Available online at www.sciencedirect.com

www.elsevier.com/locate/solener

ScienceDirect

Solar Energy 103 (2014) 70–77

Degradation evaluation of crystalline-silicon photovoltaic modulesafter a few operation years in a tropical environment

Ababacar Ndiaye a,⇑, Cheikh M.F. Kebe a, Abderafi Charki b, Papa A. Ndiaye a,Vincent Sambou a, Abdessamad Kobi b

a Centre International de Formation et de Recherche en Energie Solaire (CIFRES), Ecole Superieure Polytechnique-UCAD, BP 5085 Dakar-Fann, Senegalb University of Angers-ISTIA-LASQUO, 62 Avenue Notre Dame du Lac 49000 Angers, France

Received 27 May 2013; received in revised form 27 January 2014; accepted 1 February 2014Available online 28 February 2014

Communicated by: Associate Editor Arturo Morales-Acevedo

Abstract

This paper presents an evaluation of the performance degradation of Photovoltaic modules after few operation years in a tropicalenvironment. To this end, the International Center for Research and Training in solar energy at Dakar University and the Lasquo-ISTIAlaboratory of Angers University have put in place a research project in order to investigate the impact of the tropical climatic conditionson the PV modules characteristics. Accordingly, two monocrystalline-silicon (mc-Si) PV modules and two polycrystalline- silicon (pc-Si)PV modules are installed at Dakar in Senegal and monitored during a few operation years: Module A (16 months), Module B (41months), Module C (48 months) and Module D (48 months). After few operation years under tropical environment, the global degra-dation and the degradation rate of electrical characteristics such as I-V and P-V curves, open-circuit voltage (Voc), short-circuit current(Isc), maximum ouput current (Imax), maximum output voltage (Vmax), maximum power output (Pmax) and fill factor (FF) are evaluate atstandard test conditions (STC). This study reports on data collected from 4 distinct mono- and poly-crystalline modules deployed atDakar University in Senegal. The study has shown that Pmax, Imax, Isc and FF are the most degraded performance characteristics forall PV modules. The maximum power output (Pmax) presents the highest loss that can be from 0.22%/year to 2.96%/year. However,the open-circuit voltage (Voc) is not degraded after these few exposition years for all studied PV modules.� 2014 Elsevier Ltd. All rights reserved.

Keywords: Photovoltaic module; Degradation; Tropical environment

1. Introduction

The performance of PV modules varies according to theclimatic conditions and gradually deteriorates through theyears (Adelstein and Sekulic, 2005; Cereghetti et al., 2003;Dunlop and Halton, 2005; Osterwald et al., 2006;Sanchez-Friera et al., 2011; Som and Al-Alawi, 1992). An

http://dx.doi.org/10.1016/j.solener.2014.02.006

0038-092X/� 2014 Elsevier Ltd. All rights reserved.

⇑ Corresponding author. Tel.: +221 77 654 63 93; fax: +221 33 823 5574.

E-mail address: ababacar.ndiay@gmail.com (A. Ndiaye).

important factor in the performance of PV technologieshas always been their long-term reliability especially forthe new emerging technologies. The most important issuein long-term performance assessments is degradation.Degradation is the outcome of a power or performance lossprogression dependent on a number of factors such asdegradation at the cell, module or even system level. Inalmost all cases, the main environmental factors relatedto known degradation mechanisms include temperature,humidity, water ingress and ultraviolet (UV) intensity.All these factors impose significant stress, over the lifetime

Fig. 1. Photovoltaic test field at Dakar in Senegal.

