Wind Energy Potential of Jordan

M. A. Alghoul et al. / International Energy Journal 8 (7) 7-78 7 Wind Energy Potential of Jordan M. A. Alghoul *, M.Y.Sulaiman +, B.Z.Azmi + and M. Abd. Wahab + www.serd.ait.a.th/reri Abstrat - The daily mean wind speed data for 5 loations in Jordan over a period of 9 years are olleted and analyzed. Data are fitted to the Weibull distribution funtion. Weibull parameters are derived from the umulative funtion of the erved data reords (989-997), and used to alulate the mean wind speed and variane of the theoretial distribution. The goodness of representing the erved distribution with the Weibull distribution is determined using the Kolmogorov-Smirnov (K-S) test. At the % and 5% levels of onfidene the erved data are well represented by the Weibull distribution. The annual mean values of the wind speed of the erved and theoretial distributions are 6.ms - and 6.6ms - for Ras.Monief, 4.79ms - and 4.77ms - for Aqaba, 3.7ms - and 3.5ms - for Amman and 3.9ms - and 3.3ms - for Irbid and.34ms - and.4ms - for Der Alla respetively. Based on the annual wind speed, wind resoure for Ras.Monief, Aqaba, Amman, Irbid and Der Alla are varied from very good to poor. The annual mean power density of Ras.Monief, Aqaba, Amman, Irbid and Der Alla are 6.76 Wm -, 8.95 Wm -, 57.45 Wm -, 4.95 Wm -, and 4.97 Wm - respetively. Values of the power density obtained from the manufaturer s power distribution urve of a 3MW wind turbine at a hub height of meters are also given for omparison. The result of the analysis showed that only Ras.Monief and Aqaba have good wind energy potential. Keywords - Wind speed, Weibull distribution, Kolmogorov-Smirnov test, Power density, Wind energy potential.. INTRODUCTION Wind is an air motion aused by the rotation of the earth and the heating of the atmosphere by the sun. The total ineti energy of air movement in the atmosphere is estimated to be about 3x P 5P W h or about.% of the solar energy reahing the earth []. The maximum tehnially usable potential is estimated to be theoretially 3 trillion KWh per year or about 35% of the urrent world total energy onsumption, []. Sine the surfae of the earth is neither flat nor homogeneous, the amount of heat energy that is absorbed varies spatially as well as temporally. Consequently, this reates temperature, pressure and density (speifi mass) differenes, whih, in turn, reate fores that enable air to move from one plae to another. It is evident that, depending on the surfae feature (morphology) of the earth, some areas would be preferable to others for extrating ineti energy from the wind in the boundary layer of the atmosphere. Thus, wind energy an be used for many proesses, suh as powering, windmill, pumping water, and sailing boat. Also, wind energy is very lean but is not persistent for a long duration. Sine wind energy is renewable and environmentally benign, it has the advantage of being harnessed on loal basis for appliations in rural areas and remote areas. Water pumping for agriulture and plantations is probably the most important appliation * Solar Energy institute (SERI), Universiti Kebangsaan Malaysia (UKM), 436 Bangi, Malaysia. + Physis Department, Institute of Advaned Tehnology, University Putra Malaysia 434 Serdang, Selangor Malaysia. Corresponding author: E-mail: dr.alghoul@gmail.om; alghoul@eng.um.my that ontributes to the rural development through multiple ropping. Wind driven eletri generators ould be utilized as an independent power soure and for purposes of augmenting the eletriity supply from grids, deentralized prodution of eletriity would help loal industries. Wind annot be transported and therefore, wind turbines must be loated where wind resoures are present. The energy ontent of the wind being related to the ube of the wind speed varies signifiantly with only small hanges in wind speed. The annual wind speed at a loation is useful as an initial indiator of the value of the wind resoures. The relation between the annual mean wind speed and the potential value of the wind energy resoures is given in Table []. Table. Signifiant of wind energy aording to speed Annual mean wind speed at m Ht Below 4.5 (msp -P ) Index value of wind resoure Poor 4.5-5.4 (msp -P ) Marginal 5.4-6.7 (msp -P ) Good to very good Above 6.7 (msp -P ) Exeptional Theoretially, the relation between wind power density (Wm - ) and the wind speed is 3 P = ρv () where ρ is the air density whih is a funtion of the air pressure B and the air temperature T [3]. This is given as,

