Content uploaded by Sebastian Pingel
Author content
All content in this area was uploaded by Sebastian Pingel on Nov 24, 2014
Content may be subject to copyright.
RECOVERY METHODS FOR MODULES AFFECTED BY
POTENTIAL INDUCED DEGRADATION (PID)
S. Pingel, S. Janke and O. Frank
SOLON Energy GmbH, Am Studio 16, 12489 Berlin, Germany
Tel.: +49 30 81879-0, Fax: +49 30 81879-9999, Sebastian.Pingel@solon.com
ABSTRACT: PID of standard PV cells has been in the discussion in the last few years. The symptoms of PID have
been investigated and some ideas for the mechanism behind were published. This paper focuses on the recovery of
PID affected modules both in the field and in the lab. Recovery data of two small systems that were setup in Berlin
with PID field returns is presented. Two methods, proper grounding on one hand and applying a regeneration
potential in the night on the other hand, are compared. In the laboratory four methods are investigated: storage at
room temperature, applying a defined (reversed) potential, treatment with temperature and bias voltage. PID is
shortly introduced in case of bifacial cells. This type of cells may be affected by both negative and positive potential
since typically both sides are dielectrically passivated.
Keywords: Crystalline solar cells, Reliability, Degradation, PID
1 INTRODUCTION
In the past two years the discussion about “Potential
Induced Degradation” (PID) of standard Si-PV modules
has developed and currently effort is undertaken to define
a standard test procedure for PID, see e.g. [1].
In this paper the PID effect is shortly introduced for
n- and p-type c-Si bifacial cells. Depending on the
construction of bifacial cells these may be affected by
both negative and positive potential because both sides
are often passivated by a dielectric layer that may be
prone to PID. For standard p-type Al-BSF cells only the
front side is passivated by a dielectric layer. This layer
may depending on the cell process reduce the risk of PID
to occur in negative or positive potential (typically
negative potential versus ground).
The focus of recent papers [2-9] was to identify the
drivers and understand the mechanism of PID that lead to
degradation of modules in the field or in laboratory tests.
In case of p-type standard cells the effects of negative
potential relative to ground and environmental factors
(humidity and temperature) but also cell and module
design parameters were discussed [2],[5],[6] an important
role is attributed to sodium ions that accumulate on the
surface of the PV cell [7-9] .
This paper focuses on the recovery of PID affected
panels with methods applicable in the laboratory or in the
field. In the latter case modules may be recovered by
proper grounding of the PV system or by applying a
temporary defined potential (e.g. SMA PV Offset Box
[10]). Outdoor recovery data is shown and the two
methods are compared. In the laboratory recovery of PID
affected modules is investigated (1) by storage at room
temperature, (2) by applying a defined (reversed)
potential, (3) by treatment with temperature and (4) bias
voltage. The different methods are compared.
2 EXPERIMENTAL
For characterization of single cell coupons and panels
prior and after the PID test a flash tester and a high
resolution electroluminescence (EL) camera was utilized.
EL images were taken with high or low current and also
reverse bias images were taken to investigate PID and
recovery effects.
The setup for degradation/recovery was varied from
case to case depending on the test, in all cases +/- 1000V
was the voltage used. Single cell laminates were treated
in an oven with a temperature of 60-85°C. In case of
mono facial coupons an aluminum foil was used to
supply a conductive layer on the glass. In case of bifacial
modules the glass-glass coupons were wrapped in
aluminum foil to assure that both sides are exposed
equally to the applied potential. In case of large panels a
setup working between room temperature and 40°C with
water as conductive layer was utilized. Temperature
recovery was done in large and small ovens or climatic
chambers with a temperature of 60-85°C and dry
atmosphere. Simple storage was done at room
temperature around 25°C. For bias treatment a simple
power supply was used to provide current and voltage.
3 PID EFFECT ON BIFACIAL MODULES
Bifacial cells are using light from both sides were
front and back surface are typically passivated by a
dielectric layer (e.g. SiNx). These cells are of special
interest for their PID susceptibility since both sides may
be affected by PID.
Figure 1: IV parameters of glass coupons with p-type and
n-type solar cells that were exposed to 85°C and both
negative and positive external potential of 1000V.
Different types of bifacial solar cells were
encapsulated in EVA between two glass plates. The
coupons were wrapped in aluminum foil and tested in an
oven for PID susceptibility were both polarities (+/-
1000V) were applied between the solar cells and the
aluminum foil for three hours. Thereby a leakage current
from the back or front side of the solar cell may flow
through EVA and glass to (or from) the aluminum foil.
