Beruflich Dokumente
Kultur Dokumente
com
Available online at www.sciencedirect.com
ScienceDirect
ScienceDirect
Energy
Available
Available Procedia
online
online 00 (2017) 000–000
atatwww.sciencedirect.com
www.sciencedirect.com
Energy Procedia 00 (2017) 000–000
www.elsevier.com/locate/procedia
ScienceDirect
ScienceDirect
www.elsevier.com/locate/procedia
Energy
EnergyProcedia
Procedia124 (2017) 000–000
00 (2017) 532–539
www.elsevier.com/locate/procedia
panels [2]. Furthermore, to prevent shading of direct light, PV installations are designed to have all structural and
functional parts at the rear side. These items, like cables and racks, will influence the irradiance on the rear.
2. Experimental details
The rear side irradiance per cell was measured using solar cells, laminated between two transparent foils,
including an absorbing layer behind the solar cell, exactly blocking the rear of the solar cell. This set-up made these
measurement cells monofacial, but allowed the light incident on the inter-cell area of the bifacial module of interest
to be transmitted.
Indoor IV-curves were recorded with 1000 W/m2 front side irradiance by a steady-state solar simulator. Diffuse
rear irradiance was created by placing scattering white panels at 1 m distance behind the module. A black, white or
aluminium-coloured object of 10 cm wide and 25 cm high was placed at a variable distance between the module and
the scattering panels and positioned to shade two cells or four cells from the same string. The rear side irradiance per
cell was measured under the same conditions. Fig. 0 shows a sketch of the measurement set-up.
Fig. 0. Sketch of the measurement set-up. The distance between the steady-state solar simulator and the module ensures 1000 W/m2 irradiance.
The shade object can be placed between 1 and 30 cm behind the rear of the bifacial module. The white foam is placed about 1 metre behind the
module.
The effect of inhomogeneous irradiance was simulated with LT-spice by serial connection of sixty equivalent
circuit, using the one diode model. It was assumed that the light-generated current scaled linearly with the total
irradiance on the cell. The simulated module was divided in three strings with bypass diodes.
3. Results
Fig. 1 shows the distribution of the measured rear irradiance per cell for an equator-facing bifacial module on a
rooftop rack. The cell with the lowest rear irradiance, 47 W/m2 has only 55% of the highest irradiance, 83 W/m2.
However, due to the additional, homogeneous front irradiance of 845 W/m2 the total irradiance on the module is
only reduced by 2.2% compared to a homogeneous front + rear irradiance of 845 + 83 W/m2. Note that the
inhomogeneity is partly due to inhomogeneities in the albedo light and partly due to objects near the rear side, in
particular the lower irradiance at the fourth and fifth row of cells, counted from the top.
K.M de Groot and B.B. Van Aken / Energy Procedia 00 (2017) 000–000
534 Koen M. de Groot et al. / Energy Procedia 124 (2017) 532–539
Fig. 1. Measured rear side irradiance per cell for a 6x12 cell module.
Using the measured homogeneous front irradiance and inhomogeneous rear irradiance of Fig. 1, IV-curves are
simulated using one-diode approximation with variable current source in LT-spice [3]. The resulting IV-curves are
shown in Fig. 2. The homogeneous case corresponds to a total irradiance of 928 W/m2 for each solar cell; the
inhomogeneous case applies for each cell the measured irradiance with average of 909 W/m2; and the reference
case corresponds to the homogeneous irradiance, viz. 903 W/m2, where the Isc and Voc are identical to the
inhomogeneous case.
Fig. 2. Simulated IV-curves for three cases of rear side irradiance as explained in the text. The open circles indicate the maximum power point for
each curve.
To highlight the differences between these three cases, in Fig. 3 only the part of the simulated IV-curve with I >
7.2 A is shown. Although the inhomogeneous case shows a clear deviation from a regular IV-curve, this is only the
Koen M. de Groot et al. / Energy Procedia 124 (2017) 532–539 535
K.M de Groot and B.B. Van Aken / Energy Procedia 00 (2017) 000–000
case at currents > 99% of Isc. In the region of the maximum power point, the reference and inhomogeneous case are
virtually identical. This means that no cell is put under reverse bias and there is no risk of degradation or failure due
to hot spots [4].
Fig. 3. Detail of the IV curve of Fig.2. Only the part with I > 7.2 A is plotted to highlight the differences between the three cases.
Due to the inhomogeneous irradiance a power loss relative to the homogeneous case of 2.2% is simulated. But
the average irradiance of the inhomogeneous case is 909 W/m2. This is also 2.0% lower than for the homogeneous
case. The power loss is thus purely due to a lower irradiance and not due to inhomogeneous-irradiance induced
current mismatch. As it is not possible to measure IV-curves and the rear irradiance per cell under the exact same
outdoor conditions, the effect is replicated in the lab with a constant light source.
Fig. 4 shows the change of IV-curves by varying the distance of a black shading object behind the module. With
decreasing distance between the rear of the module and the object, losses in the IV-curves appear, typical for
modules with only one string partially shaded. The drop in the average current between 17 and 22 V with and
without shading is calculated as measure of the rear side shading.
536 Koen M. de Groot et al. / Energy Procedia 124 (2017) 532–539
K.M de Groot and B.B. Van Aken / Energy Procedia 00 (2017) 000–000
10.6
10.4
no shade
10.2
Current [A] 20 cm
10 15 cm
9.8 10 cm
9.6 8 cm
5 cm
9.4
2 cm
9.2 0 cm
9
0 5 10 15 20 25
Voltage [V]
Fig. 4. Measured IV-curves varying the 10 cm wide shading object’s distance. Only the part with I > 9 A is plotted to highlight the details.
