See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/245525765

Survey and Assessment of Wind-Speed and Windpower in Egypt, Including

Air Density Variation

Article in Wind Engineering · March 2006

DOI: 10.1260/030952406778055081

CITATIONS READS

20 4,755

2 authors:

Khaled S. M. Essa Fawzia Mubarak

Egyptian Atomic Energy Authority Egyptian Atomic Energy Authority, Nuclear and radiological safety research center
75 PUBLICATIONS 396 CITATIONS 23 PUBLICATIONS 159 CITATIONS

SEE PROFILE SEE PROFILE

All content following this page was uploaded by Fawzia Mubarak on 15 February 2018.

The user has requested enhancement of the downloaded file.

Wind 30-2-Essa 18/7/06 1:16 pm Page 1

Survey and Assessment of Wind-speed and Wind-power

in Egypt, including Air Density Variation
by

Khaled S.M. Essa and Fawzia Mubarak

R EPRINTED FROM

WIND ENGINEERING
VOLUME 30, N O . 2, 2006

M ULTI -S CIENCE P UBLISHING C OMPANY

5 WATES WAY • B RENTWOOD • E SSEX CM15 9TB • UK
T EL : +44(0)1277 224632 • FAX : +44(0)1277 223453
E-MAIL: [email protected] • WEB SITE: www.multi-science.co.uk
01_S384-Essa.qxd 18/7/06 12:33 pm Page 95

W IND E NGINEERING VOLUME 30, N O . 2, 2006 PP 95–106 95

Survey and Assessment of Wind-speed and Wind-

power in Egypt, including Air Density Variation
Khaled S.M. Essa*, and Fawzia Mubarak**
*Mathematics and Theoretical Physics Department,
**Radiation Protection Dept., Nuclear Research Center,
Atomic Energy Authority, Cairo, Egypt
*E-mail<[email protected]>

ABSTRACT
Nearly 5 years of meteorological data were analysed from 18 stations throughout Egypt,
classified as located in Mediterranean, Inland, and Red Sea zones. The national annual
average wind speed was 5.8 m/s at 10 m height. The Hurguda station (Red Sea coast) has 5.8
m/s mean annual wind speed and the largest peak wind speed there was 13.8 m/s, with 98%
of wind-speed records being in the range of 3 to 10 m/s. Dekhala station (Inland) has the least
annual average wind speed of 2.3 m/s , with 98.5% of wind speed records in the range of 1 to
5 m/s. For Mediterranean stations, Port Said has the largest mean wind speed of 4.9 m/s. A
general wind energy potential considering both wind speed and air density was derived.
Comparison are made between Weibull, Rayleigh, and actual data distributions of wind
speed and wind power of two years (2003 and 2004). A Weibull distribution is the best match
to the actual probability distribution of wind speed data for most stations. The maximum
wind energy potential was 373 W/m2 in June at El-Tor (Red Sea coast) where the annual
mean value was 207 W/m2.

Keywords: Wind energy potential / air density

1. INTRODUCTION
The wind climate in Egypt may be utilized favourably for electricity production, Hurguda in
the Red Sea coast, was selected for the National Wind Energy Technology Center (NWETC)
due to its good infrastructure as the capital of the Red Sea coast area of Egypt and its
reasonable good wind conditions [1], with mean wind speed of 5.8 m/s and energy potential of
1400 W/m2 at height of 25 m.
This paper analyses meteorological data from 18 meteorological stations as shown in Fig.1,
the data are 15-minute averages at 10 m height above ground level, as standard in Egypt.
Study sites were classified as Mediterranean Sea, Inland, or Red Sea zones. Areas for wind
power generation have been identified.

2. CALCULATION OF WIND ENERGY POTENTIAL

The power density, P, of the wind is:

1
P = ρ u3 (1)
2
01_S384-Essa.qxd 18/7/06 12:33 pm Page 96

96 S URVEY AND A SSESSMENT OF W IND - SPEED AND W IND - POWER IN E GYPT

Where ρ is the air density, u is the wind speed for a given period of time and the unit of P is W/m2
with u is in m/s [2]. In Eq. (1), consideration should be made for air density variation [3–4].
At high-altitude stations, an assumption of sea-level air density causes available wind energy
to be overestimated by nearly 30% [Reed, 5]. Reed proposed an air density correction factor for
converting the sea-level wind energy estimates to the site altitude. This density correction
factor is dependent on the site elevation and the annual cycle of monthly mean temperatures.

