You are on page 1of 12

Solar Energy 177 (2019) 471–482

Contents lists available at ScienceDirect

Solar Energy
journal homepage: www.elsevier.com/locate/solener

Electrical performance of bifacial silicon PV modules under different indoor T


mounting configurations affecting the rear reflected irradiance

Juan Lopez-Garcia , Alberto Casado, Tony Sample
European Commission, Joint Research Centre (JRC), Ispra, Italy

A R T I C LE I N FO A B S T R A C T

Keywords: Bifacial photovoltaic (PV) modules can increase the performance with respect to traditional PV modules because
Bifacial both sides of the cells, front and rear, absorb solar radiation. To assess their performance and quality, PV
PV performance modules are characterized using international standards. However, currently only a draft IEC technical speci-
Silicon modules fication exists for bifacial PV modules and research needs to be done in order to study the indoor performance
testing conditions. One of the issues that need to be addressed is how to measure bifacial PV modules correctly
and analyse the different testing approaches proposed. This work outlines the indoor performance testing of c-Si
bifacial modules under different module mounting setups including open rack, a structure with baffles and 3
modules. For each mounting method a white reflective rear panel of several dimensions was placed at various
distances behind the module as a potential approach for a double-sided illumination characterization method.
Electrical performance is also studied with a single-side illumination method with a black rear panel. The rear
irradiance measurements and non-uniformity are also studied and the performance measurements are validated
with a single-side illumination method. Additional rear irradiance allows Pmax increment up to 20% under
certain conditions. However, the rear irradiance non-uniformity needs to be improved in order to fulfil the
current requirements of the draft technical specification.

1. Introduction include in the datasheet projected specifications including rear side


irradiation contribution in Isc as a percent of STC (10%, 20%, 30%, etc).
In order to reduce the cost of solar power generation, the main focus However, it is not indicated under which conditions or environments
areas of the photovoltaic (PV) industry are improving the solar cells and these increases can be achieved. The incorrect or excessively con-
module efficiency, lifetime, cost and PV module’s energy yield. Bifacial servative evaluation of the output power can also lead to under-
PV modules can produce additional output energy in comparison to estimation of the total output power and potentially produce a safety
conventional mono-facial modules because both sides of the cell/ problem in the system.
module, front and rear, can absorb solar radiation, utilizing the scat- In absence of standards, one of the main issues that need to be
tered light from the ground and surroundings. To assess their perfor- addressed is how to measure bifacial PV modules correctly. Different
mance and quality, PV modules are measured and characterized in a testing methods have been proposed for indoor (Guerrero-Lemus et al.,
controlled environment under standard test conditions (STC) as defined 2016; Roest et al., 2017; Lopez-Garcia et al., 2016; Newman et al.,
by the International Electrotechnical Commission (IEC) (IEC, 2016a). 2017; Schmid et al., 2017; Razongles and Sicot, 2016) and outdoor
Although the IEC is currently working on the development of a draft (under natural sunlight) (Deline et al., 2017; Pyrot et al., 2017) mea-
technical specification (IEC, Draft) for the measurement of current- surement of bifacial PV devices (cells and modules) as shown in Fig. 1.
voltage characteristics of bifacial photovoltaic devices, there is cur- Two main approaches for indoor module testing and the outdoor ap-
rently no published standard for bifacial modules. As the PV modules proach under natural sunlight are considered in the draft technical
are generally sold according to the nominal power output under STC, specification. The first indoor approach is based on the independent
the lack of bifacial standards leads each manufacturer to claim different measurement of both sides at STC (Fig. 1 a1 and a2) by means of a
amounts for the “added value” of the bifaciality that makes it difficult single-side illumination or single source solar simulator with adjustable
for customers to directly compare between bifacial manufacturers. In irradiance levels and this is currently the most used method for char-
addition to the STC rated power, most of the bifacial manufacturers acterisation of bifacial PV modules (Singh et al., 2014; Comparotto


Corresponding author.
E-mail address: juan.lopez-garcia@ec.europa.eu (J. Lopez-Garcia).

https://doi.org/10.1016/j.solener.2018.11.051
Received 9 July 2018; Received in revised form 31 October 2018; Accepted 20 November 2018
Available online 24 November 2018
0038-092X/ © 2018 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY license
(http://creativecommons.org/licenses/BY/4.0/).
J. Lopez-Garcia et al. Solar Energy 177 (2019) 471–482

Fig. 1. Schematic representation of indoor (a1, a2, a3 and b) and outdoor (e) single-side and indoor (c and d) and outdoor (f) double-sided illumination methods.

