Measurements and simulations of the performance of the PV systems at the University of Gävle

Transcription

1 FACULTY OF ENGINEERING AND SUSTAINABLE DEVELOPMENT Department of Building, Energy and Environmental Engineering Measurements and simulations of the performance of the PV systems at the University of Gävle Laura Zabala Urrutia 27 Student thesis, Master degree (one year), 5 HE Energy Systems Master Programme in Energy Systems Supervisor: Björn Karlsson Examiner: Richard Thygesen i

2 ii

3 Abstract In the following years, the countries will have to face an increase in the energy demand. So far, the fossil fuels have been the main source to meet the energy demand, but they involve serious problems: they contribute to the climate change with high emissions of greenhouse gases, there is an uneven distribution of these resources and their reserves are finite. The renewable energies are the most reliable alternative, with a very low environmental impact in comparison. Among them, the photovoltaics seems to be the most promising emerging technology for the electricity generation. Its rapid growth in the last years has been due to the reduction achieved in the cost of the PV panels. When planning a PV installation, it is essential to be able to estimate the production. The power of a PV-module is given by the manufacturer at standard conditions (STC), which means that the irradiance is G=W/m 2 at normal incidence and the temperature of the module is 25 C. However, these conditions will never be reached in a real installation. Therefore, the measured power of the system has to be adjusted for the real conditions so that the real production and performance can be estimated. Today there exists no standard method for this procedure in Sweden. The main aim of this thesis is to develop a theoretical model for the four PV-systems installed at the laboratories (building 45) of the University of Gävle to estimate the performance and production, and prove its validity by comparing with real data measured with a short time resolution (second). This will also allow to know if the power generated by the modules is the promised one by the providers. Three of the studied systems have monocrystalline silicon modules, with different schemes: one system with bypass diodes, another with TIGO optimizers, and the third one with microinverters. The fourth system has thin film modules. The theoretical model considers correction factors for the cell temperature, the angle of incidence and the real irradiation reaching the modules surface; as all these aspects reduce the power obtained. When studying this model for clear sunny days, it can be observed that the theoretical model adjusts perfectly for the four systems in these conditions and almost a completely linear dependence is achieved between the measured and estimated power. The worse adjustment is obtained for the thin film system, for which the theoretical model gives lower values than the real ones. However, a better approximation can be obtained for this system by adjusting the value of the correction factor for the cell temperature. Moreover, the high values obtained for the maximum power during the clear days, very close to the peak power, indicates that the maximum power value provided by the manufacturers is in concordance with the real performance of the modules. In case of cloudy days, a small-time delay has been appreciated between the data recorded by both logger. The results have been studied with the raw data, obtained worse adjusting, and correcting this time discordance, getting again accurate results from the theoretical model. Key words: PV, PVperformance, PVPowerModel, CorrectionFactors, AngleofIncidence, Irradiation. iii

4 iv

5 Acknowledgments First of all, I want to show my gratitude to my supervisor, Björn Karlsson, for his dedication and support through all the Thesis development. I would also like to acknowledge Mattias Gustafsson for his support and commitment. I would also like to express my sincere and deep gratitude to my family. Even if they are so many kilometres away, they are always my biggest support, encouraging me to follow my dreams and giving me all their love and help. And last, but not least, I would like to thank all the friends that I made during this amazing and unforgettable experience in Sweden during the whole year. Friends that have become as a family for me, who have been everyday there caring of me, enjoying the big adventures but also the little things in life, and supporting me. v

6 vi

7 Table of contents. Introduction..... Motivation Purpose Limitations Theoretical background Solar radiation Photovoltaics (PV cells) Photovoltaic effect and semiconductors PV cells PV technologies Crystalline cells Thin Film PV installation Inverters By-pass diodes TIGO optimizer PV performance and output IV and PV curve Performance of a PV module Output of a PV module Influence of temperature in the performance Influence of the Angle of Incidence in the performance Model for real operation conditions METHOD PV-installation at HIG Location Solar panels Inverter TIGO optimizer Measurements Power measurements: new egauge logger Other data acquisition PROCEDURE AND OPERATION Calibration of the egauge logger General procedure... 2 vii

8 4. Results Calibration of the egauge logger Determination of the study hours Comparison between the theoretical model and measured power for a clear sunny day Comparison between the theoretical model and measured power for cloudy days Time delay correction Performance of the systems Influence of other parameters Influence of h (temperature correction factor) Discussion Validity of the model in a clear sunny day Validity of the model in a cloudy day Time delay correction Performance of the system Influence of h (temperature correction factor) Conclusions References APPENDIX Appendix I: Results for 4 th May Appendix II: Results for 28 th April viii