The IV-characteristics of the coupons (front & back)
were recorded initial, before changing polarity and at the
end of the test. The results are graphically visualized in
Fig.1. The p-type bifacial cell responds like a typical p-
type standard solar cell (degradation of shunt resistance,
FF drop and slightly changing Isc and Voc) when the cell
is on negative potential to the aluminum foil. After
changing the polarity the cells recover slowly, because
shunt and FF recovery are incomplete the initial power is
not reached after applying the reversed potential in the
test time. Both front and back side measurements are
affected similarly by PID.
In case if a n-type cell also the first test step (cell on
negative potential) leads to a significant degradation, but
the mechanism is different. FF is stable while Isc
degrades asymmetrically on the front side by more than
10% and Voc drops for both front and backside by ≈6%.
This result indicates that the passivation layer on the p-
type front side is affected by the negative potential
(attracting positive ions to the surface) leading to an
increase in surface recombination velocity. The p-n
junction seems to be much less affected than in case of
the investigated p-type bifacial cell since the shunt is
degrading significantly less (stable FF). After changing
the polarity the parameters recover almost to the initial
levels. These results are similar to the ones presented by
Sunpower were in case of the A300 an n-type wafer with
n+ front surface field was passivated by SiO and an AR
coating [11], major degradation was also in Isc and Voc
besides the FF, but in that case the harmful potential was
positive.
Figure 2: PID power degradation of a bifacial cell for
both polarities.
In another test of bifacial cells PID susceptibility for
both polarities were found, as shown in Fig. 2. The front
side recovers after power degradation in negative
potential but the back side degrades further after
switching to positive potential.
This investigation shows that in case of bifacial cells
two passivated sides may be affected by PID. For bifacial
modules it may not be possible to solve PID completely
in terms of grounding or defining the potential. But an
adaption of the cells passivation layers or the panel
design, e.g. the encapsulation [7], enables to use such cell
technologies also in case of larger systems and high
voltages.
Tab 1 sums up the experimental results, but results vary
significantly with the type of cells tested also within a
group of similar cells concepts. No claim to be complete.
Table 1: Degradation of different cell types in PID test
for both polarities. Affected parameters in order of
occurrence are indicated.
Cell type
Cell neg. to grd.
Cell pos. to grd.
Std. p-type
Rsh, FF, Isc, Voc
Recovery
Bif. p-type
Rsh, FF, Isc, Voc
symmetric
Recovery
Bif. n-type
Isc, Voc, Rsh, FF
some asymmetric
(most) recovery
IBC n-type
Recovery
Isc, Voc, FF
4 RECOVERY IN THE FIELD
In case PID occurs in the field, depending on the
grade of degradation, the most economical solution
maybe to recover the panels because financial invest is
comparably lower than for exchanging the panels or
changing the system concept. An important question is
how fast the recovery works and how much of the initial
power may be recovered. To investigate outdoor
regeneration within a grid connected system PID
degraded modules from the field were inspected and two
systems consisting of seven panels each were setup in
Berlin. In one case the “PV – pole” was grounded and in
the other case a “SMA PV Offset Box” [10] was used
that applies a positive potential of 300-600V (max. 2mA)
during the night.
Figure 3: Relative recovery of field degraded panels
versus time by PVO Box and PV-Grounding.
Grounding
PV-Offset
Figure 4: Example EL image of field degraded modules
at the beginning of the test (left) and after fourteen month
outdoor recovery in Berlin (right) for the two methods.
As can be seen in Fig 3&4 both methods lead to a
significant recovery over time. In fourteen months more
than 50% of the degraded power is recovered. Since the
panels were initially installed in 2007 the recovery may
be almost complete when measurement uncertainties,
initial and annual degradation are taken into account.
During summer 2011 recovery was in both cases very
fast, in the following winter PVO recovery showed a
better performance. In 2012 the recovery of STC power
seems to slowly saturate, especially for PVO. In spring
and summer 2012 the grounded system was catching up
with the PVO because of higher temperature and
irradiation.
After fourteen month the recovery is for PVO ≈95%
(+7%) and for Grounding Kit ≈94% (+6%) in average. In
the following graph (Fig.5) the maximum power voltage
and current of each tested module is shown for the
degraded modules (June 2011) and the recovered
modules (August 2012). As can be seen both MPP
current and voltage improve and the initial high
mismatch of the field degraded modules is drastically
reduced after the outdoor recovery.