7 M. A. Alghoul et al. / International Energy Journal 8 (7) 7-78 88 B ρ = ρ o () 76T where ρ o is the density of dry air at standard temperature and pressure (.6gmP -3P at 88K, 76mm Hg). In this wor ρ is taen to be. gm -3 [4]. In loations where data are not available, a qualitative indiation of a high annual mean wind speed an be inferred from geographial loations, topographial features, wind indued soil erosion, and deformation of vegetation. The standard height aording to the World Meteorologial Organization (964) is m above the ground. This height is adopted in our analysis. In this paper an evaluation of wind energy in Jordan from limati daily data is made. The stations and duration of reords are desribed in Table. Table. Geographial loations of Jordan and duration of reords of data Station Latitude (N) Longitude (E) Elevation Duration of Reords Deg Min Deg Min Meters Amman 3 59 35 59 77. 989-997 Aqaba 9 33 35 5. 989-997 Der Alla 3 3 35 37-4 989-997 Irbid 3 33 35 5 66 989-997 RasMonief 3 35 45 5 989-997 Different distribution funtions have been suggested to represent wind speed data inluding the Pearson funtion by [5], Chi-Square funtion by [6], Weibull funtion by [7], and [8], Rayleigh funtion ( whih is a speial ase of Weibull distribution) by [9], and Johnson funtion by []. Among these, Weibull distribution is the most ommonly used in appliation beause it an represent well the wind data. However, we note that the Rayleigh distribution is a speial ase of the Weibull distribution. During the year 983, an inventory and proessing of available wind data was made and it was the first assessment of the wind energy potential in the ountry. Data olleted over a ten year period from the meteorologial stations were orreted first and then the estimated frequeny distribution and the theoretial power outputs were alulated aordingly []. The study onluded that most of Jordan regions, exluding the Jordan Valley have moderate wind theoretial power of 5 to 5 W/mP P, whih is suitable for water pumping, while there is one region in the northern part of the ountry having a wind theoretial power of 98 W/mP P whih is exellent for power prodution. All the values are alulated at a height of m. However the study also onluded that for an aurate assessment of the performane of wind systems, automati weather stations reording data on temporal and spatial basis would be needed. In 984, wind speed data loggers were installed for ontinuous reording at Ras.Monief, Shomery, Rwaished, and Dieseh. Referene [4] analyzed wind speed data from eleven stations. Monthly average and seasonal wind speed, and average power density distributions were determined for eah station. The monthly average wind speed for the two most potential sites Ras.Monief and Mafraq ranged from 3.ms - to 7.4 ms - and the average power density for these two sites ranged from to 37 Wm at Ras.Monief and from 5 to 47 Wm - at Mafraq.. WEIBULL DISTRIBUTION The alulation of the output of a wind mahine at a partiular site requires nowledge of the distribution of the wind speed. Most attention has been foused on the Weibull funtion, sine this fitted well the experimental data. Analysis showed that the result of the wind speed (v) ould be represented by umulative distribution funtion T(v) using the Weibull form as follows, [,3], v T () v = exp (3) where (sale fator), (shape fator), are the parameters hosen to fit the data. The probability density funtion f(v) is f(v) = v - exp Equation (3) is equivalent to, v { ln [ T(v) ]} ln v - ln ln =, (4) (5) whih is of the form Y = a X + b. By plotting different values of ln [- ln ( - T(v))] vs ln v, a straight line is fitted to the points. The slope of the line is and the interept on the ln [-ln ( - T(v))] axis is [- ln ]. Higher value of indiates that the wind speed for the partiular month is higher than the other month. In addition, the value of indiates wind stability. One useful he of the validity of the representation of wind speed distribution is obtained by omparing the mean speed alulated from the distribution with those alulated diretly from the data. The first and seond moments of probability density funtion f (v) in terms of, [4] are: v w = Γ + (6) σ = Γ + Γ + (7) where vb wb is the mean Weibull wind speed, σ w P P are the Weibull variane of the wind speed, and Γ is the gamma funtion. Equation shows the power as a funtion of the ube of the wind speed. At this point an important aspet must be examined. Using the annual or monthly mean wind speed value v, whether atual or derived from a Weibull fit, will not yield the right piture as far as the power density is onerned. The wind varies over time; hene wind speeds are distributed over the low and high wind speed ranges.