9 List of Figures Figure Cumulative PV capacity worldwide. [6]... Figure 2 Solar radiation reaching a tilted surface. [9]... 3 Figure 3 Cross section of a solar cell. []... 4 Figure 4 Solar cell, module, panel, and array constitution. [4]... 5 Figure 5 Different types of inverters for a PV system. [6]... 7 Figure 6 Series connected cells with by-pass diodes. [9]... 8 Figure 7 IV and PV curves of a PV-module. [2]... 9 Figure 8 IV curves for series and parallel connected cells. [2]... 9 Figure 9 Efficiency of a Si module as a function of the irradiance. [22]... Figure IV curves for a cell at different operation temperatures. [23]... Figure Angle of incidence of solar radiation towards a tilted surface. [25]... 2 Figure 2 Azimuth and tilt angles of the surface Figure 3 Dependence of the correction factor with the angle of incidence for b=, Figure 4 PV-panels installed at the Laboratory of the University of Gävle Figure 5 PV systems at the Hall Figure 6 IV and VP curves for the Monocrystalline Silicon modules Figure 7 IV curves for the Thin Film modules at different temperatures and irradiances Figure 8 SUNNY BOY 2 inverter. [27]... 8 Figure 9 egauge EG3 logger. [28]... 9 Figure 2 Pyranometers installed in the Laboratory Figure 2 Power generation of the Si Diod system and irradiation for 28 th April Figure 22 Power generation of the Si Tigo system and irradiation for 28th April Figure 23 Power generation of the Si micro-inverter system and irradiation for 28 th April Figure 24 Power generation of the Thin Film system and irradiation for 28 th April Figure 25 Measured power by the new logger and Irradiance during 3rd May Figure 26 Measured power by the new logger and Irradiance during 28th April Figure 27 Measured power by the new logger and Irradiance during th May Figure 28 Measured power, theoretical power and irradiation of Si Diod system for 3rd May for the whole day Figure 29 Measured power, theoretical power and irradiation of Si Diod system for 3rd May from 8: to 4: Figure 3 Measured power and theoretical power of Si Diod system for 3rd May from 8: to 4: Figure 3 Theoretical vs. Real power for Si Diod system, 3rd May Figure 32 Irradiance vs. Real Power for Si Diod system, 3rd May Figure 33 Irradiance vs. Real Power for Si Diod system, 3rd May, with intersection in the origin Figure 34 Measured power, theoretical power and irradiation of Si Tigo system for 3rd May from 8: to 4: Figure 35 Measured power and theoretical power of Si Tigo system for 3rd May from 8: to 4: Figure 36 Theoretical vs. Real power for Si Tigo system, 3rd May Figure 37 Irradiance vs. Real Power for Si Tigo system, 3rd May Figure 38 Measured power, theoretical power and irradiation of Si micro-inverter system for 3rd May from 8: to 4: ix

10 Figure 39 Measured power and theoretical power of Si Microinverter system for 3rd May from 8: to 4: Figure 4 Theoretical vs. Real power for Si microinverter system, 3rd May Figure 4 Irradiance vs. Real Power for Si Microinverter system, 3rd May Figure 42 Measured power, theoretical power and irradiation of Thin Film system for 3rd May from 8: to 4: Figure 43 Measured power and theoretical power of Thin Film system for 3rd May from 8: to 4: Figure 44 Theoretical vs. Real power for Thin Film system, 3rd May Figure 45 Irradiance vs. Real Power for Thin Film system, 3rd May Figure 46 Measured power, theoretical power and irradiation of Si Diod system for th May for the whole day Figure 47 Measured power, theoretical power and irradiation of Si Diod system for 3rd May from 8: to 4: Figure 48 Measured power and theoretical power of Si Diod system for th May from 8: to 4: Figure 49 Theoretical vs. Real power for Si Diod system, th May Figure 5 Irradiance vs. Real Power for Si Diod system, th May Figure 5 Measured power, theoretical power and irradiation of Si Tigo system for th May from 8: to 4: Figure 52 Measured power and theoretical power of Si Tigo system for th May from 8: to 4: Figure 53 Theoretical vs. Real power for Si Tigo system, th May Figure 54 Irradiance vs. Real Power for Si Tigo system, th May Figure 55 Measured power, theoretical power and irradiation of Si Microinverter system for th May from 8: to 4: Figure 56 Measured power and theoretical power of Si Microinverter system for th May from 8: to 4: Figure 57 Theoretical vs. Real power for Si Microinverter system, th May Figure 58 Irradiance vs. Real Power for Si Microinverter system, th May Figure 59 Measured power, theoretical power and irradiation of Thin Film system for th May from 8: to 4: Figure 6 Measured power and theoretical power of Thin Film system for th May from 8: to 4: Figure 6 Theoretical vs. Real power for Si Microinverter system, th May Figure 62 Irradiance vs. Real Power for Thin Film system, th May Figure 63 Measured power and theoretical power of Si Diod system for th May from 8: to 4: with time delay correction Figure 64 Theoretical vs. Real power for Si Diod system, th May with time delay correction Figure 65 Measured power and theoretical power of Si Tigo system for th May from 8: to 4: with time delay correction Figure 66 Theoretical vs. Real power for Si Tigo system, th May with time delay correction Figure 67 Measured power and theoretical power of Si micro-inverter system for th May from 8: to 4: with time delay correction Figure 68 Theoretical vs. Real power for Si Tigo system, th May with time delay correction x