Figure 5: Vmpp/Impp of the field degraded modules (red)
and after outdoor recovery (blue).
Until now it can’t be decided if one of the two
procedures performs better in the field in the long run.
PVO seems to lead to an initially faster recovery but
saturates in the shown case earlier; especially in cold
climates PVO could have a significant advantage since
recovery in the winter was slowed down in the cases of
the grounded modules. On the other hand working with a
grounding kit may be an effective solution in a warmer
climate where module recovery is accelerated.
EL pictures taken with a current of 1A instead of 10A
reveal that PID recovery is still not complete for some
panels, an example is shown in Fig 5.2.
Figure 5.2: June 2011 – left EL image (7A), August
2012 – EL (10A) middle EL (1A) right.
As can be seen the cells recover in the high current
EL image completely but in the low current image it’s
clearly visible that the recovery is not completed yet. The
dark cells are still affected by PID. The modules were
measured also for their weak light performance and the
measurements show very good performance that is
comparable to new modules. Nevertheless the experiment
will be continued further to investigate the long time
behavior of both methods..
5 PARAMETERS INFLUENCING RECOVERY
5.1 Influence of room temperature storage
To investigate the effect of room temperature storage
on PID eight panels were exposed to a PID test (25°C,
1000V). The panels degraded in power by 50% in
average. Afterwards the modules were stored at room
temperature for eighteen month. A recovery of STC
power to 97% in average was found. The result is shown
in Fig. 6 and indicates that also under room temperature
conditions significant recovery occurs (in average here
0.1%/day). This result will be further discussed in the
section 5.3.
Figure 6: Relative power degradation in PID test and
regeneration by room temperature storage.
5.2 Reversing the potential
The obvious method of choice to recover a PID panel
in the laboratory is to apply a reversed potential.
Figure 7: Warm water PID (40°C, 1000V) test and
recovery by reversed potential.
In the case shown in Fig. 7 above a panel was
degraded by applying 1000V between the cell circuit
(negative pole) and the aluminum frame (positive pole)
while the glass surface was covered with warm water
(≈40°C). The increased temperature accelerates the PID
effect. The recovery was done by reversing the potential
that led to a close to 95% regeneration in the same time
frame – an elongation of the recovery is expected to lead
to further regeneration. In a previous paper [4] 100%
recovery was presented. The result is influenced by the
increase in test temperature; also the types of cells and
the module construction are different and impact the
result.
5.3 Influence of high temperature exposure
In the field the modules are running at a temperature
level of 50-60°C typically, depending on the irradiation,
ambient temperature and wind speed. As was shown in
caption 4.1 a storage at room temperature leads to
recovery of the modules. The degradation and recovery
rate depend both on the temperature level.
The result of a high temperature exposure of a single
cell coupon is shown in Fig. 8. After PID degradation
(-80%) in an oven test (1000V, 60°C for 24hr) the sample
was then recovered by temperature treatment with 85°C.
After 425h the coupon recovered in Voc and Isc almost
completely. FF (-7%) and hence power are not yet
completely recovered and the regeneration is slowly
saturating with the step by step increasing shunt
resistance. The shunt recovers only considerably after
150hr and reaches after 425hr only 11% of the initial
value. In the regarded time frame the average recovery
rate is almost 4% per day, initially the rate is high but
then recovery is staring to saturate and recovery is
slowing down rapidly.
Figure 8: IV parameter change of a single cell coupon
during temperature recovery after PID degradation.
Besides single cell coupons also full size panels were
exposed to high temperature (80°C) in a climatic
chamber. Prior to the recovery test the three modules
with varying PID susceptibility were degraded by
applying 1000V between frame and cells in a wet PID
setup (Module A: 120hr, wet, 25°C and module C: 268h,
wet, 25°C) or in the DH chamber (Module B: 44hr,
85%RH, 85°C). Module C degraded by ≈20%, A and B
≈100%.
Figure 9: Relative power and shunt resistance during
temperature recovery test with 80°C.
Figure 10: EL images of modules A,B and C after certain
time steps in the temperature regeneration test.
The degraded panels were placed in the climatic
chamber in a dry and hot atmosphere at 80°C for 340
hours without applying any potential. In Fig. 9 and 10
shown above the recovery process for panels A and C are
clearly visible while module B is not recovering (at least
in the regarded time frame). Compared to reversing the
potential the temperature treatment leads to a slower
recovery but depending on the severity of the prior
degradation also temperature recovery maybe fast. Yet
the difference is not clear why module B is not showing
any tendency to recover during the temperature
treatment.
