Due to the observed increase of photovoltaic installations capacity in Poland, the research on the performance of different modules became an important issue from the practical and scientific point of view. This paper is intended to help system planners to choose photovoltaic modules and inverters taking into account the actual operating conditions. The study is devoted to the assessment of four different technologies of photovoltaic modules: polycrystalline silicon (pc-Si), amorphous silicon (a-Si), copper indium gallium selenide (CIGS), and cadmium telluride (CdTe). The data was collected at a solar plant located at high latitude location, in the eastern part of Poland, during the fourth year of the plant operation. The influence of irradiance on the temperature and efficiency of modules was studied. The results show that the efficiency of the pc-Si and CIGS modules decreases with rising temperature; however, the efficiency of the a-Si and CdTe modules is more stable. The impact of changing external conditions on the inverter efficiency as well as array and system losses during various seasons of the year was shown. The inverter efficiency reaches up to 98% in summer and drops as low as 30% in winter. Small average array capture losses of 7.41 (kWh/kWp)/month (0.25 h/day) are observed for the CIGS and 10.4 (kWh/kWp)/month (0.35 h/day) for pc-Si modules. The a-Si and CdTe array losses are higher, up to 2.83 h/day for CdTe in summer. The results indicate high annual energy yields of the pc-Si and CIGS modules, 1130 kWh/kWp and 1140 kWh/kWp, respectively. This research provided new data on pc-Si and especially the performance of the thin film modules and losses in a photovoltaic installation under temperate climate.
Keywords: photovoltaics; thin film modules; PV performance at high latitude; losses in PV system; inverter efficiency
The energy supply in Poland is based on fossil fuels, mainly coal. According to the current predictions, Poland will not manage to meet European Union requirement [[
Due to the increase in the PV capacity observed in many places around the world, the research on the performance of PV installations under the actual external conditions became an important issue. The studies of this kind help determine the annual yield and environmental or economic benefits of PV investments. Therefore, this topic is widely addressed in literature; however, mainly for the sites located at low latitude. Different kinds of PV technologies were evaluated in terms of spectral changes and effect of irradiance, represented by average photon energy (APE) and angle of incidence (AOI) e.g. in India [[
The photovoltaic performance under real outdoor conditions was also assessed in Turkey, Greece, India, Italy, Spain, and Peru [[
At higher latitude, the assessment of different modules types indicates gains in winter for the copper indium gallium selenide (CIGS) and crystalline silicon modules [[
In Germany [[
The study carried in Neatherlands [[
In Ireland, where monocrystalline modules were studied in rooftop installations, the following parameters were reported: The annual average module efficiency of 7.6% [[
Unfortunately, the studies on the PV performance at high latitude are rare. Literature devoted to the studies in Poland is limited to a single report on a-Si rooftop installation [[
Various kinds of solar modules mentioned above differ in structure, color, and appearance, but most importantly, they operate in different ways under solar illumination. The type of absorber material determines the main properties of the solar cell. The bandgap, absorption coefficient and in consequence the spectral response of the cells made with different semiconductors are not the same [[
This work presents the energy rating of four different kinds of PV modules: pc-Si, CIGS, a-Si, and CdTe at high latitude, in Poland, in the fourth year of the system operation. The investigations address the performance of inverters, which is the main factor influencing the system losses on different types of PV arrays.
The analysis is carried out also in terms of the annual energy yield and performance ratio, another important property of PV installation, defined as the operating efficiency divided by the efficiency under Standard Test Conditions (irradiance of 1000 W/m
The solar power plant on which the investigations were carried out is located in East Poland (latitude 51° 51′ N, longitude 23° 10′ E). Its total installed capacity equals 1.4 MWp; the installation is grid connected and covers an area of 3.5 ha. Because of the geographical location of the facility, the installed modules are inclined 34° from the horizontal plane facing south; the rows of panels are 6.3 m apart. All types modules are placed vertically (two rows in a set) except for the CdTe modules, which are mounted horizontally (four rows in a set). The modules are installed 0.5 m above the ground in order to ensure good heat dissipation. The pc-Si modules are installed in the main part of the photovoltaic plant. The smaller experimental part, which is separated, consists of four installations with three types of thin film modules (CIGS, CdTe, a-Si) and also traditional pc-Si modules. Within the area of the PV plant, the grass and weeds were mowed twice a year to avoid partial shadowing and lowering of the module output, but the modules were left uncleaned. All the installed modules were new and the whole plant started to operate at the end of 2014.
