=Paper=
{{Paper
|id=Vol-3039/paper19
|storemode=property
|title=Photovoltaic technologies: problems, technical and economic losses, prospects
|pdfUrl=https://ceur-ws.org/Vol-3039/paper19.pdf
|volume=Vol-3039
|authors=Artur Zaporozhets,Anastasia Sverdlova
|dblpUrl=https://dblp.org/rec/conf/ittap/ZaporozhetsS21
}}
==Photovoltaic technologies: problems, technical and economic losses, prospects==
https://ceur-ws.org/Vol-3039/paper19.pdf
Photovoltaic technologies: problems, technical and economic
losses, prospects
Artur Zaporozhetsa, Anastasia Sverdlovaa
a
Institute of Engineering Thermophysics of NAS of Ukraine, 2a Marii Kapnist Str., Kyiv, 03057, Ukraine
Abstract
Photovoltaic technologies directly convert sunlight into electricity and are one of the most
common renewable energy sources. Solar power is the most efficient and economical way to
generate electricity. To continue to reduce the cost of this technology and increase profits, the
solar industry is constantly looking for ways to maximize efficiency and minimize the losses
and costs associated with the energy production process. Nevertheless, photovoltaic systems,
including solar panels, tend to accumulate dust and dirt particles. As a result, the fraction of
incident light decreases and this leads to a decrease in energy conversion efficiency to 50%. At
the moment, a universal solution to the problem of pollution of photovoltaic systems does not
exist due to their location-specific and seasonal fluctuations, differences in local energy costs,
as well as the availability and cost of resources.
Keywords 1
Photovoltaics, solar energy, soiling, concentrated solar power, dust cleaning
1. Introduction
In 2020, the share of wind power plants (WPP) and solar power plants (SPP) in the structure of
electricity production has doubled – to 6.8% (3.3% in 2019) with a total electricity production of 148.9
billion kWh (Fig.1, Table 1). The installed capacity of these renewable energy sources (RES) during
the year increased by 1.9 GW (+ 41% compared to 2019) [1].
Figure 1: Power market: Ukraine, cumulative installed renewable power capacity (GW), 2000–2030
[2]
ITTAP’2021: 1nd International Workshop on Information Technologies: Theoretical and Applied Problems, November 16–18, 2021,
Ternopil, Ukraine
EMAIL: a.o.zaporozhets@nas.gov.ua (A. 1); science.sverdlova@gmail.com (A. 2)
ORCID: 0000-0002-0704-4116 (A. 1); 0000-0001-8222-1357 (A. 2)
©️ 2021 Copyright for this paper by its authors.
Use permitted under Creative Commons License Attribution 4.0 International (CC BY 4.0).
CEUR Workshop Proceedings (CEUR-WS.org)
The installed capacity of SPP increased the most, the peak production of which in the spring-summer
period falls on the hours of daytime consumption reduction (from 12:00 to 17:00), which requires
flexible tools to balance them. The balancing of RES, mainly SPP, during the day was last year and
remains today the main problem of integration of RES into the energy system of Ukraine.
Table 1
Dynamics of installed capacity volumes and share in the total production of wind farms and power
plants
2019 2020 ∆
Installed power, MW (%) 3555.4 5362.6 +1807.2 (+50.8%)
Volumes and share in total production, billion kWh (%) 3.1 6.8 +3.7 (4.6%)
At the same time, in 2020 the transmission system operator (TSO) and distribution system operators
(DSO) issued a total of technical conditions for connection to the united energy networks of Ukraine of
the new generation of RES with an installed capacity of 1374.95 MW, of which 1035.7 MW – to OSP
networks, where SPP accounts for 78%. That is, the growth of RES capacity will continue and,
accordingly, will increase the share of RES in the structure of electricity production [3].
Also, in Ukraine there is a law "On the electricity market" (Law "On the electricity market" from
13.04.2017, № 2019-VIII) in terms of penalizing electricity producers at the "green" rate for daily
generation imbalances in the network relative to the accepted and agreed with a "guaranteed buyer"
forecast from the manufacturer. Responsibility for the accuracy of forecasting rests with the producer
"… In order to limit the impact of support for electricity producers on the "green" tariff on electricity
prices…"[4].
Producers of "green energy" are fined for deviating from the approved forecast – by 10% for SPP.
For large producers with a share of 5% or more in the total energy balance of the country, respectively,
the tolerance will be 5%. The Law defines (see paragraph 11 of the section on transitional provisions)
that
The share of compensation to the guaranteed buyer (from) producers at the "green" tariff, the cost
of settling the imbalance of the guaranteed buyer is [5, 6]:
• until December 31, 2020 – 0 percent;
• from January 1, 2021 – 10 percent;
• from January 1, 2022 – 20 percent;
• from January 1, 2023 – 30 percent;
• from January 1, 2024 – 40 percent;
• from January 1, 2025 – 50 percent;
• from January 1, 2026 – 60 percent;
• from January 1, 2027 – 70 percent;
• from January 1, 2028 – 80 percent;
• from January 1, 2029 – 90 percent;
• from January 1, 2030 – 100 percent.
According to an IEA report under the PV-power Systems Program, the use of mesoscale
meteorological models "does not appear to improve forecasting quality." Post-processing of data
(mainly spatial averaging and error correction) can relatively reduce the inaccuracy of the local estimate
by 15-25%, but still for short- and medium-term forecasting it will be at least 25-30%.
2. The main methods for making forecasts of electricity generation during the
operation of solar power plants
Cloudless sunny weather has a positive effect on the growth rate of energy generation. But on cloudy
days, there is a high density of cloud masses, which causes a decrease in the amount of production of
this type of resources. The most urgent solution to this problem is the need to forecast the amount of
electricity that a particular solar station can produce [7,8]. High-quality forecasting enables
manufacturers and operators of grid companies to competently manage the performance indicators of
solar panels, thus effectively introducing "green" energy into the overall energy system of the country.
