ribution is  . Thus, if the wind speeds tends to spike steeply at a certain value, the distribution will have a high k value.
The Weibull results of standard deviation were compared with the actual standard deviation to assess the degree of convergence. The results are displayed in Figure 4. Figure 5 also show that the Weibull scale parameter “c” and wind speed has a correlation.
Figure 6 clearly demonstrated that all the sites have a better magnitude of wind speed profile in the wet (April to September) than in the dry season (October to March). Ikeja however showed much better wind speed profile in the wet than others and this is due to itslocation which is close to the Atlantic Ocean, thus, the ocean wind blows into the city. The mean measured data distribution between the dry and wet periods gave the range of mean wind speeds as 3.5 to 4.3 m/s and 3.5 to 4.1 m/s for Abeokuta, 4.3 to 5.3 m/s and 4.1 to 5.4 m/s for Akure, 5.0 to 6.0 m/s and 3.8 to 5.6 m/s for Ibadan, 6.4 to 7.7 m/s and 5.8 to 7.5 m/s for Ikeja, 3.5 to 4.6 m/s and 2.7 to 4.7 m/s for Oshogbo respectively. Other studies have shown these seasonal signatures of winds in Nigeria   .
The seasonal and whole year’s PDF and CDF plots are presented in Figures 7-9. The figures shows that the wind speed profiles for the periods follow the same cumulative distribution pattern. These patterns were also observed in previous studies    . The difference in shapes of the CDF and PDF plots were the results of the varying values of k and c.
On the wind power density across the period and years considered (Figure 10 and Figure 11), Ikeja has the highest wind power potential with its peak in August and 2006 while Oshogbo has the least. As can be seen in
Figure 3. Frequency of wind speed data (10 m height) over Abeokuta, Akure, Ibadan, Ikeja and Oshogbo, Nigeria for the period 1961-2011.
Figure 4. Actual standard deviation and Weibull deviation of wind over Abeokuta, Akure, Ibadan, Ikeja and Oshogbo, Nigeria for the period 1961-2011.
Table 2. Frequency of occurrence of wind speed data over Abeokuta, Akure, Ibadan, Ikeja and Oshogbo, Nigeria for the period 1961-2011.
Table 3. Results of the Weibull analysis of wind data for (a) Abeokuta (b) Akure (c) Ibadan (d) Ikeja and (e) Oshogbo, Nigeria.
Figure 5. Plot of Weibull parameters and wind speed.
Figure 6. Seasonal and whole year (1961-2011) variation of average wind speed values over Abeokuta, Akure, Ibadan, Ikeja and Oshogbo, Nigeria.
Figure 7. Wet season Probability Density Function (PDF) and Cumulative Distribution Function (CDF) from the Weibull analysis for Abeokuta, Akure, Ibadan, Ikeja and Oshogbo, Nigeria.
Figure 8. Dry season Probability Density Function (PDF) and Cumulative Distribution Function (CDF) from the Weibull analysis for Abeokuta, Akure, Ibadan, Ikeja and Oshogbo, Nigeria.
Figure 9. Whole year (1961-2011) Probability Density Function (PDF) and Cumulative Distribution Function (CDF) from the Weibull analysis for Abeokuta, Akure, Ibadan, Ikeja and Oshogbo, Nigeria.
Figure 10. Monthly and seasonal wind power density (W/m2).
Figure 11. Annual wind power density over Abeokuta, Akure, Ibadan, Ikeja and Oshogbo, Nigeria.
Figure 11, the significant monthly change in power density according to Keyhani et al.  underscores the importance of distinguishing different months and periods of the year when a wind power project is assessed or designed to produce maximum power.
Wind power density directly determines electric power efficiency in using wind energy in the sense that the higher the density, the higher the electric power generating potential of the area. Thus the potential for production of electric power during this period will be much greater in Ikeja than in other sites.
The monthly and seasonal most probable and maximum energy carrying wind speeds analyses for all the sites are given by Figures 12(a)-(e). The values of ranged from 2.75 to 7.87 m/s (for January
Figure 12. The most probable and maximum energy carrying wind speed over (a) Abeokuta; (b) Akure; (c) Ibadan; (d) Ikeja; and (e) Oshogbo for the period 1961-2011.
to December), 3.61 to 6.71 m/s (dry season), 3.84 to 7.04 m/s (wet season), and 3.92 to 6.88 m/s (whole years). For the yearly analysis, it ranged from 2.46 to 10.98 m/s. Also, the values of ranged from 3.44 to 9.63 m/s (for January to December), 4.45 to 8.03 m/s (dry season), 4.93 to 8.68 m/s (wet season), and 4.69 to 8.36 m/s (whole years). For the yearly data, it ranged from 2.26 to 11.85 m/s (1961 to 2011), respectively.
The most probable wind speed corresponds to the peak of the probability density function, while the wind speed carrying maximum energy can be used to estimate the wind turbine design or rated wind speed. Prior studies have shown that wind turbine system operates most efficiently at its rated wind speed. Therefore, it is required that the rated wind speed and the wind speed carrying maximum energy should be as close as possible  .
In this study, the assessment of the potential of wind power generation at five selected locations in the southwestern part of Nigeria was carried out. Fifty-one years of monthly mean wind data at 10-m height from the Nigeria meteorological department Oshodi, Nigeria were assessed and subjected to Weibull two-parameter and other analyses. It was discovered that:
1) The annual mean wind speeds for Abeokuta, Akure, Ibadan, Ikeja and Oshogbo are 3.67, 4.79, 5.11, 6.74 and 3.88 m/s, respectively. The annual values of the wind speed carrying maximum energy for these locations are respectively 6.48, 4.33 and 3.90 m/s.
2) The mean annual value of Weibull shape parameter k is between 5.36 and 8.34, while the annual value of scale parameter c is between 3.93 and 7.21 m/s. The two-parameter Weibull probability distribution was adequate in predicting the mean wind speed distribution in all the five sites.
3) The annual mean power densities for Abeokuta, Akure, Ibadan, Ikeja and Oshogbo are 65.09, 145.07, 176.96, 387.07 and 87.34 W/m2 respectively. Both the mean wind speed and power density are generally used to classify the wind energy resource (e.g., Pacific Northwest Laboratory (PNL) wind power classification scheme, Illica et al.  ). Therefore, based on the wind data used in this study, the wind energy resource in south-west Nigeria may generally be classified into class 1. However, based on monthly mean power density, the wind resource may fall into higher class category in some cases.
c (m/s): The Weibull scale parameter
: Cumulative distribution function
: Probability distribution function
k: Weibull shape parameter or factor
: Wind Power density (W/m2)
: Wind speed (m/s)
: Mean wind speed (m/s)
: Most probable wind speed
: Wind speed carrying Max Energy
: Standard deviation
: The gamma function of x
: Air density (kg/m3)