The results and the comparison to the room
temperature storage show that recovery can be
accelerated by increasing the temperature.
For a standard PID-test as it is discussed at the
moment in the IEC committee and PID working group
led by NREL the result for temperature recovery
indicates that the temperature level should be comparable
to temperatures in the field during operation since it is
not yet clear if higher temperatures lead to activation of
further degradation mechanisms not found in the field
[12] or higher temperature may even shift the balance to
favor of regeneration versus degradation.
5.4 Bias voltage recovery
Another regeneration method for PID affected cells is
the application of a forward potential. Here a current is
flowing through the solar cell and heats the cell.
Especially when there are local shunts, as occur in case of
PID, the temperature are high since here the local current
density is even larger.
EL +1.0V/-10A
Bias -4.7V/10A
EL +1.0V/-10A
Figure 11: Effect of a short ≈15s reverse bias treatment
on EL (forward bias) image.
The effect can be further accelerated by switching
from forward (positive) to negative bias since in this case
all intact parts of the cell show the natural blocking
characteristic while areas where PID is occurring lose
their blocking characteristic and behave comparable to
ohmic shunts. In this case local heat dissipation is much
higher and as shown above temperature leads to recovery.
The pn-junction may become much hotter than the
observed cell temperature. If degradation is far advanced
in such a way that no diode like behavior is observed
anymore neither forward nor reverse bias lead to fast
recovery since then the voltage drop across the junction is
too small to lead to significant heat dissipation.
6 CONCLUSION
In this paper the PID effect on bifacial silicon solar cells
was investigated and different degradation types were
found for p- and n-type solar cells. Also cells degrading
both in positive and negative potential were tested.
Bifacial cells need to be tested for both polarities and in
case of PID counter measures need to be made on cell or
module level.
PID recovery by two methods, PVO and grounding,
that are both applicable in the field were setup in Berlin
and results were compared. The two methods lead to a
significant regeneration with good performance. The
usage of a PV Offset Box leads to an initial faster
recovery. Especially during the winter when temperature
and irradiation are low the recovery by grounding of the
negative pole is slowed down but during summer the
recovery accelerates again. After fourteen month the
results are quite similar and it is not clear yet if one of the
two methods is working better in the long run. The test
will be continued further.
Tests in the laboratory show that reversing the
external potential fast recovers PID affected modules,
also temperature treatment leads to a regeneration of the
modules but the process is slower compared to reversed
potential. As was shown also storage at room temperature
shows a significant recovery rate. For increasing
temperature it was found that recovery is accelerating.
7 REFERENCES
[1] Hacke et al.: ”Considerations for a Standardized Test
for Potential induced Degradation of Crystalline
Silicon PV Modules”, PVMRW (2012)
[2] Berghold et al.:”Potential Induced Degradation of
solar cells and panels”, 25th EU PVSEC (2010)
[3] Hacke et al.:”Test-to-failure of crystalline silicon
modules”, 35th IEEE (2010)
[4] Schütze et al.: “Laboratory study of potential induced
degradation of photovoltaic modules”,
36th IEEE (2011)
[5] Nagel et al., “Crystalline Si solar cells and modules
featuring excellent stability against potential-induced
degradation”, 26th EU PVSEC (2011)
[6] Pingel et al.: “Potential Induced Degradation of Solar
Cells and Panels”, 35th IEEE (2010)
[7] Koch et al.:”Polarization effects and tests for
crystalline silicon solar cells” 26th EU PVSEC (2011)
[8] Hacke et al.:”Characterization of Multicrystalline
Silicon Modules with System Bias Voltage Applied
in Damp Heat”, 25th EU PVSEC (2010)
[9] Bauer et al.:”On the mechanism of potential-induced
degradation in crystalline silicon solar cells”,
pssr (2012)
[10] SMA: “PV Offset Box PVO-11”, downloaded from
www.sma.de (15.09.2012)
[11] Swanson et al., “The Surface Polarization Effect in
high-efficiency Silicon Solar Cells”,15th IEEE (2005)
[12] Hacke et al., “Testing and Analysis for Lifetime
Prediction of Crystalline Silicon PV Modules
Undergoing Degradation by System Voltage Stress”,
38th IEEE (2012)