A. Ndiaye et al. / Solar Energy 103 (2014) 70–77 71

of a PV device and as a result detailed understanding of therelation between external factors, stability issues andmodule degradation is necessary. In general, degradationmechanisms describe the effects from both physicalmechanisms and chemical reactions and can occur at bothPV cell, module and system level. More specifically, thedegradation mechanisms at the cell level include gradualperformance loss due to ageing of the material and lossof adhesion of the contacts or corrosion, which is usuallythe result of water vapor ingress. Other degradation mech-anisms include metal mitigation through the p–n junctionand antireflection coating deterioration. All the abovementioned degradation mechanisms have been obtainedfrom previous experience on c-Si technologies (Dunlopand Halton, 2005; Quintana et al., 2002; Som and Al-Ala-wi, 1992). In the case of a-Si cells an important degradationmechanism occurs when this technology is first exposed tosunlight as the power stabilizes at a level that is approxi-mately 70–80% of the initial power. This degradationmechanism is known as the Staebler–Wronski effect (Stae-bler and Wronski, 1977) and is attributed to recombina-tion-induced breaking of weak Si–Si bonds by opticallyexcited carriers after thermalization, thus producing-defects that decrease carrier lifetime (Stutzmann et al.,1985). At the module level, degradation occurs due to fail-ure mechanisms of the cell and as a result of degradation ofthe packaging materials, interconnects, cell cracking, man-ufacturing defects, bypass diode failures, encapsulant fail-ures and delamination (Munoz et al., 2011; Wenhamet al., 2007). Degradation investigations using indoormethodologies are based on the acquisition of I–V curvesand power at STC (Ndiaye et al., 2013a). The electricalcharacteristics of PV modules are initially measured atSTC and then the modules are either exposed outdoorsor indoors through accelerated procedures (Carr andPryor, 2004; Meyer and Van Dyk, 2004; Osterwald et al.,2002). For each investigated PV cell or module the electri-cal characteristics are regularly acquired using the solarsimulator and the current, voltage or power differencesfrom the initial value that provides indications of thedegradation rates at successive time periods.

A wide variety of degradation rates has been reported inthe literature with respect to technologies, age, manufactur-ers, and geographic locations, and has been recently sum-marized (Jordan and Kurtz, 2011). Significant variation inthe data can be caused by different module types, age, con-struction (encapsulation, front- and back-sheet), electricalset-up (open-circuit, short-circuit, load resistor, grid-tied),and measurement uncertainty (Skoczek et al., 2009). The lit-erature contains an excellent review of long-term field test-ing based on discreet IV measurements (Sanchez-Frieraet al., 2011), but fewer reports include more comprehensiveI–V parameters investigation, including voltage and currentat maximum power point (Chamberlin et al., 2011; Granataet al., 2009; Reis et al., 2002).

In this paper we present the degradation evaluation ofelectrical characteristics of crystalline-silicon PV modules

such as I–V and P–V curves, open-circuit voltage (Voc),short-circuit current (Isc), maximum ouput current (Imax),maximum output voltage (Vmax), maximum power output(Pmax) and fill factor (FF) in a tropical environment aftera few operation years on four crystalline silicon (mc-Siand pc-Si) PV modules located at Dakar University inSenegal.

2. Experimental platform presentation

2.1. Photovoltaic test field

The photovoltaic platform shown in Fig. 1 is used in thisstudy. It is installed at Dakar in Senegal. Senegal is locatedon the extreme western Africa between 12.5� and 16.5�North latitude and 12� and 17� West longitude. It presentsa dry tropical climate characterized by two seasons: a dryseason from November to June and a rainy season fromJuly to October (ANAMS, 2012). Senegal has a significantsolar potential with annual average radiation duration ofabout 3000 h and an exposure rate of 5.7 kWh/m2/d. Thisradiation varies between the northern part which is moresunlit (5.8 kWh/m2/d in Dakar) and the southern part,the richest in terms of precipitation (4.3 kWh/m2/d inZiguinchor) (PSA, 2011). The temperature varies from16 �C around Dakar (January) to 38 �C in the South (Octo-ber). The rainfall increases from North to South with anannual average of 300 mm in the extreme North and1500 mm in the extreme South (ANAMS, 2012). The aver-age relative humidity varies between 75% and 95%(Wofrance, 2012). The platform is installed in Dakarbetween 17.28� West longitude and 14.43� North latitudeto 31 m altitude.

Platform consists of two monocrystalline (A and C) andtwo polycrystalline (B and D) photovoltaic modules. Thetechnical characteristics of PV modules provided by themanufacturers are given in Table 1. The modules haveoperated during a few years: Module A (16 months),Module B (41 months), Module C (48 months) andModule D (48 months). Thus, performance parameters(I–V and P–V curves, Voc, Isc, FF and Pmax) are measuredunder the standard test conditions (AM1.5, 1000 W/m2,

Table 1Technical characteristics of PV modules.

