THE EFFECTS OF THE TRANSIENT AND PERFORMANCE LOSS RATES ON PV OUTPUT PERFORMANCE

Abstract

Solar photovoltaic (PV) panels experience long-term performance degradation as compared to their initial performance, resulting in lower like-per-like efficiencies and performance ratios. Manufacturers of solar photovoltaic modules normally guarantee a lifespan of more than 20 years. To meet such commitments, it is important to monitor and mitigate PV module degradation during this period, as well as beyond, to recognize maintenance and repair needs. Solar PV modules degrade over time, becoming less effective, less reliable, and eventually unusable. The effects of transient and performance loss rates on the output performance of polycrystalline silicon (p-Si) solar PV modules are the focus of this study. PV modules' electrical performance and solar energy conversion efficiency change as solar irradiance and ambient temperature change. A rise in ambient temperature or a decrease in solar irradiance, for example, all result in a reduction in performance. Large variations in operating conditions due to uncontrollable external parameters such as cloud movement and wind velocity, as well as changes in factors external to PV systems such as unexpected shading, inverter problems, and control failures, may trigger transient performance changes on PV modules output. The data used in this analysis were from the Warrenpoint site location of the Electric Supply Board (ESB) for the years 2016-2020. Clear days in winter, spring, summer, and autumn were caused by a rise in daily sunshine hours in February, May, June, and September, according to the output performance. Due to the highest amount of solar irradiation at the site location, these days saw an increase in PV output generation. According to the performance loss rates, the median degradation rates in 2016 (4.5%/year to 14%/year) and 2017 (0.1%/year to 5.2%/year) are 8.40%/year and 3.87%/year, respectively. This means that the degradation rate is greater than 1%/year, the hazardous probability is between 90% and 100%, and a severity of 10 is given (With an associated failure of corrosion in solder bonds). 2018 (-7.5%/year to 2.5%/year), 2019 (-16%/year to-23%/year), and 2020 (-5.1%/year to-10% /year) had median degradation rates of-2.75%/year,-18.23%/year, and-5.2%/year, respectively. This shows that the degradation rates are less than 1% per year, and their hazardous probabilities range from severity rank 9 to 1, or 80% to 70% to 0% safety risk. All of these factors have a negative impact on PV output performance.

Full text

International Conference on Innovations in Energy Engineering & Cleaner Production IEECP21
THE EFFECTS OF THE TRANSIENT AND
PERFORMANCE LOSS RATES ON PV
OUTPUT PERFORMANCE
Chibuisi C. Okorieimoh
School of Electrical & Electronic
Engineering
Dublin Energy Lab
Technological University Dublin
Dublin, Ireland
chibuisi.okorieimoh@tudublin.ie
Brian Norton
School of Electrical & Electronic
Engineering
Dublin Energy Lab
Technological University Dublin
Dublin, Ireland
brian.norton@tudublin.ie
Michael Conlon
School of Electrical & Electronic
Engineering
Dublin Energy Lab
Technological University Dublin
Dublin, Ireland
michael.conlon@tudublin.ie
Abstract
Solar photovoltaic (PV) panels experience long-term
performance degradation as compared to their initial performance,
resulting in lower like-per-like efficiencies and performance ratios.
Manufacturers of solar photovoltaic modules normally guarantee a
lifespan of more than 20 years. To meet such commitments, it is
important to monitor and mitigate PV module degradation during
this period, as well as beyond, to recognize maintenance and repair
needs. Solar PV modules degrade over time, becoming less
effective, less reliable, and eventually unusable. The effects of
transient and performance loss rates on the output performance of
polycrystalline silicon (p-Si) solar PV modules are the focus of this
study. PV modules' electrical performance and solar energy
conversion efficiency change as solar irradiance and ambient
temperature change. A rise in ambient temperature or a decrease in
solar irradiance, for example, all result in a reduction in
performance.
Large variations in operating conditions due to uncontrollable
external parameters such as cloud movement and wind velocity, as
well as changes in factors external to PV systems such as unexpected
shading, inverter problems, and control failures, may trigger
transient performance changes on PV modules output. The data used
in this analysis were from the Warrenpoint site location of the
Electric Supply Board (ESB) for the years 2016-2020. Clear days in
winter, spring, summer, and autumn were caused by a rise in daily
sunshine hours in February, May, June, and September, according
to the output performance. Due to the highest amount of solar
irradiation at the site location, these days saw an increase in PV
output generation. According to the performance loss rates, the
median degradation rates in 2016 (4.5%/year to 14%/year) and 2017
(0.1%/year to 5.2%/year) are 8.40%/year and 3.87%/year,
respectively. This means that the degradation rate is greater than
1%/year, the hazardous probability is between 90% and 100%, and
a severity of 10 is given (With an associated failure of corrosion in
solder bonds). 2018 (-7.5%/year to 2.5%/year), 2019 (-16%/year to
-23%/year), and 2020 (-5.1%/year to -10% /year) had median
degradation rates of -2.75%/year, -18.23%/year, and -5.2%/year,
respectively. This shows that the degradation rates are less than 1%
per year, and their hazardous probabilities range from severity rank
9 to 1, or 80% to 70% to 0% safety risk. All of these factors have a
negative impact on PV output performance.
Keywords: Transient, Performance loss rates, PV output
performance, Degradation rates.
INTRODUCTION
Solar PV panels experience long-term performance degradation
resulting in lower like-per-like efficiency and performance ratios
when compared with their initial performance [1]. Also, there are
some transient effects (such as PV ambient temperature, wind
velocity, shade, and dust particles) that reduce the output
performance of the solar PV panels [2]. Reducing rates of PV
module degradation aims to maintain the efficiency of solar PV
systems [3]. As manufacturers usually guarantee the life span of PV
modules for more than 20 years [3], it is, therefore, necessary to
track and mitigate the degradation of PV modules over this period
both during and beyond this period knowing degradation behavior
is essential for operation, maintenance, and repair [1]. Most
significantly, many PV module failures and, performance losses are
caused by the gradual accumulation of damaged PV modules due to
long-term outdoor exposure in harsh environments. This outdoor
environmental stress is known as weathering [4]. To put a check on
the outdoor installed PV modules, there is a need for accelerated
tests. Outdoor testing of PV modules may take a longer time to be
accomplished and it is impossible to wait up to 20-25 years to
introduce a new PV module. Hence, it is important to develop and
use accelerated tests to quantify or measure up the new PV modules
[5]. Such accelerated stress tests are thermal cycling, humidity-
freeze, damp heat, mechanical load (both static and dynamic), and
ultraviolet exposure [5]. When a PV module fails to generate power,
such failure will be seen as a reliability issue while a decrease in
output of a PV module is caused by environmental degradation such
as corrosion and it is classified as a durability issue. Therefore, the
durability and reliability issues may eventually lead to PV module
failure [4].
IEECP’21, July 29-30, 2021, Silicon Valley, San Francisco, CA – USA
© 2021 IEECP – SCI-INDEX
DOI : https://dx.doi.org/10.6084/m9.figshare.14736087

