Available online at www.sciencedirect.

com

Solar Energy 83 (2009) 1274–1284

www.elsevier.com/locate/solener

The Sun’s apparent position and the optimal tilt angle

of a solar collector in the northern hemisphere
Tian Pau Chang *
Department of Computer Science and Information Engineering, Nankai University of Technology, Nantou 542, Taiwan

Received 23 January 2008; received in revised form 17 December 2008; accepted 26 February 2009
Available online 8 April 2009

Communicated by: Associate Editor David Renne

Abstract

In this paper, the Julian dating system is adopted to calculate the Sun’s apparent position. Both the sunshine duration and the optimal
installation angle of a fixed solar collector are obtained for different time periods and latitudes in the northern hemisphere. To simulate
different operating environments, both kinds of radiation flux are considered, i.e. extraterrestrial radiation and the global radiation cal-
culated using an empirical model. The results show that Julian dating system is precise enough to predict the Sun’s locus according to
analyses of the solar declination and the azimuth and elevation angles. The orientation that a solar collector must be installed at can be
roughly estimated by determining the sunshine duration. The majority of the Sun’s path lies in the southern sky over a year, and hence
the sunshine duration in the southern sky is longer, except for locations with latitudes below 1.5°. The yearly optimal angles are positive
(approximately 0.91 and 0.76 multiplied by the latitude for extraterrestrial and global radiation, respectively), for latitudes below 65o.
Above this, the curves are flatter, and the differences between the two types of radiation flux become greater. The ratio of the yearly
irradiation captured by a collector installed at its yearly optimal angle to that captured by the ground surface increases with latitude,
and reaches a maximum of 1.71 and 1.35 near latitudes of 65° for extraterrestrial and global radiation, respectively. The efficiency of
a solar collector decreases when operating in a cloudy environment. The amount of global radiation incident on the ground surface
is about 0.51–0.66 times that of extraterrestrial radiation throughout the hemisphere.
Ó 2009 Elsevier Ltd. All rights reserved.

Keywords: Julian date; Sunshine duration; Solar collector; Optimal angle

1. Introduction ment of a collector’s efficiency. Generally speaking, the

optimal orientation is due south in the northern hemi-
Solar energy is a free and clean natural resource which sphere. However, it may be due north during special time
can be utilized anywhere in the world by using a collector periods depending on the relative position of the Earth
or photovoltaic (PV) cell (Fig. 1). Due to the recent short- and the Sun. Thus, making a precise prediction of the Sun’s
age of fossil fuels, solar energy now plays a more important movement locus is a major premise for the development of
role when considering our energy supply. The level of solar solar energy applications.
radiation captured by a collector is mainly affected by the There are many different methodologies for the calcula-
installation azimuth and tilt angles. This is because both tion of the Sun’s apparent position. Durisch and Anglehart
of these angles influence the angle of incidence of the solar (1987) as well as Schubnell and Ries (1990) used the time
radiation upon the collector panel. Therefore, the choice of counted since a new year as a variable, and combined it
orientation and tilt angle is a critical issue for the improve- with the mean angular velocity of the Earth to accurately
calculate the Sun’s position. However, many other papers
*
Fax: +886 49 2561408. adopt the day number (dn) as a time variable (Tang and
E-mail address: [email protected] Wu, 2004; Huang and Sun, 2007; Sakonidou et al., 2008),

0038-092X/$ - see front matter Ó 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.solener.2009.02.009
 T.P. Chang / Solar Energy 83 (2009) 1274–1284 1275

Nomenclature

a0, a1, k, r0, r1, rk constants of atmospheric transmit- TC time transferring constant, 1.002738
tance, dimensionless TD time difference between LT and GMT, hour
d day of month, dimensionless Y year A.D. , dimensionless
dn day number of year, dimensionless Greek letters
E orbit eccentricity, dimensionless a longitude of the Sun on equatorial plane, degree
GMST right ascension of Greenwich Mean Sidereal k longitude of the Sun on ecliptic plane, degree
Time, second X latitude of the Sun on ecliptic plane, degree
H altitude of observer, kilometer f angle between the ecliptic and equatorial plane,
H elevation angle of the Sun, degree degree
I0 accumulated extraterrestrial radiation, J/m2 m perihelion anomaly, degree
Ic clear sky radiation on horizontal surface, J/m2 e mean anomaly deviated from the vernal equinox
Icb clear sky beam radiation on horizontal surface, , degree
J/m2 l argument of perigee, degree
Icd clear sky diffuse radiation on horizontal surface, d solar declination, degree
J/m2 hz zenith of the Sun, degree
It global radiation on tilted plane, J/m2 w azimuth of the Sun, degree
JC Julian century, dimensionless c geographic longitude, degree
JD Julian day, dimensionless / geographic latitude, degree
JD1 Julian time, dimensionless x solar hour angle, degree
JD2000 Julian day of GMT 0 on Jan. 1, 2000, dimen- b tilted angle of collector, degree
sionless h incident angle of direct beam to the collector,
LMST local mean sidereal time, hour degree
LT local time, hour q reflection coefficient (albedo), dimensionless
M mean anomaly of perihelion, degree sb atmospheric transmittance of beam radiation,
M month of year, dimensionless dimensionless
Rb beam radiation ratio, dimensionless sd atmospheric transmittance of diffuse radiation ,
Sc solar constant, 1367 W/m2 dimensionless