3. AIR DENSITY CORRECTION FACTOR

By theory of dependent random variables [5], if air density and wind speed are dependent on
each other, then the expectation, E, of both sides in Eq. (1) leads to:

1
E (P) = E (ρ u3 ) (2)
2

Essa et al. [3] show that the cross correlation coefficient, r, between the wind speed cube and
the air density is defined as:

Cov(ρ , u3 )
r= (3)
Sρ Su3

Where Sρ and Su3 are the standard deviations of air density and wind speed cube in the time
series. Then:

1
E (P) =  E (ρ)E (u3 ) + rSρ Su3 
2
(4)

This expression can be simplified for experimental applications as follows:

1. The second term makes the major difference, according to the literature. Obviously,
this term vanishes with constant air density because r = 0. and Eq. (4) is reduced to
Eq. (1). So that E (ρ) = ρ:

Figure 1: Locations of stations in Egypt.

01_S384-Essa.qxd 18/7/06 12:33 pm Page 97

W IND E NGINEERING VOLUME 30, N O . 2, 2006 97

1
E ( P ) = ρE (u3 ) (5)
2

2. There is no cross–correlation for instantaneous air density and wind speed

measurements. so Eq. (1) becomes valid. In the general random formulation of Eq. (4),
“r” plays the most important role depending on its actual value between –1 and +1. Since
moist air is lighter than dry air, increased moisture is linked to increased wind speeds
especially in climates as in Egypt. With such an effect, air density and wind speed are
inversely proportional. Hence, the cross correlation would be negative between wind
speed and air density; therefore Eq. (1) would give greater values than Eq. (4) so:

1  Sρ Su3 
E (P) = E (ρ)E (u3 ) 1 + r  (6)
2  E (ρ)E (u3 ) 

The term within the brackets is defined as the correction factor, α:

Sρ Su3
α = 1+ r (7)
E (ρ)E (u3 )

The coefficient of variation, C, is defined as the ratio of the standard deviation to the
arithmetic average. Two of such ratios appear in Eq. (7) namely the coefficient of variation of
air density C
and of wind speed cube Cu3. Hence Eq. (7) becomes:

α = 1 + rCρCu3 (8)

Eq.(8) indicates that with small variation of the coefficient (in particular smaller than one),
the second term on the right hand side becomes negligible. However, for relatively large
coefficients of variation, the second term may be significant which means that the traditional
formulation overestimates the wind energy potential. Evidently when r = 0 the relative error
becomes zero. This paper provides a basis for estimating the size of the correction factor for
air density and wind speed time series records. To evaluate E(
), we use the concept of
expectation value in the universal gas law [6], so:

E(ρ) (9)
E (ρ) =
R.E (T )

where
is the air pressure (Pascal), T is the air temperature (Kelvin) and R is the universal gas
constant.
The random variability in the actual daily temperature, pressure, air density and wind
speed measurement time series for 18 Egyptian stations is presented for about 5 years from
9/3/2000 to 31/12/2004. Each measurement is ‘ the run of the wind’ during 15 minutes. From
Table 1 we find that the two stations with the largest and smallest average wind speeds,
Hurguda and Dekhala respectively. Fig. (2) shows that the daily variation of air density time
series exhibit distinct seasonal; however the amplitude of wind speed fluctuations is
effectively stationary without explicitly observable periodicities or trends. This feature of the
wind speed supports the use of Eq. (2) as an approximation for wind energy calculation.
Table 1 shows the statistical features of wind speed at 10 m height for the 18 Egyptian
meteorological stations during 5 years (from 9/3/00 to 31/12/04). It is clear that Red Sea
zone generally had the larger wind speeds; in this zone. Hurguda had the same average wind
speed as El-Tor station (5.8 m/s), but Hurguda had larger maximum value of wind-speed
01_S384-Essa.qxd 18/7/06 12:33 pm Page 98

98 S URVEY AND A SSESSMENT OF W IND - SPEED AND W IND - POWER IN E GYPT

1.25 1.25
density (kg/m3)

density (kg/m3)
1.2 1.20

1.15 1.15

1.1 1.10
69 268 468 668 868 1066 1266 1466 1667 69 268 468 668 868 1066 1266 1466 1667 1867
Julian day Julian day

Figure 2: Daily air density time-series for Hurguda (left) and Dekhala (right) stations.

Table 1 Statistical data of 18 Meteorological stations in Egypt for wind speed (m/s) during
5 years, ( average = average wind speed, Max = maximum, and Min = minimum)
Zone Station Average Max. Min. Range Frequency%
/(m/s) /(m/s) /(m/s) /(m/s)
Medite- Alexandria 4.4 11.3 1 2–8 98.2
rranean Arish 2.3 10.0 1 1–5 95.5
Baltim 3.7 11.8 1.5 2–7 96.1
Port Said 4.9 12.3 1 3–8 96.3
Red Sea El-Kossier 4.4 10.3 1 3–8 96.7
El-Tor 5.8 12.0 1 3–9 94.8
Hurguda 5.8 13.8 1.25 3–10 98.1
Inland Aswan 4.2 9.8 1 3–7 96.6
Baharia 2.7 6.8 1 2–5 98.1
Cairo 3.7 11.5 1 2–6 95
Dekhala 2.3 7.8 1 1–5 98.5
El-Kharga 3.1 8.8 1 1–6 96.8
Farafra 3.2 7.8 1 2–6 97.6
Ismailia 3.7 12.7 1 2–7 96.4
Luxor 2.8 9.8 1 2–6 95.9
Minya 3.4 8.0 1 1–6 95.6
Siwa 2.4 8.0 1 1–5 96.3
Wadi El-Natron 3.7 9.8 1 2–6 96.5