International Conference on Innovations in Energy Engineering & Cleaner Production IEECP21
2
A. PV Durability and Reliability Issues
The best way to deal with the PV reliability issues is by the use of
the bathtub reliability curve (see Figures 1 (a) and 1 (b)) to find the
physics of failure for each mode [6]. The bathtub reliability curve
of a PV module is a graphical model made to represent the failure
rate of a group of PV modules over some time. The curve helps the
PV manufacturers to predict when failures occur on the PV module
and possibly identify the root causes of the failure and possible ways
of preventing them [4]. The bathtub reliability curve describes the
failure rate of the PV module as a function of in-service life.
Therefore, the curve consists of three essential parts, namely: Failure
mode A (infant mortality), failure mode C (normal life), and failure
mode B (end of life wear-out).
Failure mode A: Failure mode A is the early life failure (also,
known as infant mortality) that normally occurs in the first 1-2 years
of a PV module’s life. Failure mode A occurs at the initial stage of
the module's life cycle. The cause of failure mode A may be due to
fundamental design faults, processing issues, errors in
manufacturing, or inappropriate installation [4]. Therefore, Passing
IEC 61215 or 61646 qualification tests are not proof that a PV
module has been tested and shown to be durable and reliable rather
the IEC environmental stress test protocols are designed primarily
to test the period of early life failures (infant mortality) (see Figure
1 (b)) [6].
Failure mode C: Failure mode C is the constant (random) failures
(also, known as normal life). This is the second part of the failure
mode that occurs within the lifespan of the PV module. It is called
the “constant (random) failures” because the failures in this mode
are usually predictable and homogeneous. This failure mode usually
occurs within this period when the stresses of the module have
exceeded the strength of its weakest component. The cause of this
failure mode is a result of unexpected environmental stress or load
issues. For instance, when a PV module exceeds its capabilities it
can suffer from a normal life failure (failure mode C).
Failure mode B: Failure mode B is the last part of the curve known
as the end of life for the PV module (also, known as the end of life
wear-out). In this failure mode, the curve rises steeply as many of
the PV modules simply reach the point where they failed due to
simple age or wear and tear. Failures of this kind are reasonably
predictable. B. Distinguishing Transient
Performance changes from longer-
term degradation
PV module output varies with solar irradiance and module
temperature. It is also affected by shading, rain, and dust [7],[8]. All
these variations are transient on a variety of timescales and/or
reversible. Degradation refers to the loss of output due to physical
degradation or damage to the PV cell, the effects are not reversible
[1]. It refers to effects that will ultimately require the replacement of
a PV cell for the system to return to its initial performance. The
transient effect caused by an increase in PV cell ambient temperature
can lead to reductions in output and efficiencies [2]. Degradation is
measured by changes in mean efficiency and/or performance ratio
over the long term as illustrated indicatively in Figure 2 [1]. It can
also be observed in perturbation caused by cell failure in the current-
voltage (I-V) curves for an array [1].
Individual module degradation can be attributed to intrinsic property
changes in the PV materials caused by external effects such as:
▪ Potential induced degradation (PID) [9]; and
▪ Light-induced degradation (LID) [10].
The outdoor operation of cells as part of a module in an array means
mechanisms external to the solar cell such as corrosion in
interconnections and solder bonds play a significant role in
performance degradation [3]. This makes it important to determine
the degradation rates under outdoor operational conditions rather
than indoor testing of isolated modules. [3], classified the major
difficulties in evaluating degradation rates of PV modules from real
operational data into:
• Large fluctuations of the operational data due to
uncontrollable external parameters such as weather
conditions like solar radiation, rain, cloud movement,
wind velocity, and ambient temperature together with
unexpected changes of factors external to PV systems
such as unexpected shading, inverter problems, and
control failures.
• systematic ‘degradation’ in the measurement of
PV module operational performance caused by control
sensor drifting with time as a result of electronic aging of
components such as the drifting of irradiance sensors. The
energy output of a PV system depends on weather
conditions [11], [12], [3]. The degradation rate of silicon
PV modules is around -0.7% per year of maximum power
rating [11]. Reducing rates of PV module degradation aim
to maintain the efficiency of solar PV systems [3]. As
manufacturers usually guarantee the life span of PV
modules for more than 20 years [3], it is, therefore,
necessary to track and mitigate the degradation of PV
modules over this period. Both during and beyond this
period knowing degradation behavior is essential for
operation, maintenance and repair.
C. Degradation Rates of Photovoltaic
Modules
The study of annual degradation rates of recent crystalline silicon
PV modules was carried out by Tetsuyuki and Atsushi [13]. Six
crystalline silicon PV modules connected to an electric power grid
were analyzed. Three indicators were used for the annual
degradation rates of the different crystalline silicon PV: energy
yield, performance ratio, and indoor power. The performance of the
module was evaluated from electricity output measurements taken
over 3 years. The following trends were found in the three
indicators; energy yield: 0.0, -0.4% per year, 0.0, 0.1% per year,
1.5% per year and 0.5% per year, performance ratio: 0.0, -0.4% per
year, -0.1% per year, 0.0, 1.4% per year and 0.5% per year and
indoor power: 0.1% per year, -0.3% per year, 0.2% per year, 0.0,
0.7% per year and 0.6% per year were similar. The performance of
the newly installed PV modules was found to decrease by over 2%
as a result of initial light-induced degradation (LID) after installation
[13].
The power output of an outdoor PV module has been shown to
reduce as a result of thermal cycling causing crack formation
between solders and metals [14]. Dunlop and Halton [7] studied the
degradation of PV modules in outdoor conditions for 22 years. They
monitored the electrical power outputs of monocrystalline silicon
(m-Si), polycrystalline silicon (p-Si), and amorphous silicon (a-Si)
modules. They found an 8% to 12% decrease in maximum power
output of the PV modules (Pmax) after 20 years of outdoor exposure.
Their research showed that about 80% of the reduction was due to
corrosion and the remaining 20% was attributed to dust
accumulating on the PV modules. An experimental study of
degradation modes and their effects on the PV module was
conducted after 12 years of field operation [15]. Their investigation
found that degradation led to annual reductions in output power
ranging between 2.08% and 5.2%. Short circuit current (Isc) is
reduced by between 2.75% and 2.84% annually. The open-circuit
voltage (Voc) was found to be the least affected, with annual
reductions ranging between 0.01% and 4.25%. The existence of only
one highly degraded PV module in a PV system reduces daily output
from Takatoshi et al, [16]:

