ENERGY YIELD ESTIMATION OF MONOFACIAL AND BIFACIAL...
Transcript of ENERGY YIELD ESTIMATION OF MONOFACIAL AND BIFACIAL...
ENERGY YIELD ESTIMATION OF MONOFACIAL AND BIFACIAL SOLAR MODULES
Corrado Comparottoa1, Matthias Noebelsa2, Eckard Wefringhausa3, Nicoletta Ferrettib4, Giulia Mancinib5,
Juliane Bergholdb6, Roland Einhausc7, Frédéric Madonc8 aInternational Solar Energy Research Center (ISC) Konstanz, Rudolf-Diesel-Straße 15, 78467 Konstanz, Germany
bPI Photovoltaik-Institut Berlin AG, Wrangelstr. 100, 10997 Berlin, Germany cAPOLLON SOLAR, 66 cours Charlemagne, 69002 Lyon, France
1Corresponding author. Tel.: +49-7531-361-83-372; E-mail address: [email protected]
E-mail addresses: [email protected], [email protected], [email protected], [email protected], [email protected], [email protected], [email protected]
ABSTRACT: Monofacial and bifacial solar modules featuring diverse characteristics were produced and measured
indoor under standard test conditions (STC). They were installed in the desert in El Gouna, Egypt, where their IV
curves were recorded continuously for six months, together with the solar irradiance and the ambient temperature.
Simple models based on indoor measurements of power and outdoor measurements of irradiance were built in order
to estimate the modules energy yield. The estimations are compared with the measured values. Where the models do
not match the measurements, a deep analysis was carried out in order to understand the discrepancies. In the absence
of an encapsulant, the energy yield can decrease significantly, mainly due to a mismatch in refractive indices. It is
also noticed that indoor measurements at STC can deliver quite different results when performed by different parties.
Light induced degradation (LID) can also play a partial role in p-type cells. A formula calculating the outdoor bifacial
efficiency is suggested for a fair comparison between monofacial and bifacial modules.
Keywords: Bifacial, n-type, Energy performance, PV Module
1 INTRODUCTION
Bifacial solar modules have been investigated almost
since the very beginning of PVs [1], [2]. It has already
been demonstrated also at large-scale power plants that
the energy yield of a solar system can be enhanced
significantly by using bifacial modules [3], which take
advantage of the contribution of the rear side. However,
the evaluation of the energy yield of such modules is
difficult to estimate, due to the many variables involved
outdoor, viz. temperature, irradiance, albedo, modules
height or distance between modules. Understanding the
correlation between laboratory measurements under
standard test conditions (STC) and outdoor performance
is of both scientific and economic importance and plays a
crucial role for the customer. This work enriches this
understanding with the help of mathematical models.
2 DESCRIPTION OF THE MODULES
In the presented work, seven solar modules of the
following technologies are studied:
Modules 1 and 2: commercial monofacial modules
employing 60 Cz p-type Al-BSF cells each.
Module 3: glass-glass encapsulant-free bifacial
module fabricated at Vincent Industrie with the NICE
technology [4]. 60 ISC Konstanz' n-type bifacial cells
(called BiSoN [5]) with 2 busbars were used for this
module. The AR coating of the cells had not been
optimized for an encapsulant-free module. Due to a
limited number of available cells, no strict cell
binning was carried out, resulting in a slight current
and fill factor mismatch. A standard "non-solar" glass
exhibiting lower light transparency was used for the
rear side of this module, while an extra clear
photovoltaic (PV) glass, with only one AR layer on
the external surface, was employed for the front.
Modules 5 and 6: glass-glass bifacial modules
manufactured at GSS, Germany, using 60 BiSoN
cells each.
Module 7: monofacial module with a white backsheet
fabricated at Bosch Solar Energy AG within the
German publicly funded project nSolar using 120
halved bifacial n-type cells [6].
Module 8: bifacial module with transparent
backsheet, employing the same type and amount of
cells as Module 7.
All modules were of the same size (1.66 m²). It must be
noted that all modules, except from modules 1 and 2, are
not commercial modules, they were fabricated for
research purpose only. Module 4 broke during
installation. The main characteristics of the modules are
reported in Table I.