which is counted from January 1st throughout the year (1– different calendar systems, and has popularly been applied
365), then calculate the solar declination (d), which is the in the field of astronomical and historical researches
angle between the lines joining the centers of the Sun, the
Earth, and the equatorial plane. Subsequently, the zenith Zenith
(hz)) and azimuth angle (w) of the Sun can be determined.
For a day number dn, the angles are given by: Sun
d ¼ 23:45 sinð2pð284 þ dnÞ=365Þ ð1Þ
cos hz ¼ sin d sin / þ cos d cos / cos x ð2Þ
cos w ¼ ðcos hz sin /  sin dÞ= sin hz cos / ð3Þ
where / is the latitude of the installation location, and x is
the solar hour angle, which changes by 15°/h and is zero at Normal Solar
θz
solar noon, negative in the morning and positive after vector collector
noon. The expressions listed above have been previously
θ
adopted by other researchers (Braun and Mitchell, 1983;
Morcos, 1994; Chen et al., 2005; Huang and Sun, 2007;
Gunerhan and Hepbasli, 2007). East
The Julian calendar is alternative counting method
invented by Joseph Scaliger, a French chronologist in the
β
16th century who named it after his father, Julius. The
exact number of days counted from noon on 31 December
4714 B.C. is called the Julian day; it is convenient for dis-
tinguishing time sequences among long term events that South
happened during different eras or are even recorded with
Fig. 1. Geometry of solar collector.
1276 T.P. Chang / Solar Energy 83 (2009) 1274–1284

(Farrell, 1988; Duffet-Smith, 1988; Michalsky, 1988). Note that the value of m and y must be replaced with
Using the Julian date system, Reda and Andreas (2004) m + 12 and y  1, respectively, for January and February.
proposed a step by step procedure for calculating the solar The operator Int() means that the output is an integer
azimuth and zenith angles in the period from the year value.
2000 to 6000, with an uncertainty of only ±0.0003°. Shen The Julian time (JD1) and Julian century (JC), which
(2005) utilized the Julian date to calculate the Sun’s posi- are counted from the beginning of the Julian day at the
tion as well as to determine the optimal installation angle Greenwich Mean Time 0 on January 1, 2000 (JD2000),
of a south facing solar panel. The electric power output are defined to be:
from a photovoltaic cell installed at various orientations JD1 ¼ JD  JD2000 ð9Þ
and tilt angles was estimated, but the accuracy of the Sun’s
calculated position was not discussed; furthermore, Shen’s JC ¼ JD1=36525 ð10Þ
study is localized to Taiwan. In this paper, the Julian day is The latitude of the Sun in the ecliptic coordinate system
used as a time variable for predicting the Sun’s position. (x) is 0° because the Sun’s motion always lies on the eclip-
The validity of the methodology will be verified by using tic plane. The actual perihelion anomaly (t) can be
concrete data, including the solar declination and instanta- obtained after expanding the Kepler equation in series as
neous azimuth and elevation angles. Meanwhile, the time below:
duration of the Sun standing either in the southern or in
the northern sky will be accumulated for different time t ¼ M þ 360E sinðMÞ=p þ 900E2 sinð2MÞ=4p
periods. To make a detailed estimation for the installation  180E3 sinðMÞ=4p ð11Þ
of a solar collector, in addition to the extraterrestrial radi-
ation (i.e. solar radiation under air mass zero (AM0), suit- where
able for outer space applications), the global radiation M ¼el ð12Þ
(taking into account the direct beam, diffusion, and reflec- E ¼ 0:01670862  0:00004204JC ð13Þ
tion from the ground) predicted using the empirical model
2
proposed by Liu and Jordan (1960, 1963) and Hottel e ¼ 280:466457 þ 0:985647358JD1 þ 0:000304JC ð14Þ
(1976), will be considered to simulate a different operating l ¼ 282:937348 þ 0:00004707624JD1
environment. Although there are other empirical models
available in literature for calculating the global radiation þ 0:0004569JC2 ð15Þ
(Kasten and Czeplak, 1980; Bird and Hulstrom, 1981; Here, M is the mean anomaly of the perihelion, E is the
Muneer et al., 2000; Gueymard, 2003a,b), the one used in orbit eccentricity, e is the mean anomaly deviation from the
the present study has been adopted by previous researchers vernal equinox, and l is the perigee argument. The longi-
such as Morcos (1994), Huang and Sun (2007), and Chang tude of the Sun in ecliptic coordinate (k) can thus be calcu-
(in press). It is applicable to different latitudes and can be lated using:
treated as the worldwide average value of the global radia-
tion. Subsequently, the optimal installation angle of a fixed k ¼ t  360 þ l ð16Þ
solar collector for a particular period of time, e.g. 12 The angle between the ecliptic plane and the equatorial
months, can be obtained throughout the northern hemi- plane (f) is 23.4392o. The longitude of the Sun with respect
sphere according to both radiation sources. This discussion to the equatorial plane (a) is the same as that to the ecliptic
is unprecedented. plane, i.e. a = k. The latitude of the Sun with respect to the
equatorial plane (d), (also known the solar declination), is
given as:
2. Sun’s apparent position by Julian date
d ¼ sin1 ðsin X cos f þ cos X sin k sin fÞ ð17Þ
For a date during the year y, month m, and day d, its For a given local time (LT, in seconds) and geographic
Julian day (JD) is defined as (Duffet-Smith, 1988; Shen, longitude (c in seconds, 1 h/15°), the right ascension of
2005): the Greenwich Mean Sidereal Time (GMST, in seconds)
JD ¼ B þ C þ D þ d  1524:5 ð4Þ and local mean sidereal time (LMST, in seconds) can be
calculated using:
where GMST ¼ 24110:4841 þ 8640184:81266
B ¼ 2  A þ IntðA=4Þ ð5Þ  JD1=36525:24 ð18Þ
LMST ¼ GMST þ c þ ðLT  TDÞ  TC ð19Þ
C ¼ Intð365:25  ðy þ 4716ÞÞ ð6Þ
where TD (in seconds) is the time difference between the
D ¼ Intð30:6001  ðm þ 1ÞÞ ð7Þ local time and the Greenwich mean time, and the constant
TC which is used for transferring the solar time to sidereal
A ¼ Intðy=100Þ ð8Þ time is 1.002738. Lastly, the Sun‘s instantaneous azimuth
 T.P. Chang / Solar Energy 83 (2009) 1274–1284 1277