International Conference on Innovations in Energy Engineering & Cleaner Production IEECP21
3
i. 19.8 kWh to 18.7 kWh during sunny days;
ii. 11.3 kWh to 10.8 kWh during partly cloudy
sunny days; and
iii. 5.5 kWh to 5.3 kWh during cloudy days.
D. Analysis of Risk Priority Number
(RPN) on the Severity of PV Failure
Modes
Failure Modes and Effects Analysis (FMEA) finds the effect of each
failure mode and its causes on the system, according to the severity
(S), occurrence (O), and detect-ability (D) [17]. The IEC 60812
standard has assumed a different range of S, O, and D for a PV
system, which is helpful to identify the particular single failure mode
based on RPN for the particular PV system and operating
environment conditions [17]. A measurement of RPN is therefore
expressed in (1) [17]:
RPN = S×O×D (1)
Where:
S is the severity, which is a non-dimensional number. Severity
determines the single failure mode, which strongly affects the PV
system performance.
O is the occurrence, which depends on the probability of occurrence
of a defect in the PV system during the exposure time.
D is the detection, which technology or instrument can identify the
failure modes in a PV system during its exposure time.
The severity rank of failure mode depends on the degradation rate
per year and safety issues. It is very difficult to find out the severity
rank of a particular failure mode, as the degradation of a PV module
is a cumulative sum of many factors [18],[19]. The highest rank in
the severity given according to the safety issue probably insulation
resistance failure, de-lamination, and burn mark occurs in the PV
module, it is a threat or hazard to person or either property [17]. The
severity number from 9 to 10 related to safety issues and the highest
degradation factor, whereas the numbers from 8 to 1 depending on
the performance degradation factor. In the present study, the severity
rank performs according to References [20],[21]. The rank of
severity has been given by Pramod et al [17] in Table 1.
MATERIALS AND METHODS
A. Site Location and Climate
Description
The location chosen for this study is based on the Electricity Supply
Board ESB) site located at Upper Dromore Road, Warrenpoint,
Northern Ireland at 54.115551oN latitude and 6.263654oW
longitude. The City of Warrenpoint acquires its power from the ESB
public grid, which is shared with other residential and industrial
consumers.
(https://www.google.com/maps/place/Newry+BT34+3PN,+UK/@
54.1132142,-
6.2642131,248a,35y,44.92t/data=!3m1!1e3!4m5!3m4!1s0x4860da
664698253b:0xd3507b57cb2eea3!8m2!3d54.1150048!4d-
6.2630492 )is used to identify the site location.
B. System Monitoring and Data Acquisition
The data acquisition system used in this research consists of two
SMA STP-20000TL-30 inverters each with a 20 W sensor box and
a data logger. The sensor box measures the total solar radiation on
the solar PV modules in-plane. The sensor box and the inverter have
been connected to the data logger and the power injector. The data
recording was set at 15 minutes (quarter-hour) intervals in the data
logger and was extracted directly from the Excel spreadsheets to the
computer and then analyzed using the MATLAB and Excel software
tools.
RESULTS
A. Transient (Partial Shading) Effect in Solar
Cells
I. Description of System
In this study, a system description of distributed circuit simulations
of a PV module under partial shading conditions is presented. The
PV module is connected to a variable DC voltage source converter
(VSC) to quantify the I-V and P-V characteristics curves. A
MATLAB SIMULINK is used to model the circuit: (i) as three
strings of 20 series-connected cells parallel to bypass diodes which
allow current flow when cells are shaded or damaged with a standard
irradiance of 1 kWm-2 applied to String 1 (cells 1-20) while (ii)
partial shading is applied to String 2 (cells 21-40) with an irradiance
of 0.3 kWm-2 and (iii) String 3 (cells 41-60) with an irradiance of
0.6 kWm-2 as shown in Figure 4.
II. Simulation Process
The model is therefore simulated and at the end of the simulation,
the I-V and P-V characteristics curves were plotted. When the PV
module is connected to the voltage-sourced converter (VSC) it
makes it difficult for the Maximum Power Point Tracking (MPPT)
algorithm to converge at the highest peak. The global maximum
power point (GMPP) indicated in the red circle of Figure 5 is 334 W
at the maximum current of 5.29 A. The resultant characteristics of
the PV array are shown in Figure 5. The P-V curves generate three
peaks under partially shaded conditions (see Figure 5).
III. Simulation Results
Figure 6 shows the variation in the solar PV cell string and bypass
diode used to reduce the shading effect. In string 1 (i.e., cells 1-20),
the bypass diode (with blue color) has zero current. This is because
string 1 solar cells do not have any shading effects. While in String
2 (i.e., 21-40 cells) and String 3 (i.e., 41-60 cells) solar cells had
shading effects. It is noticed that the current flow is above zero. This
shows that the bypass diode works.
B. Inverter Percentage Conversion Loss
When the inverter converts the DC energy from the solar PV system
to AC energy, some energy is lost, which could be due to the cable,
PV module, or inverter. As shown in Table 2, this is estimated as
inverter percentage conversion loss using equation (2) and the
values vary according to the number of energy losses from the
inverter given by (3).
Inverter percentage conversion loss = 
 × 100%
(2)
Ƞinverter =
 ×100% (3)
Where: Ƞinverter is the inverter efficiency that is the ratio of output
energy (AC energy) to input energy (DC energy) multiply by 100%.
C. Yields, Array Capture, and System Losses
Table 3 displays the daily DC array, AC final, and reference yields,
DC array capture, and AC system losses of the PV system as
measured at quarter-hourly intervals using the ESB Warrenpoint
system. These were obtained from the system measurement and
analyzed using (4), (5), (6), (7), and (8). In December and May, the
monthly daily DC array, AC final, and reference yields, DC array
capture, and AC system losses ranged from 0.46 to 4.72 h/day, 0.45
to 4.63 h/day, 5.2 to 16.41 h/day, 4.74 to 11.69 h/day, and 0.01 to
0.09 h/day, respectively. The average annual daily DC array, AC
final, and reference yields, DC array capture, and AC system losses
were 2.32 hours per day, 2.30 hours per day, 8.83 hours per day,
6.51 hours per day, and 0.02 hours per day, respectively. Figures 7
(a) and 7 (b) show the DC array, AC final, and reference yields of a
monthly daily PV system as obtained from the ESB Warrenpoint