M. A. Alghoul et al. / International Energy Journal 8 (7) 7-78 This illustrates that the average of the ube of many different wind speeds will be muh greater than the ube of the average speed as shown in table 3 [5]. Hene one must introdue another parameter nown as the Energy Pattern Fator [E.P.F] or Cube Fator [6], whih adjusts the mean power density in eq. () by introduing a orreting fator. This fator is nown as the E.P.F and an be dedued from the following: Total amount of power available in the wind EPF = Power alulated by ubing the mean wind speed or EPF = Mean Mean power power density density at the A more realisti monthly mean power density is then given as, P = ρ ( EPF ) v 3 (8) In this wor, the Weibull monthly mean wind speed vbwb is used to alulate the monthly mean power density of eq. (8). 3. TEST OF GOODNESS OF FIT for the monthly month mean To determine the goodness of fit, it is neessary to introdue a formal statistial test that enable erved frequeny distribution to be ompared with the theoretial frequeny distribution. Kolmogorov Simrnov (K-S) test is based on the maximum differene between an empirial and a theoretial frequeny umulative distribution. Thus D = max O ( v) T ( v ) (9) where O(v) is the value of the empirial umulative frequeny distribution evaluated at v and T(v) is the orresponding theoretial umulative frequeny distribution funtion. If the value of v does not exeed the ritial value at a partiular signifiane level, one an aept the null hypothesis (that there is no differene between the erved and theoretial values). The ritial value of 73 D at 5% and % signifiane level an be estimated as follows [7],.36 Q =.5. N.63 Q = () () N where N is the sample size, (N > 3). The K-S test was applied aording to stations and months as shown in Tables 3-7. 4. RESULTS AND DISCUSSIONS Wind data from five stations in Jordan over a period of 9 years are olleted and analyzed. Information about these loations is given in Table. Weibull parameters are derived from graphial plot of eq. (5). The Weibull mean speed and variane are alulated using eqs. (6) and (7). The monthly mean power density of the wind is evaluated using eq. (8). Wind speed distribution an be fitted to a mathematial model whereby ertain harateristis of the wind regime an be determined. Suh a model is the Weibull distribution. In Tables 3-7, the erved monthly average wind speed, Weibull wind speed, the erved and simulated (Weibull) variane of wind speed, ube of the mean wind speed, mean of the ube of wind speed, EPF, the Weibull parameters (, and ) and K-S test are given for all stations. Also given are the annual mean values of all these quantities. The Weibull distribution model gives a good fit to the erved monthly wind speed data. The goodness of fit is tested using the K-S test. In Aqaba, Irbid, and Dier Alla, Weibull