NONLINEAR DYNAMICAL CHARACTERIZATION OF WIND SPEED AND WIND POWER IN KANO, JIGAWA AND KATSINA STATES OF NIGERIA

Abstract

Accurate forecasting of wind speed and wind power potential as wind based generation remains the fastest growing source of renewable energy. However this cannot be achieved except if the dynamics of the wind speed is known so as to accurately select the most appropriate models to forecast and harness the vast wind potential. The focus of this research is nonlinear dynamical characterization of wind speed and wind power for the period of thirty eight years (38) in three (3) states; Kano (11°59’47’’N, 8°31’0’’E), Jigawa (11°45’22’’N,9°20’20’’E) and Katsina (12°59’26’’N, 7°36’06’’E) in Northwestern Nigeria. The data were subjected to various components of nonlinear analyses using the MATLAB software which included; Phase Portrait, Power Spectral analysis and Poincaré Map. The results from the Phase Portraits shows a spongy bird’s nest-like structure in all the sampled locations which are made up of distinct curves indicating chaotic behavior, nonlinearity and nonstationarity in the time series. The Power Spectral Analysis shows that there are no regular sharp dominant peaks which implies an irregular nature of the time series and also indicated low predictability. The existence of the higher harmonics in the power spectra plots indicates that the processes underlying the time series are nonlinear and have a broadband noise with continuum of frequencies observed which peaks at f_o = 39 cycles/day for all the three locations. The Poincaré map affirmed the characterization which showed distinct points scattered in phase space and are more concentrated in the origin (attractor) indicating chaos in the time series data. This research can help in adapting the best strategy to harness wind energy in the northern part of Nigeria. From this research, it was observed that Katsina has the highest wind potential in terms of wind energy harnessing. Kano has the highest chaotic behaviour and jigawa has the least