International Conference on Innovations in Energy Engineering & Cleaner Production IEECP21
4
system, as well as the AC system loss. The DC array capture loss
could be due to transient effects (such as shading, dust, wind
velocity, ambient temperature, or module temperature) [2],
corrosion of solar cell connections, or degradations.
YA,day = 󰇛
󰇜
 (4)
YF,day = 󰇛
󰇜
 (5)
YR,day = 󰇛
󰇜


(6)
Lc,day = YR,day – YA,day (7)
Ls,day = YA,day-YF,day (8)
Where:
• YA,day: daily array yield, that is, the ratio of the DC output
energy (kWh) to its module capacity (kWP) from a solar
PV array over a total number of days in operation [22].
• YF,day: daily final yield, that is, the ratio of the AC output
energy (kWh) to its module capacity (kWp) from a solar
PV array over a total number of days in operation [23].
• YR,day: daily reference yield, that is, the ratio of total daily
in-plane solar irradiation (kWh/m2) its reference solar
irradiance (GSTC).
• Lc,day: daily DC array capture loss, that is, the difference
between the DC array yield (YA,day) and the reference
yield (YR,day).
• Ls,day: daily AC system loss, that is, the difference
between the final yield (YF,day) and array yield (YA,day).
DC array yield and AC final yield are plotted as a function of solar
irradiance in Figure 8 (a-b) using quarter-hourly (15-minute)
interval data. Figure 8 (c-d) shows that the DC array yield and AC
final yield are both linearly proportional to solar irradiance. Figure
8 (a-b) depicts sublinear behavior caused by a transient effect like
shade/shadow cast, overcast period, or average inverter efficiency
loss (about 0.6%) over the PV field. As a result, at low solar
irradiance levels, both DC array and AC final yields are either zero
or very low due to inverter losses as well as PV generator low
irradiance losses.
D. Measurement of Solar Irradiance
Figures 9 and 10 depict various views of the ESB Warrenpoint site
solar irradiance, as well as a solar power calendar based on the plane
of array solar irradiance averaged for each quarter-hourly period
between March and May 2016. March and May were chosen
because of their peak clearness indices. It has been observed that
March 13, 14, and 22 and May 13, 16, 27, and 31 are clear days,
whereas other days such as March 9, 10, 11, 12, 15, 16, 17, 18, and
22, and May 8, 12, 14, 17, 23, 24, 29, and 30 are partly cloudy, and
other days such as May 1, 2, 3, 4, 5, 6, 7, 9, 10, 11, 15, 18, 19, 20,
21, 25, 26, and 28, and March 1, 2, 3, 4, 5, 6, 7, 8, 10, 19, 20, 21, 23,
24, 25, 26, 27, 28, 29, 30, and 31 are overcast. As a result of the peak
daily clearness index found in May, there are clearer days in May
than in March.
E. Measurement of Output Performance
The daily incident solar radiation for any given location is
determined by the sun's path across the sky and the amount of cloud
cover in the area (Trueblood et al., 201346). Figure 11 (a-d) depicts
daytime power profiles at quarter-hourly (15-minute), half-hourly
(30-minute), and hourly (60-minute) intervals for three days in each
season: a clear day, an overcast day, and a middle day. The clear
day, as defined here, is the day of the season with the greatest
amount of solar irradiation, resulting in a parabolic curve (see Figure
11 (a-d)); the overcast day is a day with the least amount of solar
irradiation, resulting in distortions from perfect parabolic shapes
(see Figure 11 (a-d)); and a middle day is a day with the median
amount of solar irradiation, resulting in partial parabolic curves. The
chosen days of power profiles span the months of each of the four
seasons (winter, spring, summer, and autumn) (see Figure 12).
Figure 12 shows that the clear days (as seen in Figure 11 (a-d)) in
winter, spring, summer, and autumn were caused by an increase in
daily sunshine hours in February, May, June, and September.
Because of the highest amount of solar irradiation at the site
location, these days were generally characterized by an increase in
PV output generation. The median increase in daily sunshine hours
occur in January, April, July, and October during the winter, spring,
summer, and autumn. As a result, PV output generation was
moderately high. The overcast day was generally characterized by
low solar irradiation due to a decrease in daily sunshine hours, as
seen in December, March, August, and November (see Figure 12).
As a result, the overcast day generates less PV output. The autumn
and winter daily profiles, on the other hand, are more extended, with
higher output generation at midday, but they have fewer total hours
than the summer and spring profiles, which have more hours of
daylength. Because PV panels are more efficient at lower
temperatures, output generation is higher during clear days in the
spring than during clear days in the summer [24]. The middle day
demonstrates that PV output generation can vary throughout the day,
owing to cloud movement.
F. Performance Variations
I. Weather-Corrected Performance Ratio
(PR)
The performance ratio (PR) is a metric used to evaluate solar
photovoltaic installations. PR normalizes the output of the PV
system to its installed capacity and the available solar irradiance at
the site of installation, allowing a comparison of the performance of
systems with different installed capacities in different geographical
locations [25]. 3-5 years of data are required to capture seasonal
variations [25]. Because the performance ratio is affected by the
module and ambient temperature of the system's site location due to
variation with changes in meteorological conditions, it is important
to measure or quantify this variation and show how it can be
removed or reduced by using the two methods described below [26],
[27]:
o Traditional calculation of PR (uncorrected PR) using
equation (9):
PRuncor r=
󰇟󰇛
󰇜󰇠
= PRSyst. =
  (9)
o Modifications of uncorrected PR through temperature
normalization to produce a temperature-corrected PR to
become a weather-corrected PR using (10):
PRcorr = 
󰇟󰇛
󰇜󰇛 
󰇡󰇢󰇜󰇠
 (10)
RD = 
 × 100% (11)
%TLosses = PRuncorrected – PRcorrected (12)
Where: PRuncorr: uncorrected performance ratio; PRcorr: corrected
performance ratio; PAC: measured AC electrical generation (W);
PSTC: summation of installed modules (49920 Wp);
GPOA: measured plane of array (POA) irradiance (W/m2); t: data
collection time (15 mins.); GSTC: irradiance at standard test
conditions (STC) (1000 W/m2); Tref: reference temperature
(25oC);Tref: reference temperature (25oC); : Temperature
coefficient for power (-0.4%/oC);
RD: Degradation rates (%); %TLosses: Percentage temperature losses;
m and c are the slope and vertical intercept of the linear trend line of
the PR versus time (months) plot respectively.
As a result, the weather-corrected PR from 2016 to 2020 is analyzed
using the annual PR regression method, and performance data is

