Outdoor Performance Characterization of Multi-Crystalline Silicon Solar Module
The performance parameters of solar modules are characterized
under Standard Test Conditions (STC). However, solar modules when deployed outdoor
results in the output that differs from its output obtained at STC, due to the
influence of various environmental factors, leading to confusion in deciding
the appropriate electrical components and loads used for Solar Photovoltaic
(SPV) power plant. In view of the above, Outdoor performance characterization
of the multi-crystalline silicon solar module was studied. Daily solar illumination
data were measured using an optical Pyranometer and analyzed together with the
module output power. Blow of air mass was found very active at the site. Stable
voltage output and dynamic current outputs were recorded. The average module
efficiency of ~10 and 9.5% were obtained for the high and low illumination days,
respectively. However, the integrated daily efficiency was found as 9.8% irrespective
of the variation in solar illumination. Percentage of variations between the
solar illumination and its associated current output under the blow of air mass
were not found matching. Higher blow of air mass and its associated reduction
in the solar insolation condition found yielding higher current output than
Received: May 09, 2012;
Accepted: August 07, 2012;
Published: September 08, 2012
Recently, Solar Photovoltaic (SPV) technology is gaining popularity due to
the reasons that the cost of the conventional source of energy and their demand
is rising (Dupraz et al., 2011; Panwar
et al., 2011). Also, SPV technology oriented electrical conversion,
handling and management technologies are well developed and the cost is becoming
affordable (Albuquerque et al., 2010; Patrao
et al., 2011; Vivar et al., 2012;
Sharaf et al., 2000; Valdes
et al., 2012; Menti et al., 2011;
Hassaine et al., 2009; Welch
and Venayagamoorthy, 2010; Parthasarathy et al.,
2011). Even then, further improvements are awaited in terms of SPV-power
system efficiencies. Currently, the system efficiencies are found lingering
between 9~12% (Eltawil and Zhao, 2010). Therefore, various
approaches were made to improve the system efficiencies, reliability in operation,
life time etc., (Patrao et al., 2011; Menti
et al., 2011; Chekired et al., 2011;
Firat and Beyene, 2012; Su et
al., 2012; Gautam and Kaushika, 2002; He
et al., 2012). However, so far, those approaches were found resulting
incomparable results with their laboratory predictions (Eltawil
and Zhao, 2010). The direct exposure of the solar photovoltaic modules to
the sunlight results in the outputs that differs from its outputs obtained under
standard test conditions (STC) (Kurnik et al., 2011).
In detail, the outdoor solar insolation flux is found dynamically varying with
the time as well as with the blow of air mass (Su et
al., 2012; Nakada et al., 2010; Minemoto
et al., 2009), the conversion efficiency is found variable with respect
to the illumination intensity (Reich et al., 2009;
Huld et al., 2010), the UV illumination present
in the outdoor spectrum were found imposing material degradations to the EVA
(Kempe, 2010), the environmental effects like: humidity,
stray light, UV lighting, temperature, etc., were found playing effective role
in the degradation of module interconnects, development of defects, etc., (Firat
and Beyene, 2012; Skoczek et al., 2009).
Hence, in total, the aforementioned factors were found resulting in performance
degradation in SPV systems (Huld et al., 2010;
Skoczek et al., 2009; Khan
et al., 2010; Stamenic et al., 2004).
It is reported that the 20 years of exposure to the outdoor sunlight leads to
performance losses of 1.9~2.8% on crystalline Silicon SPV systems and 0.7~1.4%
on the poly-crystalline SPV systems (Ishii et al.,
2011). Apart from degradations, it is impossible to quantify exactly the
output rate or even define the daily output capacity of the solar module and
as a consequence there are confusions over its utility. Particularly the confusion
mounts over the establishment of relation between the designed capacities to
the deliverable capacities. This anomaly is solely generated by the blow of
air mass which is different for different geographical locations based on its
climate induced compositions (Krishnan et al., 2009;
Akhmad et al., 1997). To address this problem,
direct outdoor monitoring, estimation of solar irradiance and clearness index,
direct measurements under outdoor conditions are proposed (Nakada
et al., 2010; Sanchez et al., 2011;
Katsumata et al., 2011; Minemoto
et al., 2007a). On the other hand, the cost of SPV system is estimated
from module efficiency, module cost, land cost, array construction and installation
cost as well as the site dependent solar insolation as discussed (Kamel
and Muhlbauer, 1984). Nevertheless, the payback period is only depending
on the electricity produced by the SPV power system (Garcia-Valverde
et al., 2009) and thus the payback period should be clearly determined
to promote the attraction towards SPV power plant business. Moreover, it is
also essential to have an idea about the output characteristics of the SPV power
plants in order to decide appropriate electrical components and loads.