Modules Technology Manufacturer Reference Parameters Value

Module A Monocrystalline silicon Bosch SP36-145M Maximum output power (Pmax) 145 WMaximum output voltage (Vmax) 17.9 VMaximum output current (Imax) 8.1 AOpen circuit voltage (Voc) 22.7 VShort-circuit current (Isc) 8.5 AFill factor (FF) 75.14%

Module B Polycrystalline silicon Aleo S18-230 Maximum output power (Pmax) 230 WMaximum output voltage (Vmax) 29.2 VMaximum output current (Imax) 7.88 AOpen circuit voltage (Voc) 36.6 VShort-circuit current (Isc) 8.44 AFill factor (FF) 74.48%

Module C Monocrystalline silicon Waaree WS-110 Maximum output power (Pmax) 107 WMaximum output voltage (Vmax) 16.6 VMaximum output current (Imax) 6.5 AOpen circuit voltage (Voc) 21 VShort-circuit current (Isc) 7.5 AFill factor (FF) 68.00%

Module D Polycrystalline silicon BP Solar BP-110 Maximum output power (Pmax) 114 WMaximum output voltage (Vmax) 17.5 VMaximum output current (Imax) 6.5 AOpen circuit voltage (Voc) 21.8 VShort-circuit current (Isc) 7.5 AFill factor (FF) 70.00%

72 A. Ndiaye et al. / Solar Energy 103 (2014) 70–77

25 �C). These measurements are made with the test instru-mentation of photovoltaic modules known as “IV 400”

(IV-400, 2012).

2.2. Photovoltaic module analyzer “I–V 400”

“I–V 400” carries out the field measurement of the I–V

characteristic and of the main characteristic both of asingle module and of module strings. The instrumentationmeasures, together with the I–V characteristic of the devicebeing tested, also the values of its temperature and incidentirradiation. The acquired data are then processed toextrapolate the I–V characteristic at standard test condi-tions (STC) in order to proceed with the comparison withthe nominal data declared by the modules manufacturer,thus immediately determining whether or not the stringor the module being tested respects the characteristicsdeclared by the manufacturer. Output current or voltagefrom the module is measured with the 4-terminal method,which allows extending the measurement cables withoutrequiring any compensation for their resistance, thusalways providing accurate measures. Measurement of out-put voltage from module is up to 1000 V DC. Measure-ment of output current from module is up to 10 A DC.Measurement of solar irradiation (W/m2) is carries outwith reference cell. Measurement of output DC and nomi-nal power from module is performed. Numerical andgraphical display of I–V curve is available. The module fillfactor is also measured. Mechanical inclinometer is inte-grated for the detection of the incidence angle of solar irra-diation. Electrical specifications of photovoltaic moduleanalyzer are given in Table 2.

Each parameter is calculated as ±[% reading + (numberof dgts) � resolution] at 23 �C ± 5 �C,<80%HR (DatasheetI-V 400, 2012).

3. Electrical characteristics and PV module degradations

The most important electrical characteristics of a PVmodule are the I–V and P–V curves, short-circuit currentIsc, open-circuit voltage Voc, the fill factor FF and themaximum power output Pmax. They are defined andmodeled as follows.

3.1. The I–V and P–V curves

The I–V (current–voltage) curve of a PV string (or mod-ule) describes its energy conversion capability at the exist-ing conditions of irradiance (light level) and temperature.Conceptually, the curve represents the combinations ofcurrent and voltage at which the string could be operatedor ‘loaded’, if the irradiance and cell temperature couldbe held constant. Fig. 2 shows a typical I–V curve, thepower-voltage or P–V curve that is computed from it,and key points on these curves. Referring to Fig. 2, thespan of the I–V curve ranges from the short circuit current(Isc) at zero volts, to zero current at the open circuit voltage(Voc). At the ‘knee’ of a normal I–V curve is the maximumpower point (Imp, Vmp), the point at which the arraygenerates maximum electrical power. In an operating PVsystem, one of the jobs of the inverter is to constantlyadjust the load, seeking out the particular point on theI–V curve at which the array as a whole yields the greatestDC power. At voltages well below Vmp, the flow of

Table 2Electrical specifications of “I–V 400”.

Parameters Range Accuracy Resolution

Voltage (Vdc) 5.0–999.9 0.1 ±(1.0%rdg + 2dgt)Current (Idc) 0.10–10.00 0.01 ±(1.0%rdg + 2dgt)Power maximal (Wdc) 50–9999 1 ±(1.0%rdg + 6dgt)Irradiation (mVdc) 1.0–100.0 0.1 ±(1.0%rdg + 5dgt)Temperature (�C) �20.0 to 100.0 0.1 ±(1.0%rdg + 1 �C)

A. Ndiaye et al. / Solar Energy 103 (2014) 70–77 73

solar-generated electrical charge to the external load isrelatively independent of output voltage. Near the kneeof the curve, this behavior starts to change. As the voltageincreases further, an increasing percentage of the chargesrecombine within the solar cells rather than flowing outthrough the load. At Voc, all of the charges recombineinternally. The maximum power point, located at the kneeof the curve, is the (I,V) point at which the product ofcurrent and voltage reaches its maximum value.