International Conference on Innovations in Energy Engineering & Cleaner Production IEECP21
5
sorted for solar irradiance levels greater than 700 Wm-2, as proposed
by Quansah and Adaramola in their works [25]. (11) Is used to
compute the degradation rates (RD) [25].
Figures 13-17 show annual PR regression graphs for five years
(2016-2020) for both temperature-corrected PR and uncorrected PR.
Table 4 and Figure 18 show the annual uncorrected system PR,
temperature-corrected system PR, degradation rates, and percentage
of temperature losses from 2016 to 2020. Figure 17 depicts a
decrease in PV power output over time due to the performance loss
rate or degradation rate. It can be seen using error bars and the
Severity ranking of failure mode proposed by Pramod et al [17].
According to Figure 17 and Table 4.
The median degradation rates in 2016 (4.5% /year to 14%/year) and
2017 (0.1% /year to 5.2%/year) are 8.40% /year and 3.87%/year,
respectively. This demonstrates that the degradation rate is greater
than 1% per year, and the hazardous probability is between 90% and
100%. [17]. This is assigned a severity of 10 (with an associated
failure of corrosion in solder bonds) [28] and a severity of 10 (with
an associated failure of EVA discoloration) [29].
The median degradation rates in 2018 (-7.5%/year to 2.5%/year),
2019 (-16%/year to -23 %/year), and 2020 (-5.1%/year to -
10%/year) are -2.75%/year, -18.23%/year, and -5.2%/year,
respectively. This demonstrates that the degradation rates are less
than 1%/year and that their hazardous probabilities range from
severity rank 9-1 or 80% - 70% to 0% safety hazard [17]. EVA
discoloration, metallization of the front side grid, de-lamination
between EVA and solar cell, glass weathering, de-lamination
between EVA and solar cell, oxidation of antireflecting coating, cell
metallization and hotspot, surface soiling, corrosion in solder bond,
and de-lamination, junction box degraded could all be associated
failures here [17]. CONCLUSIONS
Since environmental factors such as humidity, dust accumulation,
and wind velocity are agents of transient and performance loss rates,
it is important to minimize or reduce these effects by inspecting the
proposed geographical location before the installation of solar PV
systems. Because of the diversity of climates, it is essential to
broaden the optimization considerations to achieve a more
significant result. Instead of using standard methods for installing a
solar PV system, it is important to consider dominant factors such
as wind directions and speeds, which have transient effects on solar
PV system output performance. Since solar cell output performance
degrades as cell temperature rises due to thermal degradation, it's
critical to maintain the surface of a solar panel at a temperature that
doesn't exceed its standard test conditions (25oC). Air- or water-
cooling techniques may help to alleviate the problem of overheating
caused by an increase in solar irradiance and high temperatures on a
solar panel. Therefore, using the characteristics of an anti-reflecting
material to increase the output performance of a solar PV panel is
recommended.
Figure 1 (a): Using Bathtub curve to explain PV Durability and
Reliability Issues [4].
Figure 1 (b): Multiple failure modes overlap of solar PV modules
[4].
Figure 2: Degradation of Solar PV system [1].
Figure 3: Location and Satellite view of ESB site situated at Upper
Dromore Road, Warrenpoint, Northern Ireland, UK
(https://www.google.com/maps/place/Newry+BT34+3PN,+UK/@
54.1132142,-
6.2642131,248a,35y,44.92t/data=!3m1!1e3!4m5!3m4!1s0x4860da
664698253b:0xd3507b57cb2eea3!8m2!3d54.1150048!4d-
6.2630492).
a. b.
Figure 7: Monthly daily yields, DC array capture, and AC system
losses of ESB Warrenpoint system.

International Conference on Innovations in Energy Engineering & Cleaner Production IEECP21
6
Figure 4: PV module connected to a variable DC voltage
source converter (VSC)
(https://uk.mathworks.com/help/physmod/sps/ug/partial-
shading-of-a-pv-
module.html;jsessionid=9479da359d71d0f731ea5a9a6d64)
Figure 5: I-V and P-V characteristics curves of a PV system
Figure 6: Shading effect of PV current (with yellow color line) and
diode current (with blue color).
Figure 8: Quarter-Hourly data for AC Final yield and DC Array
yield.

International Conference on Innovations in Energy Engineering & Cleaner Production IEECP21
7
Figure 9: Measured solar irradiance profiles for each day in March
2016.

International Conference on Innovations in Energy Engineering & Cleaner Production IEECP21
8
Figure 11 (a): Power output profiles of selected days during the
winter season.
Figure 11 (b): Power output profiles of selected days during the
spring season.
Figure 11 (c): Power output profiles of selected days during the
summer season.
Figure 11 (d): Power output profiles of selected days during the
autumn season.
Figure 12: A chart showing power profiles of selected days in
winter, spring, summer, and autumn in 15 minutes, 30 minutes, and
60 minutes sensor configuration period of ESB Warrenpoint system.

International Conference on Innovations in Energy Engineering & Cleaner Production IEECP21
9
Figure 10: Measured solar irradiance profiles for each day in May
2016.
Table 1: Severity ranking of failure mode [17].
S/N
Severity
Rank
1.
Degradation rate should be >1.0%/year with safety,
hazardous probability in the range <90−100>%
10
2.
Degradation rate should be <0.9–1.0>%/year with
safety, hazardous probability in the range <80–
90>%
9
3.
Degradation rate should be in the range <0.8–
0.9>%/year with safety, hazardous probability in the
range
<70–80>%
8
4.
Degradation rate should be in the range <0.7–
0.8>%/year with safety, hazardous probability in the
range
<60–70>%
7
5.
Degradation rate should be in the range <0.6–
0.7>%/year with safety, hazardous probability in the
range
<50–60%> %
6
6.
Degradation rate should be in the range <0.5–
0.6>%/year with safety, hazardous probability in the
range
<40–50> %
5
7.
Degradation rate should be in the range <0.4–
0.5>%/year with safety hazardous probability in the
range <30–40>%
4
8.
Degradation rate should be in the range <0.3–
0.4>%/year with safety hazardous probability in the
range <20–30>%
3
9.
Degradation rate should be in the range <0.2–
0.3>%/year with safety hazardous probability in the
range
<10–20>%/
2
10.
The degradation rate should be <0.1–0.2>%/year
with no safety hazard
1
Table 2: Monthly DC Energy and AC Energy, inverter efficiency,
and percentage conversion loss of quarter-hourly system
measurement obtained from the ESB Warrenpoint System.
Month
DC
Energy
[MJ]
AC
Energy
[MJ]
Inverter
Efficiency (ƞ)
(%)
Inverter
Percentage
Conversion
Loss (%)
January
2763
2720
98.40
1.6
February
1978.6
1975.3
99.80
0.17
March
8640.2
8543.1
98.90
1.12
April
17810.3
17802.16
99.95
0.046
May
26092
25595
98.10
1.51
June
20628
20606.5
99.90
0.104
July
20980
20940
99.81
0.191
August
18320
18291
99.84
0.158
September
12990
12838
98.82
1.17
October
8121
8062
99.27
0.73
November
4540
4531
99.80
0.2
December
2540
2526
99.45
0.55
Total
145,403.1
144,430.1
Average
12,116.9
12,035.8
99.42
0.629