Solar irradiance was measured by using a monocrystalline silicon cell (5 × 3.3 cm), which was tilted at the same angle as the modules. The irradiance range of the sensor is 0–1400 W/m
The specification of PV modules and the capacity of each installation are given in Table 1. The polycrystalline modules are connected to the grid using an inverter with nominal apparent AC power 20 kW, characterized by the maximum efficiency of 98%, Euroeta efficiency of 97.8%, operating temperature range −20° to 60°C and maximum input voltage of 1000 V. The inverter is equipped with two maximum power point trackers (MPPT), 51 modules are connected to the one tracker (3 strings) and 34 modules to the second tracker (2 strings). Each type of thin film modules installation is connected to inverter of AC power 3.3 kW with maximum efficiency of 96.2%, Euroeta efficiency of 94.8%, operating temperature range −25 °C to 70 °C, maximum input voltage of 600 V and one MPPT. The ratio of PV peak power to the inverter nominal power is as follows: 1.06 for pc-Si; 1.0 for a-Si; 1.04 for CIGS, and 1.13 for CdTe.
The DC and AC electric power registered by inverters as well as solar irradiance and module temperature measurements were recorded every 5 min. by the central data logging computer to achieve synchronous data collection.
The solar plant under investigation is located in the warm summer continental climate (according to Köppen's classification) which is characterized by significant differences between the warm and cold parts of the year. Among many climate elements, solar irradiation is the most important factor determining the performance of the plant. Figure 1 shows the distribution of monthly irradiation over the entire year 2018 at the location of the installation. During the warm half of the year, three times more solar energy is received than in the cold part of the year, which is directly reflected in the energy production yields (v.i.). According to the prior measurements at the studied location, the annual sum of solar irradiation varied by about 200 kWh/m
At the location of the studied solar power plant, where a clear difference between the warm and cold half of the year occurs, the back PV module surface temperature reaches maximum 40 °C in October and almost 60 °C in August. In both parts of the year, the temperature of the modules increases with irradiance. Starting from January, when the maximum irradiance is about 700 W/m
High irradiance conditions result in the increase of the module temperature because of air temperature growth and direct heating of the modules surface. The rise of module temperature is the reason for efficiency drop in the case of pc-Si and CIGS technologies (Figure 3). This kind of changes is not observed for the a-Si. The PV efficiency values were computed as the quotient between PV power output P
(
where G—solar irradiance [W/m
During the days, when irradiance is below 600 W/m
Table 3 presents the efficiency of the PV modules based on the experimental data shown in Figure 3. All the estimated efficiency values are below nominal values, but the difference is significant only for CdTe, being half of the nominal value.
Figure 4 and 5 show the inverter efficiency as a function of normalized DC input, which is the ratio of daily AC energy output E
(
During the warm part of the year, the efficiency of the inverter increases with the DC input, reaching 95%–98% at high input conditions (for DC input above 20% of the inverter rated capacity). High values of the inverter efficiency (Figure 4) are observed from April to September for the inverters of each tested installation. Figure 4 also shows the difference between the quality of inverters used, since the nominal efficiency of the inverters equals 98% and 96.2% for the pc-Si and thin film installations, respectively. Under favorable conditions, the efficiency of inverters is close to its maximum value. In the cold part of the year (October to March), when the irradiance fluctuates or long low insolation periods occur, instantaneous input power to the inverter is very low or even zero and the inverter performance deteriorates (Figure 5). An inverter also requires some portion of the energy to power itself, so at low insolation when the input power is below the threshold level, an inverter is unable to even start its operation. Varying insolation results in discontinuous operation of the inverter and a decrease of its daily efficiency to 30%–80%.
