Evaluation of Wind Speed Probability Distribution Model and Sensitivity Analysis of Wind Energy Conversion System in Nigeria

Abstract

This paper evaluates the wind potential of some specified locations in Nigeria, and then examines the response of wind energy conversion systems (WECSs) to this potential. The study employs eight probability distribution (PD) functions such as Weibull (Wbl), Rayleigh (Ryh), Lognormal (Lgl), Gamma (Gma), Inverse Gaussian (IG), Normal (Nl), Maxwell (Mwl) and Gumbel (Gbl) distributions to fit the wind data for nine locations in Nigeria viz. Kano, Maiduguri, Jos, Abuja, Akure, Abeokuta, Uyo, Warri and Ikeja. The paper then uses the maximum likelihood (ML) method to obtain the parameters of the distributions and then evaluates the goodness of fit for the PD models to characterize the locations’ wind speeds using the minimum Root Mean Square Error (RMSE). The paper analyses the techno-economic aspect of the WECSs based on the daily average wind speed; it evaluates the performance of ten 25 kW pitch-controlled wind turbines (WT1 – WT10) with dissimilar characteristics for each location, including the cost/kWh of energy (COE) and the sensitivity analyses of the WECSs. Results reveal that Ryh distribution shows the best fit for Kano, Jos, Abeokuta, Uyo, Warri and Ikeja, while the Lgl distribution shows the best fit for Maiduguri, Abuja and Akure due to their minimum RMSE. WT7 achieves the least COE ranging from $0.0328 in Jos to $4.4922 in Uyo and WT5 has the highest COE ranging from $0.1380 in Ikeja to $53.371 in Uyo. The paper also details the sensitivity analysis for the technical and economic aspects.

Keywords

• wind speed
• wind turbine
• probability density function
• cost of energy
• wind resource.