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

net/publication/270364063

Contribution to the Quantification of Solar Radiation in Algeria

Article in Energy Procedia · December 2013

DOI: 10.1016/j.egypro.2013.07.085

CITATIONS READS

37 438

4 authors:

Mohamed Nia M. Chegaar

1 PUBLICATION 37 CITATIONS
Ferhat Abbas University of Setif
56 PUBLICATIONS 2,101 CITATIONS
SEE PROFILE
SEE PROFILE

M.F. Benatallah M. Aillerie

3 PUBLICATIONS 44 CITATIONS
University of Lorraine
396 PUBLICATIONS 3,701 CITATIONS
SEE PROFILE
SEE PROFILE

All content following this page was uploaded by Mohamed Nia on 19 August 2015.

The user has requested enhancement of the downloaded file.

 TerraGreen 13 International Conference 2013 - Advancements in Renewable Energy
and Clean Environment

Contribution to the quantification of solar radiation in Algeria

M. Nia1*, M. Chegaar1, M.F. Benatallah1 and M. Aillerie2,3,#
1
L.O.C.,Department of Physics, Faculty of Sciences, Ferhat Abbas University, 19000, Setif, Algeria
Email : mohamed_nia@yahoo.fr, chegaar@yahoo.fr
2
Lorraine University, LMOPS-EA 4423, 57070 Metz, France
3
Supelec, LMOPS, 57070 Metz, France
#
Email: aillerie@metz.supelec.fr

Abstract

For an energy production optimization with a photovoltaic global system, a fundamental need is the knowledge of the
global solar irradiation at different Algerian locations (Algiers, Oran, Bechar and Tamanrasset) using available
climatological measured data. Different expressions relating the global solar irradiation to the calculated
extraterrestrial global irradiation, measured sunshine duration and temperature at these locations are dealt with. These
include the well-known Angström-Prescott linear regression, the logarithmic and the exponential relationships.
Accordingly, several other models have also been tested to choose the more suitable for each location.
The present work applied these considerations in the objective to optimize the production efficiency of photovoltaic
energy using available data of the global solar irradiation. The modeled results are compared to the measured ones
using statistical parameters tests such as the mean bias error (MBE), the mean absolute error (MAE), the root mean
square error (RMSE) and the coefficient of determination (R2). The agreement between the measured and the
computed values is remarkable and the models are recommended to predict the mean monthly global solar irradiation
in Algeria and any location of the same climatic characteristics.

© 2013The
© 2013 TheAuthors.
Authors. Published
Published by Elsevi
by Elsevier Ltd.
Selection and/or
Selection and/or peer-review
peer-review underunder responsibility
responsibility of the TerraGreen
of the TerraGreen Academy Academy.

Keywords: Global irradiation; horizontal surface; temperature; relative humidity; sunshine duration.

1. Introduction

Solar energy is one of the most important energy sources. Knowledge of solar radiation distribution at
a particular site is essential in the design and study of many solar energy applications. Unfortunately, for

*
Corresponding author. E-mail address: mohamed_nia@yahoo.fr.

© 2013 The Authors. Published by Elsevier Ltd.

Selection and/or peer-review under responsibility of the TerraGreen Academy
many developing countries like Algeria, solar radiation measurements are not easily available due to the
cost, maintenance and calibration requirements of the measuring equipment. Therefore, it is very
important to elaborate methods based on quickly available meteorological data to estimate solar radiation.
Several empirical models have been developed to calculate global solar radiation using various
parameters, as extraterrestrial radiation, sunshine hours, albedo, temperature, relative humidity and total
precipitable water, number of rainy days, altitude and latitude.

Algeria is a high insolation country; the number of sunshine hours amounts is approximately 3300h
per year. The climate is largely favorable for solar energy exploitation but the distribution of the solar
radiation is not perfectly known.The main objective of this work is to compare the results obtained by
application of some models that predict the monthly average daily global radiation on a horizontal surface
versus measured data for different sites over Algeria; consequently, the most accurate models are
selected.