The processes of solar radiation absorption and conversion are hindered by such factors as reflection of light, shadowing (soiling), degradation of modules, heating of modules that cause a decrease in voltage and, in consequence, their efficiency. The harmful role of factors of this kind are represented by the array capture losses L
(
Y
(
In addition to the losses connected with the absorption of the solar radiation, the amount of electric energy delivered to the grid is affected by system loses L
(
Y
(
The final yield and losses for each type of studied modules are shown in Figure 6; their shape follows the distribution of solar irradiation in each type of technology.
The array losses for the pc-Si modules are higher than for CIGS in the warm part of the year. The highest difference (13.3 (kWh/kWp)/month) was noticed in May, the sunniest month in 2018 at the location of the PV system. The lowest value was equal to 3.1 (kWh/kWp)/month in September. The average array losses difference between the pc-Si and CIGS systems in the warm part of the year was estimated as 8.3 kWh/kWp. CIGS is the only technology that shows higher array losses in the cold part of the year than in the warm part. The highest values of differences, comparing to pc-Si, were noticed in November (6.9 (kWh/kWp)/month) and January (6.2 (kWh/kWp)/month). Simultaneously, the CIGS system losses are high in the summer months, even up to 12.3 (kWh/kWp)/month in May. A-Si exhibits significant array losses through the entire year that ranged from 5 (kWh/kWp)/month in December to 41 (kWh/kWp)/month in May. In general, the array losses are the highest for the CdTe modules, especially in summer (with maximum value in May equal to 97.5 (kWh/kWp)/month). Thus, the energy yield of the CdTe modules is the lowest in spite of the minimum system losses in this case.
The values of energy yield for each type of technology in 2018 is presented in Table 4. In the presented study, based on the data collected in 2018 it is worth to address the degradation processes that were supposed to occur during previous years of solar plant operation. Quantity estimation of modules degradation was made possible by linear regression of the energy yield decrease [[
The performance ratio parameter PR (Figure 7) allows comparing the real system efficiency with the nominal efficiency at STC and reflects the overall losses occurring under the actual operating conditions due to the degradation, temperature influence, and soiling, of modules and inverter as well as performance of other electrical components. The performance ratio can be calculated as a ratio of the final yield Y
(
In order to avoid accidental errors, caused by e.g. inverter failure or local shading of some modules in the string, monthly performance ratio was computed as an average of daily PR values. The data was filtered to keep the standard deviation (SD) of average monthly PR below 3% in summer months and 5% in winter months for all technologies under study. The lowest value of SD was achieved for period between June and August (about 2%) which proves high precision of performance ratio estimation.
PR for the pc-Si installation is over 80% throughout the entire year, except for December, which is a good result. PR value in this case is changing from about 86% in August to 91% in April. This dependency is not observed for CIGS and a-Si technology. PR of CIGS is close to 90% for all summer months except September. The a-Si modules exhibit a performance improvement during the summer months. However, noticeable array losses for this technology result in a low overall PR value. The performance ratio of the CdTe installation does not yield the expected results, PR drops to 42%–45% in summer.
The actual performance of the photovoltaic system depends on the quality of the components and connections between them as well as the external conditions. The overlapping of these factors makes juxtaposing the results from different locations in the world difficult; however, in some aspects the comparison is possible.
The performance of the silicon modules tested in this work is similar to that observed at the same latitude in UK [[
CIGS, similarly to a-Si, exhibits better performance in the sunny and warm part of the year, which reflects the positive influence of the temperature growth on this PV technology. The bandgap value of CIGS (higher than for pc-Si) and temperature coefficient (lower than for pc-Si) foster summer gains. However, the summer system losses are largest in the case of the CIGS installation. This observation suggests the occurrence of occasional shutdowns in the inverter operation. Most of the time the system operates properly but sometimes, when the combination of very high irradiance and low temperature takes place, the saturation of the inverter can occur. PR for CIGS modules amounts to 90.7% in the summer months and 76.7% in the winter months. The annual average equals 83.7%. The CIGS modules, the mean efficiency of which is 12.5% and energy yield is similar to the value achieved by the pc-Si modules, definitely outperformed the thin film technologies in other studies.