This procedure should, first of all, take into account the amount of solar radiation that the stations
will receive. This indicator can be influenced by many different factors, including meteorological and
climatic conditions, namely, daylight hours, rainfall, wind speed and much more.
During choosing a forecasting method, a specialist needs to take into account what data he should
receive as a result of calculations. If he needs to find out the total production of electricity or change
the volume of production, then he will need various mathematical methods and approaches. Next, we
will talk about two basic techniques that allow to make a forecast of the production of this type of
resource for up to 6 hours or a day in advance.
2.1. Methods for making a forecast of sun activity for up to 6 hours
Currently, the main methods for determining solar activity are the total sky imagery and the study
of photos from space [9] (Fig.2). These methods provide an opportunity to make short-term forecasts
up to 6 hours and make the following assumptions for the future generation of electricity from solar
stations for a period of at least a day.
Figure 2: Camera locations, different FOVs of a camera [9]
This method is used to make forecasts of future volumes of electricity production by "green" plants
in almost real time. This method makes it possible to predict with high accuracy the generation of
electricity half an hour in advance.
To make a forecast, a specialist must perform the following operations:
• to get pictures of the sky in the area where the power plant is installed – for this, it is necessary
to take pictures from the earth's surface;
• to analyze the received information, identify thin and thick clouds;
• to evaluate the vector of movement of cloud masses – for these purposes, a sequence of a small
number of images will be required;
• knowing the location of the clouds, information about the vector of their movement, it is
necessary to calculate the irradiation power and create a forecast for the generation of electrical
energy.
Specialized meteorological stations provide the necessary information regarding the size, structure
and movement of cloud masses. It is these data that make it possible to make accurate predictions even
for a small-time horizon.
As for forecasting for a longer period, its accuracy is noticeably reduced. The main reason is that it
is very difficult to determine the movement of clouds and the change in their geometric shape.
Currently, there is still no exact mathematical model that allows to determine the formation of cloud
fields.
Alternatively, you can use images of cloud masses taken from the surrounding area, and then make
a forecast that will take into account the vector of movement of cloud masses. The technician should
also be aware that clouds in different layers have different properties. For example, low clouds can
move faster than high clouds. In the photo, the clouds located in the lower layers can cover the high
ones.
2.2. Analysis of cloud conditions by studying images from space
This method practically repeats the algorithm of the previous method. Except that instead of the
images of the sky and cloud fields taken from our planet, the photographs taken from satellites installed
in space will be used. This information can be transmitted by optical photography or by using infrared
sensors.
The main advantage of this method is to obtain cloud scales over a large area. It should be noted that
satellite imagery is of high quality and allows to cover almost the entire area of the Earth. As for the
previous method, Total sky imagery is suitable for implementation in a certain area. A reliable
measurement of the reflectance makes it possible to accurately calculate the cloud index. According to
the applied formula, this indicator depends on the optical depth of the cloud. This method has been
studied quite deeply, and therefore has become widely used in the study and marking of solar resources
on the map (Fig.3). Thus, insolation is determined for a specific area [9].
Figure 3: Clear Sky RBR generated [10]
Among the disadvantages of the method is the fact that most ordinary space satellites send data
exclusively through the visible spectrum, and therefore, morning forecasts are often less accurate. The
main reason is the lack of accumulation of the necessary information. Experts are trying to cover this
gap with images obtained using infrared radiation. You also need to understand that the spatial
resolution of space satellites is several times less than that of images of cloud fields taken from the
earth's surface. This method allows to photograph large clouds, while small formations remain
unnoticed, which negatively affects the accuracy of the insolation calculation. The frequency of
transmission of information is much less than in the first method. In addition, it should be taken into
account that processing satellite data requires much more time from specialists, which also affects the
accuracy of forecasting.
The use of satellites makes it possible to create an accurate forecast up to 6 hours, which in most
cases exceeds the performance of numerical methods for determining the weather.
At the moment, solar energy around the world has begun to practice forecasting the production of
"green" energy for a short time, starting with one day and ending with a week. So far, this type of
forecasting does not have a clear system, since the results show a large error. Therefore, the longer the
forecast period will be covered, the more reliable the result will be.
Do not forget about factors that are indirectly related to natural features, one of the most common is
the level of dustiness of solar panels. As a rule, this indicator increases if there is hot weather without
rain for a long time. The rain will wash away the dust from the panels and increase the generation of
electrical energy accordingly. This factor should also be taken into account when making a forecast.
The best way to predict solar energy production is to average the forecasts that are provided by
various weather services.
Today, forecasting the indicators of solar energy production is done by several large companies
located in countries where solar energy is developed. Everything is just beginning in Ukraine, there is
a lot of work ahead, the result of which will be the emergence of its own forecasting systems. In the
near future, forecasts will help not only take into account the weather conditions of our country, but
also adapt to changes in local legislation.
3. Reduction of solar energy production due to air pollution with dust and
particulate matter
Pollution can easily cause electricity loss of more than 1% per day and is a specific phenomenon that
depends on local climatic conditions [11]. The predominant type of pollution can vary significantly
depending on the location: deposits of mineral dust, bird droppings, biofilms of bacteria, algae, lichen,
mosses or mushrooms, plant debris or pollen, engine exhaust gases, industrial emissions, as well as
agricultural emissions, for example, feed dust (Fig.4) [12, 13, 14].
