Shading Analysis & Improvement for Distributed Residential Grid-

Connected Photovoltaics Systems

Peter Bulanyi, Rodd Zhang
Si Clean Energy Pty Ltd, Coffs Harbour, NSW 2450 Australia,
Phone: (02) 6651 4548, Email: [email protected]
Keywords: PV system, Bypass Diode, Partial Shading

Abstract
Despite incentive reductions, distributed residential PV Systems continue to be
installed throughout Australia in large numbers. Currently there are more than one
million residential Photovoltaic (PV) systems installed across Australia (Australian
Clean Energy Council, 2013). One of the significant challenges for residential roof
top systems is the shading effects caused by various obstructions including trees,
roof top structures and neighbouring buildings. Different levels of shading will lead to
extraordinary performance variations in both individual solar modules and the entire
PV array. This is primarily due to the operation of module bypass diodes contained
inside the junction box. This paper will explain various shading scenarios and
describe shading categories and classifications. Following the comprehensive
research and analysis of module bypass diodes functionality and their pros and
cons, a series of proposed installation solutions will be recommended to reduce
shading losses.

1. Introduction
Numerous factors can affect Photovoltaic (PV) system performance. Shading is one
of the most significant issues for designers and installers to contend with. Generally,
shading refers to a shadow on the PV panels on a rooftop that will reduce the system
energy yield. PV shading issues can present complications in addressing true PV de-
rating factors since the three primary PV panel characteristics of Power, Voltage and
Current may each be affected by varying degrees. As a consequence, there are a
limited number of documents systematically analyzing the impact of shading on PV
system performance. (Masoum, Padovan and A.S. Masoum, 2010) This paper will
explain the different types of shading, suggest a shading classification system and
offer practical solutions to improve system output in the event that the PV array will
be subjected to shading effects.

2. Shading Types

Shading is not a simple phenomenon. Haze, cloud, dust, trees, bird droppings,
buildings and roof-top structures can create infinitely variable shading. Different

Solar2014: The 52nd Annual Conference of the Australian Solar Council

 
orientations and tilts can lead to different shading shapes. These can be large or
small spots, dappled or hard edged, linear or irregular in nature. Some shading
types can cover an entire single cell of a module, some can cover an entire module
of the array and some create various shading on either strings, module or cells. To
categorize, there are two broad groups of shading and these are Objective Shading
and Subjective Shading. Refer to figure 1. Within Subjective Shading there is a
further two categories and these are Dynamic Shading and Static Shading.

1.1 Objective  and  Subjective  Shade  

Objective Shading that is formed by the weather or heavy pollution, such as heavy
cloud or haze, will inevitably reduce overall sunlight intensity. Subjective Shading is
caused by proximity objects both near and far that can block sunlight and thereby
create solid shade shapes. Objective Shade is unavoidable but Subjective Shade
can largely be improved or prevented with good design, installation and periodic
cleaning.

1.2 Static  and  Dynamic  Shade  

Subjective Shading is further divided into either Static Shading or Dynamic Shading.
Static Shading refers to shade caused by close proximity obstructions that have
typically adhered to the glass, such as bird droppings, leaves or accumulated dirt on
a module’s bottom edge. This type of shade is dependent on the initial random
positioning of the shade barrier only and does not change during the course of the
solar day. Static Shading is fixed and unaffected by the angle of the sun. Dynamic
Shading is typically caused by buildings, trees, narrow spaced array tilt frames and
other such objects that create a shadow based on sun angle. Dynamic Shading is
subject to the angle of the sun and therefore changes dynamically during the course
of the solar day.

1.3 Whole  and  Partial  Shade  

There is simply too much ambiguity when trying to use the term Partial Shading in
reference to a solar module, string or PV array. To maintain consistent language,
!!!"#"  !"#!
Whole Shading and Partial Shading should only refer to cell shading ratio .
!"##  !"#!

Whole Shading and Partial Shading references the threshold point below which there
is insufficient irradiance to generate cell voltage (Whole Shade) and above which
there is sufficient irradiance to generate cell current (Partial Shade). Whole or Partial
shading of a cell will lead to different effects on module voltage or current or both. A
system to quantify shading of modules, strings, and PV arrays will be proposed in
this paper.