(13.8 m/s during 15 minutes), and larger range (3–10 m/s) for wind-speed frequency 98%.
Therefore Hurguda was selected as the site of the National Wind Energy Technology Center
(NWETC) due to its favourable wind conditions. Inland zones (expect Aswan and Ismailia)
had the least favourable wind speeds, Dekhala had the least average wind speed (2.3 m/s)
with the range of 1–5 m/s for 98.5% frequency distribution. Amongst Mediterranean stations,
Port Said had largest mean wind speed (4.9 m/s).
Table 2 presents the statistical properties of air density and wind speed (at 10 m) cube for
18 meteorological stations in Egypt for 5 years.
The variations of the wind speed are large compared with air density variations. As a first
impression, since the variations in the air density are very small, one might ignore these
variations and assume a constant air density equal to the arithmetic mean which is 1.18 kg/m3
for Hurguda station and 1.19 kg/m3 for Dekhala station as in Table 1. However, although the
variations in the air density are small, their impact on the wind energy calculations might be
significant. In order to confirm this point, Eq. (8) was applied to the data given in Table (2) to
give the statistical data in Table (3). Table 3 shows the cross-correlation, r, between the air
density and cube of the wind velocity time series, coefficient of variation of air density (Cρ)
and velocity cube (Cu3) series, correction factor α , and power density E(P) for the 18 stations.
01_S384-Essa.qxd 18/7/06 12:33 pm Page 99

W IND E NGINEERING VOLUME 30, N O . 2, 2006 99

Table 2 Statistical data of 18 Meteorological stations in Egypt, (ρ = air density, u3 = wind

speed cube, st. dev. = standard deviation, Max = maximum, and Min = minimum)
Zone Station Parameter average Median Mode St. dev. Max. Min.
and units
Medite- Alexandria ρ (kg/m3) 1.20 1.20 1.25 0.03 1.26 1.2
rranean u3 (m/s)3 117 76 76 120 926 1.0
Arish ρ (kg/m3) 1.20 1.19 1.22 0.03 1.26 1.1
u3 (m/s)3 26 8 3 67 729 1.0
Baltim ρ (kg/m3) 1.20 1.20 1.17 0.02 1.26 1.2
u3 (m/s)3 81 34 15 136 1622 3.4
Port Said ρ (kg/m3) 1.20 1.20 1.23 0.02 1.26 1.2
u3 (m/s)3 149 107 125 153 1838 1.0
Red Sea El-Kossier ρ (kg/m3) 1.18 1.18 1.20 0.03 1.25 1.1
u3 (m/s)3 114 767 43 113 1076 1.0
El-Tor ρ (kg/m3) 1.19 1.18 1.17 0.03 1.25 1.1
u3 (m/s)3 251 186 107 237 1728 1.0
Hurgu-Da ρ (kg/m3) 1.18 1.18 1.14 0.03 1.23 1.1
u3 (m/s)3 249 190 166 214 1622 1.9
Inland Aswan ρ (kg/m3) 1.17 1.17 1.17 0.03 1.27 1.1
u3 (m/s)3 91 64 43 79 927 1.0
Baharia ρ (kg/m3) 1.19 1.19 1.24 0.03 1.27 1.1
u3 (m/s)3 28 15 8 33 307 1.0
Cairo ρ (kg/m3) 1.20 1.19 1.24 0.03 1.26 0.8
u3 (m/s)3 71 43 52 94 1520 1.0
Dekha-La ρ (kg/m3) 1.19 1.18 1.23 0.03 1.27 1.1
u3 (m/s)3 21 8 8 33 465 1.0
El-Kharga ρ (kg/m3) 1.18 1.17 1.14 0.03 1.27 1.1
u3 (m/s)3 50 21 15 69 669 1.0
Farafra ρ (kg/m3) 1.19 1.19 1.22 0.03 1.27 1.1
u3 (m/s)3 44 27 27 45 465 1.0
Ismailia ρ (kg/m3) 1.19 1.18 1.19 0.03 1.27 1.1
u3 (m/s)3 79 52 64 120 2032 1.0
Luxor ρ (kg/m3) 1.18 1.17 1.18 0.03 1.27 1.1
u3 (m/s)3 39 15 8 67 926 1.0
Minya ρ (kg/m3) 1.20 1.19 1.23 0.03 1.27 1.1
u3 (m/s)3 63 34 27 78 512.00 1.00
Siwa ρ (kg/m3) 1.19 1.19 1.15 0.03 1.27 1.1
u3 (m/s)3 27 8 3 52 512 1.0
Wadi ρ (kg/m3) 1.20 1.19 1.23 0.03 1.27 1.1
El-Natron u3 (m/s)3 66 43 27 78 941 1.0