3.2. The short-circuit current (Isc)

At normal levels of solar irradiance, the short-circuitcurrent can be considered equivalent to the photocurrentIph, i.e. proportional to the solar irradiance G (W/m2).But this may result in some deviation from the experimen-tal result, so a power law having exponent a is introducedin this paper to account for the non-linear effect that thephotocurrent depends on. The short-circuit current Isc ofthe PV modules is not strongly temperature dependent. Ittends to increase slightly with increase of the module tem-perature. For the purposes of PV module performance,modeling this variation can be considered negligible. Then,the short-circuit current Isc can be simply calculated by(Ndiaye et al., 2013b):

I sc ¼ I sc0

GG0

� �a

ð1Þ

where Isc0 is the short-circuit current of the PV module un-der the standard solar irradiance G0; while Isc is the short-circuit current of the PV module under the solar irradianceG; a is the exponent responsible for all the non-linear effectsthat the photocurrent depends on.

Fig. 2. I–V and P–V curves of a photovoltaic module.

3.3. The open-circuit voltage

The relationship of the open-circuit voltage to irradianceis known to follow a logarithmic function based on an idealdiode equation, and the effect of temperature is due to theexponential increase in the saturation current with anincrease in temperature (Luis and Silvestre, 2002). Thisconclusion causes some difficulties in replicating theobserved behavior of the tested PV modules. Additionalterms or some amendatory parameters must be introducedto account for the shunt resistance, series resistance and thenon-ideality of the diode. Based on the model given by(Van Dyk et al., 2002) and then take into account the effectof temperature, the open-circuit voltage Voc at any givenconditions can be expressed by (Ndiaye et al., 2013b):

V oc ¼V oc0

1þ b � ln G0

G

� � T 0

T

� �c

ð2Þ

where Voc and V oc0are the open-circuit voltage of the PV

module under the normal solar irradiance G and thestandard solar irradiance G0; b is a PV module technologyspecific related dimensionless coefficient (Van Dyk et al.,2002); and c is the exponent considering all the non-lineartemperature–voltage effects.

T and T0 represent the PV module temperature underthe normal solar irradiance G and the standard solarirradiance G0.

3.4. The maximum power output

The photovoltaic module performance is highly affectedby the solar irradiance and the PV module temperature. Inthis paper, we consider a simplified model maximumpower-output of PV module to estimate its performance(Ould Bilal et al., 2012). It is given by (Eq. (3)).

P max ¼ V oc � ISC � FF ð3Þ

where Isc and Voc are respectively the short-circuit currentand open-circuit voltage of solar photovoltaic module(Omer, 2008), FF (dimensionless) is the fill factor whichis defined below. It is the ratio between the nominal andmaximum power standard (Koutroulis et al., 2006).

3.5. The fill factor

The fill factor (FF) of a PV module or string is animportant performance indicator. It represents thesquare-ness (or ‘rectangularity’) of the I–V curve, and isthe ratio of two areas defined by the I–V curve, as illus-trated in Fig. 3. Although physically unrealizable, an idealPV module technology would produce a perfectly rectan-gular I–V curve in which the maximum power point coin-cided with (Isc, Voc), for a fill factor of 1. The fill factor isimportant because if the I–V curves of two individual PVmodules have the same values of Isc and Voc, the array withthe higher fill factor (squarer I–V curve) will produce more

Fig. 3. The Fill Factor, defined as the gray area divided by the cross-hatched area.

74 A. Ndiaye et al. / Solar Energy 103 (2014) 70–77

power. Also, any impairment that reduces the fill factorwill reduce the output power. The fill factor can beexpressed by (Ndiaye et al., 2013b):

FF ¼ Imp � V mp

I sc � V oc

ð4Þ

where Imp and V mp are respectively the maximum currentand maximum voltage of solar photovoltaic module.

3.6. Degradation determination

In our study, global degradation (GD) and degradationrate (DR) of photovoltaic module performance characteris-tics are determinated. The methodology used in this studyis shown in Fig. 4.