et al., 2014). In this setup, the non-illuminated side of the module is surfaces in order to change the albedo from the ground included in the
covered by a non-reflective or black absorbing material in order to draft technical specification IEC TS 60904-1-2 (Fig. 1f) (IEC, Draft;
reduce the reflected light. The bifaciality factors are calculated and then Deline et al., 2017). Basically, in the latter method, besides the STC
the measurement at equivalent irradiance levels can be performed as measurement at 1000 W/m2 (GR = 0 W/m2), AM1.5G and 25 °C, Pmax
defined in the IEC TS 60904-1-2 (Fig. 1a3). This method, the equivalent of the module shall be measured at 1000 W/m2 ± 10% on the front
irradiance method, is described in detail in Section 3.2. The second side (or corrected to this value), plus different rear side irradiance levels
approach is based on the simultaneous illumination of both sides of the GRi (i = 1, 2 ,3…, for instance, GR1 < 100 W/m2, 100 W/
bifacial device with 1000 W/m2 on the front and at least two different m2 < GR2 < 200 W/m2 and Gr3 > 200 W/m2). The rear irradiance
rear side irradiance levels. Different setups have been considered in- driven power gain yield, BiFi, is the slope derived from the linear fit of
cluding the measurement with a double-source solar simulator the Pmax versus GR data series. This linear least squares fit must be
(Fig. 1b), tilted mirrors (Fig. 1c) or by using a diffuse reflector placed at forced to cross the Pmax axis at Pmax STC. Besides BiFi, two specific Pmax
a specific distance behind the module with a known reflectivity (for values must be reported, PmaxBiFi10 and Pmax BiFi20, for GR1 = 100 W/m2
example, a reflective white rear sheet as shown in Fig. 1d). The setup and GR2 = 200 W/m2, respectively, and then by linear interpolation of
consisting of a double-source solar simulator is considered in the draft the data series Pmax versus Grear the following equations are obtained:
technical specification IEC TS 60904-1-2 as a suitable method for
PmaxBiFi10 = PmaxSTC + BiFi × 100 (1)
double-sided illumination. However, this approach presents some pro-
blems at the module scale such as the logistics of timing two flashes, PmaxBiFi20 = PmaxSTC + BiFi × 200 (2)
controlling the reflection of the environment and the added cost of
using two controlled light sources instead of one (Roest et al., 2017; This rear irradiance driven power gain yield, BiFi, calculation and
Newman et al., 2017). An alternative setup allows for a simultaneous approach is also applied for the indoor double-sided illumination
measurement of both sides with a single flash from the front and in- method. Since the paper is focused on indoor testing conditions, the
cludes two mirrors (typically aluminium or silver coated reflector) with outdoor approach also mentioned in the IEC TS 60904-1-2 will be not
metallic grid filters to attenuate the incident light on the backside of the taking into account for further discussion.
module (Schmid et al., 2017; Razongles and Sicot, 2016; Soria et al., This work presents the indoor performance testing of crystalline
2016). The third setup consisting on a reflective rear surface parallel to silicon bifacial PV modules with a white reflective rear panel at various
the module has been recently considered and studied by several groups distances behind the module and with different dimensions in order to
due to its potential simulation of real outdoor direct light, albedo and evaluate the potential of this characterization method.
self-shading (Newman et al., 2017; Zhang et al., 2018; Singh et al.,
2015). However, this setup shows several problems such as rear irra- 2. Experimental details
diance non-uniformity, specifications of the reflector material and
placement which would result in a difficult implementation. In a recent 2.1. Module characteristics
survey performed by Newman et al. among 13 research institutes, in-
dustry and manufacturers (Newman et al., 2017), most of the partici- There are currently several manufacturers selling bifacial silicon PV
pants scored this setup highly as a possible way to clearly mimic real- modules with different characteristics, dimensions, number of cells and
world conditions. Two outdoor approaches have been proposed con- designs (half cells modules, frameless and framed modules, types of
sisting on an outdoor single side illumination with equivalent irra- junction boxes, etc). The different module designs affect the transparent
diance levels measurements (Fig. 1e) (Pyrot et al., 2017) similar to the open space between cells that influence the light transmitted through
indoor method and a double-side illumination with reflective cloth or the module. It has been reported that the rear irradiance changes

472
J. Lopez-Garcia et al. Solar Energy 177 (2019) 471–482

significantly with the transparent space between cells at low ground


clearance (Deline et al., 2017). However, after analyse the dimensions
and geometry of bifacial PV modules from 16 manufacturers, it was
observed that more than 80% of the modules showed a transparent
open area below 14% that has a small impact on the rear irradiance
according to (Deline et al., 2017). This work is based on a re-
presentative type of bifacial module commercially available on the
market but this also imply some limitations on the results that need to
be taken into account by the reader to properly interpret the conclusion
presented here.
For this work, two representative commercially available n-type
bifacial c-Si glass/glass frameless modules from a single manufacturer
were characterized indoor at the European Solar Test Installation
(ESTI). The modules consisted of 60 cells (15.76 × 15.76 cm size) in
three sub-strings with a good module design without any partial “self-
shading” on the rear side due to the junction box, lips of the frame or
label. The total module area is 1.63 m2 and the effective area of the
module (without the edges of the module inserted into the mounting
structure) is 1.53 m2. The free area to allow light to pass through is
0.101 m2 (2 mm gaps between cells and border areas close to the
junction boxes) which represents the 6.6% of the effective module area.
The modules were light soaked following the standard IEC 61215-2
(IEC, 2016b) to eliminate possible light induced degradation (LID) prior
to measurements.

2.2. Measurement set-up

2.2.1. Initial set-up: single-side illumination


Fig. 3. Open rack setup for the measurement of bifacial PV modules and a white
According to the draft IEC TS 60904-1-2, the contribution from the reflective rear panel.
light incident on the opposite side of the device under test must be
eliminated completely during the measurement by creating a non-ir-
radiated background. The background is considered to be non-irra- 2.2.2. Open rack
diated if the irradiance does not exceed 3 W/m2, at any point, on the In order to ensure the reproducibility of the measurements, a special
non-exposed side of the device. A previous electrical characterization of structure was designed to hold the modules (Fig. 3). The structure al-
the modules was carried out with an opaque or non-reflective painted lows to set a gloss white wooden reflective rear panel (optical re-
wooden panel with irradiance below 3 W/m2 in order to avoid the il- flectance > 94% in the wavelength range 300–1200 nm) at various
lumination of the non-exposed side. An average optical reflectance of distances behind the module plane and parallel to it. The reflective rear
the black panel in the wavelength range corresponding to the spectral panel with area, A2, has different sizes as a multiple of the module area,
responsivity of the devices under test below 5% (Fig. 2) was measured A1, and it was placed parallel and centred. For this single module or
by mean of a Perkin–Elmer LAMBDA 9 spectrophotometer at room stand-alone type configuration, the light can be transmitted through the
temperature. In this setup, the rear panel (A2) had the same dimensions module by the spacing between the cells and through the non-covered
as the module (A1) and was placed in direct contact to the non-illu- areas around the module and then scattered or reflected back from the
minated surface of the module, thus only light that was transmitted back panel to the module contributing to the rear irradiance. According
through the module could be scattered or reflected back to the module. to the draft technical specification IEC TS 60904-1-2, the rear irra-
I-V curves have been obtained by illuminating alternatively the front or diance was measured with a calibrated secondary reference cell in at
the rear side while covering the rear with the different panels. least 9 positions and the average rear irradiance was obtained from at
least 5 symmetrical distributed points (IEC, Draft).

2.2.3. Structure with baffles


According to the IEC TS 60904-1-2 draft technical specification, it is
highly recommended to limit the size of the test area to the one of the
device under test using baffles. A scheme of the PV module setup with
baffles and non-reflective background material is shown in Fig. 4a.
Black opaque wooden panels are set around the module to block the
undesired incident light and only the light transmitted through the
module by the gaps around the cells will be reflected by the white rear
panel and contribute to the rear irradiance, as it is shown in Fig. 4b for
the setup used. Similar to the open rack configuration, the reflective
rear panel was placed behind the module at different distances and the
rear irradiance was measured with a calibrated secondary reference cell
of 2x2 cm2.