Consequently, it is clear that the test results of solar modules obtained at
STC fails to reveal the realistic output capabilities and characteristics of
them when deployed outdoor. Therefore, widespread usage of the STC results for
the field performance assessment misleads the predictions of output power and
the output characteristics. In view of the above, this study was conducted to
analyze the outdoor performance characteristics of a solar module of 5 W in
power together with an installed optical Pyranometer. The solar spectral irradiance,
module output current, module output voltage, module output power and module
efficiency were analyzed for different outdoor illumination conditions.
MATERIALS AND METHODS
Outdoor performance measurement of SPV module was conducted at the Pondicherry
University campus, Puducherry, India (Latitude: 12.0107° and Longitude:
79.856°). Multi-crystalline silicon solar module of 5 W containing 18 cells
of 25 cm2 (12.5x2 cm2) in series (450 cm2),
processed using standard industrial module making process by Maharishi Solar
Technologies Pvt. Ltd., India was employed for this study. An optical Pyranometer
was installed with computer controlled data acquisition system at the measurement
site and used to collect the spectral irradiance data. The output characteristics
of the solar module tested at AM 1.5 STC is shown in the data sheet (Table
|| Multi crystalline silicon SPV module data sheet at STC
Measurements were carried out using high precision Agilent table-top multi-meters
for the entire duration of the days from Sun-rise to Sun-set. Among the measured
data, the data obtained for two different days: 15th of September 2011 (15DS11)
and 2nd of December 2011 (02DD11) and their corresponding solar illumination
data were selected for further analysis.
The important parameters that are needed to characterize the performance of
the solar module are solar irradiance (Pirr), open circuit voltage
(Voc), short circuit current (Isc), output power (Pod)
and efficiency (ηmod). The module output current, module output
voltage and solar spectral irradiance were directly measured and used further
to obtain the output power, efficiencies by following the standard procedures.
The output power (Pod) of a solar module can be obtained from their
measured Voc and Isc values by multiplying with its Fill-Factor
(FF) as reported (Maiti et al., 2010):
On the other hand, in general the power produced by a solar module (Pod)
is dependent of the active area of the module (Aact), module efficiency
(ηmod) and the solar irradiance (Pirr) as defined
(Kamel and Muhlbauer, 1984):
The outdoor efficiency was further calculated using the following relation:
where, the active area of the module (Aact) is 450x10-4
RESULTS AND DISCUSSION
Different days have different solar insolation profile. However, days that
received better average profile of solar insolation with high (15DS11) and low
(02DD11) illumination intensities were considered for analysis. Figure
1 shows the plot of global solar irradiance data obtained from Pyranometer
for entire duration of the days of 15DS11 and 02DD11 which reveal that the solar
irradiance is affected by the blow of air mass. The disturbances were found
both in the morning as well as in the afternoon hours. This may be related to
the blow of air mass from land-to-sea and sea-to-land driven by the temperature
differences. It can be observed that the peak irradiation intensity of 15DS11
is ~ 950 and 02DD11 is ~817 W m-2, respectively.
Figure 2 shows the plot of the recorded output current (Isc)
and output voltage (Voc) with respect to the time duration of the
days 15DS11 and 02DD11, respectively. It is worth to note that the profile of
the output current (Isc) is exactly falling with the profile of the
solar irradiance as presented in Fig. 1.
||Global solar spectral irradiance data for entire duration
of the days of 15DS11 and 02DD11
The increase and decrease of solar irradiance in morning and evening generates
proportional free electrons in the solar module that leads to almost linear
increase and decrease in the output current, respectively that may be identified
in the figure. While, the module output voltage was found getting steeply build
up in the morning and attains saturated value of ~9.3 V (15DS11) for a period
of duration of ~7.5 h and ~9.15 V (02DD11) for a period of duration of ~7.3
h, respectively, above 40% of illumination level with respect to STC value (1000
W m-2). The peak output currents of 0.71 A and 0.68 A were measured
for the peak illuminations of both the days of ~950 and ~ 817 W m-2,
respectively. The measured Isc values of both the days are falling
with 97.26 and 93.15%, respectively with respect to the STC value (0.73 A).