Figure 4: Examples of Soiling [15]
As a result of natural processes and human activities, dust is formed, which is present in the air in
the form of solid suspended particles (PM). The properties of the dust particles vary depending on the
location. Dry desert areas are characterized by electrostatically attracted inorganic materials, in coastal
areas – salt and dirt caused by rain. Industrial and cooler areas are dominated by wind-borne organic
mud, evaporated rain deposits, and atmospheric pollutants from fossil fuels [16, 17]. In China, Africa,
India, peak losses can vary by 10-70%; in Europe – 1-7%. Table 2 shows the different properties of
dust particles on different continents [18].
Table 2
Comparison of dust samples in some places [19]
Location Major elements Major Oxides Origin
o
Hangzhou (30.25 N, Si, Ca, Al, Fe, K, SiO2, Al2O3, Cao, Sand, potash feldspar, straw
120.16o E), China Mg, Na Fe2O3 burning, mechanical wear
o o
Perth (56.39 N, 3.43 Si, Ca, Al, Fe, K CaO, SiO2, KAlSi3O3 Acidic and sandy soils from
E), Australia deserts
United Arabic Emirates SiO2, Fe2O3, CaO Human activities, wind-blown
dust coming from the Arabian
Peninsula
Doha (25.29o N, 51.51o Ca, Si, Fe, Mg, Al CaCO3, CaMg(CO3)2, Calcite, dolomite, building,
E), Qatar SiO2 local soil
Cairo (30.04o N, 31.24o Si, Ca, Al, Fe, Mg, SiO2, CaCO3 Cement industry, desert, fossil
E), Qatar K, Na fuel combustion
Northern Poland Si, Al, Mg, Fe, K, SiO2, Al2O3, MgO Sand, frictional elements of
Ca, P, S mechanical components
The HÜBTAM laboratory analyzed in a scanning electron microscope particle taken from the
surfaces of solar panels and determined the content and number of particles forming dust. The figures
show the monthly change of elements with a powder content of more than 5% of the number of elements
[20].
In the figures 5-7, the particles contain oxide forms of elements that occur in nature, such as Si, C,
Ca, O, Al, and F. In particular, the height of the oxygen element is indicated in Figure 6 shows. As can
be seen from the results, between the end of the winter season and the transition of the spring period,
the transitions of heavy elements carried by dust also increase. This transition also increases the
tendency of contaminants to adhere to the surfaces of the panels. Change of item depending on months
below 5% is given in Table 3.
Table 3
Amount of element in % March-December
September
November
December
October
August
Element Name
March
April
June
May
July
S 0 1.12 4.09 0.22 0 0.34 0.43 0 0 0
Mg 2.12 3.6 1.56 1.49 0.57 0.19 0.43 1.11 0.23 0.56
Fe 1.54 1.30 0.42 0.47 0.76 0.22 0.27 0.57 0.58 0.23
Na 1.49 0 1.93 1.09 0.01 0.06 0.42 0.07 0 0
In 0 0.83 1.40 0 0 0 0 0 0 0
K 1.09 0.23 0 0.36 0.12 0.11 0.2 0.11 0.16 0.03
The size of the deposited particles on the photoelectric surfaces is generally in the range of 1-50
microns. The diameter of the dust particles is less than 10 microns and less than 2.5 microns corresponds
to PM10 and PM2.5, respectively. Solid particles ranging from PM2.5 to PM10 never wash even after heavy
rain. Smaller PMs are distributed more evenly than larger dust particles due to the larger specific surface
area. Particles larger than PM10 are highly localized and tend to precipitate more than smaller particles.
However, they are more likely to be washed away after heavy rain. The reflection, scattering, and
absorption of incident light on the photocells depend significantly on the particle size [21].
The deposition of dust on photovoltaic surfaces is influenced by factors such as the material of the
upper surface of the photovoltaic module and the inclination angle of the panel, the speed and direction
of the wind, and the presence of moisture on the surface.
The upper surface of the photoelectric module includes glass for traditional modules and various
transparent acrylic materials based on polymer for lightweight structures. The acrylic material has a
larger dust build-up than the glass cover. Vertically mounted acrylic glazing showed 16% more
precipitation. It has also been reported that the surface of the photoelectric module made of glass attracts
less dust than the surface made of Tedlar [22]. On very smooth surfaces, the adhesion forces between
the dust particles and the surface are extremely high, and even at very high wind speeds (up to 100 km/h
or more), fine particles are not removed from such surfaces.
Increasing the tilt angle of the photoelectric module affects the stability of the dust particles. It has
been found that dust accumulation decreases with increasing inclination angle. Thus, a two-sided
photovoltaic system with a vertical plane will produce higher power than a single-facial and
horizontally positioned photovoltaic system. However, the cleaning cost for this type of photovoltaic
module may be higher. The inclined photovoltaic tracking system has a smaller angle of incident losses
compared to the horizontal photovoltaic system. A soil-coated tracking photovoltaic system shows only
up to 6% loss, while a horizontally positioned photovoltaic system without tracking can experience
losses of up to 10% [23].
Figure 5: Monthly change of elements Si and C [20]
Figure 6: Monthly change of O and Ca elements [20]
Figure 7: Monthly change of Al and F elements [20]
Wind speed and direction affect dust deposition on photovoltaic surfaces. High wind speed can
remove dust from the surface of photovoltaic panels, while low wind speed contributes to dust
accumulation. The sedimentological effect of wind on the work of photocells is small but is systematic.