Table I: Overview of the studied modules Module
numberManufacturer Module type Encapsulant
Wafer base
doping
1, 2 Commercial Monofacial EVA p
3Vincent
IndustrieBifacial None n
5, 6 GSS Bifacial EVA n
7 Bosch Monofacial EVA n
8 Bosch Bifacial EVA n
3 INDOOR MEASUREMENTS
Modules 1 and 2 were measured under STC in-house
by the module producer while all other modules were
measured with the same flasher at PI Berlin, Germany,
also under STC. Front and rear side of the bifacial
modules were measured separately by taping the non-
flashed side with a black foil, as shown in Figure 1. The
corresponding maximum power of front and rear side is
denoted as Pf and Pr, respectively. In order to simulate
different outdoor backgrounds, the bifacial modules were
also flashed in a bifacial mode by placing a grey curtain
(with an albedo of 0.3, as shown in Figure 2) or a white
curtain (with an albedo of 0.7) 60 cm behind the rear side
of the modules. In these cases, the power at the maximum
power point is denoted as Pg and Pw, respectively. The
outcome of all indoor measurements is reported in
Figure 3.
Figure 1: Taped rear side of a bifacial module
Figure 2: Grey curtain used for bifacial measurements
Figure 3: Indoor measurements
4 OUTDOOR MEASUREMENTS
All modules were installed in pairs in the desert in El
Gouna with an inclination of 20° and the front facing
south, as shown in Figure 4. A concrete base was used to
fix the modules to the ground. Typical albedo values for
concrete and Sahara sand are in the range of 0.2-0.55 [7]
and 0.47-0.49 [8], respectively. IV curve parameters as
well as irradiance and temperature were recorded for
each module at an interval of five minutes from 1st March
to 31st August 2014. In order to measure the in-plane
irradiance of front and rear side of the modules, sensors
were installed as shown in Figure 5. The solar irradiances
(or insolations) Ee,f and Ee,r (in W/m2) were measured in
the module plane by the sensor facing the sun and the
sensor facing the ground, respectively. Integrating Ee,f
and Ee,r over the six considered months, radiant
exposures of 1133 kWh/m² (hereafter called He,f) and 276
kWh/m² (hereafter called He,r) are respectively found.
Figure 4: Installation site in El Gouna, Egypt
Figure 5: Irradiance sensors
5 MATHEMATICAL MODELS
Taking into account the outdoor measurements of
solar irradiance and radiant exposure, i.e. Ee,f, Ee,r, He,f
and He,r, and the indoor measurements of power of each
module, i.e. Pf, Pf, Pg and Pw, three simple mathematical
models were tested in order to estimate energy and
power. Ee,STC is the irradiance used for the STC
measurements, i.e. 1000 W/m2. Yexp is the expected
energy yield (over six months) and Pexp is the expected
power (given at an interval of five minutes).
Model "front&rear"
𝑌𝑒𝑥𝑝 ,𝑓&𝑟 (kWh) =𝑃𝑓 ∗ 𝐻𝑒,𝑓 + 𝑃𝑟 ∗ 𝐻𝑒 ,𝑟
𝐸𝑒 ,𝑆𝑇𝐶 1𝑎
𝑃𝑒𝑥𝑝 ,𝑓&𝑟 (kWh) =𝑃𝑓 ∗ 𝐸𝑒 ,𝑓 + 𝑃𝑟 ∗ 𝐸𝑒,𝑟
𝐸𝑒 ,𝑆𝑇𝐶 1𝑏
Model "grey"
𝑌𝑒𝑥𝑝 ,𝑔 𝑘𝑊ℎ = 𝑃𝑔 ∗𝐻𝑒,𝑓
𝐸𝑒 ,𝑆𝑇𝐶 2𝑎
𝑃𝑒𝑥𝑝 ,𝑔 𝑊 = 𝑃𝑔 ∗𝐸𝑒 ,𝑓
𝐸𝑒 ,𝑆𝑇𝐶 2𝑏
Model "white"
𝑌𝑒𝑥𝑝 ,𝑤 𝑘𝑊ℎ = 𝑃𝑤 ∗𝐻𝑒,𝑓
𝐸𝑒,𝑆𝑇𝐶 3𝑎
𝑃𝑒𝑥𝑝 ,𝑤 𝑊 = 𝑃𝑤 ∗𝐸𝑒,𝑓
𝐸𝑒 ,𝑆𝑇𝐶 3𝑏
Sometimes [9] the equivalent peak power (Ppe), is
used to describe the peak power of bifacial modules. This
corresponds to Pexp,f&r if it is assumed that Pf = Pr and
Ee,r = 0.25 * Ee,f, i.e. Ppe = 1.25 * Pf.