distribution passes the K-S test at 5% signifiant level as shown in tables 3-5. In Ras Monief, it passed the test in all months at 5% signifiant level exept in Marh, May, and September at % signifiant level as shown in Table 6. In Amman it passed the test in all months at 5% signifiant exept in August at % signifiant level as shown in Table 7. v msp - Table 3. The main harateristis parameters of Aqaba wind speed Aqaba vbwb msp - σb σb wpb v 3 avg EPF K-S P Jan 3.5 3.39 3.99 3.77 87.7 43.4.3 3.8.8.6 Feb 3.66 3.6 5.4 4.98.3 49..4 4.3.66.4 Mar 4.73 4.66 7.7 7. 57.55 5.7.44 5.4.8.3 Apr 5.38 5.49 6.9 5.3 59.44 55.78.67 6.8.56.6 May 5.75 5.68 5.9 5.9 8.69 89.83.48 6.39.66.4 Jun 6.6 6.6 5.85 7.7 355.8 45.8.44 7.5.5.3 Jul 5. 4.98 4.3 4.6 93.5 5.8.54 5.6.48.3 Aug 5.48 5.49 3.93 3.9 3.7 64.8.4 6.5 3.3.3 Sept 6. 6. 4.6 5.5 33.85 7.97.39 6.85.96.5 Ot 4.36 4.38 4.4 5.7 4.86 83..7 4.94.4.5 Nov 3.75 3.8 4.7 4.76.3 5.87. 4.7.8.3 De 3.56 3.44 4.5 3.89 9.35 44.93.3 3.87.8.6 Annual 4.79 4.77 5. 5.6.8 3.3.79 5.37.6.4 (v Bavg )P 3

74 M. A. Alghoul et al. / International Energy Journal 8 (7) 7-78 Table 4. The main harateristis parameters of Irbid wind speed B Irbid v vbwb msp - msp - σb σb wp v 3 avg (v Bavg )P 3 EPF K-S P Jan.8.83 3.6 3.64 57.9.5.63 3.3.5.3 Feb.98 3.3 4.6 3.97 7.88 6.57.7 3.37.55. Mar 3.6 3.7 3.59 3.6 68.99 8.66.4 3.56.7.4 Apr.9.94.75.63 5.3 4.5.9 3.3.88.3 May 3.9 3.8.47.4 54.76 9.59.85 3.59.5.5 Jun 3.78 3.75.9.6 75.6 54.7.4 4.8 3.6.4 Jul 4.48 4.46.94.4 5.73 89.64.9 4.97 3.36. Aug 3.89 3.9.76.69 97.76 59..66 4.37 3.3.3 Sept.99 3..96.85 45.6 6.6.7 3.4.36. Ot..8.69.64.6 8.5.5.33.67.5 Nov.55.59 3.4 3. 48.44 6.66.9.87.49. De.57.57 3.75 3.69 53.9 6.9 3.5.8.35.3 Annual 3.9 3.3.75.65 63.45 33.53.9 3.49.4.3 Table 5. The main harateristis parameters of D.Alla wind speed v D.Alla vbwb σb msp - msp - v 3 avg (v P σb Bavg )P 3 K-S wpb EPF Jan.78.84 7.34 7.4 7.6.46 4.99.9.5.3 Feb.3.37 3.7 3.5 46..47 3.7.55.7. Mar.33.47 3.6 3.76 47.59.59 3.78.66.8.3 Apr.66.87.4.9 37.93 8.8. 3.4.98.6 May.5.5.63.53 9. 5.83.83.83.3. Jun..4.9.6 5. 7.94.89.3.7. Jul.8.9.8.3.39 5.97.7.6.96. Aug.57.6.93.89 8.76 3.84.8.8.76. Sept.66.75..5.97 4.6.38.97.77.3 Ot.7..93.8 34.3.8 3.37.4.34. Nov 3.9 3.4 5.93 5.4 4. 35.63.9 3.56.4.6 De.96 3. 6.96 6.4 8.9 5.94 4. 3.8.9.3 Annual.34.4 3.9 3. 46.86 4.6.95.63.6.3 R.Monief v msp - Table 6. The main harateristis parameters of R. Monief wind speed vb wb σb v 3 avg σb wpb Jan 6.58 6.56. 3.8 553.4 84.84.94 7.39.89.4 Feb 6.8 6.76 3.89 4.9 66.9 36.7.98 7.6.87.5 Mar 6.77 7.8.69.77 599.3 3.37.93 8..3.98* Apr 5.93 6.7 9.39 8.3 394.45 8.5.89 6.96.7.6 May 5.55 5.85 6.54 6.9 96.49 7.7.73 6.6.37.87* Jun 6.7 6.49 5.46 5.3 34.54 34.66.46 7.5 3.8.7 Jul 6.8 6.95 4.33 4.56 47.65 37.55.8 7.7 3.6.6 Aug 6.9 6. 4.58 4.76 35.67 37.53.37 6.95 3..4 Sept 5. 5.49 5.4 4.65.3 34.34.65 6.7.75.95* Ot 4.69 4.83 5. 5. 8.49 3.4.75 5.46.9.5 Nov 6.4 6.46.39.4 479.67 64.76.8 7.9.. De 6.9 6.4 3.49.94 5.9 36.8. 6.89.76. Annual 6. 6.6 8.9 8.4 4.48 35.4.75 7.3.45.5 *Signifiant at % level. (v Bavg )P 3 EPF K-S