2.2.4. “Array simulation”: Three modules configuration


The last set-up consists of three modules placed in the same plane
Fig. 2. Optical reflectance (%) of a black and a white reflective rear panels. with the device under test (DUT in Fig. 5) in the centre and the other

473
J. Lopez-Garcia et al. Solar Energy 177 (2019) 471–482

configuration” can be considered as an intermediate case between the


open rack and the structure with baffles since the incident light can pass
through the spacing between the cells of the three modules and is re-
flected by the reflective rear panel. Even if under indoor conditions
using a solar simulator most of light is direct light, whereas under
outdoor conditions there is a variable diffuse component, this config-
uration tries to simulate more realistic conditions in terms of light
transmission than those described in the draft technical specification.

2.3. I-V measurement methods

I-V curve measurements were performed with a PASAN IIIB pulsed


solar simulator (with a nominal sweep time of 10 ms) at ∼1000 W/m2,
25.0 ± 0.1 °C and corrected to STC (1000 W/m2, 25 °C and AM1.5G).
In addition to the standard forward sweep single flash method, the so-
called multi-flash (MF) method was also used as described in (Lopez-
Garcia et al., 2016). Basically, 35–40 individual points are taken to
construct the final I-V curve. In this case, each point of the I-V curve was
obtained by applying a constant voltage to the module which was then
subjected to a flash from the solar simulator at the desired irradiance
level. The resultant module I-V curve was then obtained by performing
Fig. 4. (a) Scheme of a bifacial PV module setup with baffles and (b) real image
an interpolation of the measured points (Ferretti et al., 2013; Virtuani
of the structure with baffles used in the lab.
et al., 2012). With this method, capacitive effects of these high-effi-
ciency and highly capacitive modules are avoided by considering only
stabilized current and voltage output values for each measurement
point.
The spectral responsivity of the front and rear side of the bifacial
modules was measured using a differential spectral responsivity tech-
nique with a large area pulsed solar simulator equipped with a number
of bandpass filters to obtain illumination of modules with quasi-
monochromatic light (Van Steenwinkel, 1986) and the spectral re-
sponses were used to calculate the Mismatch factor (MMF) corrections.
No significant differences between front and rear spectral responses
were found.

3. Results and discussion

3.1. Rear irradiance analysis

The draft technical specification describes in broad terms the re-


quirements of a solar simulator with adjustable irradiance levels for
double-sided illumination. The technical specification is mainly focused
Fig. 4. (continued) on double-source simulators but the requirements can be also applied in
this case under study of double-sided illumination using rear reflectors.
The draft technical specification states that the non-uniformity of ir-
radiance must be below 5% on both sides, at the irradiance levels used
for the characterization of bifacial devices and those values used for
corrections and uncertainty evaluation. On the other hand, the draft
technical specification also has the option to perform outdoor mea-
surements of the power generation gain, where the non-uniformity of
irradiance on the rear side must be below 10%. To measure the rear
irradiance with calibrated reference devices, the draft technical speci-
fication proposes to choose the average of at least 5 points from a
suggested matrix of 9 points, with symmetrical distribution, for in-
stance, P1-P3-P5-P7-P9, P2-P4-P5-P6-P8 or P1-P4-P7-P3-P6-P9 as
shown in Fig. 6. It is also mentioned that a matt reflective cloth can be
used to increase the reflection uniformity of the surface behind the
device for outdoor measurements. Note that in the draft technical
Fig. 5. Array configuration or 3 modules setup with the bifacial PV modules specification, the rear-side irradiance non-uniformity for outdoor (10%)
mounted. is different from the requirement for indoor conditions (5%). The
question that arises from this difference is whether the indoor or out-
two modules (modules A) at the sides. According to the standard IEC door irradiance non-uniformity requirement needs to be reviewed, for
60904-9 (IEC, 2007) the non-uniformity of the pulsed solar simulator example, in the case of double-sided illumination.
used is ± 0.45% on a 3 × 3 m2 area in the measurement plane (in The draft technical specification also suggests the use of baffles for
agreement with the total width of the modules). This “array the indoor power generation gain with double-sided illumination:
“Reflections between the two light sources may add irradiance non-

474
J. Lopez-Garcia et al. Solar Energy 177 (2019) 471–482

increases to a maximum then decreases with the distance for the open
rack and structure with baffles. Maximum rear irradiance with the open
rack is obtained at a distance between 20 and 50 cm, however, a shorter
distance (10 cm from the module) is observed with the baffles. The
figures also show the different behaviour of the subsets and the full set
of points which are indicative of the non-uniformity of the rear irra-
diance. The non-uniformity is also plotted in Fig. 7a and b. The rear
mismatch, observed as marked kinks on the rear side I-V curves of the
module, can be explained by the large non-uniformity on the rear side.
The non-uniformity increases to a maximum and then decreases with
the distance between module and rear panel but in nearly all cases is
found to be higher that the draft technical specification requirement of
5%. Furthermore, the choice of sensor number and position on the rear-
side can also greatly impact the measured non-uniformity of irradiance.
In this work, reference cells of 2 × 2 cm2 have been used. Indeed one
subset closely follows the rear irradiance and non-uniformity of the full
Fig. 6. Proposed points to measure the non-uniformity (NU) of irradiance. set, while the other two greatly underestimates or overestimates the
rear irradiance and underestimates the non-uniformity. The results
uniformity. This may generate significant offsets between single-side and have shown that the 3 modules setup is an intermediate case between
double-sided measurement methods results. In this case, double-sided illu- the open rack and the structure with baffles (data are not shown for
mination results must be corrected. Using black masking around the module simplicity).
is recommended to avoid unwanted reflections in double-sided measure- Fig. 8a and b show the opposite case with a rear panel of size equal
ments” (IEC, Draft). to the module (A2 = A1). Note that some points (at 5 cm distance) are
In order to analyse the rear irradiance values, Grear, and distribution, missed in the figures due to technical difficulties to measure under
a 2 × 2 cm2 monocrystalline silicon calibrated reference cell with cer- those conditions. However, the structure with baffles follows a similar
tified uncertainty UC (k = 2) = 0.48% was placed behind the module trend to that of the open rack, therefore, the general information given
facing the reflective rear panel using a holder as small as possible to in the figures are still valid. The rear irradiance is drastically reduced
avoid interferences. At least 9 symmetrically distributed points were for the open rack and to a lesser extent for the structure with baffles. In
measured for representative setups and distances tested. The PASAN both cases, the rear irradiance decreases constantly with increasing
IIIB pulsed solar simulator used for the measurements shows a very distance to the module. Similar rear irradiance values in both setups for
good temporal stability with changes in the irradiance lower than 0.1% this panel size are found which indicates that the reflected light is
at 1000 W/m2 from flash to flash. The rear irradiances Grear and the I-V limited to that passing through the gaps around the cells. This is a
curves were measured separately but without changing the setup or limiting factor for a double-sided illumination with the rear panel with
measurement conditions. The front side irradiance was set to the same size as the module. The non-uniformity of the rear irradiance
1000 ± 1 W/m2 without further corrections for the proposal of rear decreases slightly with the distance (for the open rack setup) with the
irradiance measurement only. The uncertainty on the Grear is assumed longest distances to the module, exhibiting values below 10%. Similar
to be that of the reference cell used. The rear irradiance non-uniformity to the previous configuration, one subset closely follows the non-uni-
in the module plane is defined as: formity of the full set, while the other two follow a similar trend but
underestimate the value.
Grearmax − Grearmin Therefore, the position and number of points measured on the rear
Grear Non − uniformity(%) = × 100
Grearmax + Grearmin (3) side can have an impact in the determination of the module measuring
conditions and subsequently on the modules performance for double-
where the Grear min is the minimum rear irradiance and Grear max is the sided illumination. It is also shown that the technical specification re-
maximum rear irradiance. The average rear irradiance and the rear quirements of rear irradiance (100–200 W/m2) and non-uniformity
irradiance non-uniformity of several subsets of symmetrically dis- cannot be fulfilled at the same time with the current setup using re-
tributed points are plotted as a function of the distance to the rear panel flective rear panels and improvements need to be made.
for the open rack and the structure with baffles in Fig. 7a and b, re- A rough calculation of the incident light passing through the gaps
spectively, using the largest rear panel size. The rear irradiance