The cross-correlation, r, between the air density and cube of the wind speed time series for
Hurguda station is equal –0.153 and –0.081 for Dekhala.
The coefficient of variation of air density and wind speed cube series (Table 3) are
Cρ = 0.03/1.18 = 0.0254 and Cu3 = 214.14/249.43 = 0.859, for Hurgada station and Cρ = 0.03/
1.19 = 0.029 and Cu3 = 32.96/21.1 = 1.56, for Dekhala station. Hence Eq. (8) gives, E (P) = 147
W/m2 = 528.23 kWh/m2 for Hurguda station and α = 0.996, E (P) = 12.48W/m2 = 44.9 kWh/m2
and α = 0.997 for Dekhala station.
It is clear that the average wind energy production is not only a function of the coefficient
of variation of the air density and wind speed cube, but also on the cross-correlation
coefficient r. If air density and wind speed variations are independent, wind energy is
calculated with average air density measurements instead of a standard constant air density.
01_S384-Essa.qxd 18/7/06 12:33 pm Page 100

100 S URVEY AND A SSESSMENT OF W IND - SPEED AND W IND - POWER IN E GYPT

Table 3 Statistics data of 18 Meteorological stations in Egypt, (coefficient of variation of air

density (C
) and wind-speed cube (C 3u ) series, a correction factor α, and power
density E(P))
Zone Station r Cρ C u3 α E(P) E(P)
W/m2 kWh/m2
Medite- Alexandria –0.101 0.021 1.021 0.998 70.7 254.52
rranean Arish 0.192 0.022 2.523 1.011 16.14 58.117
Baltim 0.121 0.021 1.669 1.004 49.19 177.1
Port Said 0.144 0.020 1.025 1.003 89.72 323.02
Red Sea El-Kossier –0.015 0.022 0.991 1.000 67.39 242.62
El-Tor –0.303 0.022 0.944 0.994 148.29 533.83
Hurguda –0.153 0.025 0.859 0.997 146.78 528.4
Inland Aswan –0.181 0.029 0.874 0.995 53.33 191.98
Baharia –0.051 0.027 1.168 0.998 17.13 61.67
Cairo 0.009 0.025 1.319 1.000 42.77 153.98
Dekhala –0.081 0.029 1.562 0.996 12.48 44.92
El-Kharga –0.127 0.029 1.375 0.995 29.42 105.91
Farafra –0.091 0.029 1.025 0.997 26.16 94.18
Ismailia 0.03 0.03 1.51 1.00 47.25 170.09
Luxor –0.114 0.029 1.691 0.994 23.25 83.69
Minya 0.008 0.027 1.229 1.000 37.99 136.76
Siwa 0.052 0.027 1.955 1.003 16.14 58.11
Wadi El-Natron –0.070 0.024 1.175 0.998 39.58 142.48

4. Frequency Distribution functions of wind speed

Knowledge of the frequency distribution of wind speeds at a specific site is essential for
estimating wind turbine energy production. Mathematically, there are many statistical
models for representing the frequency distribution of wind speed in a compact form. The most
common one is the Weibull Distribution, then the Rayleigh distribution which is a form of
Weibull distribution [7]. Justus et al., [8] claim that the Weibull distribution is the best. Cortis et al.,
[9] preferred the Rayleigh distribution. Hennessey [10] compared both distribution and found
that they agree to within 10%. However, the particular function depends on the particular wind
and weather characteristics of the particular site. The shape of a typical wind speed
frequency distribution at a site calculated by Weibull distribution is given as:

k u
k −1   u k 
f (u) =   exp  −    (10)
c c   c 

where f(u) is the frequency of occurrence of wind speed u. The two Weibull parameters thus
defined are usually referred to as the scale parameter c and the shape parameter k. For k > 1
the maximum (modal value) lies at values u > 0, while the function decreases monotonically
for 0 < k  1. The Weibull distribution can degenerate into two special distributions, namely for
k = 1 the exponential distribution and for k = 2 the Rayleigh distribution. The important
characteristics of Weibull distribution can be derive as [1]:
mean value of wind speed: cΓ(1+1/κ)
mean mth power: cm Γ(1+m/κ)
variance (σ2): c2[Γ (1+2/κ) − (Γ2(1+1/κ)]
modal value: c [(κ−1)/κ]1/κ
median: c [In 2]1/κ
01_S384-Essa.qxd 18/7/06 12:33 pm Page 101