2- Models used

In the present work, data of monthly mean of daily global solar radiation and sunshine duration from
four Algerian meteorological stations (Algiers, Oran, Bechar and Tamanrasset) are used [1]. The
geographical locations of stations are presented in table 1. The duration of records of sunshine duration,
air temperature and relative humidity is 25 years and of global solar radiation is approximately 10 years.

Table1. Geographical locations of stations

Location Latitude (°) N Altitude (m) Longitude (°)
Algiers 36.43 25 3.15 E
Oran 35.38 99 0.37 W
Bechar 31.38 806 2.15 W
Tamanrasset 22.47 1378 5.31 E

2.1. Angstrom–Prescott model (model 1)

Angstrom–Prescott model [2, 3] is the most generally used model it correlates the clearness index
(G/G0) as.
      (1)
Where (G) is the monthly average daily global solar radiation, (G0) is the monthly average daily
extraterrestrial radiation, (S) is the sunshine duration, (S0) is the maximum possible sunshine duration, (a)
and (b) are empirical coefficients, (G0) was calculated from the following equation [4].
         (2)
Where (Ics) is the solar constant (1367Wm-2), (n) is the day number (starting from 1 January) and the
parameter (Z) is given by the following equation.

              (3)

(L) is the latitude of the site; ( ) is the sun declination and (w) is the hour angle.
2.2. Non-linear equation models
Polynomial types of non-linear models are proposed in solar energy literature [2, 3]. Most of these
non-linear solar irradiation estimation models are defined as the modifications of the Angström
expression.

2.2.1. Second order correlation (model 2)

The ratio of global solar radiation to extraterrestrial radiation (G/G0) was expressed by a function of
the ratio of sunshine duration (S/S0) [2, 3, 5-8] as follows.
         )2 (4)

2.2.2. Third order correlation (model 3)

The following third order Angström type correlation has been proposed by different authors [5, 6, 9]

         )2 3 (5)

2.2.3. Logarithmic and exponential models

The following correlations have been also proposed by different authors [5, 7]

2.2.3.1. Logarithmic model (model 4)

The ratio of global solar radiation to extraterrestrial radiation was proposed as a logarithmic function
of sunshine duration as follows.

       (6)

2.2.3.2. Exponential model (model 5)

In this model (G/G0) was correlated with (S/S0) in the form of an exponential function as.
       (7)

2.2.4. Other correlations

A number of nonlinear regression equations were developed to predict the relationship between global
solar radiation using combinations of several measured climatic parameters as the mean temperature, the
relative humidity and the relative sunshine duration [6, 10].

2.2.4.1. Correlation with sunshine duration and temperature (Model 6)

In this model a combination of sunshine duration and mean monthly temperature was used as follows.

          (8)

Where (T) is the mean monthly temperature and (Tmax) is the maximum of the mean monthly
temperature.
2.2.4.2. Correlation with sunshine duration and humidity (Model 7)
In this model a linear correlation combination of sunshine duration and mean relative humidity was
used as.

         (9)

Where (RH) is the mean relative humidity.

2.2.4.3. Correlation with sunshine duration, temperature and humidity (Model 8)

In this model a combination of three measured parameters, sunshine duration, mean monthly
temperature and mean relative humidity was used in a correlation as the following equation.

            (10)

a, b, c and d are site constants determined by means of a least squares method.

3- Application and discussion

The monthly averages data for the four stations processed for the correlations are presented in Table 2
and Table 3.

Table 2. Monthly average measured data for Algiers and Bechar

Algiers Bechar
Month G/G0 S/S0 T (°C) RH G/G0 S/S0 T(°C) RH
1 0.452 0.48 16.6 0.764 0.706 0.78 15.53 0.479
2 0.479 0.56 17.6 0.779 0.684 0.80 18.53 0.4
3 0.512 0.60 18.8 0.763 0.708 0.84 21.83 0.326
4 0.497 0.61 20.7 0.746 0.672 0.84 25.71 0.259
5 0.542 0.71 23.7 0.743 0.650 0.82 30.4 0.221
6 0.533 0.71 27.7 0.7 0.685 0.83 36.17 0.165
7 0.624 0.78 31.2 0.685 0.661 0.84 39.61 0.115
8 0.618 0.80 32.1 0.688 0.650 0.83 39.07 0.141
9 0.591 0.74 29.6 0.700 0.672 0.82 34.1 0.229
10 0.524 0.62 25.7 0.725 0.690 0.80 27.28 0.351
11 0.529 0.51 21.3 0.747 0.679 0.77 20.93 0.455
12 0.465 0.49 17.7 0.771 0.729 0.76 16.56 0.51