Fig. 7. Rear irradiance Grear (W/m2) and Grear non-uniformity of rear irradiance (%) for symmetrical distributed points for the largest rear panel (A2 = 2.7A1) with, a)
the open rack setup and b) the structure with baffles.

475
J. Lopez-Garcia et al. Solar Energy 177 (2019) 471–482

Fig. 8. Rear irradiance Grear (W/m2) and Grear non-uniformity of rear irradiance (%) for symmetrical distributed points for the smallest rear panel (A2 = A1) with, (a)
the open rack setup and (b) the structure with baffles.

between the cells and the transparent free area reaching the reflective observed for the open rack setup increasing concentrically. Average
rear panel can be performed. It should be noted a non-uniform trans- Grear up to 188 W/m2 with a standard deviation of ± 52 W/m2 and a
mission of the light due to the different area of the free spaces in the maximum value of 270 W/m2 was obtained. In the other hand, the rear
module, greater at the borders than between cells would be expected. irradiance for the structure with baffles presents a different distribution
For the rear panel with area A2 = A1, assuming a homogeneous with significantly lower but more homogeneous values between
1000 W/m2 front irradiance and the structure with baffles (to ensure 30–50 W/m2 with standard deviation between ± 2 and ± 20 W/m2 and
that only the light passing through the free areas of the module con- maximum value of 75 W/m2. It should be noted the different spatial
tributes to the reflected light), the maximum irradiance reaching the distribution of the rear irradiance in Fig. 9c. The particular design of the
rear panel would be around 67 W/m2 (see Section 2.1). This is higher structure with baffles in the area around the reference cell and in case
than the effective average rear irradiance measured for this size of real for the largest rear panel size, led to a higher irradiance contribution in
panel (30–15 W/m2). the area close to the reference cell, due to difficulties to completely
Fig. 9a exhibits the minimum, maximum and average rear irra- blocks the area around the cell with baffles. Fig. 10a shows the rear
diance Grear as a function of the distance for an open rack and a irradiance as a function of the distance to the reflective panel for the
structure with baffles for A2 = 2.7A1. Fig. 9b and c exhibit the spatial smaller area (A2 = A1). Similar rear irradiance values and irradiance
distribution for the corresponding setup with the reflective panel placed distribution with the values decreasing concentrically (Fig. 10b and c)
at 50 cm. A lower rear irradiance in the central part of the module was were found for the different configurations.

Fig. 9. (a) Average, maximum and minimum rear


irradiance Grear (W/m2) as a function of the dis-
tance module-rear reflective panel for the open rack
and structure with baffles for the largest panel and
the spatial distribution of the rear irradiance with
the reflective rear panel placed at ∼50 cm from the
module for the (a) open rack setup and (b) the
structure with baffle.

476
J. Lopez-Garcia et al. Solar Energy 177 (2019) 471–482

Fig. 10. (a) Average, maximum and minimum


rear irradiance Grear (W/m2) as a function of the
distance module-rear reflective panel for the
open rack and structure with baffles for the
smallest panel and the spatial distribution of the
rear irradiance with the reflective rear panel
placed at ∼50 cm from the module for the (a)
open rack setup and (b) the structure with baffle.

3.2. Initial module characterisation: Single side illumination and (≤10 ms) at 1000 W/m2 exhibited lower Pmax values ranging from
measurements at equivalence irradiance levels −3.3% to −3.6% on average, in comparison to the multi-flash method
due mainly to a decrease in the FF. According to these results, the multi-
Conventional flash light simulators are generally used for the flash was the preferred method and it was therefore used for all of the
measurement of module performance due to the possibility to have very experiments reported in this paper.
controlled environmental conditions and their ease of use. Moreover, As described in the IEC draft for bifacial modules IEC TS 60904-1-2,
class AAA solar simulators nowadays lead to a spectral mismatch factor bifaciality refers to the ratios between the main I-V characteristics of
close to unity for standard c-Si modules. However, for high-efficiency c- the rear and front side of a bifacial device, typically at Standard Test
Si modules with increased capacitance, the effects of the highly-capa- Conditions (STC). This is quantified with reference to bifaciality coef-
citive devices can lead to inaccurate measurements of the PV module ficients, namely the short-circuit current bifaciality coefficient φIsc, the
electrical parameters. Fig. 11 shows the I-V curve corrected to STC for a open-circuit voltage bifaciality coefficient φVoc and the maximum
module under test measured using a forward sweep single flash and a power bifaciality coefficient φPmax. This last one is defined as:
multi-flash method, both using a black non-reflective rear panel in di- PmaxRear
rect contact to the non-illuminated side. Single I-V measurements made, φPmax (%) = × 100
PmaxFront (4)
on the nominally front surface of the modules, as a forward sweep
where Pmax Rear and Pmax Front are the module output power measured
when illuminating only the rear and the front side at STC, respectively,
with a black rear panel. These coefficients are required in order to
calculate the Pmax at an equivalent irradiance level when a single-side
illumination is used (IEC, Draft). Fig. 12 shows the I-V curves, under
STC conditions, for the module under test measured with a black rear
panel according to the standard for the front and rear side. A high bi-
faciality for Pmax and Isc of 98% was obtained. Fig. 12 also shows the I-V
curves of the module measured with a white reflective panel of the
same dimensions as the module placed in direct contact to the rear
surface or non-illuminated side of the module. Assuming the black rear
panel as the reference or monofacial configuration, the Pmax or the Isc
change due to the reflective panel can be calculated by means of the
equation:
Pmaxwhitepanel −Pmaxblackpanel
Pmax change(%) = × 100
Pmaxblackpanel (5)
Fig. 11. I-V curves of a bifacial PV module measured with a pulsed solar si- where the Pmax white panel and Pmax black panel are the maximum output
mulator with forward sweep and multiflash method, respectively. power of the bifacial module with a white and a black panel on the non-