Table 1 2000 180

Radiation parameters for different climate types. azimuth angle of the Sun
1800 elevation angle of the Sun
Climate type ro r1 rk 160
extrat. radi. on tilted collector
extrat. radi. on horizon
Tropical 0.95 0.98 1.02 1600 global radi. on tilted collector
140
Mid-latitude summer 0.97 0.99 1.02 global radi. on horizon

Instantaneous radiation (W/m2)

Sub-arctic summer 0.99 0.99 1.01 1400
120
Mid-latitude winter 1.03 1.01 1.00
1200

Angle (degree)
100

1000
(w) and elevation (h) observed at a location with the geo-
80
graphic longitude c and latitude / are determined by the 800
following equations: 60
600
sinðLMST  aÞ
w ¼ tan1 ð Þ ð20Þ 40
sin / cosðLMST  aÞ  cos / tan d 400

h ¼ sin1 ðsin / sin d þ cos / cos d cosðLMST  aÞÞ 200 20

¼ 90o  hz ð21Þ 0 0
0 2 4 6 8 10 12 14 16 18 20 22 24
Here, the elevation is the angle between the Sun’s direc- Hour of the vernal equinox day
tion and the horizontal plane, whose complement is the
Fig. 3. The Sun’s instantaneous position and radiation observed in Taipei
zenith angle. on the vernal equinox day.

3. Global radiation on tilted collector by empirical model

where h is the instantaneous angle between the direct beam
and the normal of collector, it can be calculated by:
According to the model developed by Liu and Jordan
(1963), the radiation incident on a south facing collector
h ¼ cos1 ðsin hz cos w sin b þ cos hz cos bÞ ð24Þ
tilted at an angle b to the horizontal surface, as shown in
Fig. 1, is made up of three components: the direct beam,
diffusion, and reflection from the ground. The sum of these q is the ground reflection coefficient (albedo), which is
three terms is called global radiation (It): about 0.7 for snow covered ground and 0.2 for snow free
I t ¼ ðI cb Rb þ I cd ð1 þ cos bÞ=2 þ I c qð1  cos bÞ=2 ð22Þ ground. It was assumed to be 0.2 in this study (Yakup
and Malik, 2001; Cheng et al., 2006). The clear sky radia-
The tilt angle b is positive for collectors facing due south tion on a horizontal surface, Ic, is the sum of Icb and Icd:
but negative for due north and zero for the horizon. The
geometric factor, Rb, is the ratio of beam radiation on a I c ¼ I cb þ I cd ð25Þ
tilted surface to that on a horizontal surface and is given by:
Rb ¼ cos h= cos hz ð23Þ 2000 180
azimuth angle of the Sun
1800 elevation angle of the Sun
160
extrat. radi. on tilted collector
30 extrat. radi. on horizon
1600 global radi. on tilted collector
25 present study 140
global radi. on horizon
general model
Instantaneous radiation (W/m2)

20 1400
120
15
1200
Angle (degree)
Solar declination (degree)

10 100

1000
5
80
0 800

-5 60
600
-10
40
400
-15

-20 200 20

-25
0 0
-30 0 2 4 6 8 10 12 14 16 18 20 22 24
0 30 60 90 120 150 180 210 240 270 300 330 360 Hour of the summer solstice day
Day of year
Fig. 4. The Sun’s instantaneous position and radiation observed in Taipei
Fig. 2. Solar declinations throughout a year in 2007. on the summer solstice day.
1278 T.P. Chang / Solar Energy 83 (2009) 1274–1284

where Icb is the clear sky beam radiation on a horizontal 2000 180
azimuth angle of the Sun
surface which can be calculated using the expression de- elevation angle of the Sun
1800 160
rived by Hottel (1976): extrat. radi. on tilted collector
extrat. radi. on horizon
1600
I cb ¼ I o sb cos hz ð26Þ global radi. on tilted collector
global radi. on horizon
140

Instantaneous radiation (W/m2)