Solar2014: The 52nd Annual Conference of the Australian Solar Council

 
 Shade

Improvable?
Yes No

Subjective Objective

Variable?
Yes No Weather  

Dynamic Static

Partial  -­‐  Whole Partial  -­‐  Whole

e.g.  Buildings e.g.  Bird  Droppings

Fig. 1 Shade categories

3. Shading Effects

Conventional thinking on shading effects is generally limited to the ensuing Power

reduction. However, the specific effects of shade on either PV voltage or current or
both can be significant and an understanding is essential when tasked with installing
PV arrays, reviewing system performance, optimizing energy production and
increasing system reliability. (Chin, Neelakantan, Yang, etc. 2010)

When Objective Shading appears, on a heavily cloudy day for instance, irradiance
will be significantly reduced. However, once irradiance reaches the minimum
required threshold (typically around 50W/m2) the PV array open circuit voltage (Voc)
may be significant. Thus under an Objective Shading condition, even though the
output current may be extremely low or even zero, it doesn’t mean the system is
dormant. The PV array may still generate a high voltage. Figure 2 shows how
Objective Shading affects a PV IV curve: the current is strongly influenced by the
irradiance whilst voltage is comparatively not.

Solar2014: The 52nd Annual Conference of the Australian Solar Council

 
Figure 2. Objective Shading effect on IV curve of Kyocera KD240GH-2PB solar module

The effects of Partial and Whole Shading are very interesting as the ratio of shade covering a
cell determines whether the cell is operating in a current or voltage mode. As a cell becomes
Partially Shaded, the unshaded portion remains operational with virtually no change to
voltage and it continues generating albeit at a reduced current. As the partial shade increases
there is a threshold point at which the cell transitions to Whole Shading and the cell voltage
collapses. The practical implication here can be very important. For example, let’s say there
is a bird dropping, (and it’s a large Pelican one!), circular in shape, approximately 100mm,
diameter and it is positioned at the centre of four cells. A corner of each cell is partially
shaded. In this case each cell is partially shaded only 25% and there is a proportional drop in
cell current, but each cell’s voltage remains virtually unchanged. If the same bird dropping
was positioned over a single cell however, the effect couldn’t be more different. In this case
not only current is proportionally reduced but the cell has transitioned to a Whole Shade
voltage mode in which cell voltage collapses. This can have serious consequences on
module output power as will be explained in detail later.

In order to fully understand Subjective Shading effects, it is necessary to introduce

the module’s built in protection device, the “Bypass Diode”.

4. Bypass Diode Effects

The purpose of Bypass Diodes is to shunt module current around Whole Shaded
cells. This is so as to minimise possible overheating, cracking or potentially burning
due to localised cell hot-spot effects. (Pachpande and Zope, 2012). The operating
principle of a solar cell with a Bypass Diode (Wenham, Green, Watt and Corkish) is
such that during normal operation without shade, the Bypass Diode is reverse biased
and has high impedance, therefore it is inactive and no current flows through it.
When the cell is Whole Shaded, neighbouring cell voltages forward bias its Bypass
Diode which then transitions to an active state. In the active state, the Bypass Diode
impedance becomes lower than the shaded cell and module current shunts around
the shaded cell through the diode. This prevents the unshaded cells from forcing a
current against the reverse biased state of the shaded cell; therefore hot spot
overheating is prevented.

Solar2014: The 52nd Annual Conference of the Australian Solar Council

 
Unfortunately, for c-Si modules it is not cost effective to provide a Bypass Diode
across each cell. Instead, the cells are grouped into three sub-strings and one
Bypass Diode per sub-string is installed as per Figure 3. The above described
Bypass Diode operation is now grouped into each sub-string rather than each cell.
Therefore in the event of Whole Shading of a single cell, the neighbouring sub-
strings forward bias the Bypass Diode associated with the shaded cell’s sub-string.
As a result, the voltage of the entire Whole Shaded cell sub-string collapses down to
the forward bias voltage of the Bypass Diode (approximately 0.6V). This represents
the equivalent loss in module voltage of approximately 1/3. This is certainly not good,
but by far a lesser evil than an overheating, cracking or burning cell. (Figure 4)

Figure 3: Bypass Diode arrangement in a 60 cell module

Figure 4: Voltage variation when Whole Shading apply

In the case of Partial Shading of a cell, the effect is vastly different. According to
Masoum, Padovan, and A.S. Masoum, when irradiance is above the minimum
required, only current will be significantly affected. This means that cell voltage will
remain normal, but current will significantly drop according to the cell shade ratio.
Refer to Figure 5, it is only when irradiance cannot meet the minimum requirement

Solar2014: The 52nd Annual Conference of the Australian Solar Council

 
and cell voltage collapses that module current is shunted through its associated
Bypass Diode.