11 Figure 69 Measured power and theoretical power of Thin Film system for th May from 8: to 4: with time delay correction Figure 7 Theoretical vs. Real power for Si Tigo system, th May with time delay correction Figure 7 Measured power vs. Theoretical power for different h values, 3rd May, Si Diod system Figure 72 Measured power vs. Theoretical power for different h values, 3rd May, Si Tigo system Figure 73 Measured power vs. Theoretical power for different h values, 3rd May, Si Microinverter system Figure 74 Measured power vs. Theoretical power for different h values, 3rd May, Thin Film system Figure 75 Measured power and theoretical power of Si Diod system for 4th May from 8: to 4: Figure 76 Theoretical vs. Real power for Si Diod system, 4th May Figure 77 Measured power and theoretical power of Si Tigo system for 4th May from 8: to 4: Figure 78 Theoretical vs. Real power for Si Tigo system, 4th May Figure 79 Measured power and theoretical power of Si Microinverter system for 4th May from 8: to 4: Figure 8 Theoretical vs. Real power for Si microinverter system, 4th May Figure 8 Measured power and theoretical power of Thin Film system for 4th May from 8: to 4: Figure 82 Theoretical vs. Real power for Thin Film system, 4th May Figure 83 Measured power and theoretical power of Si Diod system for 28 th April from 8: to 4: Figure 84 Theoretical vs. Real power for Si Diod system, 28 th April Figure 85 Measured power and theoretical power of Si Tigo system for 28 th April from 8: to 4: Figure 86 Theoretical vs. Real power for Si Tigo system, 28 th April Figure 87 Measured power and theoretical power of Si Microinverter system for 28 th April from 8: to 4: Figure 88 Theoretical vs. Real power for Si microinverter system, 28 th April Figure 89 Measured power and theoretical power of Thin Film system for 28 th April from 8: to 4: Figure 9 Theoretical vs. Real power for Thin Film system, 4th May xi

12 List of Tables Table PV systems installed in the Laboratory at the University of Gävle Table 2 Electrical characteristics of the EOPLLY 25 Monocrystalline silicon modules at Standard conditions Table 3 Temperature coefficients for the EOPLLY 25 Monocrystalline silicon modules Table 4 Electrical characteristics of the Thin Film CIGS 85W modules at Standard conditions Table 5 Temperature coefficients of the Thin Film CIGS 85W modules Table 6 Characteristics of the SUNNY BOY 2 inverter Table 7 Characteristics of the TIGO optimizer Table 8 Data from the installation for the calculation of the angle of incidence Table 9 Peak power and maximum recorded power for each system xii

13 Nomenclature A AC b CIGS CO 2 DC E G G b G d GHG h hh I b I d I MP IR I SC K b(θ) K d L L L ST mm MPP MPPT n ɳ OECD P Area of the module Alternating current Correction factor for the K b(θ) calculation Copper indium gallium diselenide Carbon dioxide Direct current Output energy Total solar irradiation Beam or direct irradiation Diffuse irradiation Greenhouse gases Increase rate for ambient temperature Hours Beam or direct irradiation Diffuse irradiation Maximum power current Infrared radiation Short circuit current Correction factor for direct irradiation Correction factor for diffuse irradiation Local meridian Standard meridian Minutes Maximum power point Maximum power point tracking Day number Efficiency Organisation for Economic Co-operation and Development Power xiii

14 PLC P MAX P PEAK PV STC T T amb T cell UV VIS V MP V OC α β β γ γ δ θ λ ϕ ω Power line communications Maximum power Peak power Photovoltaics Standard conditions Temperature Ambient temperature Cell temperature Ultraviolet radiation Visible light Maximum power voltage Open circuit current Temperature coefficient of I SC Temperature coefficient of V OC Tilt Azimuth Temperature coefficient of MPP Declination Angle of incidence Latitude Correction factor for operating temperature Hour angle xiv