ROC: Real Operating Conditions.STC: Standard Test Conditions.The global degradation of a PV module parameter

expresses the degradation of parameter considered fromthe first putting into service of the PV module to the expe-rience date and can be expressed by:

GDð%Þ ¼ X ðtnÞ � X ðt0ÞX ðt0Þ

� 100 ð5Þ

where X(tn) and X(t0) represent the value of the parameterconsidered in the STC conditions respectively at time tn

and t0; t0 represents the initial time corresponding to the

Fig. 4. Methodology for determin

first putting into service of PV modules and tn is the instantof carrying out the tests.

Thus, the degradation rate (DR) of photovoltaic moduleis given by the following (Eq. (6)).

DRð%Þ ¼ GD

Dtð6Þ

Dt (years) is the exposure duration in the field of the PVmodules from their initial operation until of the tests.

Table 3 summarizes the global degradation (GD) andthe degradation rate (DR) for the different performanceparameters of PV module (Pmax, Vmax, Imax, Voc, Isc andFF) for each of the four PV modules studied in this paper.

4. Results and discussion

The following section presents and discusses the param-eters measurements at standard test conditions (STC) ofthe different photovoltaic modules studied (I–V and P–Vcurves, open-circuit voltage (Voc), short-circuit current(Isc), maximum output current (Imax), maximum outputvoltage (Vmax), maximum power output (Pmax) and fill fac-tor (FF)) after few operation year under tropicalenvironment.

4.1. Degradation of I–V and P–V curves

Fig. 5 shows a comparison of I–V and P–V curves of thefour PV modules studied between the initial state (firstputting under exposition) in black and after a few yearsof operation in tropical environment in red.

The measurements are performed under STC conditions(standard test conditions). One can notice that the IV andPV curves are consistent with that of a PV module duringnormal operation (Munoz et al., 2011). However, we canalready note a change after a few years of exposure.Indeed, Isc, Imax, Vmax and Pmax (red curves) are shiftedrelative to the initial values (black curves). This shiftreflects a degradation (decrease) of the parametersconcerned. This degradation noted is much higher formodules C (mc-Si) and D (pc-Si) that have been longer

ing PV modules degradation.

Table 3A summary of degradation parameters of crystalline-silicon PV modules under tropical environment.

Modules Technology Yearsfielded

Parameters Initialvalue

Clean moduleafter a fewexposition year

Differenceabsolute

Globaldegradation(%)

Degradationrate (%)

A Monocrystalline 1.3 Pmax (W) 145 144,59 �0,41 �0.28% �0.22%Vmax (V) 17.9 17.83 �0.07 �0.39% �0.30%Imax (A) 8.1 8.06 �0.04 �0.49% �0.38%Voc (V) 22.7 22.71 0.01 0.04% 0.03%Isc (A) 8.5 8.47 �0.03 �0.35% �0.27%FF (%) 75.14 74.64 �1.50 �2.00% �0.51%

B Polycrystalline 3.4 Pmax (W) 230 217.37 �12.63 �5.49% �1.62%Vmax (V) 29.2 28.04 �1.16 �3.97% �1.17%Imax (A) 7.88 7.75 �0.13 �1.65% �0.49%Voc (V) 36.6 36.56 �0.04 �0.11% �0.03%Isc (A) 8.44 8.33 �0.11 �1.30% �0.38%FF (%) 74.48 72.09 �2.39 �3.21% �0.94%

C Monocrystalline 4 Pmax (W) 107 94.19 �12.81 �11.97% �2.99%Vmax (V) 16.6 15.14 �1.46 �8.80% �2.20%Imax (A) 6.5 6.22 �0.28 �4.31% �1.08%Voc (V) 21 21.01 0.01 0.05% 0.01%Isc (A) 7.5 7.14 �0.36 �4.80% �1.20%FF (%) 68.00 62.77 �5.23 �7.69% �1.92%

D Polycrystalline 4 Pmax (W) 114 100.50 �13.50 �11.84% �2.96%Vmax (V) 17.5 16.46 �1.04 �5.94% �1.49%Imax (A) 6.5 6.11 �0.39 �6.00% �1.50%Voc (V) 21.8 21.81 0.01 0.05% 0.01%Isc (A) 7.5 7.27 �0.23 �3.07% �0.77%FF (%) 70.00 63.38 �6.62 �9.46% -2.36%

A. Ndiaye et al. / Solar Energy 103 (2014) 70–77 75

exposed for 4 years. The following section presents the glo-bal degradation (GD) and the degradation rate (DR) forthe different performance parameters of PV module (Pmax,Vmax, Imax, Voc, Isc and FF) for each of the four PVmodules.