The study of dust deposition on the photovoltaic collector was carried out using wind tunnel simulations
and field experiments, and the authors concluded that the wind direction and orientation of the collector
affect dust deposition. Wind speed above 2 m/s has less effect on the distribution of dust deposits. At a
speed of 25 m/s and relative humidity of 40%, the wind can carry away about 80% dust particles with
a diameter of ≥50 μm, about 50% particles with a size of 25 μm, and < 5% particles with a size of 10
μm. In another work, it has been demonstrated that particles of 10 μm or less can only be removed at
an air velocity above 25 m/s. The wind at a speed of 0.57 m/s could attach 1334 μg/cm 2 of dust to the
surface of the photovoltaic panel with an inclination of 29° and a north direction of 10° east [24].
The presence of moisture on the surface of the photoelectric module creates an adhesion force
between the particles that causes the particles to adhere to the surface of the photoelectric module. The
deposition and accumulation of dust are highly dependent on the bonding force between the
photovoltaic surfaces and the dust elements. These include gravitational, capillary, electrostatic, and
Van der Waal coupling forces. The presence of humidity in the atmosphere helps to increase the
adhesion force. Under high humidity conditions, capillary forces provide 98% adhesion [23]. For a dry
atmosphere, Van der Walla forces prevail. Increasing humidity from 40% to 80% increases adhesion
by about 80%. Small precipitates significantly accelerate the dust deposition process. However, heavy
rain can remove contaminants from the surface of photovoltaic panels.
4. Effects of Pollution and Techno-Economic Loss Assessment
Contamination of photovoltaic modules has a negative effect on their characteristics. Incident solar
radiation, short-circuit current, idling voltage, and filling factor are four parameters that determine the
output power of photovoltaic modules. Generally, deposited dust reduces light transmission and
therefore reduces incident solar radiation. The short circuit current is directly proportional to the
incident light. Consequently, the power reduction is apparent when using dusty photovoltaic modules.
Reducing incident radiation can slow the temperature rise of the photovoltaic system; thus, this does
not greatly affect the idling voltage. With a high ash content such as 0.4 mg/cm 2, the density of ash
deposition on poly-Si photovoltaic surfaces reduces the output power by 30% compared to a similar
transparent photovoltaic panel in Greece. Relatively small ash deposition (i.e. 0.06 mg/cm 2) reduced
2.5% of the generated electricity. The output power of the contaminated concentrating photovoltaic
system is more adversely affected because the incident sunlight is scattered in the other direction and
therefore a large number of beams are lost and not received.
To assess the global impact and cost of pollution, the optimum between clean-up costs and pollution
revenue loss between clean-up activities was identified for the top twenty photovoltaic markets.
According to the data and estimates of the articles [15], pollution led to a reduction in global solar
energy production by at least 3-4% in 2018, which led to a loss of income in the world of at least 3-5
billion euros (Fig.8).
Based on the assumptions made, global pollution losses could rise significantly to 4-7% of annual
electricity generation, leading to economic losses of more than 4-7 billion euros by 2023. This
development is mainly due to the increased deployment of photovoltaic systems under high insolation
conditions. Additional factors that increase the impact of pollution are an increase in the efficiency of
photovoltaic modules and a projected increase in the share of solar installations on roofs (from ~ 29%
in 2018 to ~ 35% in 2023) [25]. Other factors, such as improved air quality in some parts of the world,
can reduce anthropogenic sources of pollution, although air quality policies are usually in place for a
long time [26]. On the other hand, rising temperatures and changes associated with climate change can
cause an increase in global soil dryness and the risk of droughts and forest fires, worsening PV and CSP
pollution due to higher concentrations of aerosols and other factors. uneven precipitation [27].
At this stage, an overview of panel surface cleaning methods is considered necessary.
5. Cleaning of contaminated photovoltaic modules
As already mentioned, removing dust from photovoltaic modules is important for increasing output
power (Fig.9).
Figure 8: Effects of pollution on solar power generation [15]: (A) – installed photovoltaic capacity by
2018 and average estimate by 2023, sorted by the country for the top 22 and global CSP capacity; (B)
– relevant pollution indicators; (C) – specified cleaning costs per cleaning and square meter; (D) –
typical energy output in kWh/kWp; (E) – estimated range of an optimal number of annual cleaning
cycles (columns) and actual range of typical annual cleaning cycles; (F) – minimum expected financial
losses due to contamination calculated on the basis of optimal cleaning cycles
5.1. Passive pollution protection
The removal may be natural, mechanical, or chemical. Natural methods of dust removal are based
on wind energy, gravity, and rainwater. Of course, precipitation is free, but it is seasonally variable,
which makes this method unreliable. Moreover, after cleaning, rinsing the surface, and drying the glass
is required, otherwise, the glass will become an ideal surface for depositing dust particles [28]. In the
early morning, late evening, night, and on a rainy day, the photovoltaic module can be rotated to a
vertical or inclined position to remove dust; however, these methods are not very effective. In addition,
rotation of the array of photovoltaic modules is not feasible. Mechanical methods include cleaning,
blowing, vibration, and ultrasonic treatment. A broom or brush is typically used for a cleaning method
that is driven by a machine. Due to the small size and strong adhesion of dust, this method is not very
effective. For blowing methods, the wind from the blower is used, but this requires a lot of energy [29].
Figure 9: Schematic Illustration of Soiling Mitigation Technologies [15]: (A) – important soiling
mechanisms which could be addressed by anti-soiling coatings (ASCs); (B) – single-axis tracking and
optimization of night stowing position; (C) – working principle of EDS (standing wave version); (D) –
dew mitigation by low-ε coatings and active and passive heating; (E) – PV module design approaches
for soiling loss reduction: the red overlay indicates lost cell strings dew to soiling; (F) – site adaption
5.2. Electrodynamic Shield (EDS)
Transparent electrodynamic shields (EDS), also called electrodynamic dust shields, repel dust
particles, creating a time-varying (dynamic) electric field above the surface. The basic electrodynamic
shield (EMF) consists of a transparent or thin layer of electrodes of direct or complex shape, deposited
on a substrate. The electrodes are coated with a thin transparent dielectric sheet which is glued to the
electrodes with an adhesive to insulate the electrodes from the air. EDS is activated by means of
alternating high voltage supplied to electrodes. Both standing and traveling waves can be used on EDS.
