PROGRESS IN PHOTOVOLTAICS: RESEARCH AND APPLICATIONS

Prog. Photovolt: Res. Appl. 2006; 14:553–566

Published online 16 May 2006 in Wiley InterScience (www.interscience.wiley.com). DOI: 10.1002/pip.692

Applications Evaluation of Solar Energy

Potential and PV Module
Performance in the Gobi
Desert of Mongoliaz
Amarbayar Adiyabat1*,y , Kosuke Kurokawa1, Kenji Otani2, Namjil Enebish3, Garmaa Batsukh3,
Mishiglunden Battushig3, Dorjsuren Ochirvaani3 and Bathuu Ganbat3
1
Tokyo University of Agriculture and Technology (TUAT), Naka-cho 2-24-16, Koganei, Tokyo 184-8588, Japan
2
National Institute of Advanced Industrial Science & Technology (AIST), Tsukuba Central 2, Tsukuba, Ibaraki 305-8568,
Japan
3
Inter-University Solar Energy Research Net-Center, National University of Mongolia, P.O. Box 46A/436, Ulaanbaatar
210546, Mongolia

Here, we present the results of evaluation of solar energy potential and photovoltaic
(PV) module performance from actual data measured over a period of more than
2 years in the Gobi Desert of Mongolia. To allow estimation of solar energy potentials
and durability of PV systems in the Gobi Desert area, a data acquisition system,
including crystalline silicon (c-Si), polycrystalline silicon (p-Si) modules, and two
sets of precision pyranometers, thermometers, and anemometer, was installed at
Sainshand City in October 2002. This system measures 23 parameters, including
solar irradiation and meteorological parameters, every 10 min. High output gain
was observed due to operation at extremely low ambient temperatures and the module
performance ratios (PRs) were high (>1
0) in winter. In summary, the present study
showed that a PV module with a high temperature coefficient, such as crystalline sili-
con, is advantageous for use in the Gobi Desert area. Copyright # 2006 John Wiley
& Sons, Ltd.

key words: PV module performance; solar energy potential; field tests; Gobi Desert

INTRODUCTION

T
he Gobi Desert, Mongolia, is one of the most promising candidate sites for introduction of the 100 MW
class very large scale photovoltaic systems (VLS-PV) specified by Task 8 ‘Very Large Scale Photovol-
taic Power Generation Systems’ conducted as part of the IEA photovoltaic power systems program
(IEA PVPS).1 Within the framework of the IEA PVPS Task 8 activity, a conceptual design has been developed

* Correspondence to: Amarbayar Adiyabat, Tokyo University of Agriculture and Technology (TUAT), Naka-cho 2-24-16, Koganei, Tokyo 184-
8588, Japan.
y
E-mail: amar@cc.tuat.ac.jp
z
Amarbayar Adiyabat and Namjil Enebish are Ph.D. Students; Kosuke Kurokawa and Garmaa Batsukh are Professors; Kenji Otani,
Mishiglunden Battushig, Dorjsuren Ochirvaani and Bathuu Ganbat are Reseachers.

Received 23 August 2005

Copyright # 2006 John Wiley & Sons, Ltd. Revised 12 November 2005
554 A. ADIYABAT ET AL.

and a trial calculation of the costs associated with power generation and construction of a VLS-PV system in the
Gobi Desert area has been performed.2,3
The meteorological environmental characteristics of the Gobi Desert may affect the PV system performance
and design specifications. Therefore, it is necessary to clarify the factors that will affect the system design,
operation, and maintenance. However, no useful reference data are available and there have been no case studies
analyzing solar energy resources or performance for PV system installation in the Gobi Desert area.
In the Gobi Desert, Mongolia, we set up two types of photovoltaic (PV) modules and checking devices (e.g.,
I–V curve tracer, etc.) as well as meteorological devices to study the characteristics of PV system operation
under such severe environmental conditions. The present study was performed to verify the output simulation
technique for the VLS-PV to confirm the efficiency of using a large-scale concentrated PV system in this area,
and to clarify the specific requirements for system design. To clarify the actual environmental capabilities (loss
analysis) in the Gobi Desert, we measured meteorological data, such as the amount of solar irradiation and
temperature, and the I–V characteristics of the PV modules.

EXPERIMENTAL SETUP
To determine the potential of VLS-PV in the Gobi Desert area, we installed two types of crystalline silicon PV
module and checking devices (e.g., I–V curve tracer, etc.) as well as a new data acquisition system at the field
site, Sainshand City (44 54’ N and 110 07’ E) (Figure 1), which is located in the southeastern part of Mongolia.
The data acquisition system (Figure 2) is switched on automatically every 10 min and records the total solar
irradiation received on the horizontal and 45 tilted surfaces, site meteorological data, and measures PV module
current-voltage (I–V) curves.
The measurement items are listed below:
1. Global irradiance on the horizon
2. In-plane irradiance at 45
3. Wind speed & direction
4. Air temperature
5. Humidity
6. Albedo
7. Short-circuit current (Isc)
8. Open-circuit voltage (Voc)

