Conceptualizing wind variability in Delaware
Date
2017
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
University of Delaware
Abstract
This study explores ways to conceptualize the variability of winds in southern Delaware by analyzing annual wind observations from 29 weather observation stations. Wind speeds were interpolated to a standard height of 114 meters using the Log Law by determining the surface roughness at each station using the Danish Wind Industry Land Roughness Criteria. This criteria specifies the height of vegetation and various terrain aspects surrounding a weather observation station and provides a corresponding roughness length to aid in the calculation of winds aloft accounting for local wind speed influences. ☐ These station data then were analyzed using several different techniques and subsequently interpolated to a 4.73-km grid using inverse distance weighting. Values of the mean, median, mean-median difference, standard deviation, upper and lower quartile, and quartile deviation were calculated for each grid point and mapped. This led to the characterization of four distinct regions of wind in southern Delaware – the Bay, Coastal, Inland, and Southwest corridors – the identification of which was assisted by auto- and cross-correlations between and among the stations to identify station-to-station similarity. The applicability of the Normal and Weibull distributions, as well as wind roses, for each of the four regions also was examined. ☐ Results indicate in the assessment of wind power resource availability, particularly in areas with highly variable winds, reliance on the mean wind speed and the assumptions implicit in the mean about the normal distribution of the wind speeds may cause some biases, due to the degree of skewness in wind speed data. Consequently, the mean-median difference may prove to be a more useful tool in assessing the applicability of the accuracy of the mean wind speed in representing winds over a given area. Moreover, the Weibull distribution better represents wind speeds than the Normal distribution and the quartile deviation is better than the standard deviation at describing variable wind patterns. As expected, the Coastal and Bay corridors represented the best sites for potential wind farms while the southwest corner of the state may also be useful for on-shore wind farms.
Description
Keywords
Applied sciences, Earth sciences, Wind energy, Wind farm, Wind pattern, Wind power, Wind variability