International Conference on Innovations in Energy Engineering & Cleaner Production IEECP21
10
Table 3: DC array, AC final, and reference yields, DC array capture,
and AC system losses of quarter-hourly measured ESB Warrenpoint
system.
Month
YA
(h/day)
YF
(h/day)
YR
(h/day)
Lc,day
(h/day)
Ls,day
(h/day)
January
0.5
0.49
1.39
0.89
0.01
February
1.59
1.58
2.94
1.35
0.01
March
1.56
1.54
4.81
3.25
0.02
April
3.33
3.32
13.31
9.98
0.01
May
4.72
4.63
16.41
11.69
0.09
June
3.86
3.85
15.12
11.26
0.01
July
3.8
3.79
13.31
9.51
0.01
August
3.31
3.3
11.09
7.78
0.01
September
2.43
2.4
9.88
7.45
0.03
October
1.47
1.46
8.66
7.19
0.01
November
0.85
0.84
3.87
3.02
0.01
December
0.46
0.45
5.2
4.74
0.01
Average
2.32
2.30
8.83
6.51
0.02
Table 4: Annual uncorrected system PR, temperature-corrected
system PR, degradation rates, and percentage of temperature losses
from 2016-2020.
Year
RD
(%/year)
PRuncorrected
(%)
PRcorrected
(%)
Tavglosses
(%)
2016
4.5 to 14
86.5
58
28.5
2017
0.1 to 5.2
91
61
30
2018
-7.5 to 2.5
99.8
67
32.8
2019
-16 to -23
100
67
33
2020
-5.1 to -10
89
59
30
ACKNOWLEDGMENTS
Our thanks go to the Fiosrai Research Fund of Technological
University Dublin, Ireland for funding this research work. We also
thank the Electric Supply Board (ESB) of Ireland for permitting us
to use data from their solar plant to carry out this research study.
REFERENCES
[1] C.C. Okorieimoh, B. Norton, and M. Conlon, “Long-Term
Durability of Solar Photovoltaic Modules” In Scott L., Dastbaz M.,
Gorse C. (eds) Sustainable Ecological Engineering Design.
Springer, Cham, 2020, pp. 317-325.
https://doi.org/10.1007/978-3-030-44381-8_24.
[2] C.C. Okorieimoh, B. Norton, and M. Conlon, “Effect of
transient performance changes on photovoltaic modules output”
Poster. 10TH Annual GRS Symposium, 2019,
10.13140/RG.2.2.28516.94083.
[3] X.Y. Li, “Degradation analysis of photovoltaic modules based
on operational data: effects of seasonal pattern and sensor drift”, IOP
Conf. Series: Earth and Environmental Science, 2016, Vol. 40, pp.
012063. DOI:10.1088/1755-1315/40/1/012063.
[4] F.Z. Allen, and P.D. David, “Photovoltaic Module, Weather
Durability and, Reliability” “Will my module last outdoors?”. Solar
Energy Competence Centre, Atlas Material Testing Technology
LLC, 2015. www.atlas-mts.com.
[5] J. Wohlgemuth, “Reliability of PV Systems, Reliability of
Photovoltaic Cells, Modules, Components, and Systems”, Proc. of
SPIE, 2008, Vol. 7048, pp. 704802-1.
[6] Z. Allen, “PV Durability and Reliability Issues. Atlas Material
Testing”, Journal of Renewable Energy World, 2009, Vol. 1, Issue
5.
[7] E.D. Dunlop, and D. Halton, “The performance of crystalline
silicon photovoltaic solar modules after 22 years of continuous
outdoor exposure. Prog Photovoltaics”, 2006, Vol.14, Issue 1,
pp.53–64. DOI: 10.1002/pip.627.
[8] G.N. Tiwari, R.K. Mishra, and S.C.Solanki “Photovoltaic
modules and their applications: A review on thermal modeling”,
Applied Energy, 2011, Vol. 88, pp. 2287-2304.