6 ENERGY YIELD ESTIMATION
Using equations (1a), (2a) and (3a), the estimated
energy yields in the considered period of six months are
compared with the values measured outdoor, as shown in
Figure 6.
Figure 6: Measured and estimated energy yields from
1st March to 31st August 2014
The model "front&rear" estimates accurately the
energy yield of modules 5 to 8, with a difference between
estimated and measured values within 1.5%. However,
Modules 1 to 3 performed worse in El Gouna than
estimated. The difference between the measured energy
yield of module 1 (module 2) and the estimation of the
model "front&rear" is around 7% (6%). As modules 1
and 2 are p-type cell modules, LID could explain part of
this difference. Another source of error could have been
introduced by incongruence in the indoor measurements
under STC (viz. temperature, irradiance or lamp
spectrum), as modules 1 and 2 were measured by the
producer with a different flasher as all modules, which
were characterized at PI Berlin. The poor performance of
module 3 is explained in detail in section 6. The model
"white" also gives similar values as the model
"front&rear", with a difference within 1.7% between
estimated and measured energy yields for modules 5 to 8.
The highest difference between the estimations of the
model "grey" and the outdoor measurements is 6.7%.
7 POWER PERFORMANCE RATIO
We define the power performance ratio PR* as the
ratio of the measured power to the expected power:
𝑃𝑅∗ =𝑃𝑚𝑝𝑝
𝑃𝑒𝑥𝑝 4
where Pmpp is the power output at the maximum power
point measured in El Gouna at an interval of five
minutes. In order to understand under which conditions
module 3 performed worse than expected (and differently
than the other modules), PR* is plotted versus the day of
the year (Figure 7a), the local time (Figure 7b), Ee,f + Ee,r
(Figure 7c), the azimuth (Figure 7d) and the ambient
temperature (Figure 7e), being Pexp calculated with the
model "front&rear". Each point in the graph is given at
an interval of five minutes. Module 2 and module 6
behaved similarly to module 1 and module 5,
respectively, and have been omitted from the graph for
simplicity.
Figure 7a: PR* versus the day of the year
Figure 7b: PR* versus the local time
Figure 7c: PR* versus Ee,f + Ee,r
Figure 7d: PR* versus the azimuth
Figure 7e: PR* versus temperature
As can be seen in Figures 7a, 7b, 7c and 7d, PR* of
module 3 almost never reaches the unity. Moreover,
unlike for all other modules, PR* drops in the early
morning or late afternoon, for low values of irradiance
and for azimuth values lower than 90° or greater than
270°. This is an effect caused by a mismatch in refractive
index between the glass and the solar cells, whose AR
coating had not been optimized for an encapsulant free
module. The module was measured indoor only by
flashing it perpendicularly, while the sun shined mostly at
angles different from 90°, where the light trapping of the
module was less effective and this lowered the energy
yield. In order to limit the optical discontinuity, an
improved new series of NICE modules has been
developed, which features an AR layer not only on the
outer surface of the front glass, but also on the inner
surface.
8 OUTDOOR EFFICIENCY
Besides indoor front side and rear side module
efficiencies, calculated as:
ƞin,front =𝑃𝑓
𝐸𝑒 ,𝑆𝑇𝐶 ∗ 𝐴 5
ƞin,rear =𝑃𝑟
𝐸𝑒 ,𝑆𝑇𝐶 ∗ 𝐴 6
where A is the module area (1.66 m2), an outdoor pseudo
module efficiency and an outdoor bifacial module
efficiency can be respectively defined as:
ƞout,pseudo =𝑌
𝐸𝑒 ,𝑓 ∗ 𝐴 7
ƞout,bif = 𝑌
𝐻𝑒,𝑓 + 𝐻𝑒,𝑟 ∗ 𝐴 8
where Y represents the yield of the considered module
from 1st March to 31st August 2014, expressed in kWh.