477
J. Lopez-Garcia et al. Solar Energy 177 (2019) 471–482

Fig. 12. I-V curves of a bifacial PV module under STC for the front and rear side
Fig. 13. I-V curves of a bifacial PV module with a white reflective rear panel
with a black and a white reflective rear panel, respectively.
placed at different distances from the module.

illuminated side of the module, respectively. Even though the rear ir-
respectively, was obtained for both calculations (Table 1), below the
radiance contribution due to reflected light from the white panel is low,
measurement uncertainty (Müllejans et al., 2015; Müllejans et al.,
an Isc and Pmax gain of +2.5% and +2.3%, respectively, was obtained
2009).
for the front side illumination and a gain of +2.8% and +2.6% ob-
tained for the rear side illumination. The current best ESTI module
expanded combined uncertainties for Isc and Pmax for the PASAN IIIB 3.3. I-V measurements for open rack setup
pulsed solar simulator are ± 0.86% and ± 1.1%, respectively, for ab-
solute measurements and 0.40% and 0.85% for relative measurements. Fig. 13 shows the I-V curves (corrected to STC) measured with the
The latter applies here and the uncertainties are well below the in- multi-flash method for the front side of the bifacial module at different
creases observed. It is clear that the use of non-reflective rear material is distances between the white reflective rear panel and the module for
recommended to obtain reproducible and reliable results under STC in A2 = 2.7A1. It is also shown inset as a zoom of the curves between 0 and
compliance with the recommendations in the draft technical specifi- 35 V. The non-uniformity of the reflected light lead to a mismatch on
cation. the rear irradiance which produces kinks in the I-V curves. The mis-
By using the setup described in Section 2.2.1 for single-side illu- match in rear irradiance for the different setups will be discussed in
mination, the IEC TS 60904-1-2 states that the Pmax of the device must Section 3.5, but for the open rack, contributes up to 3%, particularly at
be measured at equivalent irradiance levels GEi corresponding to large panel distances. Similar I-V curves were obtained for the different
1000 W/m2 on the front side plus different rear side irradiance levels sizes of rear panels used and set-up configurations.
GRi. The equivalent irradiance levels are determined as functions of the Fig. 14a and b show the Isc and Pmax as a function of the distance
bifaciality coefficient φ according to Eq. (6) and Eq. (7): between the modules and the rear reflective panel for different areas of
this panel. The dotted lines indicate the Isc and Pmax rated values and
G Ei = 1000W/m2 + φ ·GRi (6) the estimated bifacial gain or projected specifications including addi-
tional backside irradiation contribution in Isc (Pmax) as a percent of STC.
φ = Min (φIsc,φPmax ) (7) The grey area in Fig. 14b represents the manufacturer tolerance range
given in the datasheet for the front side illumination under STC and the
where φ is the minimum value between the Isc and the Pmax bifaciality black line is the front side Pmax measured with the rear black non-re-
coefficients φIsc and φPmax. Besides the STC measurement at 1000 W/ flective panel. Note that this value differs from the rated value given by
m2 (for GR = 0 W/m2), at least two different equivalent irradiance le- the manufacturer in the datasheet. The Isc and Pmax increased as the
vels are required, so two specific Pmax values must be reported: module-rear panel distance and reflective panel area increased. For
PmaxBiFi10 and PmaxBiFi20, for GR1 = 100 W/m2 and GR2 = 200 W/m2, A2 > A1, the electrical parameters seem to be saturated at a distance
respectively. For the modules under test, φ = 0.98 and the equivalent from the module of ∼50 cm. The maximum gain in Pmax of 20% was
irradiance levels for GR1 = 100 and GR2 = 200 are GE1 = 1098 W/m2 obtained for the largest panel size A2 = 2.7A1 at 50 cm from the
and GE2 = 1196 W/m2, respectively. module. It should be noticed that the Pmax and Isc for rear panel area
The differences between Pmax measured from the front side at A2 = A1 showed the maximum between 5 and 10 cm from the module.
1000 W/m2 and the Pmax measured in the rear side at 100 W/m2 (or This means that the greater the distance, the larger the solid angle
200 W/m2) multiplied by the bifaciality, and the Pmax at equivalent between the rear panel and the module that is not able to “see” or
irradiance level GE1 (and GE2), that is, illuminating the front side only at collect all the reflected light. Furthermore, for the largest panel area, as
1098 W/m2 (or 1186 W/m2) in a single flash are listed in Table 1. A the distance increases the shadow originated by the module decreases
−0.04% and +0.78% difference for 100 W/m2 and 200 W/m2, (so increasing in theory the reflected light) but the rear irradiance or

Table 1
Comparison of Pmax at equivalent irradiance levels GEi and the Pmax obtained by the sum of the front and rear output power for a single-side illumination method
according to the draft of the technical specification.
Single-side illumination

Pmax front STC Pmax rear STC Pmax rear Pmax rear 1000 + φPmax * 100 1000 + φPmax * 200 Pmax GE1 Pmax GE2 Difference (%) Difference (%)
(W) (W) 100(W) 200(W) GE1 GE2

250.5 246.5 23.2 48.3 273.3 298.1 273.4 295.8 −0.04 +0.78

478
J. Lopez-Garcia et al. Solar Energy 177 (2019) 471–482

Fig. 14. (a) Short-circuit current, Isc (A) and b) Pmax (W) as a function of the distance of the reflective rear panel to the module for the open rack setup. Grey area
represents the manufacturer Pmax tolerance given in the datasheet.