[9] S. Pingel, O. Frank, M. Winkler, S. Daryan, T. Geipel, H.
Hoehne, and J. Berghold, “Potential induced degradation of solar
cells and panels”, Proc. 35th IEEE Photovolt. Spec. Conf., 2010, pp.
pp. 2817-22.
[10] B. Sopori, P. Basnyat, S. Shet, V. Mehta, J. Binns, and J.
Appel, “Understanding light-induced degradation of c-Si solar
cells”, Proc. IEEE Photovolt. Spec. Conf., 2012, pp. 5200-54200.
[11] C.R. Osterwald, A. Anderberg, S. Rummel, and L. Ottoson,
“Degradation analysis of weathered crystalline-silicon PV
modules,” Proc. 29th IEEE Photovolt. Spec. Conf., 2002, pp. 1392-
5.
[12] G.N. Tiwari, R.K. Mishra, and S.C. Solanki, “Photovoltaic
modules and their applications: A review on thermal modeling”,
Applied Energy, 2011, Vol. 88, pp.2287-2304.
[13] I. Tetsuyuki, and M. Atsushi, “Annual degradation rates of
recent crystalline silicon photovoltaic modules”, Progress in
photovoltaics: Research and Applications, 2017, Vol. 25, pp. 953-
967. DOI: 10.1002/pip.2903.
[14] P. Nochang, J. Jaeseong, and H. Changwoon, “Estimation of
the degradation rate of multi-crystalline silicon photovoltaic module
under thermal cycling stress. Microelectronics Reliability”, 2014.
DOI: http://dx.doi.org/10.1016/j.microrel.2014.03.021
[15] M. Saadsaoud, A.H. Ahmeda, Z. Er, and Z. Rouabah,
“Experimental study of degradation modes and their effects on the
reliability of photovoltaic modules after 12 years of field operation
in the steppe region”, Special issue of the 3rd International
Conference on Computational and Experimental Science and
Engineering (ICCESEN 2016), Journal of Acta Physica Polonica A,
2017, Vol. 132, pp.3-11. DOI: 10.12693/APhysPolA.132.930.
[16] H. Takatoshi, N. Tomoya, T. Tadashi, and I. Yoshitaka,
“Influence of degradation in units of PV modules on the electric
power output of PV system”, Journal of International Council on
Electrical Engineering, 2018, Vol. 8, No. 1, pp.118–126. DOI:
https://doi.org/10.1080/22348972.2018.1477095
[17] R. Pramod, M. Maria, M.K. Nallapaneni, O.S. Sastry, and G.N.
Tiwari, “Risk priority number for understanding the severity of
photovoltaic failure modes and their impacts on performance
degradation,” Case Studies in Thermal Engineering, 2019, Vol. 16.
https://doi.org/10.1016/j.csite.2019.100563
[18] S. Chattopadhyay, R. Dubey, V. Kuthanazhi, J.J. John, C.S.
Solanki, A. Kottantharayil, B.M. Arora, K.L. Narasimhan, V.
Kuber, J. Vasi, A. Kumar, and O.S. Sastry, “Visual degradation in
field-aged crystalline silicon PV modules in India and correlation
With electrical degradation”, IEEE J. Photovolt., 2014, Vol 4, pp.
1470-1476.
[19] S. Kumar, S. Roy, and R. Gupta, “Comparison of reliability
tests by characterization of degradation in photovoltaic modules”,
MOJ Sol. Photoenergy Syst., Article 00006, 2017.
10.15406/mojsp.2017.01.00006
[20] D.C. Jordan, T.J. Silverman, J.H. Wohlgemuth, S.R. Kurtz, and
K.T. VanSant, “Photovoltaic failure and degradation modes”, Prog.
Photovolt. Res. Appl., 2017, Vol. 25, pp. 318-326.

