Statistical methodologies for wind resource analysis, case: Catatumbo region - Norte de Santander, Colombia

Abstract

In this article, we analyze 4 probability distribution functions to characterize the wind resource at northeast zone of Colombia, based on data of taken from existing meteorological stations in three towns in the zone of Norte de Santander, Colombia. The data is sampled each 10 minutes, obtaining average values per hour every day, for each month of the year. The probability distributions studied are Weibull, Rayleigh, Log-Logistics and Gamma. In the case of the Weibull distribution, 5 methods were used to estimate two principal parameters of the distribution; shape and scale. For other distributions only was used one. The distributions for each data set were determined, exposing the best fit for each zone according to the 3 performance indices considered (RMSE, , R2). The main result obtained is the identification of the probability distribution that presents a better adjustment to the existing wind speed data in the zone where the meteorological stations are located. Specifically, it was identified that the probability distributions that best fit the data, for each month of the year, are Log-Logistics and Gamma. Further, the identification of the probability distribution that best fits to the wind speed data, allows to characterize with a greater degree of accuracy the wind potential of a zone.