1400
where sb is the atmospheric transmittance of beam radia- 120
tion, and is equal to the ratio of the beam radiation on a 1200

Angle (degree)
horizontal surface to the air mass zero (AM0) extraterres- 100

trial radiation, it can be approached empirically (Hottel, 1000

1976): 80
800
sb ¼ ao þ a1 expðk= cos hz Þ ð27Þ 60
600

The constants ao ¼ ro ao ,
a1 ¼ r1 a1
and k ¼ rk k for a stan- 40
dard atmosphere with a visibility of 23 km can be calcu- 400

lated from the following relationship by assuming that 200 20

the observation altitude is less than 2.5 km:
2 0 0
ao ¼ 0:4237  0:00821ð6:0  H Þ 0 2 4 6 8 10 12 14 16 18 20 22 24

2 Hour of the autumnal equinox day

a1 ¼ 0:5055 þ 0:00595ð6:5  H Þ ð28Þ
Fig. 5. The Sun’s instantaneous position and radiation observed in Taipei
k  ¼ 0:2711 þ 0:01858ð2:5  H Þ2 on the autumnal equinox day.

where H is the altitude of the observer in kilometers and is

assumed to be 0.006 km in this study. The correction axis). For easy comparison, the azimuth is set to 0o at solar
parameters, ro, r1 and rk are related to climate conditions noon and increases towards the dawn and dusk. Both the
as summarized in Table 1. Note that the parameters suit- azimuth and elevation are anticipated, and are symmetric
able for higher latitude zones are not offered by the paper with respect to solar noon (however, they are not exactly
of Hottel (1976), on the calculation these three parameters symmetric). Furthermore, these angles are consistent with
are, respectively, set to the approximate values of 0.99, 0.99 those calculated using Eqs. (2)–(3); we can conclude that
and 1.00 for latitudes near the Arctic zone.The accumu- the mathematical expressions introduced in this study are
lated extraterrestrial beam radiation at AM0 for a period sufficient for predicting the Sun’s position.
of time can be obtained from the integral: In addition, the instantaneous solar radiation received is
Z plotted on these figures (the left ordinate axis). It can be
I o ¼ S c ð1 þ 0:033 cosð2pdn=365:25ÞÞdt ð29Þ seen that the radiation incident upon a tilted collector at
its daily optimal angle is stronger than that upon a hori-
where, Sc is the solar constant (1367 W/m2). The diffusion zontal surface. The optimal tilt angle was calculated by
component of clear sky radiation on a horizontal surface,
Icd, can be estimated using the model developed by Liu
and Jordan (1960): 2000 180
azimuth angle of the Sun
I cd ¼ I o sd cos hz ð30Þ 1800 elevation angle of the Sun
160
extrat. radi. on tilted collector

sd ¼ 0:2710  0:2939sb ð31Þ 1600

extrat. radi. on horizon
global radi. on tilted collector
140
global radi. on horizon
where sd is the atmospheric transmittance of diffuse
Instantaneous radiation (W/m2)

1400
radiation. 120

1200
Angle (degree)

100
4. Results and discussions 1000
80
Fig. 2 compares the solar declinations throughout a year 800
calculated using different expressions. The declination from 60
600
the present study obtained using the Julian date system and
sampled at solar noon everyday in 2007 matches the gen- 400
40

eral model given by Eq. (1). The annual mean of the dis-
20
crepancy between these two curves is about 0.3o. 200

To examine the appropriateness of the methodology 0 0

mentioned, Figs. 3–6 demonstrate the Sun’s apparent posi- 0 2 4 6 8 10 12 14 16 18 20 22 24

tion observed in Taipei (c = 121.5o, / = 25.1o) during four Hour of the winter solstice day

particular days in 2007 in terms of the instantaneous azi- Fig. 6. The Sun’s instantaneous position and radiation observed in Taipei
muth and elevation angles (referring to the right ordinate on the winter solstice day.
 T.P. Chang / Solar Energy 83 (2009) 1274–1284 1279

Table 2
Solar parameters for Taipei on four specified days in 2007.
Period Radiation type Optimal tilt angle (degree) Irradiation on Irradiation on tilted Sunshine time in
horiz. surface (Wh/m2) collector (Wh/m2) south/north sky (hour)
Vernal equinox (3/21) Extrat. 24.5 9541 10482 11.85/0.13
Global 25.0 6309 6842
Summer solstice (6/22) Extrat. 13.1 11213 11511 2.95/10.75
Global 7.9 7569 7629
Autumnal equinox (9/23) Extrat. 25.4 9363 10368 11.95/0
Global 25.9 6183 6745
Winter solstice (12/22) Extrat. 56.3 6274 11296 10.42/0
Global 53.3 3817 5851

searching for the value at which the solar radiation upon 600 90
the collector is maximized for a specific period of time. February
Table 2 summarizes the corresponding data for these days. southern sky duration
northern sky duration
80
500
The optimal angles during the vernal equinox, autumn optimal angle (extra.)
optimal angle (global) 70
equinox, and winter solstice days are positive values, irre-

Monthly optimal tilt angle (degree)

