Refining Wind Power Assessment in Essaouira, Morocco: The Comparison Study for Four Methods for Weibull Distribution Parameter Estimation

Abstract

Quantitatively modelling wind speeds at a given location is necessary optimize wind power use for energy system planning purposes and assist in selecting structural material for wind turbine designs. This research analysed whether probability distribution functions (PDFs) approximate wind speeds within the means of renewable energy engineering and materials performance. Four techniques were analysed: Lysen empirical method (EML), Justus method (EMJ), a graphical method (GRAPH), and the four-moment mixture method (MFMM). The four techniques estimate the two Weibull parameters, k (shape) and c (scale), which are used to describe how the wind behaves structurally, predict the loads on the structures, and guide materials for durability of turbine component parts. The data set included one year's worth of wind speed collections for Essaouira, Morocco, which has an extensive coastal history of high wind speeds. All parameter estimations and other calculations were performed using MATLAB. The accuracy of the techniques and comparisons of accuracy were evaluated using statistical measures, including RMSE, Chi- square (χ 2 ), MAE, and R 2 . The findings indicate the strengths and weaknesses for all four methods, including considerations for selecting the best method to analyse prototype wind speed data in both energy engineering applications and materials-based design.