15 . Introduction.. Motivation One of the main challenges of the last years for all the countries is facing the increase of energy demand. The reports state that the energy consumption in OECD countries will grow by 4%, while for the non-oecd countries the energy consumption will increase by 84%. [] The main current source for the heat and energy production are the fossil fuels (oil, carbon, and natural gas), that constitute around the 8% of the energy consumption worldwide. [2] However, the use of fossil fuels involves important problems: they account for the 65% of the total greenhouse gas (GHG) emissions worldwide [3], they are not a reliable source since there is an uneven distribution of these resources being most of reserves located in unsecure regions of Middle East, and they are finite resources (the proved reserves would last for years in the case of the carbon, 53 for the oil and 54 years for natural gas). [4] So, it is crucial to find alternative ways for the energy generation which do not contribute to the climate change and turn out to be a reliable source. In this search, the renewable energies are the most promising sources as their environmental impact is significantly less than the fossil fuels and they use natural resources (wind, sun, water) available almost everywhere. In order to promote the use of renewable energies, different policies have been settled, such as the Kyoto protocol (997), an international agreement which commits the participating countries to reduce their emission of GHG, specially the CO 2 release. Likewise, the 2/2/2 Directive from the European Union (29) is another measure which has a great importance in the future role of the renewable energies in Europe as it stablishes that every country should achieve a reduction of 2% of their CO 2 emissions and a growth by 2% of their electricity consumption from renewable sources by 22. [5] Among the renewable energies, the PV is one of the most promising emerging alternatives. By 23, the installed capacity worldwide was of 35 GW in total. Although Europe was the pioneer leading the first PV installations, the last years Asia has taken over that role. More precisely, China has been the country which has installed more capacity these last years, followed by Japan and United States, as it can be seen in the figure. Figure Cumulative PV capacity worldwide. [6]

16 The rapid growth of this technology is mainly due to the great reduction in the cost of PV modules that has been achieved (divided by five in the last six years). [7] However, it still presents some challenges, such as the options of including interconnections, demand-side response, flexible generation, and storage. When planning a PV installation, it is essential to be able to estimate the production. The power of a PV-module is measured during standard conditions (STC), which means that the irradiance is G=W/m 2 at normal incidence and the temperature of the module is 25 C. However, these conditions will never be reached in a real installation. Therefore, the measured power of the system has to be adjusted for the real conditions so that the real production and performance can be estimated. Today there exists no standard method for this procedure in Sweden..2. Purpose The main aim of this thesis is to develop a theoretical model for the four PV-systems installed at the laboratories (building 45) of the University of Gävle to estimate the performance and production. The validity and accuracy of the model will be analysed by comparing the results with the measurements taken in operation conditions with a short time resolution. In this way, the developed model can be used to determine the estimated production of an installed PV system. Its performance is dependant of the solar intensity, the modules temperature, and the angle of incidence. In addition, it can also be used to prove if the generated power matches the one promised by the supplier of the modules. The Thesis Project include the following tasks: Installing a measurement system to record the power of the systems with a short time resolution ( second). The irradiance, and ambient temperatures are monitored in an existing logger. Monitoring the performance during a short period. Evaluating the performance of the system. Developing a model of the performance. Comparing measurements and simulations..3. Limitations There are some limitations in the development of this project. Firstly, when defining the theoretical model, some approximations have been made. The most remarkable ones are that it has been assumed that the efficiency of the modules remain constant with the irradiation (simplifies the calculations) and the use of estimated values for the system efficiency, the diffuse correction factor, and the cell temperature. Moreover, the data of the power of the modules acquired by the new egauge logger is very accurate because of this device s resolution. However, the data of the ambient temperature and irradiation is recorded by the old logger, which has a lower resolution (minute resolution), which implies a loss of accuracy when using this data compared to the power data. 2