Standing waves move dust-up and down, and traveling waves move dust horizontally. A traveling wave
system requires a complex electrical circuit and high voltage (> 800 V) [30]. For the production of
photovoltaic energy on a utility-scale, a standing wave design is suitable. It has been demonstrated that
98% of the dust from a transparent conveyor consisting of transparent indium and tin oxide electrodes
printed on a glass plate can be removed using this approach while electrostatic traveling waves are
generated by a four-phase rectangular applied voltage. To replace the expensive indium and tin oxide,
a glass plate with repellent sand and a high-voltage single-phase rectangular power supply system based
on the power supply were later used. This system generated an inverted motion of sand particles that
were carried down by gravity. The effectiveness of the EDS approach can be reduced over time and, in
cyclical operation, at low dust levels [31]. Moreover, dust particles that settle on photovoltaic systems
over a longer period of time are less sensitive to the EDS removal process.
5.3. Cleaning
So far, no passive pollution protection technology (e.g. surface coatings) has completely eliminated
the need for cleaning. In addition, there is no universally recommended method of cleaning, as cost-
effectiveness and efficiency vary depending on local conditions, available resources, and frequency of
cleaning. In general, cleaning methods can be divided into manual, semi-automatic, and fully automatic.
The market for fully autonomous cleaning, which is only 0.13% of the current global solar power
capacity, will grow from 1.9 GW to 6.1 GW in 2022 [32].
Robotic cleaning uses an object-based sensor that can remove dust from photovoltaic systems to
maximize the output of photovoltaic solar panels (Fig.10). PVCleaner Robot V1.0. T. is the first robotic
device for cleaning photovoltaic systems [33]. Robot vacuum cleaner E4 ‐ Water Free (DCR) (Ecoppia
Empowering Solar, Herzliya, Israel) is able to clean 99% of the contaminants. This anhydrous
microfiber wipes clean the surface of the photoelectric module, and the downward movement of this
DCR depends on gravity. E4 ‐ Energy Independent DCR (manufactured by Ecoppia Empowering Solar)
does not depend on energy, since it consumes energy from batteries during the cleaning process. DCRs
work well at a tilt angle from 5° to 35°. The SMR ‐ 640AD model (Miraikikai Inc, Hayashi ‐ Cho,
Takamatsu, Kagawa, Japan) is portable, can move in any direction, is easy to handle, and is powered
by a lithium-ion battery, has an 80% reduction potential. from the cost of cleaning. PV ‐ ROB12 (Zero
One Mechatronics) uses air and water as a medium and requires an AC power supply for operation [34].
All of them are suitable for traditional photovoltaic systems but are crucial for vertical integrated
photovoltaic facades of buildings (BIPV). For offices in the UAE, an anhydrous robotic dust cleaning
system was developed. This robot had its own battery, which for work was charged from solar panels.
This system consisted of soft brushes made of microfibre at the extreme ends, four wheels, a hail sensor,
a control system, and a three-step engine [35].
Figure 10: Robotic solar panel cleaner
The choice of optimal treatment technology is influenced by a variety of factors, including the type
of contamination and deposition rate, water availability, site availability, and system configuration (e.g.
tracking versus fixed angle of inclination, installation of roof or ground), as well as labor costs,
necessary equipment, and supply contract terms. Efforts are also being made to determine the optimal
cleaning schedule based on the determination of the degree of pollution and weather, as well as dusty
forecasts.
5.4. Anti-pollution coating (self-cleaning)
Hydrophobic and hydrophilic are two types of self-cleaning coatings available to protect against
contamination and protect the photovoltaic system from dust deposition. The water wetting angle
(WCA) above 90° is known as hydrophobic and below 90° is hydrophilic. WCA above 150° is called
superhydrophobic and below 5° is called superhydrophobic [36]. High surface energy materials have
wettability and are suitable for hydrophilic/super hydrophilic surfaces. Low surface energy materials
such as silanes, silicones, nanoparticles, and polymers are used because of their water-repellent
properties for hydrophobic surfaces. Hydrophilic coatings reduce contamination by photocatalytic
reaction, and superhydrophobic coatings allow water droplets to roll down and remove dirt from the
surface. Self-cleaning activities are predominant in nature. For example, lotus leaves, rice leaves, and
butterfly wings exhibit superhydrophobicity, while pitcher, shark skin, fish scales, and snail shells have
super hydrophilic properties [37]. A popular super-hydrophilic film is TiO2, which has hydrophilicity
and photocatalytic activity. However, TiO2 reduces glass transmittance and rapidly loses hydrophilicity
by restoring the angle of contact with water in the dark. Composite films TiO2/ SiO2 can overcome all
these limitations [38].
6. Estimating Costs of Pollution Reduction Technologies
The research [15] provides approximate calculations of the estimate of positive net present value (NPV)
for which pollution reduction technologies become economically feasible.
In order to estimate the maximum allowable process costs, photovoltaic plants of a communal scale,
the optimal number of cleaning cycles, purchase prices for electricity in the amount of 0.03 euro/kWh
and a 10-year payback period for investments in technologies at a discount rate of 5% are taken into
account.
The economic benefits of reducing pollution resulting in fewer cleaning cycles can easily increase
with higher cleaning costs (e.g., installation on the roof and in remote locations) or in areas with severe
pollution.