the irradiance produced by the reflective rear panel also decreases with case between the open rack and the structures with baffles where the
increasing distance (as in the case for a solar simulator where increasing light transmitted between the cell gaps of the 3 modules and around the
the distance between light source and the test plane reduces the total edges of the modules is reflected by the rear panel contributes to the
irradiance). Therefore, the maximum Pmax corresponds to the trade-off rear irradiance. Fig. 16a shows the Isc as a function of the distance
between these two factors. The FF increased slightly with the distance between the modules and the rear panel. Isc increased as the distance
of the rear panel and decreased after 20 cm showed no significant module-rear panel and the area of the panel increased but the beha-
variations with the distance (less than 1.4% of total variation). viour is different depending on the rear panel area. The saturation
distance seems to be shorter as the rear panel area decreases with a
3.4. I-V measurements for structure with baffles maximum between 20 and 50 cm for the greatest area and of 5 cm for
the smallest. Pmax (Fig. 16b) has a similar behaviour as the Isc but the
Fig. 15a and b show the dependence on the distance between the distance of the maximum value is located around 10–20 cm. There is no
module and the rear reflective panel for the structure with baffles of Isc further improvement of Pmax for a distance module-rear panel beyond
and Pmax, respectively. Both parameters increase to a maximum and 20 cm. The Pmax for the greatest area of the rear panel correspond to a
then decrease as the distance between the module and the rear panel 10% gain of the nominal output power given in the datasheet. No sig-
increase. The Isc and Pmax reach the maximum at 5 cm from the module nificant effects (lower than 0.5% in current change) was observed due
with a gain in Pmax that hardly reaches the 9% of the rated value given to the mismatch of the rear irradiance. The FF changed with the dis-
in the datasheet at STC. The electrical parameters are slightly influ- tance to the module of the rear panel with a no clear trend and below
enced by the rear panel area only for the smallest area used. As will be 1%.
shown later, the small dependence of the electrical parameters of the
modules with the distance and area is due to the small variation on the 3.6. Comparison between three set-ups
average rear side irradiance (32–49 W/m2). This is mainly due to the
use of baffles, so that only the light transmitted through the gaps As mentioned previously, the non-uniformity of the light reaching
around the cells contributes to the rear irradiance. Differences lower the rear side lead to a mismatch in the rear irradiance that produce
than 1% in current due to mismatch in rear irradiance was observed. No kinks in the I-V curves. In general, the structure with baffles and the 3
significant variations were observed in the FF with the distance to the module setup showed similar kinks with the same mismatch in the I-V
rear panel with a change lower than 0.5% for the different panel areas. curves at similar voltages with some variation for the smallest panel
(A2 = A1). However, the position of the kinks depended on both the
3.5. I-V measurements for 3 modules setup distance to the panel and the panel size in the open rack setup. This
indicates that the rear irradiance values and spatial distribution is very
The 3 module configuration can be considered as an intermediate dependent on the reflective panel size. The change in the current due to

Fig. 15. (a) Short-circuit current, Isc (A) and b) Pmax (W) as a function of the distance of the reflective rear panel to the module for the structure with baffles. Grey area
represents the manufacturer Pmax tolerance given in the datasheet.

479
J. Lopez-Garcia et al. Solar Energy 177 (2019) 471–482

Fig. 16. (a) Short-circuit current, Isc (A) and b) Pmax (W) as a function of the distance of the reflective rear panel to the module for the 3 modules setup. Grey area
represents the manufacturer Pmax tolerance given in the datasheet.

Fig. 17. Change (%) in the current due to the kinks as a function of the distance Fig. 18. Pmax gain (%) with respect to the measurement of the front side only at
to the rear reflective panel due to the rear irradiance non-uniformity for the STC with the black rear panel at 0 cm from the module for the open rack,
configurations and panels. structure with baffles and 3 modules setup, respectively, as a function of the
distance module-rear reflective panel.
the kinks, ΔI (%), as a function of the distance between the module and
the rear reflective panel is plotted in Fig. 17. The magnitude of the
kinks, measured as the change in the current, was lower than 1% and
0.5% (for all real panel sizes) for the baffles and 3 modules setup, re-
spectively. The magnitude of the kinks increased slightly with the size
of the panel in these cases. However, for the open rack setup, a change
in current due to the mismatch of rear irradiance up to 3% was shown
increasing with the distance between module and rear panel and with
the size of the panel.
Fig. 18 shows the change in the Pmax, defined as Pmax gain with
respect to the front side only measurement at STC for the open rack
configuration, structure with baffles and 3 modules setup as a function
of the distance module-rear reflective panel for two panel sizes. Pmax
increased up to +17.7%, 7.5% and 5.5%, for the open rack, 3 modules
and structure with baffles, respectively. It also shows that the Pmax gain Fig. 19. Pmax (W) measurements at rated STC, with a black panel and a white
remain, in general, below 5% for the structure with baffles and prac- reflective panel and their increase with respect to the rated STC value. Grey
area represents the manufacturer Pmax tolerance given in the datasheet.
tically independent of the rear reflective panel size and slightly higher
than 5% for the 3 modules configuration. The rear panel size has a
significant influence on the Pmax gain in open rack configuration in- with a black panel attached to the rear side and the Pmax measured with
dicating that the Pmax gain (and then rear irradiance) can be increased the white reflective panel also at 0 cm from the module rear surface are
just by increasing the rear panel size. One of the requirements of the plotted in Fig. 19. The grey area represents the tolerance range given by
technical specification IEC TS 60904-1-2 for a double-sided illumina- the manufacturer. A difference of +2.7% in Pmax with respect to the
tion solar simulator (rear irradiances close to 200 W/m2) can be rated value when the module is measured according to the draft spe-
reached under the setup proposed. It should be noted that the gain in cification is observed, but it is still with the stated uncertainties. This
Pmax is obtained with reflected light from the rear panel whereas under difference increases up to +5.4% when using the white rear panel in
outdoor real conditions the incident light on the rear is a combination comparison with the manufacturer rated STC output power. There may
of ground reflection and a diffuse sky component. be a number of reasons for the higher Pmax value measured in com-
The rated Pmax at STC (given in the datasheet from the manu- parison with the manufacturer rated values such as the positive binning
facturer), the Pmax measured according to the draft IEC TS 60904-1-2 of the modules, but one possible explanation could be our use of the

480
J. Lopez-Garcia et al. Solar Energy 177 (2019) 471–482

Table 2
Comparison between the Pmax measured with single side illumination at equivalent irradiance levels GRi and double-side illumination with the white reflective rear
panel at different distances.
Double-sided illumination Single-side illumination ΔPmax (%)

Configuration Panel size Distance panel (cm) Avg GRi (W/m2) Min. GRi (W/m2) Max. GRi (W/m2) Pmax GRi (W) Pmax GEi (W)

3 modules 2.7A1 70 95 66 122 270.1 273.4 −1.20


Open rack 2.7A1 10 105 42 102 271.3 273.4 −0.76
Open rack 2.7A1 50 190 131 266 296.3 295.8 +0.17

Multiflash method instead of a single flash approach. Since the char-


acterisation method used by the manufacturer is not indicated in the
datasheet, we cannot state conclusively that this is the cause.