17 2. Theoretical background 2.. Solar radiation The sun radiates as a blackbody at a temperature of 5777 K. This solar radiation towards the Earth has an average value of 367 W/m 2 outside the atmosphere, which is called the solar constant. So, this can be defined as the energy from the sun per unit time received on a unit area of surface perpendicular to the direction of propagation of the radiation at mean earthsun distance outside the atmosphere. [8] The solar radiation can be divided into three parts: Ultra violet radiation (UV) with a wavelength λ<.38 µm Visible light (VIS) with.38< λ<.78 µm Infrared radiation (IR) with λ>.78 µm Before entering the atmosphere, the energy content of each part is: UV 6.5%, VIS 47.9% and IR 45.6%. After crossing the atmosphere, as part of the radiation is scattered, the distribution changes and almost all the solar radiation is in the range.3 to 2.5 µm. The total radiation reaching the Earth s surface (G) is the sum of the direct and diffuse radiation. The direct or beam radiation (G b) is the solar radiations that reaches the Earth s surface without being scattered by the atmosphere, and thus, with no change in its direction. The diffuse radiation (G d) is the part of the solar radiation that is scattered by the atmosphere so that its direction is changed. [8] When considering the solar radiation reaching a tilted surface, the ground reflected radiation (G g) also needs to be considered Photovoltaics (PV cells) Figure 2 Solar radiation reaching a tilted surface. [9] Photovoltaic effect and semiconductors The photovoltaics refers to the generation of electricity from light. This is based on the photovoltaic effect, discovered by Becquerel in 839, which implies that sunlight striking certain materials is able to generate a measurable electrical current. [] This photovoltaic effect occurs in the semiconductor materials, which are used to manufacture solar cells. These semiconductor materials behave as insulators at low temperatures, while they are good conductors when there is available heat or energy. [] 3

18 PV cells The conventional PV cells consist on a junction of two thin layers of different semiconductor materials. These are known as p(positive)-type semiconductor and n(negative)-type semiconductor. The p-type semiconductor is made of crystalline Silicon with an amount of an impurity (usually Boron) that makes the material to have a deficit of electrons. Thus, a p-semiconductor has many holes (missing electrons), and few electrons. The n-type semiconductors are also made of crystalline Silicon, but they are doped with impurities (basically phosphorus) that leads to a surplus of electrons. When these two different semiconductors are joined, and exposed to a light source with a suitable wavelength, the holes move from the p to n-semiconductor, and the electrons on the opposite direction, generating an electric field. [2] Figure 3 Cross section of a solar cell. [] A typical PV cell produces a voltage of around.5-.6 V and a current of about 3 A at Standard conditions (temperature of 25 ºC, normal irradiance of W/m 2 ) [3], which means just few watts of power. The solar cells are connected in series to obtain the desired voltage (usually 2-24 V), and in parallel to reach the desired current. Several solar cells mounted together form a module. Likewise, several modules connected constitute a solar panel, and a group of panels form an array, as it can be seen in figure 4. 4

19 Figure 4 Solar cell, module, panel, and array constitution. [4] 2.3. PV technologies The PV cells can be made of different materials. According to their development and presence in the market, there are three generations of PV cells technologies: First generation: It encompasses the crystalline silicon modules, both monocrystalline and polycrystalline. This technology is the most common one, accounting for 9% of the PV cells, and it is fully commercial. [5] Second generation: It comprehends the thin film technologies. The main ones are: Amorphous silicon, Cadmium telluride, Copper indium selenide (CIS) and Copper indium gallium diselenide (CIGS). Third generation: refers to organic photovoltaics technologies, which are still being tested. Just the two first groups will be studied in detail, as at the moment they are the most relevant ones Crystalline cells Monocrystalline Silicon cells The Monocrystalline Silicon cells are made of Silicon with a single, ordered, and continuous crystal lattice structure (with no defects or impurities). This crystal structure is obtained by a process called Czochralski. [5] The cells made of this type of silicon are highly efficient, but it is a very expensive technology since the manufacturing cost is very high. The typical efficiency for the monocrystalline Si cells is around -2%. [5] Polycrystalline Silicon cells In the pursuit of finding new ways to manufacture Silicon crystal cheaper, the multicrystalline or polycrystalline Silicon cells appear. These cells consist on small grains of monocrystalline Silicon put together. This leads to some separation or grain boundaries between the regions of the crystal. 5

20 The manufacturing of this material is cheaper and easier, and it requires less electricity and material consumption. However, the efficiency of these cells is lower Thin Film The thin films are formed by thin layers of semiconductor materials that are applied to a solid backing material. So, they can be of various kinds, but as it has been mentioned above, the most common ones are: Amorphous Silicon, Copper indium (gallium) diselenide (CIGS) and Cadmium telluride. The main characteristic of this technology is the cost reduction in their manufacturing process. This aspect, together with other advantages (high flexibility and easy installation) is increasing their use. Nevertheless, they have worse efficiencies: 5-3%. [5] 2.4. PV installation A PV installation must include a series of components so that the conversion of sunlight energy into electricity to consume is ensured. The PV modules are responsible for absorbing the sunlight and convert it into electricity, as it has been explained above. The PV modules generate DC electric current, so an inverter must be installed to convert it to AC. Moreover, the corresponding mounting, cabling and electrical accessories are also necessary. The PV installations may also include a solar tracking system to orientate the panels in the optimal position so that the maximum amount of solar radiation reaching the surface can be exploited and converted into electricity. Finally, the PV installations can include batteries to store the excess of produced energy so that it can be used during the nights or at periods with low irradiation. Below, a more detail explanation is included for the installation equipment that is part of the studied PV installation at the University Inverters The figure 5 shows the main inverters schemes for the PV installations, where a) corresponds to the centralized technology, b) to the string inverter, c) to the multi-string technology and d) to the AC-module and AC- cell technologies. 6