On the other hand, their respective solar irradiances were found as 95 and 81.7%,
respectively to the STC illumination. From this, it is noticeable that the relatively
lower percentage of light received under the influence of the blow of air mass
yields higher percentage of current output (02DD11) than its linear dependence.
This inconsistency in the correlation between the illumination and the current
yield reveals that the module receives more active photons during the heavy
float of air mass. This fact of the observation of higher current output at
lower illumination intensity is found supported by the report which stated the
presence of higher density blue photons in the outdoor light than in the STC
AM 1.5 spectrum (Katsumata et al., 2011; Minemoto
et al., 2007b).
In addition to the above, it can be sighted from Fig. 2 that
the output voltage builds up primarily before the current output at low illumination
||Module output voltage (Voc) and output current
(Isc) with respect to the time duration of the days 15DS11 and
As module position is fixed, it is not directly facing the sun and hence the
insolation falls with wide angle of incidence boosting the reflection losses.
As a consequence, the current generation may be dominantly influenced only by
the weak diffused light than the direct light. Hence, there may be a quasi-static
dynamism in the build-up of Voc.
This visible build-up of voltage with respect to the increase in illumination
level is also an indication of the presence high resistive material across the
junction. It was reported that the Voc of the n+p solar
cell rises slowly when there is high resistive material that elucidate difference
between the quasi-Fermi potentials of the two active space charge regions in
the solar cell (Fossum and Lindholm, 1977).
Meanwhile, it was also reported that the solar cells having high density of
trap states due to the multi-crystalline structure can result in an additional
voltage build-up at low illumination intensities (Khan et
al., 2010). Therefore we may state that the visible voltage build-up
at low illumination level may be due to the presence of non-uniform doping in
the multi-crystalline solar cells due to the impact of its grain boundaries
as discussed (Halder and Williams, 1983). This situation
of voltage build-up or voltage drop was found either in the early hours of the
day or in the late evening hours of the day, whenever the light injection levels
are falling less than 20% of its STC value as observable from the Fig.
From the Fig. 2, the saturated mean values of Voc
were found as 9.3 and 9.15 V for those days of 15DS11 and 02DD11, respectively.
This voltage saturation was found stable for the durations of 8:00 and 7:30
h, respectively. These differently observable values of saturated Voc
with respect to the solar illumination states that the Voc saturation
requires critical illumination intensity as stated earlier. Moreover, it was
also interestingly noted from the Fig. 2 that the timely short
fall of illumination after the establishment of saturated Voc did
not impose any major drop in Voc. On this basis, we may imply that
the dynamic fall in outdoor illumination intensity due to the dynamic movement
of air mass may not be effective to the cut-off voltage of inverter operation
in the case of SPV power system.
Figure 3 shows the plots of module output power (Pod)
with respect to the time duration of the days 15DS11 and 02DD11 obtained using
the Eq. 1. It was observed that the 5 W capacity module results
in the total output powers of 221.6 and 112.8 watts, respectively which are
equivalent to the 3.6936 and 1.88 W h day-1.
The comparative analysis of the Fig. 2 and 3,
further implies that there are two types of output performances exist from the
module deployed under outdoor illumination.
||Module output power (Pod) with respect to the time
duration of the days 15DS11 and 02DD11
They are: (1) the stabilized voltage output above certain illumination level
and (2) The fluctuating power output with respect to the solar illumination
of the entire days. It is reported that this dynamic change in output power
pumps in harmonic distortion in the power distribution system and depreciate
the power quality (Hengsritawat et al., 2012).
Thus, we can state that the observed air mass induced output fluctuations as
shown in Fig. 3 drives the development of harmonics and the
consequent power depreciation. Meanwhile, the shown stable output voltage in
Fig. 2 also indicates that the outdoor deployed SPV module
is able to provide the operational voltage stability index (Vsi)
for further system integration (Charkravorty and Das, 2001).