Module A (mc-Si, 1.3 years fielded)

Module C (mc-Si, 4 years fielded)

Fig. 5. Comparison between I–V and P–V curves befo

4.2. Degradation of Pmax, Vmax, Imax, Voc, Isc and FF

Degradation parameters were determined from (Eq. (5))for the global degradation and from (Eq. (6)) for the deg-radation rate. The results for the different performance

Module B (pc-Si, 3.4 years fielded)

Module D (pc-Si, 4 years fielded)

re and after a few exposition years of PV modules.

Fig. 6. Degradation rates for individual performance parameters of thefour PV modules.

76 A. Ndiaye et al. / Solar Energy 103 (2014) 70–77

parameters of PV modules (Pmax, Vmax, Imax, Voc, Isc andFF) are summarized in Table 3. All tests are performedin the standard test conditions (STC) corresponding toAM 1.5, 25 �C and 1000 W/m2. In Fig. 6 is shown acomparison of the degradation rate of the four PV modulesfor all parameters studied. Each module is represented by adifferent symbol. Monocrystalline-Silicon (mc-Si) are in redwhile polycrystalline-Silicon (pc-Si) modules are in blue. Anegative value implies decreased performance with time. Isnoted that the open-circuit voltage (Voc) has a zero degra-dation rate for all PV modules. The modules A (mc-Si) andB (pc-Si) have lowest degradation rates for all parameters.Indeed, they have the lowest operating time (1.3 and3.4 years respectively). The modules C (mc-Si) and D(pc-Si) have the highest degradation rates for all parame-ters; they have a longer operating time (4 years). However,the PV modules with the larger Pmax degradation are char-acterized by large contributions due to FF degradation.Module A (mc-Si) has the smallest degradation rate in Pmax

of less than 0.5%/year. All other PV modules have degra-dation higher than 1.5%/year with Module B (1.62%/year),modules C and D (2.96%/year). After four operation yearsthe two modules with different technologies C (mc-Si) andD (pc-Si) present the same degradation rate of 2.96%/year.Nonetheless the others performance parameters such asVmax, Imax, Isc and FF, degradation rates are differentand show no correlation one technology to another.Indeed, the degradation is higher in Vmax and Isc for themodule C (mc-Si) while Imax and FF are more degradedwith module D (pc-Si).

However, there are only a few long-term studies on thedegradation of PV modules published. (Skoczek et al.,

2009) have measured the performance of 204 field-agedcrystalline Si based PV modules (53 module types). Expos-ing started in 1983 at the Joint Research Center in northItaly with a moderate subtropical climate (�10–35 �C, >90% RH). They find that these high performance losses(>20%) are related to losses of the fill factor, caused by anincreased series resistance. The moderate performancelosses (<20%) can be related to losses of the Isc, caused bydegradation of the optical properties. The long term lossesare determined to be between 0.2% and 1.0% per annum.

For comparison, a recent study has shown that on aver-age the historically reported degradation rates of differentPV technologies was 0.7%/year while the reported medianwas 0.5%/year (Jordan et al., 2010). More specifically,investigations performed on outdoor exposed mono andmulti-c-Si PV modules showed performance losses ofapproximately 0.7%/year (Osterwald et al., 2002). Mostof these studies were conducted in Florida, USA.

However, our studies in sub-Saharan Africa on monoand multicrystalline PV modules show degradation ratesranging from 0.3%/year to 3%/year. The reason for thislarge difference is the different environmental conditionssuch as temperature, humidity, UV radiation and dust(Ndiaye et al., 2013c).

All data will be implemented in PVMODRELTM soft-ware developed by LASQUO laboratory of University ofAngers in order to evaluate the lifetime of PV modules(Charki et al., 2012) and (Laronde et al., 2012).

5. Conclusion

The performance degradation of Photovoltaic modulesafter few operation years under tropical environment wasstudied in this paper. Four PV modules (monocrystalline-silicon and polycrystalline-silicon) were exposed during afew years on the site of Dakar University in Senegal. Thedegradation impact on the I–V and P–V characteristicsof PV modules after a few exposition years under tropicalenvironment was highlighted. In the study, it was alsoshown that the relative differences of PV module perfor-mance parameters between before the first putting serviceof PV modules and after a few exposition year under trop-ical environment. Pmax, Imax, Isc and FF are the mostdegraded performance characteristics for all PV modules.The maximum power output (Pmax) presents the highestloss that can be from 0.22%/year to 2.96%/year. However,the open-circuit voltage (Voc) is not degraded after thesefew exposition years for all studied PV modules. In per-spective, we will work with many more PV modules fromdifferent manufacturers over longer exposition durationunder tropical environment. This project also aims to eval-uate the impact of dust deposited on the PV modules ofvarious technologies in order to investigate on the bestadapted technologies to tropical environment.