The drop in current is plotted against the distance between the shading object and the rear of the module in Fig. 5.
A continuous decrease of the drop in current is observed. At a distance larger than 20 cm, the drop in current
becomes negligible. Note that the Impp of the unshaded module with the additional diffuse rear irradiance is
~10.0 A. A drop in current smaller than 0.6 A will therefore hardly effect the Pmax of the rear shaded module.
Centring the 10-cm wide shade object between two cells, instead of at the centre of a single cell, roughly halves the
drop in current (see Fig. 4) and also halves the reduction in rear irradiance (see Fig. 5) relative to a reference cell far
from the shading object.
1
shaded cell
Low Grear ‐‐> high Grear
0.8 half‐shaded cell
Drop in current [A]
measured Grear
0.6
0.4
0.2
0
0 5 10 15 20
distance [cm]
Fig. 5. The drop in current versus the shade distance. Triangles indicate the measured irradiance, with numbers in reverse direction.
Koen M. de Groot et al. / Energy Procedia 124 (2017) 532–539 537
K.M de Groot and B.B. Van Aken / Energy Procedia 00 (2017) 000–000
The irradiance on the rear side as function of the distance has been measured using the same laminates as
discussed above, see Fig. 6. At a distance of 0 cm, less than a third of the diffuse light reaches the shaded cell. With
increasing distance, the irradiance on that cell increases. At the same time, the cell next to the shaded cell starts to
receive less light. As the thickness of the shading object is only a few cm, placing the object next to a cell hardly
blocks any diffuse light. However, from the viewpoint of the neighbouring cell, the object becomes ‘more visible’
with increasing distance, thus blocking an increasingly larger part of the diffuse light on the neighbouring cell.
160
Rear irradiance [W/m2]
140
120
100
80
reference
60
neighbour
40 half‐shaded cell
20 shaded cell
0
0 5 10 15 20
distance [cm]
Fig. 6. Measured rear irradiance as function of the shade distance for three neighbouring cells.
Fig. 7 shows the effect of the colour of the object placed behind the module. The IV curves for this graph were
taken with the shade object placed 2 cm behind the module. Clearly, the black object gives a larger reduction of the
module current at 20 V. Obviously, these objects reduce the irradiance due to blocking of the diffuse irradiance in
the same way. The difference is how transmitted light interacts with the shading object. For white and Al objects, a
significant part of the light transmitted through the solar cells is scattered back to the solar cells. in this way, the
reduction of the rear incident light is partially compensated and thus a smaller drop of the module current is
observed.
538 Koen M. de Groot et al. / Energy Procedia 124 (2017) 532–539
K.M de Groot and B.B. Van Aken / Energy Procedia 00 (2017) 000–000
10.8
10.6
10.4
10.2
Current [A]
10
9.8
White
9.6
9.4 Aluminium
9.2 Black
9
0 10 20 30
Voltage V]
Fig. 7. Measured IV curves for three colours of the shaded object at 2 cm from the module.
The results presented here represent a rear irradiance of about 10% (outdoor) or 17% (indoor) of the front
irradiance. For even higher fractions of rear irradiance, of the order of 30%, due to high albedo locations or
conditions, the effect of the shading becomes stronger. Correspondingly, the measures to prevent current mismatch
and hot spots should be stronger. That means more distance between shading objects and the rear of the module,
smaller dimensions of these objects or higher reflection by these objects.
On the other hand, conditions when the rear irradiance is dominant, e.g. during the early/late hours of the summer
months at high latitude or when the front is fully covered by snow or dust, do occur. But under those conditions the
(total) rear irradiance is low anyway, not more than 0.2 or 0.3 suns. Although this could still lead to current
mismatch for poorly designed modules or systems, the effect on the already low power output is small. The low
system current, at these low total irradiances, will not allow hot spots to form as the dissipated power scales with the
square of the current.
4. Conclusions
To conclude, the measured, large variation in rear irradiance per cell due to real outdoor conditions does not lead
to increased risk of hot spots as the current mismatch is too low to cause solar cells to get under reverse bias.
Furthermore, the power loss is as small, -2%, as the small reduction in irradiance at about 10% rear irradiance.
By replicating these effects indoors at about 17% rear irradiance, it is shown that with decreasing distance the
observed drop in current due to indirect light shading increases. For thin objects close to the module, no current drop
is observed for cells not directly behind the object, but a small reduction is observed when the object is 10-20 cm
from the module. The drop in current is in good quantitative agreement with the measured reductions in rear side
irradiance. Finally, we have shown that increasing the reflectance of the near-field object, strongly reduces the drop
in current.
Acknowledgements
This work was funded by the Dutch TKI-toeslag project BING. http://www.tki-urbanenergy.nl/.
Koen M. de Groot et al. / Energy Procedia 124 (2017) 532–539 539
K.M de Groot and B.B. Van Aken / Energy Procedia 00 (2017) 000–000
References
[1] Duffie JA and Beckman WA. Solar Engineering of Thermal Processes. 2nd ed. New York: John Wiley & Sons; 1991.
[2] Yusufoglu UA, Lee TH,. Pletzer TM, Halm A, Koduvelikulathu LJ, Comparotto C, Kopecek R and Kurz H. Simulation of Energy Production
by Bifacial Modules with Revision of Ground Reflection. En Proc 2014; 55:389-95.
[3] LTspice software package: http://www.linear.com/designtools/software/#LTspice.
[4] Jansen MJ, Okel LAG, Van der schilden RA, Romijn IG, Geerligs BJ and Van Aken, BB. Improved heat dissipation for hot spots in MWT
laminates. Proc. 22nd Int. PV Sci. Eng. Conf. Hangzhou, China; 2012.