W IND E NGINEERING VOLUME 30, N O . 2, 2006 101

Where Γ is the gamma function. The values of k and c can be evaluated using the following
formulae [11]:

k = (σ / um )
− 1 . 086
(11)

c = um / Γ (1 + 1 / k ) (12)

The mean value of u3 becomes (due to the mean mth power):

u−3 = c 3 Γ (1 + 3 / k ) (13)

The available wind power density is proportional to the mean cube of the wind speed, so if E is
the mean wind energy potential of airflow through a unit of surface area perpendicular to the
air stream during unit time (W/m2), and ρ is the air density [2–3]:

1
E = ρ c3Γ(1 + 3 / k ) (14)
2
The cumulative Weibull distribution F (u) gives the probability of the wind speed
exceeding the value u and is given by the simple expression:

  u k 
F (u) = 1 − exp  −    (15)
  c 

The annual cumulative distribution Weibull function is:

 (
T(u) = h 1 − exp − ( u/c ) 
k
 ) (16)

where h is the number hours of the year, i.e. h = 8760 hours

For many sites, it is adequate to reduce equation (10) to the one parameter Rayleigh
distribution by setting k = 2, [12]. Hence equation (10), (12), and (13) become:

2u   u 2 
f (u) = exp  −    (17)
c2  c 

C = 1 . 13 um (18)

(u 3)1 / 3 = 1 . 24um (19)

The hourly wind speed data for the two complete years, 2003 and 2004 for 18 stations in Egypt
were used to evaluate the frequencies of wind speeds as well as the annual, monthly, daily and
mean wind speed and potential power. For Hurguda, the monthly average wind speed of 15
minute data (2004) show that July has the maximum value, while for Dekhala, the maximum
value occur on September. Figure (3) represents examples for the annual wind-speed
frequency (2004) for Hurguda, as the station with highest-speed data, and Dekhala, as the
station with the lowest-speed data, respectively.
Figure (4) shows comparisons between average monthly wind energy potential for
Hurguda and Dekhala stations (2004) respectively calculated by using actual data, Weibull
and Rayleigh distribution parameters. It is clear that, Weibull distribution is the best match
with the actual data.
01_S384-Essa.qxd 18/7/06 12:33 pm Page 102

102 S URVEY AND A SSESSMENT OF W IND - SPEED AND W IND - POWER IN E GYPT

20 70
60
Frequency, %

Frequency, %
15 50
40
10
30
5 20
10
0 0
1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12
Wind speed, m/s Wind speed, m/s

Figure 3: Wind-speed frequency for Hurguda (left) and Dekhala (right) Stations (2004).

500
Actual data
Wind energy potential (w/m2)

450
Weibull
400
Raleigh
350
300
250
200
150
100
50
0
n

ar

ay

ct

ov

ec
Ju
Ap
Ja

Fe

Ju

Au

Se

O
M

D
Month
14
Actual data
Wind energy potential (w/m2)

12 Weibull
Raleigh
10
8

6
4
2
0
n

ar

ay

ct

ov

ec
Ju
Ap
Ja

Fe

Ju

Au

Se

O
M

Month

Figure 4: Average monthly wind energy potential for Hurguda (upper Figure) and Dekhala (lower
Figure) Stations (2004).

A statistical analysis for five years data gave the annual average wind speed as 5.8 m/s, in
the range of 1.25 to 13.8 m/s (with 98% of wind speed records being in the range of 3 to 10 m/s)
for Hurguda station (Red Sea coast) which has the highest-speed. The annual average wind
speed for Dekhala station (which has the lowest-speed) was 2.3 m/s, with 98.5% of records in
the range of 1 to 5 m/s. Comparisons of Weibull, Rayleigh, and the actual data distribution
were carried out for all stations (18 stations) in Egypt for two years, 2003 and 2004. Figure (5)
represent two examples; for Hurguda and Dekhala stations respectively. It is clear from the
figure that the Weibull distribution is the best match with the actual data, where the fitting
equation for Weibull parameter is: –0.008x2 + 0.1059x – 0.152 with correlation factor 0.3586, the
best fitting equation for Rayleigh parameter is: 0.0081x2 + 0.109x – 0.1856 with correlation
factor 0.3798 (for Dekhala, 2004). While the fitting equation for Weibull parameter is:
–0.0054x2 + 0.0646x – 0.04551 with correlation factor 0.3393; and the best fitting equation for
01_S384-Essa.qxd 18/7/06 12:33 pm Page 103

W IND E NGINEERING VOLUME 30, N O . 2, 2006 103

0.6 0.6
Actual data Actual data
Probability function

0.5 0.5

Probability function
0.4 0.4
0.3 0.3
0.2 Weibull 0.2 Rayleigh
Rayleigh Weibull
0.1 0.1
0 0
1 2 3 4 5 6 7 8 9 10 11 12 2 3 4 5 6 7 8 9 10
Hourly wind speed m/s Hourly wind speed m/s

Figure 5: Comparison of Wind speed frequency for Hurguda( left) and Dekhala (right) Stations
(2004), Weibull, Raleigh and actual wind speed potential.