This study shows that, in sub-Saharan Africa, silicon PVmodules present mean degradation rates ranging from0.3%/year to 3%/year.

A. Ndiaye et al. / Solar Energy 103 (2014) 70–77 77

This work provides new information on the degradationissue of studied solar panel in sub-Saharan Africa area.Indeed, we have not identified such a study in this regiondespite the fact that the region has one of the best solarpotential in the world with 3000 h of sunshine per year.

In future works, we will study the major causes and fac-tors of degradations noted in this paper. It is necessary toincrease the number of modules studied over longer expo-sure duration in our study area.

Acknowledgments

The authors would like to thank the French cooperationin Senegal which allowed the purchase of equipment usedin this work through the U3E project.

References

Adelstein, J., Sekulic, B., 2005. Performance and reliability of a 1-kWamorphous silicon photovoltaic roofing system. In: Proceedings of the31st IEEE Photovoltaics Specialists Conference. Lake Buena Vista,USA, pp. 1627-1630 (ISBN 0-7803-8707-4)

ANAMS: National Agency for Meteorology of Senegal, 2012. <http://www.meteo-senegal.net>.

Carr, A.J., Pryor, T.L., 2004. A comparison of the performance ofdifferent PV module types in temperate climates. Solar Energy 76 (1–3), 285–294.

Cereghetti, N., Bura, E., Chianese, D., Friesen, G., Realini, A., Rezzo-nico, S., 2003. Power and energy production of pv modules statisticalconsiderations of 10 years activity. In: Proceedings of the 3rd WorldConference on Photovoltaic Energy Conversion. Osaka, Japan, pp.1919–1922 (ISBN 4-9901816-0-3).

Chamberlin, C.E., Rocheleau, M.A., Marshall, M.W., Reis, A.M.,Coleman, N.T., Lehman, P.A., 2011. Comparison of PV moduleperformance before and after 11 and 20 years of field exposure. In:IEEE PV Specialists Conference, Seattle, WA, USA.

Charki, A., Laronde, R., Bigaud, D., 2012. The time-variant degradationof a photovoltaic system. J. Solar Energy Eng. 135 (2). http://dx.doi.org/10.1115/1.4007771.

Datasheet I-V 400, 2012. <www.ht-instruments.com/pdf_I-V400>.Dunlop, E.D, Halton, D., 2005. The performance of crystalline silicon

photovoltaic solar modules after 22 years of continuous outdoorexposure. Prog. Photovoltaics: Res. Appl. 14 (1), 53–64.

Granata, J.E., Boyson, W.E., Kratochvil, J.A., Quintana, M.A., 2009.Long-term performance and reliability assessment of 8 PV arrays atSandia national laboratories. In: Proceedings of the 34th IEEE PVSpecialists Conference, Philadelphia, PA, USA, pp. 1486–1491.

IV-400 Datasheet, 2012. <www.ht-instruments.com/pdf/IV-400>.Jordan, D.C., Kurtz, S.R., 2011. Photovoltaic degradation rates – an

analytical review. Prog. Photovoltaics: Res. Appl. (80).Jordan, D.C., Smith, R.M., Osterwald, C.R., Gelak, E., Kurtz, S.R., 2010.

Outdoor PV degradation comparison. In: Proceedings of the 35thIEEE Photovoltaics Specialists Conference. Honolulu, USA, pp. 2694–2697 (ISBN 978-1-4244-5890-5).

Koutroulis, E., Kolokotsa, D., Potirakis, A., Kalaitzakis, K., 2006.Methodology for optimal sizing of stand-alone photovoltaic/windgenerator systems using genetic algorithms. Solar Energy 80, 1072–1088.

Laronde, R., Charki, A., Bigaud, D., 2012. Lifetime estimation of aphotovoltaic module subjected to corrosion due to damp heat testing.J. Solar Energy Eng. 135 (2). http://dx.doi.org/10.1115/1.4023101.

Luis, C., Silvestre, S., 2002. Modelling photovoltaic systems using PSpice.John Wiley & Sons Ltd., Chichester.

Meyer, E.L., Van Dyk, E.E., 2004. Assessing the Reliability andDegradation of Photovoltaic Module Performance Parameters. Trans.Reliability 53(1), 2264–2274.