Figure 5: I-V curve and P-V curve characteristics of 12 series connected solar cells
(with bypass diode) experiencing non-uniform partial shading of 25 (green), 50 (red),
75 (mint) and 100 (purple) percent, respectfully. (Masoum, Padovan and A.S.
Masoum, 2010)

Hence, module current drop may be caused by Objective Shading due to weather, or
Partial Shading due to an object (Figure 6c Partial Shaded cell) and a module
voltage drop may be caused by Whole Shading as well as the more widely
remembered cause of high ambient temperature (Refer to Figure 6, a. Module with
unshaded cells, b. Whole Shaded cell and c. Partial Shaded cell).

Figure 6: a. Module with unshaded cells, b. Whole Shaded cell and c. Partial Shaded cell

From a practical sense, it should never be forgotten that Bypass Diodes can and do
fail. This can be due to long periods at high current and high temperature when they
are actively bypassing shaded cells, or due to their Peak Inverse Voltage rating
being exceeded such as when a nearby lightning strike occurs. (Haerberlin, 2007).
The practical way to determine the health of Bypass Diodes is with an I-V curve
tracer such as the Solmetric PVA-1000 (Solmetric Inc). By analysing PV array,

Solar2014: The 52nd Annual Conference of the Australian Solar Council

 
string, or module I-V curves, it is possible to quickly identify anomalies due to short
circuit Bypass Diodes.

Conventional thinking around Bypass Diodes tends to focus on “Bypass Diodes can
prevent power dissipated by heat in shaded cells and greatly improve system
performance and ensure output efficiency”, which of course is true, but the fact is
that Bypass Diodes can also lead to unexpected performance based on the above
described points.

Consider for example the 60 cell portrait oriented modules in Figure 7 (60 cell
modules with Whole Shaded cells. Figure 7a indicates a single Whole Shaded cell
on the bottom row. The active Bypass Diode disables the entire 20 cell sub-string
and module output voltage will drop by approximately 1/3. Even more dramatic is
Figure 7b in which all bottom row cells are shaded causing all three Bypass Diodes
to be active, and the entire module to be effectively disabled. This disproportional
effect scales upwards from cells to modules to strings to arrays and ultimately results
in I-V curves that may be severely distorted and include multiple maximum power
points.

Figure 7: 60 cell modules with Whole Shaded cells

A distorted PV array IV curve can have serious consequences when inverter

Maximum Power Point Tracking is introduced into the PV system. If sufficient Bypass
Diodes become active, it is possible that the string voltage collapses below the
inverter’s minimum tracking voltage. As a consequence, during the duration of the
shade effect, inverter on/off cycling may enter a loop that reduces the life of the
inverter’s grid protection relays. Also, depending on the MPPT algorithm employed,
the inverter may lock onto an incorrect sub-power point for significant periods of time
resulting in significantly reduced energy yields. Refer to Figure 8, Partial Shade
effects on a PV array with two parallel strings.

Solar2014: The 52nd Annual Conference of the Australian Solar Council

 
Figure 8: Partial Shade effects on a PV array with two parallel strings – with multi MPPs

Whilst it is undeniable the benefits and safety that Bypass Diodes provide, the
problem is that no matter whether one cell or all of the module sub-string is Whole
Shaded, the consequence is actually the same. Therefore it is not the number of
Whole Shaded cells that determine output power, but the number of active Bypass
Diodes.

5. Shade Classification System

To allow PV installers, electricians, technicians and engineers to communicate

effectively with regards to shade effects a new system of shade classification is
proposed. The system speaks in terms of Active Bypass Diodes. For example a
shaded module that has one active Bypass Diode is referred to as 1BP, two active
bypass diodes is described as 2BP, and three as 3BP. Refer to Figure 9.