Figure 4a and b present the plot of efficiency
versus the time duration of the days 15DS11 and 02DD11, respectively. Figure
4a and b clearly indicate that the average efficiency
varies with the growing hours of the days. The average efficiency for the day
15DS11 is ~10% (Fig. 4a) with the variable values found between
the lowest of ~6.95% (early hours) to the highest value of ~13.47% (late hours)
with fluctuations. Similarly, the average efficiency of 02DD11 is ~9.5% (Fig.
4b) with minimum and maximum values of ~5.2% (early hours) to ~13% (late
hours), respectively. Overall, it may be noticed from Fig. 4a
and b that the module efficiency is increasing with the time
duration of the day. Normally, at an outdoor exposure of modules, the increase
in illumination intensity was found resulting an increase in module efficiency
(Firat and Beyene, 2012; Eikelboom
and Reinders, 1997). In the present case, the result is in analogous to
the reports until noon hours but differs for after the noon as shown in both
Fig. 4a and b.
|| (a) Efficiency vs. the time duration of the day 15DS11 and
(b) Efficiency vs. the time duration of the day 02DD11
This observation of dual behavior is different from the report (Khan
et al., 2010). In lieu of this, it is also reported that at the lower
illumination flux, a solar cell results in higher efficiency due to the presence
of higher shunt resistance (Sanchez et al., 2011).
Further, the daily performance of efficiency of the module was determined for
the high and low illumination days of 15DS11 and 02DD11. Overall module efficiency
of the day (ηday) was obtained from the ratio between the daily
module output powers to the illumination input as stated earlier in Eq.
3. The values are found as 9.84 and 9.81% for the days of 15DS11 and 02DD11,
respectively. Therefore, we can state that the solar module deployed under outdoor
exposure to the sun results in an identical integrated daily efficiency (ηday)
irrespective of the illumination intensity.
Beside the efficiency factor, the measurements revealed that the blow of air
mass disturbs the leniency of the module output profile. Moreover, the overall
increase or decrease in solar irradiance up-shift or down-shift the SPV output
profile. Hence, the dynamic variation in the solar illumination and its associated
up or down shift of the bell-shaped output profile of the SPV module inflict
complication on the output design and forces to work below the peak power specified
either from STC or peak of day illumination. The output power of the solar module
is found highly variable with time as well as with the blow of air mass. Hence,
the power rating also varies with the time and blow of air mass. Therefore,
it may be stated from the measurements and analysis that the specified operational
hours of SPV system highly depends on the output rating that is fixed. Lesser
fixed power rating leads to the longer the duration of operation.
In concise, a solar photovoltaic module with 5 W output capacity was deployed
for outdoor measurements simultaneously with an installed optical pyranometer.
The measurements and the analysis of the measured data revealed that the site
of this measurement has heavy blow of air mass that in turn found greatly affecting
the linearity of the output profile. However, the heavy blow of air mass found
increasing the percentage of performance than the expected linear performance
variation. The daily module efficiency was calculated as 9.8% irrespective of
the solar illumination level.
Outdoor performance characteristics of the multi-crystalline silicon solar
module was studied at the site of Pondicherry University, Puducherry, India
(Latitude: 12.0107° and Longitude: 79.856°) near the sea coast. The
site is found with heavy float of air mass that in-turn found hampering the
leniency of the output profile, increasing the percentage of current yield at
heavy float condition and so on. The time dependent efficiency of the SPV module
was found increasing with the time duration but the daily efficiency found remaining
identical at all illumination levels.
The authors wish to acknowledge the enthusiastic technical support provided
by Mr. N. Tamilselvan and Ms. B.S. Akila, research scholars of the Centre for
Green Energy Technology, Pondicherry University, in taking measurements.
Akhmad, K., A. Kitamura, F. Yamamoto, H. Okamoto, H. Takakura and Y. Hamakawa, 1997. Outdoor performance of amorphous silicon and polycrystalline silicon PV modules. Solar Energy Mater. Solar Cells, 46: 209-218.
Albuquerque, F.L., A.J. Moraes, G.C. Guimaraes, S.M.R. Sanhueza and A.R. Vaz, 2010. Photovoltaic solar system connected to the electric power grid operating as active power generator and reactive power compensator. Solar Energy, 84: 1310-1317.
Chakravorty, M. and D. Das, 2001. Voltage stability analysis of radial distribution networks. Int. J. Electr. Power Energy Syst., 23: 129-135.