21 Figure 5 Different types of inverters for a PV system. [6] Below, the main characteristics of each inverter type are explained: Centralized inverters: through this type of inverter a large number of PV modules can be connected to the grid directly. These modules are divided into different series connections called strings, and these strings connected in parallel with diodes, as it can be seen in the figure 5 a). Although this technology was the most common one in the past, it has important limitations, such as the appearance of many current harmonics, a low-quality power and several losses. [6] String inverters: the centralized system has been replaced by configurations in which each single string is connected to the inverter (figure 5, b), or maintaining the connection of several strings to a single inverter, but having each string its own DC/DC converter (figure 5, c). Usually, the scheme with the DC/DC converter is used together with a maximum power point tracking (MPPT) algorithm and power line communications (PLC) for each module. These devices are known as optimizers and they allow establishing a different working point for each module, reducing the losses. [7] Micro-inverters or module-integrated converters are converters of low power rating (ranged between 5 4 W) that are connected only to one module, in contrast to the previous technologies. They require an additional DC-DC converter because of the low power rating before connecting to the grid. The main advantage of the microinverters is that they reduce the effect of shading. Regarding the strengths of this type of inverters, the high-power quality and safety can be highlighted. The microinverters include a Maximum Power Point Tracking system, which is an electronic device that detects at each moment which is the MPP and makes the best use of the PV modules array so that the electricity production is maximized. It works thanks to different algorithms implemented on it. There exist two types of algorithms: conventional algorithms for uniform irradiance conditions (such as perturb and 7

22 observation or incremental conductance) and stochastic techniques for partial shading conditions (Particle swarm optimization or Differential evolution). [8] By-pass diodes The by-pass diodes are included in the PV installation to mitigate the effect of shading. These diodes are connected in parallels to cells connected in series. If there are no shadows in the system, the diodes are reverse biased and the current circulates through all the cells that are generating. If one of the cells is shadowed, the diode of this cell will start conducting. Thus, the current does not circulate through the shaded cell and the shading impact is alleviated. For example, in the figure 6, the fourth cell would be bypassed. Figure 6 Series connected cells with by-pass diodes. [9] In real installations, not every cell is equipped with a by-pass cell. This diode is shared by a group of cells connected in series. For example, a string of 6 cells usually have 3 diodes, one for each 2 cells. [] TIGO optimizer The TIGO optimizer is an alternative solution to the micro-inverters which allows extracting the maximum possible energy from each module, removing the negative effect of weaker modules from the rest of the PV array, such as the ones that are partially or completely shaded. The TIGO Optimizer uses the Impedance matching approach to extract the maximum possible power from each of the modules of the array. [2] Its operation is based in three steps: - Analog sensing by measuring the voltage, current and temperature of the modules - Transmitting this input data to the TIGO Management Unit and receiving the operation point - Controlling the output of each module by the impedance matching components in the Optimizer 2.5. PV performance and output IV and PV curve The IV and PV curves characterize the performance of the PV-modules. In figure 7, both are represented. 8

23 Figure 7 IV and PV curves of a PV-module. [2] The main characteristics points (shown in figure 7) are described hereafter: Short circuit current I SC: current at which the voltage is. It represents the maximum possible current. Open circuit voltage V OC: voltage at which the current is. It is the maximum possible voltage. Maximum power current I MP and Maximum power voltage V MP: voltage and current values at which the maximum power is achieved. Maximum Power Point MPP (P MAX): point at which the product of I and V is maximum so that the highest possible value of power for that module is obtained. When the cells are connected in series, the voltage is added while the current remains the same. Thus, if two cells are connected in series, the voltage is doubled and so is the V OC (figure 8). In the case of connecting the cells in parallel, the current is added and the voltage does not change. So, the I SC is the one that is increased in the parallel connection. [2] Figure 8 IV curves for series and parallel connected cells. [2] 9