Figure 9 Active Bypass Diode shade classification (e.g. three active bypass diodes = 3 BP)

To scale the shade classification up to a string of modules, the number of Bypass

Diodes in the string is tallied and becomes a dividend in the description. For
example, in Figure 10 a string of five modules has 5×3 = 15 Bypass Diodes,
therefore the dividend is 15. The active Bypass Diode are then tallied as the divisor,
in this case 3 and the resulting shade classification becomes 3BP/15, which is much

Solar2014: The 52nd Annual Conference of the Australian Solar Council

 
easier and more accurate for anyone to describe than “there’s quite a lot of shade
on the bottom panel, and a bit also on the next one up”.

FIGURE 10 String shade classification

To scale the shade classification up to a PV array of multiple strings, each string has
its active Bypass Diodes tallied, then separately listed in the divisor. The dividend
remains as the tally of string Bypass Diodes, which of course remains constant
across the PV array. For example, referring to Figure 11, there is one string with no
active Bypass Diodes, and one string with 7 active Bypass Diodes, resulting in a
shade classification of (0BP + 7BP)/15. Again, this is an accurate and simple way to
record and communicate shade effects. In turn, this becomes useful to installers to
consider how an array may be configured to an inverter, or for an engineer to model
expected performance.

Figure 11 PV array shade classification

Solar2014: The 52nd Annual Conference of the Australian Solar Council

 
6. Improvement Methods

Currently there are no economical methods to significantly improve Objective

Shading due to weather. Static Shading due to bird droppings, leaves or dirt can be
easily removed with a regular cleaning schedule. For sites affected by Dynamic
Shading the following methods are recommended to minimize losses.

1. A site inspection with a shade analysis device such as the Solmetric SunEye
210 (Solmetric inc.) is a desirable method to accurately quantify the shading.
Its built in instrumentation and wide-angle fisheye lens capture the PV array’s
skyline including any Dynamic Shade objects. Image analysis and
computation of the clear sky to shade ratio may then be directly displayed in
various formats. Energy yields may be calculated with optional modeling
software.
2. Installation orientation is another significant factor. For example, if shading
along the bottom edge of a module is inevitable, using landscape orientation
!
instead of portrait may reduce the module losses by as much as .
!
3. If Dynamic Shading affects up to 10% of the array modules, the shaded
modules may be evenly distributed across each string and the inverter MPPT
functionality will minimise array output losses. If Dynamic Shading affects up
to 20% of the array modules it is recommended to quarantine the shade
affected modules into the same string and utilize a multi-MPPT inverter
keeping “healthy and unhealthy” strings separated. It is not recommended to
wire heavily shaded strings to unshaded strings due to voltage mismatch. If
Dynamic Shading affects 20% or more of the PV array modules it is time to
consider switching topology to Microinverters or Power Optimizers.
4. For a non-shaded east-west facing roof top system, in order to maximize daily
energy production, it is critical to keep string modules on the same orientation.
Strings must not be wired across the gable to include both east and west
facing modules. It is preferable to utilize a dual MPPT inverter with one input
connected to the east facing strings, and the other input connected to the
west facing strings. In the event that a single MPPT inverter must be installed
this is possible. The shade effects in this case with the rising or setting sun
are akin to Objective Shading on a whole string and are affecting current
output rather than voltage output. In all east plus west roof top systems, the
North/South aligned gable orientation should not be more than a few degrees
off True North/South, otherwise one of the arrays will have a non-Equator
orientation and East-West sloping tilt frames may be the best option for the
installation.
5. Front row Dynamic Shading is always a serious problem for tilted rooftop or
ground mounted arrays. (Passias and Kallback, 1984) It is recommend to
mathematically calculate the unshaded row spacing before installation by
𝑐𝑜𝑠(𝑎𝑧𝑖𝑚𝑢𝑡ℎ  𝑎𝑛𝑔𝑙𝑒)
𝐷!"#  !"#$%&' = sin(  𝑡𝑖𝑙𝑡  𝑎𝑛𝑔𝑙𝑒)×𝐿!"#$%& × ×1.2
tan  (𝑎𝑙𝑡𝑖𝑡𝑢𝑑𝑒  𝑎𝑛𝑔𝑙𝑒)
Also it is recommended to simulate the row spacing with professional software
to double check and improve the design accuracy.
6. Inverter selection cannot be neglected. Inverters with excellent flexibility
including a broad input DC voltage window can improve the power loss due to
shade induced voltage mismatch. Single MPPT inverters with multiple parallel