The estimates provided can be compared with assumptions about the potential to reduce pollution
and the current costs of various technologies (Table 4). Automated cleaning systems, ASC, optimized
photovoltaic module design and tracking solution are expected to achieve an acceptable utility-wide
cost range. On the contrary, electrodynamic shields and heating solutions seem too expensive or
technology is underdeveloped.
Table 4
Pollution reduction potential and costs of selected pollution reduction technologies [15]
Potential Optimum
Mitigation Most Reasonable
Reduction of Costs Potential Limitations
Technology Application Scenario
Soiling Rates
Fully automated >95% 2.4–8.2 integration in plant PV utility scale, ground
cleaning €/m2 design mounted
Anti-soiling ≪80% (literature <2 €/m2 performance utility scale, residential,
coatings review) dependent on ground-mounted and
• Applied by <20%–50% (authors location and season, rooftop, BiPV, CSP
glass estimate) degradation by +
manufacturer 32% reported for cleaning and extra benefit from AR
• Retro-fit commercial coating environmental property
stresses
Tracking <40%–60% n.a. integration in plant utility scale, ground
planning, additional mounted, state of the art
costs in CSP
Electrodynamic ≪98% (laboratory) <30 €/m2 expensive, large- BiPV, island systems,
screen/shield 32% reported for 2- scale application street lighting, rooftop,
(EDS) needs to be proven CSP
year study in Saudi
Arabia
Heating <20%–60% <80 expensive, large- BiPV, island systems,
• PCM €/m2 (PCM) scale application street lighting, rooftop
• Active cell n.a. needs to be proven installations +
heating extra benefit from
• PVT cooling during day for
PCM + PVT
Optimized PV <65% ≤0 €/Wp integration into mass utility scale, rooftop
module design production installations
and orientation
Site adaption unknown, site n.a. little experience, utility scale PV and CSP
specific research needed
Based on the technical and economic assessment, it can be concluded that automated cleaning
machines, ASC, inverted-lay tracker modifications and optimized photovoltaic module designs are
potentially applicable on a large scale in the medium term. For these technologies, reduced pollution
can result in significantly lower cleaning costs, hence the estimated investment costs become
reasonable, especially in highly contaminated areas [39-45]. But it is worth noting that economic
conditions are very difficult, since reducing the pollution level by 50% can lead to additional costs only
in the range of 2 euros/m2 for PV, for example. Accordingly, earlier stage technologies, such as EDS
and night heating, are currently too expensive and insufficiently tested in the field, but their
development is far from exhausted and should be continued.
Moreover, more research is still needed on the effectiveness of all proposed technologies depending
on location, their possible environmental impact and the long-term reliability of CSP photovoltaic
modules or mirrors, as well as operating methods to ensure efficient and safe use of the technology.
7. Conclusions
Pollution can become a serious problem for photovoltaic systems worldwide, which is even more
worrying due to the rapid expansion of the photovoltaic system market. Mitigation strategies should be
implemented to eliminate or mitigate its effects and should be tailored to the specific conditions and
configurations of each photovoltaic facility. In addition, the complexity and variability of pollution still
make forecasting difficult. For these reasons, it must be constantly monitored.
Pollution can also be assessed directly on the basis of photovoltaic data or environmental parameters,
without the need for special pollution monitors.
An economic analysis was presented to estimate the maximum allowable monitoring costs. The
results show that the allowable costs of monitoring pollution vary depending on the size of the system
and the effectiveness of the pollution reduction strategy.
8. Acknowledgments
The work is supported by “Development of models, methods and methodology for determining the state
of industrial structures according to the data of monitoring system with forecasting the residual
resource” (2021-2025, 0121U110307), “Development of methods and ways to improve the
environmental efficiency and durability of chimneys of heat power plants. Stage 2” (2021,
0121U109243), “Development of a system for monitoring the level of harmful emissions of TPP and
diagnosing the equipment of power plants using renewable energy sources on the basis of Smart Grid
with their collaboration” (2019-2021, 0119U101859) which are financed by National Science of
Ukraine.
9. References
[1] In 2020, the installed capacity of wind farms and power plants increased by 41%, and their share
in the structure of electricity production – doubled, 2021. URL: https://ua.energy/zagalni-
novyny/u-2020-rotsi-vstanovlena-potuzhnist-ves-ta-ses-zrosla-na-41-a-yihnya-chastka-u-
strukturi-vyrobnytstva-elektroenergiyi-vdvichi/
[2] Renewable power to boost energy security in Ukraine till 2030, 2019. URL: https://www.power-
technology.com/comment/ukraine-energy-growth-to-2030/
[3] Zaporozhets, Artur. Correlation Analysis Between the Components of Energy Balance and
Pollutant Emissions. Water Air Soil Pollut, 232, 114 (2021). doi: 10.1007/s11270-021-05048-9
[4] Law of Ukraine "On Electricity Market", 2019. URL: https://zakon.rada.gov.ua/laws/show/2019-
19/page#Text
[5] Law of Ukraine "On Electricity Market", Section XVII Final and transitional provisions, 2019.
