Modeling wind speed and wind power distributions in Rwanda

Utilization of wind energy as an alternative energy source may offer many environmental and economical advantages compared to fossil fuels based energy sources polluting the lower layer atmosphere. Wind energy as other forms of alternative energy may offer the promise of meeting energy demand in the direct, grid connected modes as well as stand alone and remote applications. Wind speed is the most significant parameter of the wind energy. Hence, an accurate determination of probability distribution of wind speed values is very important in estimating wind speed energy potential over a region. In the present study, parameters of five probability density distribution functions such as Weibull, Rayleigh, lognormal, normal and gamma were calculated in the light of long term hourly observed data at four meteorological stations in Rwanda for the period of the year with fairly useful wind energy potential (monthly hourly mean wind speed v ¯ ≥ 2     m    s − 1 ). In order to select good fitting probability density distribution functions, graphical comparisons to the empirical distributions were made. In addition, RMSE and MBE have been computed for each distribution and magnitudes of errors were compared. Residuals of theoretical distributions were visually analyzed graphically. Finally, a selection of three good fitting distributions to the empirical distribution of wind speed measured data was performed with the aid of a χ2 goodness-of-fit test for each station.