International Conference on Innovations in Energy Engineering & Cleaner Production IEECP21
11
[21] J. Kuitche, R. Pan, and G. Tamizhmani, “Investigation of
dominant figures modes for field-aged c-Si modules in desert
climatic conditions,” IEEE J. Photovolt, 2014, Vol. 4, pp. 814-826.
[22] Sustainable Energy Authority of Ireland (SEAI), Irish energy
balances: Statistics publications, 1990-2009, Available:
www.seai.ie
[23] J.D. Mondol, Y.G. Yohanis, and B. Norton, “The effect of low
insolation conditions and inverter oversizing on long-performance
of a grid-connected photovoltaic system,” Prog. Photovolt. Res.
Appl,2006, Vol. 15, pp. 353-368.
[24] C. Trueblood, S. Coley, T. Key, L. Rogers, A. Ellis, C. Hansen,
and E. Philplot, “PV measures up for fleet duty, Data from a
Tennessee Plant Are Used to Illustrate Metrics That Characterise
Plant Performance”. DOI: 10.1109/MPE.2012.2234405.
[25] D.A. Quansah, and M.S. Adaramola, “Assessment of early
degradation and performance loss in five co-located solar
photovoltaic module technologies installed in Ghana using
performance ratio time-series regression,” Journal of Renewable
Energy, 2019, Vol. 131, pp. 900-910.
[26] S. Kurtz, J. Luis Becerra Cruz, E. Riley, and C. Hansen,
“Weather-Corrected Performance Ratio,” National Renewable
Energy Laboratory Technical Report, NREL/TP-5200-57991,2013.
[27] D.C. Jordan, C. Deline, S.R. Kurtz, G.M. Kimball, M.
Anderson, “Robust PV Degradation Methodology and Application”,
IEEE Journal of Photovoltaics, 2018, Vol. 8, pp. 525-531. DOI:
10.1109/JPHOTOV.2017.2779779.
[28] P. Rajput, G.N. Tiwari, B. Bora, and O.S. Sastry, “Visual and
electrical degradation of 22 years field age monocrystalline silicon
PV module in the composite climate of India”, IEEE, 42nd
Photovoltaic Specialists Conference (PVSC), 2015, pp. 1-3.
[29] D. Berman, and D. Faiman, “EVA browning and the time-
dependence of I-V curve parameters on PV modules with and
without mirrors-enhancement in a desert environment”, Sol. Energy
Mater. Sol. Cells, 1997, Vol. 45, pp. 401-412.