Table 4 Numerical values of Weibull parameters for annual wind speed

distribution at the selected stations in Egypt. (u = annual average wind speed
m/s, c = scale parameter, k = shape parameter)
Zone Station Year u c k
Mediterranean Alexandria 2003 4.69 5.25 3.04
2004 4.32 4.82 3.35
Arish 2003 2.56 2.88 1.8
2004 2.27 2.56 1.96
Baltim 2003 3.84 4.33 2.41
2004 3.51 3.97 2.42
Port Said 2003 4.7 5.25 3.26
2004 5.2 5.79 3.92
Red Sea El-Kossier 2003 4.4 4.87 3.60
2004 4.4 4.88 2.32
El-Tor 2003 5.5 6.02 3.98
2004 6.7 7.52 3.25
Hurguda 2003 5.56 6.17 3.60
2004 5.97 6.11 3.17
Inland Aswan 2003 4.12 4.55 4.02
2004 3.85 4.26 3.85
Baharia 2003 2.75 3.07 3.07
2004 2.61 2.92 3.07
Cairo 2003 3.78 4.23 3.04
2004 3.68 4.13 2.96
Dekhala 2003 1.99 2.24 2.67
2004 1.91 2.15 2.97
El-Kharga 2003 3.03 3.43 2.14
2004 3.22 3.64 2.32
Farafra 2003 3.43 3.81 3.49
2004 3.35 3.72 3.45
Ismailia 2003 3.9 4.44 2.62
2004 3.91 4.38 2.93
Luxor 2003 3.04 3.43 2.18
2004 2.44 2.76 2.23
Minya 2003 3.2 3.66 2.19
2004 3.8 4.31 3.06
Siwa 2003 2.9 3.3 2.28
2004 2.5 2.87 2.29
Wadi El-Natron 2003 3.7 4.1 3.31
2004 3.5 3.82 3.91
01_S384-Essa.qxd 18/7/06 12:33 pm Page 104

104 S URVEY AND A SSESSMENT OF W IND - SPEED AND W IND - POWER IN E GYPT

Rayleigh parameter is: Raleigh –0.0049x2 + 0.068x – 0.0947 with correlation factor 0.3576 (for
Hurguda, 2004)
Table 4 shows the numerical values of Weibull parameters for annual wind speed
distribution at the selected stations in Egypt for two complete years, 2003 and 2004: the red sea
stations had the largest-mean wind speed and Weibull parameters. While the inland stations
had the smallest-value parameters. Hurguda and El-Tor have the largest-mean wind speed
and Weibull parameters. Dekhala has the smallest-mean wind speed and Weibull parameters.
Mediterranean stations (expect El-Arish) had suitable mean wind speed and Weibull
parameters for wind power. The previous statistical analysis indicate that the Red Sea zones,
Mediterranean (expect El-Arish) and some inland zones (Aswan and Ismailia) can be chosen
as favourable locations for wind power generation. Cairo station has modest wind energy
potential and, in practice, there are less suitable sites for turbines.
Tables (5–8) show also that the Red sea coast stations have the monthly and annual
highest-mean wind energy potential for the two years, 2003 and 2004 while the inland stations
had the lowest-potential.

Table 5 Monthly and annual mean wind energy potential (W/m2) at the selected stations at
the Mediterranean. (J-D = from January to December)
Station Alexandria Arish Baltim Port Said
Month 2003 2004 2003 2004 2003 2004 2003 2004
J 28.11 59.20 8.70 21.06 25.43 63.98 43.23 92.54
F 96.94 39.96 41.84 8.79 71.44 35.64 115.41 76.62
M 59.31 54.38 31.50 8.33 44.50 56.16 111.84 90.19
A 97.13 53.17 15.53 19.81 63.63 46.88 109.66 120.12
M 61.29 53.22 6.19 11.86 39.97 30.04 92.54 101.27
J 62.85 57.69 5.65 6.39 27.99 28.42 62.29 82.18
J 136.87 61.08 3.08 3.49 29.58 22.34 46.41 88.50
A 79.44 62.36 2.48 2.75 18.38 23.30 25.73 83.04
S 69.16 53.67 5.30 3.32 18.61 14.13 34.02 82.43
O 38.17 36.44 5.47 2.27 25.55 10.07 49.55 74.52
N 28.38 37.33 5.30 6.98 40.27 15.35 53.37 99.61
D 44.76 25.94 32.62 5.02 30.43 7.82 64.96 56.28
Annual 66.87 49.54 13.64 8.34 36.32 29.51 67.42 87.28