The accuracy of the considered models was tested by calculating the mean bias error (MBE), the mean
absolute error (MAE), the root mean square error (RMSE) and the coefficient of determination (R2).
The relative percentage error is defined as:

       (11)

Ycal and Ymes are the ith calculated and measured values of global solar radiation.
 Table 3. Monthly average measured data for Oran and Tamanrasset
Oran Tamanrasset
Month G/G0 S/S0 T °C RH G/G0 S/S0 T °C RH
1 0.517 0.53 16.6 0.780 0.712 0.78 19.88 0.245
2 0.547 0.53 17.65 0.772 0.751 0.81 22.06 0.205
3 0.628 0.64 19.56 0.739 0.665 0.84 25.12 0.205
4 0.593 0.63 21.3 0.691 0.747 0.77 29.78 0.165
5 0.577 0.69 24.24 0.684 0.655 0.76 33.39 0.164
6 0.649 0.69 27.34 0.666 0.653 0.69 35.8 0.165
7 0.654 0.80 30.31 0.671 0.664 0.75 35.42 0.175
8 0.649 0.79 31.58 0.662 0.679 0.77 34.97 0.213
9 0.652 0.72 28.3 0.700 0.683 0.73 33.31 0.215
10 0.565 0.66 24.47 0.719 0.719 0.77 29.75 0.227
11 0.519 0.53 20.36 0..745 0.708 0.81 24.8 0.238
12 0.483 0.53 17.43 0.787 0.662 0.79 21.13 0.253

Statistical parameters are given by the following relations.



   




   


  (12)

     


          
       with     

m is the number of measurements.

The following tables contain summaries of results, obtained from the application of all models to the
four stations. It is clear that statistical parameters, R2, MBE, MAE and RMSE vary from one model to
another. Generally, coefficients of determination (0.91-0.99) is high for all the models, in another hand
the pair (MBE, RMSE) vary respectively in the intervals [-0.04,-0.25] and [1.91, 4.88].This implies that,
there are statistical significant relationships between the clearness index, relative sunshine duration,
relative humidity and the monthly average daily temperature.

Table 4. Models’ constants for Algiers and Oran

Algiers Oran
a b c d a b c d
Model 1 0.244 0.451 0.239 0.538
Model 2 0.709 -1.052 1.181 -0.273 2.146 -1.235
Model 3 -1.459 9.506 -15.672 8.825 -1.061 5.806 -6.819 2.804
Model 4 0.661 0.277 0.743 0.349
Model 5 0.075 0.240 0.055 0.277
Model 6 0.282 0.127 0.227 0.225 0.614 -0.048
Model 7 0.916 0.256 -0.746 0.546 0.410 -0.312
Model 8 -0.526 0.082 0.431 0.936 0.606 0.532 -0.100 -0.402
 Table5. Models’ constants for Bechar and Tamanrasset
Bechar Tamanrasset
a b c d a b c d
Model 1 1.067 -0.475 0.457 0.303
Model 2 9.640 -21.857 13.315 -2.611 8.340 -5.248
Model 3 -119.431 461.466 -589.568 250.510 63.314 -251.46 335.178 -148.331
Model 4 0.601 -0.384 0.753 0.238
Model 5 1.158 -0.211 0.393 0.137
Model 6 0.788 -0.066 -0.075 0.713 0.056 -0.081
Model 7 0.343 0.344 0.197 0.458 0.300 0.008
Model 8 -0.216 0.827 0.122 0.472 0.813 0.040 -0.120 -0.272