In a PV installation, the system losses, which are mainly caused by the inverter, are influenced by high input current, power threshold, voltage threshold, excessive inverter power or voltage, inverter efficiency during operation, maximum power point tracking and also night consumption of energy. In this study, the efficiency of inverters exceeds 95% under high irradiation and drops significantly, even to 30% at low irradiation. All installations under study reveal a decrease of the performance ratio caused by an inverter that does not start convert the energy under the threshold energy level. This behavior reflects the strong differences between the warm and cold part of the year at the studied location. High inverter efficiency of 92%–93% in summer (August) was reported in Poland also by Pietruszko [[
Recently, the cloud edge effect was also indicated as one of the factors causing losses in the PV system. This kind of effect occurs on the days with blue skies when the cumulus clouds cover the Sun partially and sun shines strongly [[
The array losses are the lowest for the CIGS technology modules in summer (0.2 h/day). In winter, they are slightly higher than for pc-Si (0.29 h/day for CIGS and 0.21 h/day for pc-Si). The array losses are the highest for the CdTe modules (2.83 h/day in summer and 1.07 h/day in winter) which certainly results in low performance of this technology; however, the system losses are very low (0.09 h/day in summer and 0.06 h/day in winter). In literature, the array losses for c-Si range from 0.22 h/day [[
In the PV system under investigation, CdTe modules, in spite of a favorable temperature coefficient and small system losses, exhibit very low annual average performance ratio of 42.8%, which can be the effect of degradation. The reason for the degradation may be the residual snow occurring on this type of modules for extended periods of time. Infrared thermography inspection allowed observing that the CdTe modules used have a frame that dam up the snow. The residues such as snow may contribute to the degradation process as they shade the surface of the module and generate hot spots [[
Degradation processes occur in all types of modules; however, they involve different mechanisms: chemical reactions, mechanical damage, corrosion of metallic parts, encapsulate discoloration or delamination and antireflection coating failure. In general, PV modules degrade faster in hot and dry climates [[
In terms of the annual energy yield, the advantageous results are indicated by the comparison of the solar plant under study with other plants located in different places of the world. In this study, the annual energy yields were of 1031–1116 kWh/kWp for different PV technologies in 2015 and even 1130 kWh/kWp for the pc-Si and CIGS modules in 2018. In other countries, the following yields were achieved in the plants consisting of silicon modules (in order of increasing latitude): 990 kWh/kWp in Sweden, 1047 kWh/kWp in UK, 1030-1095 kWh/kWp in Poland, 1000 kWh/kWp in Germany, 1445 kWh/kWp in France and 1181 kWh/kWp in Italy [[
According to the obtained results, not only the most popular pc-Si modules but also the thin film CIGS modules are suitable for use under the climatic conditions of Poland. This interesting observation encourages testing other types of solar modules e.g. HIT (Heterojunction with Intrinsic Thin Layer) technology, which was already examined at low latitude (in India [[
Due to significant differences of weather during the warm and cold part of the year observed at high latitude, photovoltaic modules as well as other parts of the installation are exposed to a broad range of irradiance and temperature changes. The rise of irradiation in the summer months is beneficial for the electric energy production but leads to the module temperature growth, which can decrease the efficiency of modules; however, it is beneficial for the a-Si modules. The performance of inverters is also influenced by the external conditions. During the warm part of the year, when high input occurs, the efficiency of inverters reaches the highest values.
The energy rating (in kWh/kWp) of different kinds of modules was analyzed to show the technology-specific differences. The CIGS modules exhibit specific yield comparable to the pc-Si technology. In 2018, which was a sunny year, the yield of the CIGS modules reached 1140 kWh/kWp, similarly to the pc-Si yield equal to 1130 kWh/kWp. Other studied technologies (a-Si and CdTe) performed much worse mainly due to significant array capture loses that are probably the consequence of degradation processes.
Nowadays, when the growing interest in photovoltaic applications is observed even in high latitude countries, the results of the presented experimental research are valuable at the planning stage of a system. The provided data can be useful in the prediction of realistic power generation pattern for four different types of PV modules under changing external conditions.
Graph: Figure 1 Distribution of irradiation on the modules plane in 2018.