Table 6 Monthly and annual mean wind energy potential (W/m2) at the selected
stations at the Red Sea. (J-D = from January to December)
Station El-Kossier El-Tor Hurguda
Month 2003 2004 2003 2004 2003 2004
J 57.16 30.87 56.26 66.29 71.90 86.42
F 38.66 67.55 77.44 80.77 81.19 87.56
M 53.12 81.42 105.34 111.84 97.50 131.70
A 72.59 44.98 116.06 114.07 100.58 80.14
M 40.77 57.06 87.62 257.27 81.08 224.10
J 58.40 69.44 142.80 372.69 144.54 170.97
J 41.83 43.74 120.25 357.86 123.60 227.12
A 46.76 45.28 160.93 368.18 128.04 193.03
S 74.25 55.90 154.84 346.60 164.78 174.96
O 40.05 38.12 78.31 186.66 98.08 45.86
N 56.52 30.23 48.54 121.83 100.58 147.66
D 38.56 70.31 71.78 98.04 73.56 74.28
Annual 51.56 52.91 101.68 206.84 105.45 136.98
01_S384-Essa.qxd 18/7/06 12:33 pm Page 105

W IND E NGINEERING VOLUME 30, N O . 2, 2006 105

Table 7 Monthly and annual mean wind energy potential (W/m2) at the selected stations in
inland in 2003. (J-D = from January to December)
Station Aswan Baharia Cairo Dekhala El- Farafra Ismailia Luxor Minya Siwa Wadi El-
Month Kharga Natron
J 26.99 11.59 22.44 3.58 9.14 16.46 18.86 10.02 13.84 13.26 15.68
F 33.72 17.54 69.13 4.01 7.17 22.16 84.63 8.69 15.21 26.92 46.06
M 42.73 10.77 41.83 4.33 12.19 20.55 56.66 12.26 19.20 32.88 27.48
A 49.67 17.87 55.59 8.91 42.63 32.06 67.62 64.53 38.16 42.88 49.10
M 53.41 15.14 46.32 6.01 17.59 30.10 47.47 13.22 19.99 16.11 30.50
J 55.90 15.32 31.15 4.66 19.36 26.47 37.68 23.93 16.20 13.13 24.94
J 59.72 7.62 30.04 3.65 11.00 32.60 52.64 25.90 6.40 14.08 61.08
A 52.36 9.65 28.11 5.47 15.68 30.04 35.46 28.30 8.93 12.70 39.57
S 55.90 14.89 28.98 8.12 46.33 38.88 27.80 29.58 91.02 10.96 29.18
O 37.71 9.56 19.25 2.96 19.69 17.13 21.69 9.47 18.67 3.73 18.67
N 32.35 11.88 22.22 4.49 21.89 19.30 12.03 11.06 34.91 5.30 17.12
D 19.99 11.21 18.52 3.00 5.92 14.32 24.33 3.93 12.70 12.70 22.33
Annual 43.37 12.75 34.46 4.93 19.05 25.01 40.57 20.07 13.84 17.05 31.81

Table 8 Monthly and annual mean wind energy potential (W/m2) at the selected stations in
inland in 2004. (J-D = from January to December)
Station Aswan Baharia Cairo Dekhala El- Farafra Ismailia Luxor Minya Siwa Wadi El-
Month Kharga Natron
J 21.37 10.90 74.64 4.03 5.47 19.84 63.46 8.75 21.75 22.33 33.09
F 32.09 8.04 35.40 3.07 13.73 13.95 48.54 10.60 13.73 14.39 18.00
M 40.05 10.22 31.69 5.22 39.65 23.09 36.66 29.45 57.67 11.31 18.19
A 35.14 6.03 50.24 5.04 27.41 23.41 44.88 16.18 29.18 15.80 25.18
M 41.73 19.44 44.79 4.17 22.94 34.32 54.68 13.04 69.20 23.76 33.49
J 49.91 20.69 36.36 4.20 34.75 33.64 50.88 7.57 65.12 6.00 30.22
J 41.50 13.89 22.98 5.73 20.15 24.61 41.26 7.28 30.79 5.72 27.63
A 39.10 5.19 21.37 4.83 15.87 28.63 36.57 8.10 16.10 5.22 26.30
S 36.98 15.42 27.46 6.12 35.66 27.57 29.98 .26 43.51 5.62 31.28
O 24.63 12.40 16.61 4.65 28.42 21.32 15.84 4.14 56.10 5.76 21.27
N 31.31 6.33 15.90 2.50 10.57 12.66 18.34 4.56 20.85 11.19 18.54
D 25.54 7.38 13.60 2.33 10.51 14.61 16.52 3.22 24.67 5.46 18.96
Annual 34.95 11.33 32.59 4.33 22.09 23.14 38.13 9.93 37.39 11.05 25.18

It is clear that the wind is strongest in the summer in both Red Sea and inland zones. This
may be explained by the oscillations of the subtropical front during summer in these zones. At
the Mediterranean coast zone, wind is strongest in winter and spring seasons, (expect
Alexandria) this may be due to the Mediterranean Sea secondary depressions which
characterize these seasons.