Solar2014: The 52nd Annual Conference of the Australian Solar Council

 
 strings should be avoided where Dynamic Shading may create multiple power
points on the array P-V curve. The MPPT tracker can lock onto the wrong
peak power point and significant yield losses may occur. According to
Masoum, Padovan and A.S. Masoum’s testing results, several popular
algorithms (P&O, CMPPT and VMPPT) show limitations in correctly locating
the true Maximum Power Point. Otherwise module level MPP tracking devices
like power optimisers and microinverters can be applied to effectively resolve
this issue.

7. Conclusion
Bypass diodes are a necessity to minimize c-Si “hot spot” damage. They can
however have a disproportionally negative impact in that one active Bypass Diode
can take out an entire module sub-string when Subjective Shading is affecting as
little as one cell. This disproportional problem scales up as modules affect strings
and strings affect arrays. Therefore if shade affects a single solar module, or
modules in a string, or strings in an array it is important to think not in terms
of cells that are shaded, but to think in terms of Bypass Diodes that are active.

Shading can be a seriously challenging factor when designing and installing a roof
top PV system. Shade diversity and variability can make performance unpredictable.
PV systems are difficult, if not physically, impossible to relocate once installed which
highlights the critical nature of shade affects. This paper has explained and
demonstrated different types and effects of shading. The recommended mitigation
methods can assist to effectively increase PV system energy production and the
proposed shade classification system can assist all who are associated with shade
affected PV to communicate with increased speed and accuracy.

In conclusion, when shade affects a string there is no formula to overcome the

shading de-rating. However, by combining theoretical knowledge and practical
experience a responsible and satisfactory outcome for many installations that are
affected by shade is possible.

 
 

Solar2014: The 52nd Annual Conference of the Australian Solar Council

 
Reference  
 

“Analytical Modeling of Partially Shaded Photovoltaic Systems”, Energies 2013, 6, 128-144;

doi:10.3390/en6010128

A. Masoum, F. Padovan, M. A.S. Masoum 2010, “Impact of Partial Shading on Voltage- and
Current-Based Maximum Power Point Tracking of Solar Modules”, Institute of Electrical and
Electronics Engineers

C.Chin, P. Neelakantan, S. Yang, etc. 2010, “Effect of Partially Shaded Conditons on

Photovoltaic Array’s Maximum Power Point Tracking”, Vol.12, No.3, Institute of Electrical
and Electronics Engineers

Clean Energy Council, “How Solar PV Works”,

“https://www.solaraccreditation.com.au/consumers/purchasing-your-solar-pv-system/how-
solar-pv-works.html” access:19/12/2013

D. Passias and B. Kallback (1984), “Shading Effects in Rows of Solar Cell Panels”, Institute
of Microwave Technology, Sweden, page 281-291

H. Chang, C.Chen, L. Huang, C. Lin (2010), “Implementation of a Power Supply with the
Characteristics of Solar Panels under Partially Shaded”, Institute of Electrical and
Electronics Engineers

H. Haeberlin (2007), “Damages at Bypass Diodes by Induced Voltages and Currents in PV

modules Caused by Nearby Lightning Currents”, 22nd European Photovoltaic Solar Energy
Conference, Milano, Italy

Solmetric, Inc. “http://www.solmetric.com/”  access: 15/12/2013

S. Pachpande and P. Zope (2012), “Studying the effect of shading on Solar Panel using
MATLAB”, International Journal of Science and Applied Information Technology, Volume 1,
No. 2

S.R. Wenham, M.A. Green, M.E. Watt & R. Corkish 2007, “Applied Photovoltaics (2nd
Edition)”, Earthscan, London

Solar2014: The 52nd Annual Conference of the Australian Solar Council