3.7. Comparison with single-side illumination measurements

A rough validation of the double-sided illumination method can be


done by comparing the Pmax obtained measuring the front side at the
equivalent irradiance level GE1 and GE2, for instance, 100 W/m2 and
200 W/m2 with the proposed double-side illumination method based on
a reflective rear panel. The comparison of the double-sided illumination
method is done by comparing those values measured with single-side
illumination with the Pmax obtained using the appropriate size of the
white reflective rear panel at a certain distance in order to obtain an
average rear irradiance close to 100 W/m2 and 200 W/m2. The results
are listed in table 2. The draft technical specification requirement for Fig. 20. Pmax as a function of average irradiance level on the rear side GR (for
rear irradiance non-uniformity of 5% can be only achieved in a very few double-side illumination) and its single-side equivalent irradiance. The red line
is the linear fit of the values and the green circles are the measured values at
specific configurations using the white rear panel. However, average
equivalent irradiance level of 100 and 200 W/m2.
rear irradiances around 100 ± 10 W/m2 and 200 ± 10 W/m2 can be
easily achieved in more configurations of the reflective panel. Re-
presentative examples of these configurations and Pmax for double-sided Table 3
illumination method are listed in table 2. Assuming the average rear Maximum peak power, Pmax, measured at equivalent front irradiances, GE and
irradiance as the rear equivalent level, the results are in good agree- the calculated from the rear irradiance driven power gain yield, BiFi, Pmax,
BiFi100 and Pmax, BiFi200Pmax at GR = 100 and 200 W/m2, respectively,
ment with the single-side illumination at equivalent irradiance levels
(double-side with GFront = 1000 W/m2) for the open rack with A2 = 2.7A1.
within the measurement uncertainty. Table 2 shows the results for 3
module and open rack configuration with the reflective panel of 2.7A1 φ (%) GR (W/ GE (W/ Pmax GE BiFi Pmax BiFiGR ΔPmax (%)
size. It was not possible to reach the minimum rear irradiance (100 W/ m2) m2) (W) (W)

m2) required by the draft technical specification with the structure with 98 0 1000 251.7 0.218 251.7 +0.0
baffles. 100 1098 273.4 273.5 +0.0
According to the technical specification and as described in the 200 1196 295.8 295.2 +0.2
introduction, the rear irradiance driven power gain yield, BiFi, can be
calculated from the slope derived from the linear fit of the Pmax versus
GR data series for a double-sided illumination approach as the rear 4. Conclusions
reflective panel and two values, PmaxBiFi10 and Pmax BiFi20, for
GR1 = 100 W/m2 and GR2 = 200 W/m2, respectively, have to be given. This work outlines the indoor performance testing of c-Si bifacial PV
The Pmax as a function of the average rear irradiance GR measured with modules under different module setups including open rack, a structure
the open rack and the largest reflective panel and the equivalent irra- with baffles and 3 modules, with a white reflective rear panel of several
diance level values, GE, is shown in Fig. 20. The red line corresponds to dimensions placed at various distances behind the module as a potential
the linear fitting of the squares. A validation approach is to compare approach for a double-sided illumination characterization method.
these values with those obtained with the equivalent irradiance method Single-side illumination measurements under STC conditions following
with single-side illumination. As it can be shown in Table 3, a very good the draft technical specification IEC TS 60904-1-2 were also performed.
agreement is obtained for both methods using the rear-irradiance va- A simple forward sweep method led to incorrect measurement of Pmax
lues of the open-rack structure. for highly capacitive modules in comparison with the multiflash
Despite the high rear non-uniformity, similar Pmax values for single- method. This indicates that capacitive effects in high-efficiency bifacial
side illumination and double-sided illumination have been obtained. modules need to be considered in order to use the appropriate char-
This result agrees with previously reported works that showed that the acterisation system and method. Furthermore, in agreement with IEC
non–uniform irradiance affects Isc and the voltage region from 0 to Vmpp TS 60904-1-2, parasitic reflections from the rear side of the PV module
but to a much lesser extent the Pmax (Hermann and Wiesner, 2000; increase Pmax and subsequently the uncertainty in measurements. The
Hishikawa et al., 2010; Monokroussos et al., 2013). However, the non- use of an opaque black cover attached to the rear side of the module is
uniformity in the rear side of the module under test needs to be im- recommended to obtain reproducible and reliable results for measure-
proved in order to make this double-sided method a suitable method for ments under STC conditions.
the characterization of bifacial PV modules. The use of reflective rear panels behind the module increases the
rear irradiance, and in doing so increases the Pmax output power of the
module. In general, Pmax increases with the distance between the rear