Table 6. Statistical parameters MBE, MAE, RMSE and R 2 for Algiers and Bechar
Algiers Bechar
MBE (%) MAE (%) RMSE (%) R2 MBE (%) MAE (%) RMSE (%) R2
Model 1 -0.18 3.33 4.34 0.99 -0.08 2.20 2.75 0.99
Model 2 -0.16 3.28 3.92 0.99 -0.06 2.17 2.53 0.99
Model 3 -0.14 3.08 3.64 0.99 -0.06 2.19 2.49 0.99
Model 4 -0.21 3.74 4.63 0.98 -0.08 2.19 2.74 0.99
Model 5 -0.17 3.20 4.17 0.99 -0.08 2.21 2.76 0.98
Model 6 -0.10 2.56 3.30 0.99 -0.05 1.89 2.21 0.99
Model 7 -0.14 2.84 3.79 0.99 -0.04 1.67 2.03 0.99
Model 8 -0.09 2.43 3.10 0.99 -0.04 1.57 1.91 0.99

Table 7. Statistical parameters MBE, MAE, RMSE and R 2 for Oran and Tamanrasset
Oran Tamanrasset
MBE (%) MAE (%) RMSE (%) R2 MBE (%) MAE (%) RMSE (%) R2
Model 1 -0.24 4.18 4.75 0.98 -0.20 3.83 4.47 0.91
Model 2 -0.20 3.47 4.50 0.99 -0.18 3.62 4.23 0.92
Model 3 -0.20 3.45 4.50 0.99 -0.16 3.23 3.95 0.94
Model 4 -0.22 3.92 4.60 0.98 -0.20 3.81 4.45 0.91
Model 5 -0.25 4.36 4.88 0.98 -0.20 3.85 4.48 0.91
Model 6 -0.23 4.11 4.73 0.98 -0.19 3.36 4.29 0.92
Model 7 -0.22 3.93 4.59 0.98 -0.20 3.82 4.47 0.91
Model 8 -0.21 3.77 4.51 0.98 -0.18 3.65 4.25 0.92

Figures 1 and 2 illustrate the measured global solar irradiation values and the estimated ones for the two
sites of Oran and Tamanrasset using model 3 and for the sites of Algiers and Bechar using model 8.

Fig 1. Global solar irradiation measured and predicted with model 3 for Oran and Tamanrasset.
 Fig 2. Global solar irradiation measured and predicted with model 8for Algiers and Bechar.

It is observed that root mean square error for all the models ranges between 3.10% and 4.88% for the
considered sites. Good agreement between the predicted and the measured values of global solar
irradiation. The coefficient of determination is greater than 0.98 for Algiers, Oran and Bechar and not
less than 0.91 for Tamanrasset. The peak of solar radiation occurs in June and July for Algiers, Oran and
Bechar and in May to July for Tamanrasset.