URL: https://zakon.rada.gov.ua/laws/show/2019-19/page10#Text
[6] Kyrykenko, O.V., Blinov, I.V., Parus, Y.V., Ivanov, H.A. Simulation model of day ahead market
with implicit consideration of power systems network constraints. Technical Electrodynamics,
2019 (5):60-67. doi: 10.15407/techned2019.05.060
[7] Rana, M., Korpinska, I., Agelidis, V.G. Univariate and multivariate methods for very short-term
solar photovoltaic power forecasting. Energy Conversion and Management, 121:380-390 (2016).
doi: 10.1016/j.enconman.2016.05.025
[8] Sayenko, Y., Baranenko, T., Liubartsev, V. “Forecasting of Electricity Generation by Solar Panels
Using Neural Networks with Incomplete Initial Data”. In: 2020 IEEE 4th International Conference
on Intelligent Energy and Power Systems (IEPS), 2020, pp. 140-143. doi:
10.1109/IEPS51250.2020.9263085
[9] Dissawa, L. H., Godaliyadda, R. I., Ekanayake, P. B., Agalgaonkar, A. P., Robinson, D.,
Ekanayake, J. B., & Perera, S. (2021). Sky Image-Based Localized, Short-Term Solar Irradiance
Forecasting for Multiple PV Sites via Cloud Motion Tracking. International Journal of
Photoenergy, 2021.
[10] Ghonima, M. S., Urquhart, B., Chow, C. W., Shields, J. E., Cazorla, A., & Kleissl, J. (2012). A
method for cloud detection and opacity classification based on ground based sky imagery.
Atmospheric Measurement Techniques, 5(11), 2881-2892.
[11] K.K. Ilse, B.W. Figgis, V. Naumann, C. Hagendorf, J. Bagdahn “Fundamentals of soiling
processes on photovoltaic modules” Renew. Sustain. Energy Rev., 98 (2018), pp. 239-254.
[12] L. Micheli, M.G. Deceglie, M. Muller “Mapping photovoltaic soiling using spatial interpolation
techniques” IEEE J. Photovoltaics, 9 (2019), pp. 272-277.
[13] Iatsyshyn, A., Artemchuk, V., Zaporozhets, A., Popov, O., Kovach, V. Mathematical Approaches
for Determining the Level of Impact of Ash-Slag Dumps of Energy Facilities on the Environment.
In: Babak V., Isaienko V., Zaporozhets A. (eds) Systems, Decision and Control in Energy I. Studies
in Systems, Decision and Control, vol. 298, pp. 1-13 (2020). doi: 10.1007/978-3-030-48583-2_1
[14] Popov, O., Iatsyshyn, A., Kovach, V., Artemchuk, V., Kameneva, I., Radchenko, O., Nikolaiev,
K., Stanytsina, V., Iatsyshyn, A., Romanenko, Y. Effect of Power Plant Ash and Slag Disposal on
the Environment and Population Health in Ukraine. Journal of Health and Pollution 11 (31):
210910 (2021). doi: 10.5696/2156-9614-11.31.210910
[15] Ilse, K., Micheli, L., Figgis, B. W., Lange, K., Daßler, D., Hanifi, H., ... & Bagdahn, J. (2019).
Techno-economic assessment of soiling losses and mitigation strategies for solar power
generation. Joule, 3(10), 2303-2321.
[16] Babak, V.P., Babak, S.V., Eremenko, V.S., Kuts, Y.V., Myslovych, M.V., Scherbak, L.M.,
Zaporozhets, A.O. Models and Measures for Atmospheric Pollution Monitoring. In: Models and
Measures in Measurements and Monitoring. Studies in Systems, Decision and Control, vol. 360,
pp. 227-266 (2021). doi: 10.1007/978-3-030-70783-5_8
[17] Zaporozhets, A.O., Khaidurov, V.V. Mathematical Models of Inverse Problems for Finding the
Main Characteristics of Air Pollution Sources. Water Air Soul Poll, 231, 563 (2020). doi:
10.1007/s11270-020-04933-z
[18] Chanchangi, Y.N.; Ghosh, A.; Sundaram, S.; Mallick, T.K. Dust and PV performance in Nigeria:
A review. Renew. Sustain. Energy Rev. 2020, 121, 109704.
[19] Chen, J.; Pan, G.; Ouyang, J.; Ma, J.; Fu, L.; Zhang, L. Study on impacts of dust accumulation and
rainfall on PV power reduction in East China. Energy 2020, 194, 116915.
[20] Altıntaş, M., & Arslan, S. (2021, August). Analysis of Atmospheric Transported Particulate Matter
and Investigation of Its Effects on Solar Panel. In Proceedings of the 9th European Conference on
Renewable Energy Systems, Istanbul, Turkey (pp. 24-25).
[21] Adigüzel E. Farklı boyutlardaki toz partiküllerinin fotovoltaik panellerin verimliliği üzerine etkisi
[Effect of dust particles in different sizes on the efficiency of the photovoltaic panels], MSc,
University of İstanbul, Turkey, 2018.
[22] Kalogirou, S.A.; Agathokleous, R.; Panayiotou, G. On‐site PV characterization and the effect of
soiling on their performance. Energy 2013, 51, 439–446.
[23] Ghosh, A. (2020). Soiling losses: A barrier for India’s energy security dependency from
photovoltaic power. Challenges, 11(1), 9.
[24] Said, S.A.M.; Hassan, G.; Walwil, H.M.; Al‐Aqeeli, N. The effect of environmental factors and
dust accumulation on photovoltaic modules and dust‐accumulation mitigation strategies. Renew.
Sustain. Energy Rev. 2018, 82, 743–760.
[25] Michael, S., Beauvais, A., Chevillard, N., Paredes, M. G., Heisz, M., Rossi, R., & Schmela, M.
(2019). SolarPower Europe et al.,“. Global Market Outlook for Solar Power/2018–2022,”
Accessed: 19th July.
[26] Zaporozhets, A., Babak, V., Isaienko, V., Babikova, K. Analysis of the Air Pollution Monitoring
System in Ukraine. In: Babak V., Isaienko V., Zaporozhets A. (eds) Systems, Decision and Control
in Energy I. Studies in Systems, Decision and Control, vol. 298, pp. 85-110 (2020). doi:
10.1007/978-3-030-48583-2_6
[27] A. Berg, K. Findell, B. Lintner, A. Giannini, S.I. Seneviratne, B. van den Hurk, R. Lorenz, A.