M. A. Alghoul et al. / International Energy Journal 8 (7) 7-78 75 Amman v msp - Table 7. The main harateristis parameters of Amman wind speed vb wb σb v 3 avg σb wpb Jan 3. 3. 5.4 5.65 9.4 6.94 3.39 3.4.7. Feb 3.58 3.59 7.4 7.4 5.93 45.99 3.33 3.9.35. Mar 3.43 3.4 5.54 5.3 7.63 4.49.66 3.79.5.4 Apr 3.9 3.4 4.88 4.8 86.8 9.49.9 3.46.45.3 May 3.7 3.36 3.6.99 68.4 3.78.5 3.79.3.6 Jun 3.6 3.63.78.57 79.39 47.38.68 4..4.4 Jul 3.88 4.4.9.73 84.74 58.56.45 4.5 3.39.7 Aug 3.9 3.35.9.48 49.7 9.49.69 3.78.5.88* Sept.39.5.8.43 33.4 3.7.4.8.65.7 Ot.99...3 3.58 7.9.98..34.3 Nov.89.96 7.56 6. 65.57 4.5 6.86 3.5..5 De.74.8 6.8 6.8 95.34.58 4.63.93.3.4 Annual 3.7 3.5 4.8 4.5 86.48 3.37 3. 3.47.75.4 *Signifiant at % level. (v Bavg )P 3 EPF K-S The erved and theoretial Weibull mean wind speed for R.Monief and Aqaba stations are shown in Figs. and. From these figures it is lear that Weibull model an fit the data for the months that pass the K-S test at 5% signifiant level better than the months that pass the K-S test at % signifiant level. The annual mean values of the wind speed of the erved and theoretial distributions are 6.msP -P and 6.6 msp -P for Ras.Monief, 4.79 msp -P and 4.77msP -P for Aqaba, 3.7 msp -P and 3.5 msp -P for Amman and 3.9 msp -P and 3.3 msp -P for Irbid and.34 msp -P and.4msp -P for Der Alla respetively. (m/s) 8 R.Monief 7 6 5 4 3 Jan Feb Mar Apr May Jun Jul Aug Sept Ot Nov De Observed mean wind speed Theoretial (Weibull) mean wind speed Fig.. Observed and theoretial (Weibull) monthly mean wind speed for R.Monief. (m/s) 7 6 5 4 3 Aqaba Jan Feb Mar Apr May Jun Jul Aug Sept Ot Nov De Observed mean wind speed Theoretial (Weibull) mean wind speed Fig.. Observed and theoretial (Weibull) monthly mean wind speed for Aqaba. Table 8 illustrates the values of the Weibull mean wind speed over the month, season, and year. From Table 8 and Fig. 3, it an be seen that R.Monief has the highest seasonal, and annual mean wind speed followed by Aqaba, Irbid, Amman, and finally Der.Alla. For Ras.Monief, wind speed is high during all the seasons and varies from 5.59 msp -P in Autumn to 6.55 msp -P in Summer. For Aqaba, wind speed is high during Spring (5.7 msp -P ) and Summer (5.58 msp -P ) and is low during the Winter (3.48 msp -P ) and Autumn (4.76 msp -P ). For Amman, Irbid and Der Alla, in all seasons wind speed is less than 4 msp -P exept in summer for Irbid whih is 4.4 msp -P. Aording to annual wind speed shown in Table, wind resoure for Ras. Monief, Aqaba, Amman, Irbid and Der Alla are varied from very good to poor. Table 8: Monthly, seasonally, and annual (Weibull) mean wind speed (ms - ) of all loations Month R.Monief Aqaba Amman Irbid D.Alla Jan 6.56 3.39 3..83.84 Feb 6.76 3.6 3.59 3.3.37 Mar 7.8 4.66 3.4 3.7.47 Apr 6.7 5.49 3.4.94.87 May 5.85 5.68 3.36 3.8.5 Jun 6.49 6.6 3.63 3.75.4 Jul 6.95 4.98 4.4 4.46.9 Aug 6. 5.49 3.35 3.9.6 Sep 5.49 6..5 3..75 Ot 4.83 4.38..8. Nov 6.46 3.8.96.59 3.4 De 6.4 3.44.8.57 3. Winter 6.49 3.48 3.3.8.74 Spring 6.4 5.7 3.3 3..6 Summer 6.55 5.58 3.67 4.4.85 Autumn 5.59 4.76.5.57.4 Annual 6.6 4.77 3.5 3.3.4