Figure 1. Overview of the experimental setup and horizontal pyranometer

Copyright # 2006 John Wiley & Sons, Ltd. Prog. Photovolt: Res. Appl. 2006; 14:553–566
EVALUATION OF SOLAR ENERGY POTENTIAL AND PV MODULE PERFORMANCE 555

Pyranometer
horizontal
PV
PV Pyranometer module-2
module-1 tilted
Vane anenometer
Wind speed, direction

Albedo meter
Storage
Data Logger Module 4M I-V Curve Tracer
Thermometer ch-2
PV-3 12V
Hydrometer
Relay COM
power COM controller
supply Thermo
couple Thermometer
PV module-1
DC-DC
I-V Curve Tracer 5V 12V
Charge Thermo
Thermometer 5V
ch-1 couple
controller PV module-2
DC-DC

Thermo 12V 12V

Battery couple
Thermometer
12V 12V PV module-3
12V

Figure 2. Schema of the data acquisition system

9. Current at maximum power (Ipm)

10. Voltage at maximum power (Vpm)
11. Temperature of modules (T1, T2)
The management system software package is installed in the data acquisition system, which is supplied by a
12 V battery at measurement time and data are loaded into the 4 MB storage module (SM4M).
The crystalline silicon (c-Si), polycrystalline silicon (p-Si) modules used in this exposure test, and the data-
sheet showing the electrical characteristics under standard test conditions (STC: 1000 W/m2, AM 1
5 and mod-
ule temperature of 25 C4,5) are shown in Table I. PV module 1 (80 W) uses multi-crystal silicon solar cells
measuring 125 mm2 with 12
6% module conversion efficiency. In addition, use of an anti-reflective coating
and back surface field (BSF) structure improved cell conversion efficiency to 14%. White tempered glass,
EVA resin, and a weatherproof film along with an aluminum frame were used. PV module 2 (75 W) was pro-
duced by a multistage proprietary texture optimized pyramidal surface (TOPS) process, which maximizes
photon absorption from direct and diffused light. Ultra-clear tempered glass, torsion, and a corrosion-resistant
anodized aluminum module frame were used.

Table I. Electrical characteristics of photovoltaic (PV) modules on STC (as manufacturer

datasheets)
PV module name and Type PV module-1 p-Si PV module-2 c-Si

Parameters Unit

Short-circuit current (Isc) A 5
3 4
8

Open-circuit voltage (Voc) V 21
3 21
7
Current at maximum power (Ipm) A 4
7 4
4
Voltage at maximum power (Vpm) V 17
1 17
0
Maximum power rating (Pmax) W 80 75
Temperature coefficient W/ C 0
373 0
321

Copyright # 2006 John Wiley & Sons, Ltd. Prog. Photovolt: Res. Appl. 2006; 14:553–566
EVALUATION OF SOLAR ENERGY POTENTIAL AND PV MODULE PERFORMANCE 565

Figure 13. Monthly average performance ratio, temperature, and other loss modification factors of PV module 2

The performance of module 2 was 0
75–0
93, which was 0
1 lower than that of module 1. In addition, the
value of Koth2 was lower than that of module 1. The nominal power values (according to the manufacturers)
were used for yield and performance calculations. However, the average measured power values, Pmax,real,
of modules 1 and 2 were below the respective nominal power values of 76
3 W (95
4%) and 68
6 W
(91
4%) within the allowed tolerance range  10% (JIS C8990). This difference (4
6%, 8
6%) was included
in the performance differences between module 1 and 2. In addition, module performance difference may have
been due to differences in the materials used for module construction: for example, differences in glass reflec-
tion coefficient, etc.
Sandstorms, dust, and snow coverage did not appear to have significant effects on module performance.

CONCLUSIONS

Accurate determination of the performance of PV modules is crucial for the design of VLS-PV systems and
their economic evaluation. To develop a conceptual design and calculation of the power generation and con-
struction costs for a VLS-PV system for the Gobi Desert, an autonomous PV module performance monitoring
and data acquisition system was developed. Outdoor performance tests of two types of PV module were con-
ducted in Sainshand City, Mongolia.
The results described here indicated high output gain due to the extremely low ambient temperature and the
module performance ratio showed very high values of >1
0 in winter. In summary, the results of the present
study show that PV modules with high temperature coefficients, such as crystalline silicon, are advantageous for
use in the Gobi Desert area. We are obtaining useful information for installation of the VLS-PV system in the
Gobi Desert area, and will continue our field study to further evaluate the long-term performance of PV modules
exposed to the severe environmental conditions in the Gobi Desert.

Acknowledgements
The study was carried out by the Japanese-Mongolian Research team within the cooperative framework
between the National Institute of Advanced Industrial Science & Technology, Tokyo University of Agriculture
& Technology, and Inter-University Solar Research Net-Center of the National University of Mongolia.

REFERENCES
1. IEA, Photovoltaic Power Systems Programme, Task 8: Very large scale photovoltaic power generation systems in remote
areas, http://www.oja-services.nl/iea-pvps/tasks/index.htm.

Copyright # 2006 John Wiley & Sons, Ltd. Prog. Photovolt: Res. Appl. 2006; 14:553–566