Pitman, S. Hagemann, A. Meier, et al. Land–atmosphere feedbacks amplify aridity increase over
land under global warming Nat. Clim. Change, 6 (2016), pp. 869-874.
[28] Appels, R.; Lefevre, B.; Herteleer, B.; Goverde, H.; Beerten, A.; Paesen, R.; De Medts, K.;
Driesen, J.; Poortmans, J. Effect of soiling on photovoltaic modules in Norway. Sol. Energy 2013,
96, 283–291.
[29] Al‐Housani, M.; Bicer, Y.; Koç, M. Assessment of various dry photovoltaic cleaning techniques
and frequencies on the power output of CdTe‐type modules in dusty environments. Sustainability
2019, 11, 2850, doi:10.3390/su11102850.
[30] Guo, B.; Javed, W.; Khoo, Y.S.; Figgis, B. Solar PV soiling mitigation by electrodynamic dust
shield in field conditions. Sol. Energy 2019, 188, 271–277.
[31] Guo, B.; Figgis, B.; Javed, W. Measurement of electrodynamic dust shield efficiency in field
conditions. J. Electrost. 2019, 97, 26–30.
[32] Gallagher, B. Rise of the machines: solar module-washing robots. Wood Mackenzie, GTM Report,
2018 URL: https://www. woodmac. com/our-expertise/focus/Power–Renewables/rise-of-the-
machines-solarmodule-washing-robots.
[33] K.K. Ilse, M.Z. Khan, K. Lange, N. Voicu, V. Naumann, C. Hagendorf Advanced performance
testing of anti-soiling coatings. Part II: particle-size dependent analysis and determination of
adhesion forces Sol. Energy Mater. Sol. Cells, 202 (2019), p. 110049.
[34] Qasem, H., Mnatsakanyan, A., & Banda, P. (2016, June). Assessing dust on PV modules using
image processing techniques. In 2016 IEEE 43rd Photovoltaic Specialists Conference (PVSC) (pp.
2066-2070). IEEE.
[35] Yap, W. K., Galet, R., & Yeo, K. C. (2015, December). Quantitative analysis of dust and soiling
on solar pv panels in the tropics utilizing image-processing methods. In 2015 Asia-Pacific Solar
Research Conference.
[36] Freyman, T., & Tran, T. (2019). Renewable Energy Discount Rate Survey Results—2018. Grant
Thornton UK LLP, https://www. grantthornton. co. uk/globalassets/1.-member-firms/united-
kingdom/pdf/documents/renewable-energy-discountrate-survey-results-2018. pdf, Accessed Feb,
1, 2020.
[37] Stridh, B. (2012, June). Evaluation of economical benefit of cleaning of soiling and snow in PV
plants at three European locations. In 2012 38th ieee photovoltaic specialists conference (pp.
001448-001451). IEEE.
[38] Cordero, R. R., Damiani, A., Laroze, D., Macdonell, S., Jorquera, J., Sepúlveda, E., ... & Torres,
G. (2018). Effects of soiling on photovoltaic (PV) modules in the Atacama Desert. Scientific
reports, 8(1), 1-14.
[39] Sverdlova, A., Zaporozhets, A., Bohachev, I., Popov, O., Iatsyshyn, A., Iatsyshyn, A., Kovach, V.,
Artemchuk, V., Hrushchynska, N. Self-Organizing Network Topology for Autonomous IoT
Systems. In: CEUR Workshop Proceedings, vol. 2850, pp. 57-70 (2021). http://ceur-ws.org/Vol-
2850/paper4.pdf
[40] Zaporozhets, A., Babak, V., Sverdlova, A., Isaienko, V., Babikova, K. Development of a System
for Diagnosing Heat Power Equipment Based on IEEE 802.11s. In: Zaporozhets A., Artemchuk
V. (eds) Systems, Decision and Control in Energy II. Studies in Systems, Decision and Control,
vol 346, pp. 141-151 (2021). doi: 10.1007/978-3-030-69189-9_8
[41] Sverdlova, A., Zaporozhets, A. Predicting Anomaly Conditions of Energy Equipment Using
Neural Networks. In: E3S Web of Conferences 280, 09005 (2021). doi:
10.1051/e3sconf/202128009005
[42] Babak, V.P., Babak, S.V., Myslovych, M.V., Zaporozhets, A.O., Zvaritch, V.M. Principles of
Construction of Systems for Diagnosing the Energy Equipment. In: Diagnostic Systems For
Energy Equipments. Studies in Systems, Decision and Control, vol 281, pp. 1-22 (2020). doi:
10.1007/978-3-030-44443-3_1
[43] Chine, W., Mellit, A., Lughi, V., Malek, A., Sulligoi, G., Massi Pavan, A. A novel fault diagnosis
technique for photovoltaic systems based on artificial neural networks. Renewable Energy, 90, pp.
501-512 (2016). doi: 10.1016/j.renene.2016.01.036
[44] Chao, K.-H., Ho, S.-H., Wang, M.-H. Modeling and fault diagnosis of a photovoltaic system.
Electric Power Systems Research, 78(1), pp. 97-105 (2008). doi: 10.1016/j.epsr.2006.12.012
[45] Aghaei, M., Grimaccia, F., Gonano, C.A., Leva, S. Innovative Automated Control System for PV
Fields Inspection and Remote Control. IEEE Transactions on Industrial Electronics, 62(11), pp.
7287-7296 (2015). doi: 10.1109/TIE.2015.2475235