24 Performance of a PV module The efficiency of a cell with a surface area A is defined as: ɳ= P G A () where P is the power [W], G is the irradiance reaching the module [W/m 2 ] and A is the area of the module [m 2 ]. The power of the module is measured at Standard conditions (STC): irradiance G= W/m², cell temperature 25 and normal incidence θ =. This power also corresponds to the maximum power W p or Watt-peak. So, the Peak power P peak of a module with an area A and an efficiency ɳ can be defined as: P peak (W p)= A G= W/m², T=25, Normal incidence (2) Output of a PV module The output E(kWh) of a module with an area of A(m²) and an efficiency during a period with an irradiation G(kWh/m²) is: E(kWh)= A G = P peak G (3) As the output E [kwh] can be calculated as the product of P peak and the irradiance G [kwh/m 2 ], P peak has the dimensions of kw /(kw/m 2 )= m 2. This corresponds to the PV-area with an efficiency of %, which gives P peak = kw. In practice, the output is influenced by the operating temperature of the module (higher than 25 C) and the real angle of incidence θ of the sunlight (larger than ), so the following formula is used for the calculation of the output: E(kWh)= P max G = ϕ*η*a*g (4) where is the correction factor for the operating temperature and it takes the value.85.9 The determination of the output of the module at real operation conditions involve some difficulties: the determination of the angle of incidence θ at each instant and the estimation of the system efficiency and the solar cell efficiency. The efficiency of the cell varies with the irradiance as it can be seen in figure 9. At high irradiances, the efficiency can be considered constant, while at low irradiances the variation of the performance with G is more significant. [22]

25 Figure 9 Efficiency of a Si module as a function of the irradiance. [22] Influence of temperature in the performance The operation temperature of the cell increases compared to the ambient temperature due to the irradiance reaching the cell. This dependence can be considered linear and the increase rate is the same for a wide variety of modules. [6] The cell temperature can be calculated by the following expression: T cell=t amb+h G (5) Where h is the rate of increase and it takes the value h=3 C m 2 /Kw This increase in the cell temperature causes a rise in the output current of the cell, but a decrease in the output voltage. Thus, the short circuit current will be higher for a greater temperature, while the open circuit voltage will diminish significantly. [23] Figure IV curves for a cell at different operation temperatures. [23] As a result, the efficiency of the cell is lower for a higher operation temperature. The physical explanation for this fact is that the internal carrier recombination rates increase with temperature. So, more electrons recombine with holes, not producing electric current. [24] Influence of the Angle of Incidence in the performance The power generated by the PV modules is also influenced by the angle of incidence, so it is important to study the solar angles.

26 Solar angles The performance of a PV module is affected by the angle of incidence θ of the solar radiation towards a surface, which can be defined as the angle between the sun beam and the vector perpendicular to the surface. This angle provides the available sun energy that will be converted into electricity in the module. [] The angle of incidence θ is a function of the declination δ, the hour angle ω, the tilt β, the azimuth γ and the latitude λ. It can be calculated by the following expression: Cos(θ)=cos(δ)sin(ω)sin(β)sin(γ)+cos(δ)cos(ω)sin(λ)sin(β)cos(γ)- sin(δ)cos(λ)sin(β)cos(γ)+cos(δ)cos(ω)cos(λ)cos(β)+sin(δ)sin(λ)cos(β) (6) Where θ is the angle of incidence [ ], δ the declination [ ], ω the hour angle [ ], β the tilt angle of the surface [ ], γ the azimuth angle of the surface [ ] and λ the latitude [ ]. The figure shows the angle of incidence, while the figure 2 shows the tilt and the azimuth. Figure Angle of incidence of solar radiation towards a tilted surface. [25] Figure 2 Azimuth and tilt angles of the surface. The declination angle δ is the angle between the plane of the equator and the solar beam, so it varies during the year among the values < δ< It can be calculated as: δ = sin(36 (284+n)/365) [ ] (7) where n is the day number. 2

27 Correction factor Kb(θ) The hour angle ω determines the degrees of Earth rotation before or after solar noon at a given longitude. [] It can be calculated from the solar time as follows: ω =((hh-2) + mm/6)*5 [ ] (8) where hh corresponds to the the solar hours and mm the solar minutes. The difference between the solar and normal time can be obtained as: Solar time Normal time = 4(L ST-L L)+E [min] (9) Where L ST is the standard meridian, L L is the local meridian and E is a correction factor of the solar time Correction factor for the angle of incidence The angle of incidence affects just the direct or beam irradiation reaching the module, and it is considered by introducing a correction factor K b(θ) when calculating the power. This correction factor is defined as: Where b is a factor with a value between, and. K b(θ)= -b ( ) () cos (θ) In figure 3, this dependence is shown. Due to geometry, this relation is valid for angles smaller than 9º.,4, Angle of incidence θ Figure 3 Dependence of the correction factor with the angle of incidence for b=, Model for real operation conditions The power of a PV-module is measured at Standard conditions (STC), this is, with an irradiance G= W/m 2 at normal incidence and a module temperature of 25ºC. This can be expressed as: P (T=25, θ=)= η*g(25,)**a () Where θ is the angle of incidence [ ], η is the efficiency of the module, G is the irradiance [W/m 2 ], and A the module area [m 2 ]. 3