Monthly sunshine duration (hour)
spective of the type of solar radiation considered. This is 400 60
because the Sun’s apparent position is mostly in the south-
ern sky. The collector must thus be installed facing south to 50
collect the maximum amount of solar energy. On the con- 300
trary, the optimal angle during the summer solstice is neg- 40

ative, implying that the collector can collect more energy if

200 30
it is installed due north. Indeed, this can be realized by the
fact that the Sun’s azimuth on the summer solstice day is
20
larger than 90o for most of the day, e.g. before 10:30 and 100
after 13:30 as shown in Fig. 4. 10
To investigate more characteristics regarding the instal-
lation of a solar collector, Figs. 7–18 show the variation in 0 0
sunshine duration with geographic latitude for all 0 10 20 30 40 50 60 70 80 90
Latitude (degree, N)
12 months. The monthly optimal tilt angles of a solar
collector for both radiation sources are also shown. These Fig. 8. The monthly sunshine duration and optimal tilt angle of collector
calculations were carried out for the entire northern hemi- in February.

600 90 600 90

March
80 southern sky duration 80
northern sky duration
500 500 optimal angle (extra.)
70 optimal angle (global) 70
Monthly optimal tilt angle (degree)

Monthly optimal tilt angle (degree)

Monthly sunshine duration (hour)

Monthly sunshine duration (hour)

400 60 400 60

50 50
300 300
40 40

200 30 200 30

January
southern sky duration 20 20
northern sky duration
100 100
optimal angle (extra.)
optimal angle (global) 10 10

0 0 0 0
0 10 20 30 40 50 60 70 80 90 0 10 20 30 40 50 60 70 80 90
Latitude (degree, N) Latitude (degree, N)

Fig. 7. The monthly sunshine duration and optimal tilt angle of collector Fig. 9. The monthly sunshine duration and optimal tilt angle of collector
in January. in March.
1280 T.P. Chang / Solar Energy 83 (2009) 1274–1284

600 80 600 50

April June
southern sky duration
70 southern sky duration 40
northern sky duration northern sky duration
500 500 optimal angle (extra.)
optimal angle (extra.)
60
optimal angle (global) optimal angle (global) 30

Monthly optimal tilt angle (degree)

Monthly optimal tilt angle (degree)

Monthly sunshine duration (hour)

Monthly sunshine duration (hour)

50
400 400 20

40
10
300 30 300
0
20

200 200 -10

10

-20
0
100 100

-10 -30

0 -20 0 -40
0 10 20 30 40 50 60 70 80 90 0 10 20 30 40 50 60 70 80 90
Latitude (degree, N) Latitude (degree, N)

Fig. 10. The monthly sunshine duration and optimal tilt angle of collector Fig. 12. The monthly sunshine duration and optimal tilt angle of collector
in April. in June.

sphere at a longitude of 120o east. We can see how the opti- to simplify the figures, the lines of best fit are not shown.
mal angle depends on the Sun’s movement locus. As for the The R-square determination coefficients of linear regres-
northern hemisphere, most of the Sun’s locus lies in the sion are greater than 0.998, except for higher latitudes in
southern sky with some exceptions from April to Septem- June (R-square = 0.958). In general, the slopes of the
ber, especially in lower latitudes or in higher latitudes regression line for the spring season are very similar to
inside the Arctic Circle. For January, February, October, those in autumn, i.e. about 0.96 and 0.91 for the extrater-
November, and December, the sunshine duration in the restrial and global radiation, respectively. For example,
northern sky is always 0. in March, as shown in Fig. 9, the duration when the Sun
Table 3 summarizes the correlations between the opti- is in the southern sky is longer than when it is in the north-
mal tilt angle and the geographic latitude for a longitude ern sky, resulting in a positive tilt angle for both the extra-
of 120o east for different periods obtained by finding the terrestrial and global radiation. Furthermore, the angles
line of best fit, based on the least squares method. In order estimated from both radiation types are very close to each

600 60 600 60

May July
southern sky duration 50 southern sky duration 50
northern sky duration northern sky duration
500 optimal angle (extra.) 500 optimal angle (extra.)
optimal angle (global)
40
40 optimal angle (global)
Monthly optimal tilt angle (degree)
Monthly optimal tilt angle (degree)

Monthly sunshine duration (hour)

Monthly sunshine duration (hour)

30
400 30 400

20
20
300 300 10
10
0

200 0 200
-10

-10
-20
100 100
-20 -30

0 -30 0 -40
0 10 20 30 40 50 60 70 80 90 0 10 20 30 40 50 60 70 80 90
Latitude (degree, N) Latitude (degree, N)

Fig. 11. The monthly sunshine duration and optimal tilt angle of collector Fig. 13. The monthly sunshine duration and optimal tilt angle of collector
in May. in July.
 T.P. Chang / Solar Energy 83 (2009) 1274–1284 1281

600 70 600 90

August October
southern sky duration 60 southern sky duration 80
northern sky duration northern sky duration
500 500 optimal angle (extra.)
optimal angle (extra.)
optimal angle (global) 50 optimal angle (global) 70

Monthly optimal tilt angle (degree)

Monthly optimal tilt angle (degree)

Monthly sunshine duration (hour)

Monthly sunshine duration (hour)

400 40 400 60

30 50
300 300
20 40

200 10 200 30

0 20
100 100
-10 10

0 -20 0 0
0 10 20 30 40 50 60 70 80 90 0 10 20 30 40 50 60 70 80 90
Latitude (degree, N) Latitude (degree, N)