Munoz, M.A., Alonso-Garcia, M.C., Nieves, V., Chenlo, F., 2011. Earlydegradation of silicon PV modules and guaranty conditions. SolarEnergy 85, 2264–2274.

Ndiaye, A., Charki, A., Kobi, A., Kebe, C.M.F., Ndiaye, P.A., Sambou,V., 2013a. Degradations of silicon photovoltaic A literature review.Solar Energy 96, 140–151. http://dx.doi.org/10.1016/j.solener.2013.07.005.

Ndiaye, A., Kebe, C.M.F., Ndiaye, P.A., Charki, A., Kobi, A., Sambou,V., 2013b. A novel method for investigating photovoltaic moduledegradation. Energy Proc. 36, 1222–1231. http://dx.doi.org/10.1016/j.egypro.2013.07.138.

Ndiaye, A., Kebe, C.M.F., Ndiaye, P.A., Charki, A., Kobi, A., Sambou,V., 2013c. Impact of dust on the PV modules characteristics after anexposition year in Sahelian environment: the case of Senegal. Int. J.Phys. Sci. 8 (21), 1166–1173. http://dx.doi.org/10.5897/IJPS2013.3921.

Omer, A.M., 2008. On the wind energy resources of Sudan. Renew.Sustain. Energy Rev. 12, 2117–2139.

Osterwald, C.R., Adelstein, J., Del Cueto, J.A., Kroposki, B., Trudell, D.,Moriarty, T., 2006. Comparison of degradation rates of individualmodules held at maximum power. In: Proceedings of the 4th IEEEWorld Conference on Photovoltaic Energy Conversion, Waikoloa,USA, pp. 2085–2088 (ISBN 1-4244-0017-1).

Osterwald, C.R., Anderberg, A., Rummel, S., Ottoson, L., 2002. Degra-dation analysis of weathered crystalline-silicon PV modules. In:Proceedings of the 29th IEEE Photovoltaic Specialists Conference,New Orleans, USA, pp. 1392–1395 (ISBN 0-7803-7471-1).

Ould Bilal, B., Sambou, V., Kebe, C.M.F., Ndiaye, P.A., Ndongo, M.,2012. Methodology to size an optimal stand-alone PV/wind/diesel/battery system minimizing the levelized cost of energy and the CO2emissions. Energy Proc. 14, 1636–1647.

Quintana, M.A., King, D.L., McMahon, T.J., Osterwald, R.C., 2002.Commonly observed degradation in field-aged photovoltaic modules.In: Proceedings of the 29th IEEE Photovoltaic Specialists Conference,New Orleans, USA, pp. 1436–1439 (ISBN 0-7803-7471-1).

Reis, A.M., Coleman, N.T., Marshall, M.W., Lehman, P.A., Chamberlin,C.E., 2002. Comparison of PV module performance before and after11-years of field exposure. In: IEEE PV Specialists Conference, NewOrleans, LA, USA, pp. 1432–1435.

Sanchez-Friera, P., Piliougine, M., Pelaez, J., Carretero, J., Sidrach deCardona, M., 2011. Analysis of degradation mechanisms of crystallinesilicon PV modules after 12 years of operation in Southern Europe.Prog. Photovoltaics: Res. Appl.. http://dx.doi.org/10.1002/pip.1083.

Skoczek, A., Sample, T., Dunlop, E.D., 2009. The results of performancemeasurements of field-aged crystalline silicon photovoltaic modules.Prog. Photovoltaics: Res. Appl. 17, 227–240.

Som, A.K., Al-Alawi, S.M., 1992. Evaluation of efficiency and degrada-tion of mono and polycrystalline PV modules under outdoor condi-tions. Renew. Energy 2 (1), 85–91.

Staebler, D.L., Wronski, C.R., 1977. Reversible conductivity charges indischarge-produced amorphous Si. Appl. Phys. Lett. 31 (4), 292–294.

Stutzmann, M., Jackson, W.B., Tsai, C.C., 1985. Light-induced metasta-ble defects in hydrogenated amorphous silicon: a systematic study.Phys. Rev. B 32 (1), 23–47.

Van Dyk, E.E. et al., 2002. Long-term monitoring of photovoltaic devices.Renew. Energy 22, 183–197.

Wenham, S.R., Green, M.A., Watt, M.E., 2007. Applied Photovoltaics,second ed. Earthscan, London, UK (ISBN 978-1-84407-401-3).