481
J. Lopez-Garcia et al. Solar Energy 177 (2019) 471–482

reflective panel and the module up to a maximum value and then re- Guerrero-Lemus, R., Vega, R., Kim, T., Kimm, A., Shephard, L.E., 2016. Bifacial solar
mains constant or decreases. This increase can be as high as 20% under photovoltaics – a technology review. Renew. Sustain. Energy Rev. 60, 1533–1549.
Roest, S., Nawara, W., Van Aken, B.B., Garcia-Goma, E., 2017. Single side versus double
specific setup conditions (open rack setup with the largest rear white side illumination method IV measurements for several types of bifacial PV modules.
panel) and distances of the rear panel and as such compliance with one In: 33rd European Photovoltaic Solar Energy Conference and Exhibition, pp.
of the draft technical specification requirements. However, the rear 1427–1431.
Lopez-Garcia, J., Haile, B., Pavanello, D., Pozza, A., Sample, T., 2016. Characterisation of
irradiance non-uniformity needs to be improved considerably and the n-Type Bifacial Silicon PV Modules. In: 32nd European Photovoltaic Solar Energy
approach to measure it (number of points measured on the rear, posi- Conference and Exhibition, pp. 1724–1729.
tions, etc) need to be well defined to make this approach a suitable Newman, B., Carr, A., Groot, K., Dekker, N.J.J., Van Aken, B.B., Vlooswijk, A., Van de
Loo, A., 2017. Comparison of bifacial module laboratory testing methods. In: 33rd
double-sided illumination method for bifacial PV module character- European Photovoltaic Solar Energy Conference and Exhibition, pp. 1632–1635.
isation. The rear irradiance non-uniformity led to significant kinks in Schmid, A., Dulger, G., Baraah, G., Kräling, U., 2017. IV measurement of bifacial modules:
the module I-V curves (and then higher uncertainty in the module bifacial vs. monofacial illumination. In: 33rd European Photovoltaic Solar Energy
Conference and Exhibition, pp. 1624–1627.
performance) due to an irradiance mismatch. Following the re-
Razongles, G., Sicot, L., Joanny, M., Gerritsen, E., Lefillastre, P., Schroder, S., Lay, P.,
commendation within IEC TS 60904-1-2 to use black masking or baffles 2016. Bifacial Photovoltaic Modules: Measurement Challenges. In: 6th International
around the modules does improve the rear non-uniformity. Un- Conference on Crystalline Silicon Photovoltaics, SiliconPV 2016 92, pp. 188–198.
fortunately, this method also limits the total rear irradiance below Deline, C., Macalpine, S., Marion, B., Toor, F., Asgharzadeh, A., Stein, J.S., 2017.
Assessment of bifacial photovoltaic module power rating methodologies-inside and
100 W/m2 and therefore is not sufficient to generate the required rear out. IEEE J. Photovolt. 7, 575–580.
irradiance levels. Pyrot, L., Razongles, G., Sicot, L., Joanny, M., Hladys, B., Lefillastre, P., 2017. Bifacial
Finally, the comparison of the double-sided with a single-side illu- module measurements with GE method, 4th. Bifacial Workshop BIFIPV.
Singh, J.P., Aberle, A.G., Walsh, T.M., 2014. Electrical characterization method for bi-
mination method with equivalent irradiance levels showed a good facial photovoltaic modules. Solar Energy Mater. Solar Cells 127, 136–142.
agreement in the measurements showing that the non-uniformity of Comparotto, C., et al., 2014. Bifacial n-Type Solar Modules: Indoor and Outdoor
irradiance has a smaller influence in the Pmax than in the Isc. Further Evaluation. In: 29th European Photovoltaic Solar Energy Conference and Exhibition,
pp. 3248–3250.
work is being carried out to understand the limitations of the rear ir- Soria, B., Gerritsen, E., Lefillastre, P., Broquin, J.E., 2016. A study of the annual perfor-
radiance spatial distribution on the module output power as well on the mance of bifacial photovoltaic modules in the case of vertical facade integration.
improvement of the quality, roughness, size and type of reflective rear Energy Sci. Eng. 4, 52–68.
Zhang, Y., Gao, Q., Yu, Y., Liu, Z., 2018. Comparison of double-side and equivalent single-
panel. It seems that the rear panel size has a decisive role in the rear side illumination methods for measuring the I-V characteristics of bifacial photo-
non-uniformity and rear irradiance and increasing the size more than voltaic devices. IEEE J. Photovolt. 8, 397–403.
the area used in this work, a significant improvement in rear irradiance Singh, J.P., Guo, S., Peters, I.M., Aberle, A.G., Walsh, T.M., 2015. Comparison of glass/
glass and glass/backsheet PV modules using bifacial silicon solar cells. IEEE J.
could be obtained but probably not in rear non-uniformity. The rear
Photovolt. 5, 783–791.
reflective area is somehow limited by the current solar simulators. IEC 61215-2 Crystalline silicon terrestrial photovoltaic (PV) modules: design qualification
However, it was shown that the rear non-uniformity is not so relevant and type approval; 2016b.
for the Pmax calculation and the rear-side irradiance non-uniformity IEC 60904-9, Photovoltaic devices – Part 9: Solar simulator performance requirements;
2007.
requirement proposed by the draft technical specification may be re- Ferretti, N., Pelet, Y., Berghold, J., Fakhfouri, V., Grunow, P., 2013. Performance Testing
laxed (in agreement with the outdoor requirement, for example) in of High-Efficient PV Modules Using Single 10 ms Flash Pulses. In: 28th European
order to make this method a suitable approach for the characterisation Photovoltaic Solar Energy Conference and Exhibition, pp. 3184–3187.
Virtuani, A., Rigamonti, G., Beljean, P., Friesen, G., Pravettoni, M., Chianese, D., 2012. A
of bifacial PV modules. fast and accurate method for the performance testing of high-efficiency C-Si photo-
voltaic modules using A 10 Ms single-pulse solar simulator. In: 38th IEEE Photovoltaic
Acknowledgements Specialists Conference, pp. 496–500.
Van Steenwinkel, R., 1986. Measurements of spectral responsivities of cells and modules.
In: 7th European Photovoltaic Solar Energy Conference and Exhibition, pp. 325–329.
Authors would like to thank Dr. Cecilia Guillen for the reflectance Müllejans, H., Zaaiman, W., Dunlop, E.D., 2015. Reduction of uncertainties for photo-
measurements, Dr. Harald Müllejans for the comments and review of voltaic reference cells. Metrologia 52, 646–653.
Müllejans, H., Zaaiman, W., Galleano, R., 2009. Analysis and mitigation of measurement
the manuscript, Francesco Noseda for the technical support and Diego
uncertainties in the traceability chain for the calibration of photovoltaic devices.
Pavanello in the setup and software development. The project (PV- Meas. Sci. Technol. 20, 075101.
ENERATE) leading to this application has received funding from the Hermann, W., Wiesner, W., 2000. Modelling of PV modules: The effects of non-uniform
irradiance on performance measurements with solar simulators. 16th European
EMPIR programme co-financed by the Participating States and from the
Photovoltaic Solar Energy Conference.
European Union’s Horizon 2020 research and innovation programme. Hishikawa, Y., Shimura, H., Tsuno, Y., 2010. Influence of nonuniformity of irradiance
within a cell on the accurate I-V curve measurement under 1 sun illumination. In:
References 35th IEEE Photovoltaic Specialists Conference, pp. 2684–2687.
Monokroussos, C., Etienne, D., Morita, K., Fakhfouri, V., Bai, J., Dreier, C., Therhaag, U.,
Herrmann, W., 2013. Impact of Calibration Methodology into the Power Rating of c-
IEC 61836, Solar photovoltaic energy systems – Terms, definitions and symbols; 2016a. Si PV Modules under Industrial Conditions. In: 28th European Photovoltaic Solar
IEC TS 60904-1-2 Measurement of current-voltage characteristics of bifacial photovoltaic Energy Conference and Exhibition, pp. 2926–2934.
(PV) devices; 2018 Draft.

482

You might also like