Chekired, F., C. Larbes, D. Rekioua and F. Haddad, 2011. Implementation of a MPPT fuzzy controller for photovoltaic systems on FPGA circuit. Energy Procedia, 6: 541-549.
Dupraz, C., H. Marrou, G. Talbot, L. Dufour, A. Nogier and Y. Ferard, 2011. Combining solar photovoltaic panels and food crops for optimising land use: Towards new agrivoltaic schemes. Renewable Energy, 36: 2725-2732.
Eikelboom, J.A. and A.H.M.E. Reinders, 1997. Determination of the irradiation dependent efficiency of multicrystalline Si PV modules on basis of IV curve fitting and its influence on the annual performance. Proceedings of the 14th European PV Solar Energy Conference, June 30-July 4, 1997, Barcelona, Spain, pp: 293-296.
Eltawil, M.A. and Z. Zhao, 2010. Grid-connected photovoltaic power systems: Technical and potential problems-A review. Renewable Sustainable Energy Rev., 14: 112-129.
Firat, C. and A. Beyene, 2012. Comparison of direct and indirect PV power output using filters, lens and fiber transport. Energy, 41: 271-277.
Fossum, J.G. and F.A. Lindholm, 1977. The dependence of open-circuit voltage on illumination level in p-n junction solar cells. IEEE Transac. Elec. Devic., 24: 325-329.
Direct Link |
Garcia-Valverde, R., C. Miguel, R. Martinez-Bejar and A. Urbina, 2009. Life cycle assessment study of a 4.2 kWp stand-alone photovoltaic system. Solar Energy, 83: 1434-1445.
CrossRef | Direct Link |
Gautam, N.K. and N.D. Kaushika, 2002. Reliability evaluation of solar photovoltaic arrays. Solar Energy, 72: 129-141.
Direct Link |
Halder, N.C. and T.R. Williams, 1983. Grain boundary effects in polycrystalline silicon solar cells II: Numerical calculation of the limiting parameters and maximum efficiency. Solar cells, 8: 225-238.
Hassaine, L., E. Olias, J. Quintero and M. Haddadi, 2009. Digital power factor control and reactive power regulation for grid-connected photovoltaic inverter. Renewable Energy, 34: 315-321.
He, F., Z. Zhao and L. Yuan, 2012. Impact of inverter configuration on energy cost of grid-connected photovoltaic systems. Renewable Energy, 41: 328-335.
Hengsritawat, V., T. Tayjasanant and N. Nimpitiwan, 2012. Optimal sizing of photovoltaic distributed generators in a distribution system with consideration of solar radiation and harmonic distortion. Elect. Power Energy Sys., 39: 36-47.
Huld, T., R. Gottschalg, H.G. Beyer and M. Topic, 2010. Mapping the performance of PV modules, effects of module type and data averaging. Solar Energy, 84: 324-338.
Ishii, T., T. Takashima and K. Otani, 2011. Long-term performance degradation of various kindsof photovoltaic modules under moderate climatic conditions. Prog. Photovoltaics: Res. Applic., 19: 170-179.
Kamel, F.M. and A. Muhlbauer, 1984. Impact of solar cell efficiency on the design of photovoltaic systems. Solar Cells, 11: 269-280.
Katsumata, N., Y. Nakada, T. Minemoto and H. Takakura, 2011. Estimation of irradiance and outdoor performance of photovoltaic modules by meteorological data. Solar Energy Mater. Solar Cells, 95: 199-202.
Kempe, M.D., 2010. Ultraviolet light test and evaluation methods for encapsulants of photovoltaic modules. Solar Energy Mater. Solar Cells, 94: 246-253.
Khan, F., S.N. Singh and M. Husain, 2010. Effect of illumination intensity on cell parameters of a silicon solar cell. Solar Energy Mater. Solar Cells, 94: 1473-1476.
Krishnan, P., J.W.A. Schuttauf, C.H.M. van der Werf, B.H. Hassanzadeh, W.G.J.H.M. van Sark and R.E.I. Schropp, 2009. Response to simulated typical daily outdoor irradiation conditions of thin-film silicon-based triple-band-gap, triple-junction solar cells. Solar Energy Mater. Solar Cells, 93: 691-697.
Kurnik, J., M. Jankove, K. Brecl and M. Topic, 2011. Outdoor testing of PV module temperature and performance under different mounting and operational conditions. Solar Energy Mater. Solar Cells, 95: 373-376.