4. Conclusion

The monthly mean daily global solar radiation, relative sunshine duration, monthly average daily
temperature and relative humidity have been used in this study to test several correlation equations. It was
observed that for Oran and Tamanrasset model 3 has the highest value of coefficient of determination
which gives good results when considering statistical parameters, MBE, MAE and RMSE.
For Algiers and Bechar the highest value of coefficient of determination was obtained by applying
model 8 that gives good statistical parameters. We judge that these two models could be employed in
estimating global solar radiation in different Algerian locations and for location with same climatic
characteristics.
References
[1] Capderou M. Atlas solaire de l'Algérie. Office des publications universitaires, T1-3, 1988.
[2] Zekai ¸Sen. Solar Energy Fundamentals and Modeling Techniques, Springer 2008.
[3] Viorel Badescu. Modeling solar radiation at the earth’s surface, Springer 2008.
[4] M. Chegaar, A. Chibani. A simple method for computing global solar radiation. Rev. Energ. Ren. Chemss (2000), 111-115.
[5] M. Salmi, M. Chegaar, P. Mialhe. Modèles d'estimation de l'irradiation solaire globale sur une surface horizontale au sol.
Revue internationale d'héliotechnique, énergie environnement n° 35 (2007), 19-24.
[6] Can Ertekin, Osman Yaldiz. Comparison of some existing models for estimating global solar radiation for Antalya (Turkey).
Energy Conversion and Management 41 (2000), 311-330.
[7] NA Elagib, SH Alvi, MG Mansell. Correlationships between clearness index and relative sunshine duration for Sudan.
Renewable Energy 17 (1999) 473-498.
[8] Augustine C, Nnabuchi M. N. Relationship between global solar radiation and sunshine hours for Calabar, Port Harcourt and
Enugu, Nigeria. International Journal of Physical Sciences Vol. 4 (4), pp. 182-188, April, 2009.
[9] Zekai Sen. Simple nonlinear solar irradiation estimation model. Renewable energy 32 (2007) 342–350.
 [10] Falayi E. O, Adepitan J. O, Rabiu A. B. Empirical models for the correlation of global solar radiation with meteorological
data for Iseyin, Nigeria. International Journal of Physical Sciences , 3 (9), (2008), 210-216.
[11] A. A. El-Sebaii; A. A. Trabea. Estimation of global solar radiation on horizontal surfaces over Egypt. Egypt. J. Solids, 28(1)
(2005),163-175.
[12] Joseph C. Lam, Kevin K.W. Wan, Chris C.S. Lau, Liu Yang. Climatic influences on solar modelling in China. Renewable
Energy, 33, (2008), 1591–1604.
[13] B C Cuamba, M L Chenene, G Mahumane, D Z Quissico, J Lovseth, P O’Keefe. A solar energy resources assessment in
Mozambique. Journal of Energy in Southern Africa , 17 (4), (2006),76-85.
[14] J. Mubiru, E.J.K.B. Banda. Estimation of monthly average daily global solar irradiation using artificial neural networks.
Solar Energy, 82 (2008), 181–187.
[15] M.Drif, M.Chikh. Estimation de l’irradiation solaire par la logique floue, Rev. Energ.
Ren. Chemss (2000), 105-110.
[16] Bent Sørensen. Renewable energy its physics, engineering, use, environmental impacts, economy and planning aspects. (3 rd
ed), Elsevier 2004.
[17] M.S. AI-Ayed, A.M. AI-Dhafiri, M.Bin Mahfoodh. Global, direct and diffuse solar irradiance in Riyadh, Saudi Arabia.
Renewable Energy, 14(.1-4), (1998), 249-254.
[18] Jose´ Leonaldo De Souza, Rosilene Mendonc¸ Nica´cio, Marcos Antonio Lima Moura. Global solar radiation measurements
in Maceio, Brazil, Renewable Energy 30, (2005) 1203–1220.
[19] Zhou Jin, Wu Yezheng, Yan Gang. Estimation of daily diffuse solar radiate in China. Renewable Energy, 29 (2004) 1537–
1548.
[20] Zekai Sen, Elcin Tan. Simple models of solar radiation data for northwestern part of Turkey. Energy Conversion and
Management, 42 (2001) 587-598.
[21] M. Tiris, C. Tiris. Analysis of solar radiation data for Gebze, Turkey. Energy Conversion and Management, 38(2), (1997)
179-186.
[22] K. Gairaa, S. Benkaciali. Modélisation numérique des irradiations globale et diffuse au site de Ghardaïa. Revue des Energies
Renouvelables, 11 (1), (2008), 129 –136.
[23] F. Ahmad, I. Ulfat. Empirical models for the correlation of monthly average daily global solar radiation with hours of
sunshine on a horizontal surface at Karachi, Pakistan. Turk J Phys, 28 (2004), 301-307.
[24] S.M. Robaa. Validation of the existing models for estimating global solar radiation over Egypt. Energy Conversion and
Management, 50 (2009), 184–193.
[25] K. Kaygusuz, T. Ayhan. Analysis of solar radiation data for Trabzon, Turkey. Energy Conversion and Management, 40
(1999), 545-556.
[26] M. Chegaar, A. Chibani. Global solar radiation estimation in Algeria. Energy conversion and management, 42; (2001) 967-
973.

View publication stats