far from the coast and that all the existing WPPs are located in regions with high potential wind power. The proposed statistical approach is simple and efficient for preliminary spatial planning. Furthermore, it can be applied to other regions with sparsely distributed wind speed data, including mountainous coastal regions.