An efficiency comparison of numerical methods for determining Weibull parameters for wind energy applications: A new approach applied to the northeast region of Brazil

摘要

This work introduces a new approach for the evaluation of numerical methods used in calculating the Weibull parameters for the prediction of wind resource. In this evaluation, power is a key issue since it proves to be quite relevant in predicting the feasibility of a given site for efficient use of wind energy. Previous knowledge of such information is important for the decision process on the technical feasibility of installing industrial wind farms. The paper performed a statistical analysis of seven mathematical methods to estimate the shape parameter k and the scaling parameter c of the Weibull distribution. Wind speed and wind power data collected in two coastal cities (Icapui and Camocim), State of Ceara, in the northeast region of Brazil were used. The methods used in the development of this research are the graphical method, maximum likelihood method, modified maximum likelihood method, empirical method, moment method, energy pattern factor method and the equivalent energy method. Comparative analysis of efficiency and accuracy involves the application of the following statistical tests: analysis of variance (R~2), root mean square error (RMSE), and chi-square (X~2). The use of power as a parameter to be evaluated in this study by analyzing the Weibull distribution, is something different from what has been discussed in the literature, and indicates that methods which are well evaluated when the wind speed is used as a comparison factor can change position when ranked by power. It is important to emphasize that the values of k and c were calculated directly from the values of wind power (not the wind speed, as usual), and subsequently it was statistically tested the adequacy of the values in the prediction of wind speed and wind power. The presented results indicate that the values generated for the shape and scale parameters provided differences considering the databases formed by wind speed and wind power, namely the graphical method. The wind power data produced accurate results when submitted to statistical tests.