Fig. 14. The monthly sunshine duration and optimal tilt angle of collector Fig. 16. The monthly sunshine duration and optimal tilt angle of collector
in August. in October.

other and increase with latitude. The characteristics during 65o. Thus, the optimal tilt angles are negative for latitudes
September, as shown in Fig. 15, are quite similar to those below 35o but are almost positive for latitudes between 35o
during March except for latitudes below 4o. and 65o. As for latitudes above 65o, the angles are negative
In summer, the curves reveal more complicated varia- when we only consider the incidence of the direct beam, i.e.
tions. For regions outside the Arctic Circle, the slopes of the extraterrestrial radiation, whereas those angles are
the regression line are close to 0.91 and 0.85 for the extra- positive if the diffusion and reflection components are
terrestrial and global radiation, respectively. As shown in included. As we know, for regions at latitudes above 65o,
Fig. 12, in June, the duration when the Sun is in the south- the Sun rarely sets down to the horizon in June. Its dura-
ern sky is zero for latitudes below 22o and increases grad- tions in the southern and northern sky are approximate,
ually for latitudes greater than this. The duration when and thus the orientation of the collector becomes less
the Sun is in the northern sky is longer than when it is in important. This is why the optimal tilt angles estimated
the southern sky except for latitudes between 35o and using different types of radiation reveal a bigger discrep-

600 90 600 90

September
southern sky duration 80 80
northern sky duration
500 optimal angle (extra.) 500
optimal angle (global)
70
70
Monthly optimal tilt angle (degree)

Monthly optimal tilt angle (degree)

Monthly sunshine duration (hour)

Monthly sunshine duration (hour)

60
400 400 60

50
50
300 40 300
40
30

200 200 30
20
November
southern sky duration 20
10
100 100 northern sky duration
optimal angle (extra.)
0 optimal angle (global) 10

0 -10 0 0
0 10 20 30 40 50 60 70 80 90 0 10 20 30 40 50 60 70 80 90
Latitude (degree, N) Latitude (degree, N)

Fig. 15. The monthly sunshine duration and optimal tilt angle of collector Fig. 17. The monthly sunshine duration and optimal tilt angle of collector
in September. in November.
1282 T.P. Chang / Solar Energy 83 (2009) 1274–1284

600 90 6000 90

Yearly
80 southern sky duration 80
northern sky duration
500 5000
optimal angle (extra.)
70 optimal angle (global) 70

Monthly optimal tilt angle (degree)

Monthly sunshine duration (hour)

Yearly optimal tilt angle (degree)

Yearly sunshine duration (hour)
400 60 4000 60

50 50
300 3000
40 40

200 30 2000 30

December
southern sky duration 20 20
100 northern sky duration
optimal angle (extra.)
1000
optimal angle (global) 10 10

0 0 0 0
0 10 20 30 40 50 60 70 80 90 0 10 20 30 40 50 60 70 80 90
Latitude (degree, N) Latitude (degree, N)

Fig. 18. The monthly sunshine duration and optimal tilt angle of collector Fig. 19. The yearly sunshine duration and optimal tilt angle of collector.
in December.

On the whole, the angles presented by the present study

ancy. In fact, according to the computation of the present are in agreement with those of Yakup and Malik (2001) as
study, for most of the regions where / >65o, the irradiation well as Nijegorodov et al. (1994). Fig. 19 shows the results
captured by a solar collector installed at its monthly opti- from the calculations of an entire year. The sunshine dura-
mal angle does not exceed the irradiation incident on a hor- tion in the southern sky is longer than in the northern sky
izontal surface by even 1%. except for latitudes below 1.5o. The yearly optimal angles
In winter, the slope is only 0.86 for both radiation present an obvious change near latitudes of 65o. Below this
sources. This is the smallest of the four seasons. As shown latitude, the angles are about 0.91 and 0.76 times latitude
in Fig. 18, in December, the duration in the southern sky for extraterrestrial and global radiation, respectively.
decreases with latitude and becomes zero for latitudes Above 65o, the curves of both radiation types become flat-
greater than 68o whereas the duration in the northern sky ter and the differences between them are larger.
is always zero throughout the hemisphere. The optimal Fig. 20 demonstrates the variation in yearly irradiation
angles estimated using the global radiation are less than throughout the hemisphere. The radiation incident on a col-
those estimated using the extraterrestrial radiation by
about 3o.
6000 2
extrat. irradi. on tilted collector
extrat. irradi. on horizon
Table 3 global irradi. on tilted collector
1.9
Correlations between the optimal tilt angle and the geographic latitude (/) 5000
global irradi. on horizon
irradi. ratio of tilted to horizon (extrat.)
for different time periods. 1.8
irradi. ratio of tilted to horizon (global)

Period Optimal tilt angle Optimal tilt angle 1.7

Yearly irradiation (kWh/m2)

(extraterrestrial radiation) (global radiation) 4000

0:853/ þ 31:45o ; / 6 72o 0:864/ þ 28:17o ; / 6 72o

Irradiation ratio

January 1.6
February 0:898/ þ 20:62o ; / 6 82o 0:889/ þ 18:58o ; / 6 82o
March 0:975/ þ 3:05o ; / 6 90o 0:923/ þ 3:99o ; / 6 90o 3000 1.5