Thus, ƞin,front and ƞin,rear indicate a ratio between powers,
while ƞout,pseudo and ƞout,bif indicate a ratio between
energies. In equation (8) it may be discussed to divide
additionally by a factor of 2 (for the two module sides),
resulting in a true physical value accounting for both the
true available radiant exposure and the true module area.
Indoor and outdoor efficiencies and pseudo efficiencies
of each module are shown in Figure 8.
Figure 8: Indoor and outdoor module efficiencies and
pseudo efficiencies
As expected, ƞout,bif always lies between ƞin,front and
ƞin,rear. Module 7 has the highest front side efficiency,
when measured indoor, however modules 5 and 6 show a
higher ƞout,bif, thanks to the big contribution of the rear
side. In the case of the monofacial modules the rear side
contribution is zero, which reduces considerably ƞout,bif.
ƞout,bif is suggested for a fair comparison between
monofacial and bifacial modules.
10 CONCLUSION
Mathematical models based on indoor measurements
of power and outdoor measurements of irradiance were
built in order to estimate the outdoor performance of
monofacial and bifacial modules. The model
"front&rear" estimates accurately most of the studied
modules, proving that, for a given installation site with
known irradiances, the potential energy yield of modules
with different indoor Pf and Pr can be calculated by a
simple formula. Where the model does not match the
outdoor measurements, a detailed analysis of the
disagreements between measured and expected values
was carried out. LID could only explain part of the
disagreements. The main reasons are incongruence in
indoor measurement parameters under STC (viz.
temperature, irradiance or lamp spectrum) or a mismatch
in refractive indices within the solar module. The absence
of an encapsulant can significantly decrease the energy
yield, if the AR coating of the cells is not optimized for
an encapsulant-free module or if the glass on the front
side features only one AR layer on the outer surface. In
order to limit this optical discontinuity, an improved new
series of NICE modules has been developed, which
features an AR layer also on the inner surface of the front
side glass.
A formula calculating the outdoor bifacial module
efficiency allows for a fair comparison between
monofacial and bifacial modules.
11 REFERENCES
[1] A. Cuevas, A. Luque, J. Eguren, J. Del Alamo, 50 per
cent more output power from an albedo-collecting flat
panel using bifacial solar cells, Sol. Energy, vol. 29, pp.
419–420 (1982).
[2] G. Sala, J. Eguren, A. Luque, S. Lopez Romero, M.
C. Gonzalez, M. A. Tapia, M. J. Calleja, Albedo
collecting photovoltaic bifacial panels, Procedings oh the
5th European Photovoltaic Solar Energy Conference,
Athens, pp. 565–569 (1983).
[3] S. Goda, Cell mass production and array field
demonstration of n-type bifacial 'EarthON', nPV
Workshop, s' Hertogenbosch (2014).
[4] F. Madon, O. Nichiporuk, R. Einhaus, NICE bifacial
module technology, 6th World Conference on
Photovoltaic Energy Conversion, Kyoto (2014).
[5] J. Libal, V.D. Mihailetchi, R. Kopecek, Low-cost,
high-efficiency solar cells for the future ISC Konstanz’s
technology zoo, PV international, issue 23 (2014).
[6] D. Kania, T.S. Böscke, M. Braun, P. Sadler,
C. Schöllhorn, M. Dupke, D. Stichtenoth, A. Helbig,
R. Carl, K. Meyer, A. Grohe, J. Lossen,
H.J. Krokoszinski, Pilot Line Production of Industrial
High-Efficient Bifacial n-type Silicon Solar Cells with
Efficiencies Exceeding 20.6%, Proceedings of the 28th
European Photovoltaic Solar Energy Conference, pp.
1383-1385, Paris (2013).
[7] Tom Markvart, Luis Castañer, Practical Handbook of
Photovoltaics: Fundamentals and Applications, Elsevier,
p. 78 (2003).
[8] Tetzlaff, G., Albedo of the Sahara, Cologne
University Satellite Measurement of Radiation Budget
Parameters. pp. 60–63 (1983).
[9] MegaCell, [online]. Available: http://www.mega-
group.it/megacell/en/the-bison-solar-cell/the-bison-solar-
module/