76 M. A. Alghoul et al. / International Energy Journal 8 (7) 7-78 (m/s) 7. 6. 5. 4. 3.... R.Monief Aqaba Amman Irbid D.Alla Fig. 3. Annual Weibull mean wind speed of loations of Jordan. 5.77 WmP -P to 8.8 WmP -P. Ras.Monief has seasonal mean wind power density varies from 95.3WmP -P in autumn to 339.54 WmP -P in Winter. Aqaba has seasonal mean wind power density variation from 54.6WmP -P in Winter to 6.3 WmP -P in Spring. For Amman, Irbid and Der Alla, seasonal mean wind power density varies from 8.4 WmP -P to 7.49 WmP -P. In Table 9, the power density inferred from an atual turbine power urve is also given. The wind turbine hosen is of the horizontal type with a rated power of 3W. The power urve of the wind turbine is given in Fig. 4 [8]. The wind speed is determined for a m hub height. The rotor diameter is 33m giving a swept area of 875 m. As shown in Table 9, the agreement between the pratial and theoretial values of the power density is satisfatory. The onept of E.P.F is useful in alulating the available energy in the wind along with the nowledge of the annual or monthly wind speed. It is also useful while hoosing a loation with limited wind data, beause long-term data from neighboring sites an be orrelated with one-site short-term measurements. The annual mean values of E.P.F for Ras.Monief, Aqaba, Amman, Irbid, and Dier.Alla are.75,.79, 3.,.9, and.95 respetively as given in tables 3-5. Table 9 illustrates the values of the theoretial (Weibull) mean power density available during the month, season, and year. The monthly mean wind power density varies from.98wmp -P in Otober to 435.57 WmP -P in Marh for Ras.Monief; and from 48.34WmP -P in January to 5.8 WmP -P in June for Aqaba. For Amman, Irbid and Der Alla, monthly mean wind power density varies from Power density wm - 4 3 3 5 7 9 3 5 7 9 3 5 Wind speed ms - Fig. 4. Manufaturer s power urve of a 3W wind turbine. Table 9: Monthly, seasonally, and annual mean power density WmP -P of all loations Month R. Monief Aqaba Amman Irbid D. Alla Weibull Atual Weibull Atual Weibull Atual Weibull Atual Weibull Atual Jan 334.78 39. 48.34 49. 55.9 3. 36.7 5. 7.6 4. Feb 373.3 355. 63.8 55. 93.5 5. 45.89 3. 9.96 4. Mar 435.57 347. 5.5 9. 64.78 46. 46.96 3. 34.69 4. Apr 7.8 34. 67.8 75. 54.99 33. 3.3 8. 9.6. May.57 9. 66.4 3. 49.6 36. 36.6 33. 7.63 8. Jun 4.87 63. 5.8 75. 49.3 53. 45.3 6. 9.78 9. Jul 63. 355. 6.48 4. 58.8 66. 69.9. 8.86 7. Aug.35 66. 4.7 84. 38.49 33. 6.77 66. 5.77 5. Sep 66.4 5. 94.7 44. 3.3 6. 8.84 3. 7.8 5. Ot.98 6. 88. 93. 5. 9. 3.84 9..48. Nov 98.6 96. 7. 59. 8.8 7. 3.85 9. 6.63 4. De 3.8 66. 5.35 5. 6.3 3. 3.7 9. 68.89 9. Winter 339.54 3. 54.6 5. 7.49 34. 38.5 4. 56.3. Spring 35.98 5. 6.3 66. 56.46 38. 38.5 3. 7.3 8. Summer 35.8 9. 58. 95. 48.6 5. 58.57 75. 8.4 7. Autumn 95.3 78. 7.77 7. 48.98 6. 4.5 8. 3.3 5. Annual 6.76 54. 8.95 3. 57.45 33. 4.95 33. 4.97 5.