Graph: Figure 2 The dependence of the module temperature T on the solar irradiance G during the warm (red color) and cold (blue color) part of the year. Linear equations for the warm and cold half of the year, respectively, are as follows: T = 0.0159 × G + 34.112, T = 0.0110 × G + 20.26. The red and blue points overlap partially and not all of them are visible.
Graph: Figure 3 PV efficiency for different kinds of the modules vs irradiance.
Graph: energies-13-00196-g003b.tif
Graph: Figure 4 Inverter conversion efficiency as a function of normalized DC power in each studied installation. The plots show exemplary efficiency in June.
Graph: Figure 5 Inverter conversion efficiency as a function of normalized DC power in each studied installation. The plots show exemplary efficiency in February.
Graph: energies-13-00196-g005b.tif
Graph: Figure 6 Normalized yields and losses for the analyzed PV installation in 2018.
Graph: energies-13-00196-g006b.tif
Graph: Figure 7 Performance ratio for each studied PV technology, based on the data collected in 2018.
Table 1 Technical data of the modules and installation.
Parameter PV Module Type pc-Si a-Si CIGS CdTe Efficiency (%) 15.4 6.0 12.6 10.6 Max. power (W) 250 95 155 75 Number of modules 85 36 24 44 Installed capacity (kWp) 21.25 3.42 3.72 3.3 Total area (m2) 138.3 56.7 29.3 31.68 8.31 1.3 1.88 1.82 30.1 73 82.5 42 8.83 1.62 2.2 2.15 37.4 100 109 59.6 Temperature coefficient of power (%/°C) −0.40 −0.2 −0.31 −0.25 Temp. coefficient of ISC (%/°C) 0.04 0.09 0.01 0.02 Temp. coefficient of VOC (%/°C) −0.112 −0.34 −0.3 −0.24
Table 2 Fraction of solar irradiance at different levels.
In-Plane Fraction of Solar Irradiance 10–100 37.1% 100–200 15.0% 200–300 9.6% 300–400 6.7% 400–500 4.9% 500–600 4.7% 600–700 5.0% 700–800 4.6% 800–900 5.1% 900–1000 5.0% 1000–1100 1.9% 1100–1200 0.4%
Table 3 Efficiency of the PV modules based on the experimental data presented in Figure 3. Pc-Si: polycrystalline silicon; a-Si: amorphous silicon; CIGS: copper indium gallium selenide; CdTe: cadmium telluride.
PV Technology pc-Si a-Si CIGS CdTe All range of irradiance Average efficiency value (%) 14.6 4.4 11.0 4.4 Median efficiency value (%) 14.5 4.7 11.9 4.7 Standard deviation (%) 1.84 1.44 3.8 1.4 Irradiance Average efficiency value (%) 14.5 4.9 12.5 4.8 Median efficiency value (%) 14.4 5.0 12.4 4.8 Standard deviation (%) 1.98 1.18 2.6 1.1
Table 4 Energy yield for each PV technology in 2018.
Type of Modules pc-Si a-Si CIGS CdTe Energy yield in 2018 (kWh/kWp) 1130.54 855.98 1139.72 574.58 Energy yield in the warm half of 2018 (kWh/kWp) 855 679 885 448 Energy yield in the cold half of 2018 (kWh/kWp) 275 177 255 126
Table 5 Energy yields in 2015 [[
pc-Si a-Si CIGS CdTe Energy yield in 2015 (kWh/kWp) 1080 1031 1116 1058 Energy yield in 2016 (kWh/kWp) 1009 858 1046 732 936 764 970 439
Conceptualization, A.Z.; methodology, S.G., A.Z.; software and computing, S.G.; data curation, S.G.; formal analysis, A.Z., S.G.; investigation, S.G.; writing, A.Z.; supervision, A.Z. All authors have read and agreed to the published version of the manuscript.
This work was supported by Polish Ministry of Science and Higher Education. The solar plant was partially financed by EU project RPLU.06.02.00-06-086/12-00.
The authors declare no conflict of interest.
Authors would like to thank Piotr Dragan representing the government of partnership "Valley of Zielawa" for providing access to the solar plant.
By Agata Zdyb and Slawomir Gulkowski
Reported by Author; Author