April 0:997/  15:81o ; / 6 90o 0:921/  11:86o ; / 6 90o

1.4
May 0:962/  29:72o ; / 6 90o 0:857/  23:17o ; / 6 90o
June 0:841/  33:86o ;/ 6 65o 0:834/  28:70o ; / 6 74o 2000
1.3
0:964/  87:43o ; otherwise 0:916/ þ 81:75o ; otherwise
July 0:934/  32:64o ; / 6 90o 0:860/  26:40o ; / 6 79o
1.2
1:331/ þ 119:38o ; otherwise 1000
o o
August 0:989/  21:32 ; / 6 90 0:908 /  16:51o ; / 6 90o 1.1
September 0:997/  4:11o ; / 6 90o 0:931/  1:87o ; / 6 90o
October 0:920/ þ 15:20o ; / 6 86o 0:900/ þ 13:95o ; / 6 86o 0 1
November 0:863/ þ 28:78o ; / 6 75o 0:869/ þ 25:80o ; / 6 75o 0 10 20 30 40 50 60 70 80 90
December 0:849/ þ 34:30o ; / 6 68o 0:865/ þ 30:78o ; / 6 68o Latitude (N, degree)
o o o o
Yearly 0:908/ þ 0:73 ; / 6 65 0:764/ þ 2:14 ; / 6 65 Fig. 20. The yearly solar irradiations captured by tilted collector and by
0:360/ þ 32:40o ; otherwise 0:224/ þ 33:65o ; otherwise horizontal surface as well as their ratios.
 T.P. Chang / Solar Energy 83 (2009) 1274–1284 1283

lector installed at its yearly optimal angle (as shown in and global radiation, respectively. The analogous
Fig. 19) is greater than the radiation incident on a horizon- slopes in summer are 0.91 and 0.85 for latitudes out-
tal surface for both types of radiation flux. The ratios of the side the Arctic Circle for extraterrestrial and global
irradiation captured by a tilted collector to that captured by radiation, respectively. The slope in winter is the
a horizontal surface increase gradually with latitude, and smallest one among the four seasons. It is approxi-
reach a maximum value near latitudes of 65o, i.e. 1.71 and mately 0.86 for both sources of radiation.
1.35 for the extraterrestrial and global radiation, respec- (5) For an annual period, the sunshine duration in the
tively. As compared with the situation of AM0 (i.e. extrater- southern sky is longer except for latitudes below
restrial radiation), the efficiency of a collector decreases 1.5o. For latitudes below 65o, the yearly optimal
when it operates in cloudy or polluted environments where angles are positive and are about 0.91 and 0.76 mul-
the diffusion component (isotropic radiation) increases. On tiplied by the latitude for extraterrestrial and global
the other hand, as shown in Fig. 20, the amount of global radiation, respectively. Above this, the curves are flat-
radiation incident on a horizontal ground surface is about ter, and the differences between the two radiation
0.51–0.66 times that of extraterrestrial radiation through- types become larger.
out the hemisphere. This is equivalent to the value of the (6) The ratios of the yearly irradiation captured by a col-
clearness index (the ratio of global radiation transmitted lector installed with its yearly optimal angle to that
to the ground surface to extraterrestrial radiation). captured by a horizontal ground surface increase with
It is worth mentioning that, as stated in previous latitude. Near latitudes of 65o, the ratios have a max-
researches (Yakup and Malik, 2001; Shu et al., 2006; Chang, imum of 1.71 and 1.35 for extraterrestrial and global
in press), the optimal tilt angle of a collector becomes smaller radiation, respectively. As compared to the situation
(flatter) when the sky clearness index decreases. However, of air mass zero (AM0), the efficiency of a solar col-
the clearness index changes from location to location and lector decreases when it operates in a cloudy environ-
it is very hard to produce a general expression to precisely ment. The amount of global radiation incident on the
evaluate its level all over the world. According to the simula- ground surface is about 0.51–0.66 times that of extra-
tion results proposed by Chang (in press), the yearly optimal terrestrial radiation throughout the hemisphere.
angle in Taiwan increases by about 2o when the clearness
index rises by 0.1 in the range of 0.3–0.6. Acknowledgments
5. Conclusions The author appreciate reviewers for their treasure com-
ments, and deeply thanks Dr. Wu C.F., adviser of Central
In this paper, the validity of the Julian dating system is Weather Bureau and Dr. Huang M.W., researcher of the
examined and recommended for use in determining the Institute of Earth Sciences, Academia Sinica, Taiwan, for
optimal tilt angle for a solar collector. Both the sunshine their useful suggestions. This work was supported partly
duration and the optimal tilt angle of a fixed solar collector by the Nankai University of Technology.
are obtained for different time periods and different lati-
tudes in the northern hemisphere along the longitude of
120o east, simultaneously taking into account both the References
extraterrestrial and global radiation. The conclusions are
Bird, R.E., Hulstrom, R.L., 1981. A Simplified Clear Sky Model for
summarized below:
Direct and Diffuse Insolation on Horizontal Surfaces. SERI/TR-642–
761. Solar Energy Research Institute, Colorado, USA.
(1) The Julian dating system introduced in this paper is Braun, J.E., Mitchell, J.C., 1983. Solar geometry for fixed and tracking
precise enough to predict the Sun’s movement path surfaces. Solar Energy 31 (5), 439–444.
based on the analyses, including the solar declination Chang, T.P., in press. Performance evaluation for solar collectors in
Taiwan. Energy. doi:10.1016/j.energy.2008.09.016.
and the azimuth and elevation angles.
Chen, Y.M., Lee, C.H., Wu, H.C., 2005. Calculation of the optimum
(2) The orientation that a solar collector must be installation angle for fixed solar-cell panels based on the genetic
installed at can be roughly estimated by determining algorithm and the simulated-annealing method. IEEE Transactions on
the sunshine duration to see whether the Sun’s locus Energy Conversion 20 (2), 467–473.
is mainly in the northern or southern sky. Cheng, C.L., Chan, C.Y., Chen, C.L., 2006. An empirical approach to
estimating monthly radiation on south-facing tilted planes for building
(3) The majority of the Sun’s locus lies in the southern
application. Energy 31, 2940–2957.
sky with the exception of the period from April to Duffet-Smith, P., 1988. Practical Astronomy with your Calculator, third
September in lower latitudes or in latitudes within ed. Cambridge University Press. ISBN 0-521-35699-7.
the Arctic Circle. For the months of January, Febru- Durisch, W., Anglehart, J., 1987. Einfache Berechnung der Sonneneleva-
ary, October, November, and December, the sun- tion. Tech. Rep., Paul Scherrer Institut TM 13-87-06.
Farrell, J.P., 1988. Date computations into the third millennium.
shine duration in the northern sky is zero.
Computer Assurance Compass 88, 13–18.
(4) The slopes of the regression line that best fits the opti- Gueymard, C.A., 2003a. Direct solar transmittance and irradiance
mal tilt angle in spring are close to those in autumn; predictions with broadband models. Part I: detailed theoretical
the slopes are about 0.96 and 0.91 for extraterrestrial performance assessment. Solar Energy 74 (5), 355–379.
1284 T.P. Chang / Solar Energy 83 (2009) 1274–1284