M. A. Alghoul et al. / International Energy Journal 8 (7) 7-78 As seen in Figs. 5 and 6, it is lear that the two most potential sites in this study for wind power density are Ras.Monief and Aqaba with annual power densities of 6.76 and 8.95 WmP respetively, and annual wind speed of 6.6 and 4.77msP -P respetively. W/m 45 4 35 3 5 5 5 R.Monief Aqaba Amman Irbid D.Alla Jan Feb Mar Apr May Jun Jul Aug Sep Ot Nov De Fig. 5. The monthly Weibull mean power density WmP - P of loations of Jordan. W/m^ 35 3 5 5 5 R.Monief Aqaba Amman Irbid D.Alla Winter Spring Summer Autumn Annual Fig. 6. The annual and seasonal Weibull mean power density (WmP - P ) of loations of Jordan. The speed range over a windmill should be designed to operate the maximum wind and its struture that has to withstand depends on frequeny distribution of the wind speed. Attempts have been made to fit simple distribution to the erved frequeny distribution. Analysis is arried out for all sites. Table. The Annual Perentage Frequeny Distribution of wind speed Speed RMonief Amman Aqaba Irbid D.Alla.. 3.5 4.53.37.39. 4..84 7.48. 8.63 3. 8.87.39.87.4.5 4..9 5.64 4.94 8.53.75 5. 3.73.7 5.6 5.7 6.45 6. 3.8 7.36 5.4 8.94 4.3 7. 3. 4.47.65 3.68.83 8..67.95 8.6.5.3 9. 6.6.3 4.47.4.49. 5.7.49.98.5.5. 3.7.49.58.9.4..7.7.43.8 3..59.3.6.9 4..8.9.6. 5..64.3..3 6..4.6. 7..8.3 8..9 9..6..3..3 4-85.9 75. 43.7 49. 7.84 5-73.63 6.7 8.8 3.48 5.9 6-59.9 45. 6.37 4.78 8.64 77 Table lists the perentage frequeny of days in the wind speed range of,, 3 (m/s). The distributions are based on the measurements made at the sites over the period of study using daily data. The time for whih wind speed is in the range of 4 - ms - of the whole year are 85.9% for Ras.Monief, followed by Aqaba 75.%, Irbid 49.%, Amman 44.%, and Dier.Alla 8.%. 5. CONCLUSIONS The monthly mean wind speed data of five stations in Jordan are fitted to the Weibull distribution. For Aqaba, Irbid, and Dier.Alla, the Weibull distribution passed the K-S test at 5% signifiant level. Ras Monief Weibull distribution passed the test for all months at 5% signifiant level exept in Marh, May, and September at % signifiant level. Amman Weibull distribution passed the test for all months at 5% signifiant level exept in August at % signifiant level. Ras Monief and Aqaba have good wind energy potential; annual wind speed of 6.6msP -P and 4.77msP -P respetively, and annual mean wind power density of 6.76WmP - P and 8.95WmP -P respetively. Wind power density for all stations was inferred from an atual turbine power urve. The horizontal type turbine hosen has a rated power of 3W. The rotor diameter is 33m resulting in a swept area of 875m. The power density was obtained for wind speed at a hub height of m. The results are found to be satisfatorily when ompared to the theoretial values. ACKNOWLEDGEMENTS The wor is finanially supported by the Ministry of Siene, Tehnology and Environment, Malaysia under the IRPA (Intensifiation of Researh in Priority Areas). We hereby wish to anowledge the finanial assistane of the government of Malaysia. REFERENCES [] Ramahandra, T.V.; Subramanian, D.K. and Joshi, N.V. 997. Wind Energy Potential Assessment in Uttara Kanada, Distrit of Karnataa, India, Renewable Energy (4): 585-6. [] Wilbur L. C. 985. Handboo of energy systems engineering prodution and utilization. John Wiley & Sons. [3] John F. W. and Niholas J. 997. Wind Energy Tehnology, New Yor: JohnWiley and Sons In. [4] Habali, S.A.; Hamdan, M.A.S.; Jubran, B.A. and Zaid, Adnan.I.O. 987. Wind Speed and Wind Energy Potential of Jordan, Solar Energy 38(): 59-7. [5] Sherlo,R.H. 95. Analyzing winds for frequeny and duration. Meteor.Monogr, Am.Meteor.So (4): 7-79. [6] Corotis, R.; Sigl, A. and Klein, J. 978. Probability models of wind veloity magnitude and persistene, Solar Energy : 483-493. [7] Justus, C.G.; Hargraves, W.R.; Mihail, A. and Graver, D. 978. Methods for estimation wind speed frequeny distributions, Am.Meteor.So 7: 35-353.

78 M. A. Alghoul et al. / International Energy Journal 8 (7) 7-78 [8] Hennessey, J.P. 978. A omparison of the Weibull and Raleigh distribution for estimating wind power potential, Wind Energy : 56-64. [9] Corotis, R. 979. Simulation of orrelated wind speeds for sites and arrays, Sun, Pro.Int. Solar Energy So, Silver Jubilee Cong 3: 57-6. [] Mage, D. 98. Frequeny distribution of hourly wind speed measurements, Atmos.Envir 4: 367-374. [] The Potential of Solar Energy Appliation in Jordan. 983. Assessment and Analysis of Available Energy Resoures, Royal Sientifi Soiety, Vol.3, Amman- Jordan. [] Stevens, M. and Smulders, P. 979. The estimation of the parameters of the Weibull wind speed distribution for wind energy utilization purposes, Wind Engng 3: 3-45. [3] Hennessy, J.P. 977. Some Aspets of Wind Power Statistis,Appl.Meteorol, 6. [4] Serri, E. and Farrugia, R. 996. Wind data evaluation in the Maltese Islands, Renewable Energy 7(): 9-4. [5] Gipe, P. 993. Wind Power for Home and business. Chelsea, Green Publishing Co,USA. [6] Farrugia, R. and Serri, E. 997. Analysis of Wind Charateristis in the Maltese Arhipelago, Renewable Energy (4): 339-35. [7] Massey, F.J. 95. The Kolmogorov-Smirnov test for goodness of fit. J.AM. Stat.Ass. 46: 68-78. [8] Retsreen International, http://www.retsreen.net/