5. CONCLUSION
A statistical analysis for five years data gave the annual average wind speed as 5.8 m/s, with
98% of wind speed records being in the range of 3 to 10 m/s for Hurguda station (Red Sea
coast) which has the largest wind speed. While the annual average wind speed for Dekhala
station (which has the smallest wind speed) was 2.3 m/s, with 98.5% of wind speed records
being in the range of 1 to 5 m/s. Hurguda was selected as the site of the National Wind Energy
Technology Center (NWETC) due to its favourable wind condition. Amongst Mediterranean
stations, Port Said had larger mean wind speed (4.9 m/s).
 01_S384-Essa.qxd 18/7/06 12:33 pm Page 106

106 S URVEY AND A SSESSMENT OF W IND - SPEED AND W IND - POWER IN E GYPT

It is clear that the wind is strongest in the summer season in both Red Sea and inland zones.
At the Mediterranean coast zones wind is strongest in winter and spring seasons. The random
wind energy potential formulation shows that the average wind energy production is not only
a function of the coefficient of variation of the air density and wind speed cubed, but also of the
cross-correlation coefficient, r.
Comparisons of Weibull, Rayleigh, and the actual data distribution show that the Weibull
distribution, with appropriate parameters; is the best match with the actual data. It is also
clear that the numerical values of Weibull parameters for annual wind speed distribution
at the selected stations in Egypt for two complete years, 2003 and 2004, show that the
Red Sea stations had the largest-mean wind speed and Weibull parameters. The Inland
stations had the smallest-parameters. Hurguda and El-Tor have the largest-mean wind
speed and Weibull parameters. Dekhala has the smallest-mean wind speed and Weibull
parameters. Mediterranean stations (expect El-Arish) had favourable mean wind speed and
Weibull parameters for wind power generation. The statistical analysis of this study indicates
that the Red Sea, Mediterranean (expect El-Arish) and some Inland zones (Aswan and
Ismailia) can be chosen as favourable locations for wind energy as renewable energy
resources.

REFERENCES
[1] Said, U., and Mortensen, N., (1997): Wind resources of the Gulf of Suez and Northern Red
Sea of Egypt. Proceedings of the first international Conference and trade fair on
environmental management and technologies, Cairo, Egypt, 16–18 February.
[2] Ib Troen and Erik Lundtang Petersen (1989): European Wind Atlas. Commission of the
European Communities, Directorate-General for Science, Research and Development
Brussles, Belgium, Riso National Laboratory, Roskilde, Denmark, pp. 75–78.
[3] Khaled S. M. Essa, Soad M. Etman and M. Embaby (2004): The effect of Air Density
Variation on Wind Power Flux. Wind Engineering 28(3): 305–310.
[4] Papoulis A., (1969): Probability Random Variables and Stochastic Processes, McGraw-
Hill, New York, p. 583.
[5] J.W. Reed, (1979): Wind power climatology of the United States, SAND 78-1620, pp. 3–20.
[6] Egbert Boeker and Rienk Van Grondelle, (1994): Environmental Physics, John Wiley &
Sons, p. 45.
[7] Petersen, E. I. Troen, I., Frandsen, S., and Hedegaard, K. (1981): Wind Atlas for Denmark.
RISO. Denmark.
[8] Justus, C. G., Hargraves, W.R., and Yalcin, A. (1976): Nationwide Assessment of Potential
Output from wind Powered Generators. J. Appl. Meteor., (15): pp. 673–678.
[9] Cortis R. B., Sigl A. B., and Klein J. (1978): Probability Models of Wind Velocity
Magnitude and Persistence. Solar Energy (20): pp. 483–493.
[10] Hennessey J. P. Jr. (1978): A comparison of the Weibull and Rayleigh Distribution for
Estimating Wind Power Potential. Wind Eng., (2): pp. 156–164.
[11] Khaled S. M. Essa, and M. Embaby (2005): Statistical Evaluation of Wind Energy Using
Average Velocity at Inshas, Egypt. Wind Engineering, 29(1): pp. 83–88.
[12] Twidell, J. W. and Weir, A. D., (1987): Renewable Energy Resources. English Language
Book Society/ E & F. N. Spon.

View publication stats