Gueymard, C.A., 2003b. Direct solar transmittance and irradiance Muneer, T., Gul, M.S., Kubie, J., 2000. Models for estimating solar
predictions with broadband models. Part II: validation with high- radiation and illuminance from meteorological parameters. Journal of
quality measurements. Solar Energy 74 (5), 381–395. Solar Energy Engineering 122 (3), 146–153.
Gunerhan, H., Hepbasli, A., 2007. Determination of the optimum tilt Nijegorodov, N., Devan, K.R.S., Jain, P.K., Carlsson, S., 1994. Atmo-
angle of solar collectors for building applications. Building and spheric transmittance models and an analytical method to predict the
Environment 42, 779–783. optimum slope of an absorber plate, variously orientated at any
Hottel, H.C., 1976. A simple model for estimating the transmittance of latitude. Renewable Energy 4 (5), 529–543.
direct solar radiation through clear atmospheres. Solar Energy 18 (2), Reda, I., Andreas, A., 2004. Solar position algorithm for solar radiation
129–134. applications. Solar Energy 76 (5), 577–589.
Huang, B.J., Sun, F.S., 2007. Feasibility study of one axis three positions Sakonidou, E.P., Karapantsios, T.D., Balouktsis, A.I., Chassapis, D.,
tracking solar PV with low concentration ratio reflector. Energy 2008. Modeling of the optimum tilt of a solar chimney for maximum
Conversion and Management 48, 1273–1280. air flow. Solar Energy 82, 80–94.
Kasten, F., Czeplak, G., 1980. Solar and terrestrial radiation dependent Schubnell, M., Ries, H., 1990. Velocity-controlled tracking of the sun.
on the amount and type of cloud. Solar Energy 24 (2), 177–189. Solar Energy Materials 21, 207–212.
Liu, B.Y.H., Jordan, R.C., 1960. The interrelationship and characteristic Shen, C.H., 2005. The effects of the solar module installing angles on the
distribution of direct, diffuse and total solar radiation. Solar Energy 4 output power. Journal of Technology 20 (1), 15–20.
(3), 1–19. Shu, N., Kameda, N., Kishida, Y., Sonoda, H., 2006. Experimental and
Liu, B.Y.H., Jordan, R.C., 1963. The long-term average performance of theoretical study on the optimal tilt angle of photovoltaic panels.
flat-plate solar energy collectors. Solar Energy 7 (2), 53–74. Journal of Asian Architecture and Building Engineering 5, 399–405.
Michalsky, J.J., 1988. The Astronomical Almanac’s algorithm for Tang, R., Wu, T., 2004. Optimal tilt-angles for solar collectors used in
approximate solar position (1950–2050). Solar Energy 40 (3), 227– China. Applied Energy 79, 239–248.
235. Ab, M., Yakup, H.M., Malik, A.Q., 2001. Optimum tilt angle and
Morcos, V.H., 1994. Optimum tilt angle and orientation for solar orientation for solar collector in Brunei Darussalam. Renewable
collectors in Assiut, Egypt. Renewable Energy 4 (3), 291–298. Energy 24, 223–234.