Maiti, S., K. Vyas and P.K. Ghosh, 2010. Performance of a silicon photovoltaic module under enhanced illumination and selective filtration of incoming radiation with simultaneous cooling. Solar Energy, 84: 1439-1444.
Menti, A., T. Zacharias and J. Milias-Argitis, 2011. Harmonic distortion assessment for a single-phase grid-connected photovoltaic system. Renewable Energy, 36: 360-368.
Minemoto, T., M. Toda, S. Nagae, M. Gotoh and A. Nakajima et al., 2007. Effect of spectral irradiance distribution on the outdoor performance of amorphous Si//thin-film crystalline Si stacked photovoltaic modules. Solar Energy Mater. Solar Cells, 91: 120-122.
Minemoto, T., S. Fukushige and H. Takakura, 2009. Difference in the outdoor performance of bulk and thin-film silicon-based photovoltaic modules. Solar Energy Mater. Solar Cells, 93: 1062-1065.
Minemoto, T., S. Nagae and H. Takakura, 2007. Impact of spectral irradiance distribution and temperature on the outdoor performance of amorphous Si photovoltaic modules. Solar Energy Mat. Solar Cells, 91: 919-923.
Nakada, Y., H. Takahashi, K. Ichida, T. Minemoto and H. Takakura, 2010. Influence of clearness index and air mass on sunlight and outdoor performance of photovoltaic modules. Curr. Applied Phys., 10: S261-S264.
Panwar, N.L., S.C. Kaushik and S. Kothari, 2011. Role of renewable energy sources in environmental protection: A review. Renew. Sustainable Energy Rev., 15: 1513-1524.
Parthasarathy, S., P. Thilakan, P. Neelamegam and R.N. Srinath, 2011. Studies on the assembly induced performance losses in Multi-crystalline silicon solar PV module. Eur. J. Sci. Res., 62: 248-256.
Patrao, I., E. Figueres, F. Gonzalez-Espin and G. Garcera, 2011. Transformerless topologies for grid-connected single-phase photovoltaic inverters. Renewable Sustainable Energy Rev., 15: 3423-3431.
Reich, N.H., W.G.J.H.M. van Sark, E.A. Alsema, R.W. Lof, R.E.I. Schropp, W.C. Sinke and W.C. Turkenburg, 2009. Crystalline silicon cell performance at low light intensities. Solar Energy Mater. Solar Cells, 93: 1471-1481.
Sanchez, E., J. Torreblanca, T. Carballo, V. Parra and J. Bullon et al., 2011. Outdoor monitoring of the energy yield and electrical parameters of standard polysilicon based and new umg-Si PV modules. Energy Procedia, 8: 503-508.
Sharaf, A.M., M.M. AboulNaga and R. El Diasty, 2000. Building-integrated solar photovoltaic systems-a hybrid solar cooled ventilation technique for hot climate applications. Renewable Energy, 19: 91-96.
Skoczek, A., T. Sample and E.D. Dunlop, 2009. The results of performance measurements of field-aged crystalline silicon photovoltaic modules. Prog. Photovoltaics: Res. Applic., 17: 227-240.
Direct Link |
Stamenic, L., E. Smiley and K. Karim, 2004. Low light conditions modelling for Building Integrated Photovoltaic (BIPV) systems. Solar Energy, 77: 37-45.
Direct Link |
Su, Y., L.C. Chan, L. Shu and K.L. Tsui, 2012. Real-time prediction models for output power and efficiency of grid-connected solar photovoltaic systems. Applied Energy, 93: 319-326.
Valdes, R., L.R. Rodriguez and J.H. Lucio, 2012. Procedure for optimal design of hydrogen production plants with reserve storage and a stand-alone photovoltaic power system. Int. J. Hydrogen Energy, 37: 4018-4025.
Vivar, M., M. Fuentes, N. Dodd, J. Scott, I. Skryabin and K. Srithar, 2012. First lab-scale experimental results from a hybrid solar water purification and photovoltaic system. Solar Energy Mater. Solar Cells, 98: 260-266.
Welch, R.L. and G.K. Venayagamoorthy, 2010. Energy dispatch fuzzy controller for a grid-independent photovoltaic system. Energy Convers. Manage., 51: 928-937.