28 For the calculation of the power at real conditions, the influence of the cell temperature and angle of incidence must be considered. Moreover, a more accurate model is obtained if the direct and diffuse irradiation are considered. In the case of the direct irradiance, its contribution depends on the angle of incidence, so this irradiation will be multiplied by a correction factor K b(θ). The diffuse irradiation does not depend on this factor, so it will be multiplied by a constant correction factor K d. [26] Thus, the real power of the PV module can be calculated as: P (T, θ) = P(25,) [K b(θ) I b + K d I d ] [+(T cell 25) α] (2) where K b is the correction factor for direct irradiance, I b is the beam or direct solar radiation, K d is the correction factor for diffuse irradiance, I d is the diffuse solar radiation and α is the temperature coefficient for the power of the module. The efficiency of a module is influenced by the same factors in analogous way: η (T, θ) = η(25,) [K b(θ) I b + K d I d ] [+(T cell 25) α] (3) 4

29 3. METHOD 3.. PV-installation at HIG 3... Location The studied PV panels are located in the south façade of the Laboratory (Hall 45) of the University of Gävle. The exact coordinates are: 6 4 N and 7 6 E. The panels are installed tilted 45. Figure 4 PV-panels installed at the Laboratory of the University of Gävle Solar panels In the Laboratory, four different systems of PV modules can be found according to the type of module and the scheme of the elements that compromise them. In the figure 5, the four systems are pointed out. Figure 5 PV systems at the Hall 45. The first system on the left below consist on 6 silicon monocrystalline modules connected in series with bypass diodes, a DC/AC inverter and a MPP tracking system. 5

30 Next to this system, the Silicon modules with Tigo optimizer are placed. Each of the modules have a Tigo-optimizer, so they work independently, and they are connected in series. Later, the characteristics of this optimizer will be described. In the same level as these systems, the schemes of Silicon modules with bypass diodes and module micro-inverter are installed. There are 3 micro-inverters for 6 PV modules, which means that the panels are connected in parallel so that there is a micro-inverter for every two panels. Above all these Silicon modules, a system with 9 thin film GICS modules is placed. In the table, the number of modules and total peak power of each system is shown. Table PV systems installed in the Laboratory at the University of Gävle. Panel type Number of Peak power P peak per Total Peak power modules module [W] P peak [W] Silicon modules with bypass diodes Silicon modules with bypass diodes and TIGO-optimizer Silicon modules with bypass diodes and micro inverter Thin film CIGS modules with bypass diodes and module inverter So, the three systems installed below have the same type of PV modules (Monocrystalline silicone), and the PV panels above are Thin Film ones made of CGIS. Below, the main characteristics of these are described Silicon modules The silicon modules are EOPLLY 25 Monocrystalline series solar modules made of 72 pcs 25 25mm monocrystalline solar cells connected in series with high efficiency, high transmission rate and low iron tempered glass. The area of eah module is 58 mm 88 mm. The table 2 shows the main electrical characteristics of this type of PV modules at Standard conditions. Table 2 Electrical characteristics of the EOPLLY 25 Monocrystalline silicon modules at Standard conditions. Characteristics of the EOPLLY 25 Monocrystalline modules at STC (Cell temperature 25 ºC, Irradiance W/m 2, AM.5 G SPECTRUM) Open circuit voltage V oc V Short circuit current I sc 5.84 A Rated maximum power P MPP 95 W Voltage at maximum power V MPP V Current at maximum power I MPP A Module efficiency ɳ 5.27 % 6

31 In the table 3, the temperature coefficients are presented for the following conditions: Irradiance W/m 2 and AM.5 G SPECTRUM. Table 3 Temperature coefficients for the EOPLLY 25 Monocrystalline silicon modules. Temperature Coefficient of I SC α Temperature Coefficient of V OC β Temperature Coefficient of P MPP γ.6 %/ C -.39 %/ C -.44 %/ C In the figure 6, the IV and VP curves for these modules at different irradiances are shown. Figure 6 IV and VP curves for the Monocrystalline Silicon modules Thin Film CIGS modules The thin film modules are made of CIGS [Cu(In, Ga) Se2], and their model is Q.SMART UF L 85W. The area of each module is 9 mm 789,5 mm. The main electrical characteristics of these modules are shown in the table 4: Table 4 Electrical characteristics of the Thin Film CIGS 85W modules at Standard conditions. Characteristics of the CIGS Thin Film modules at STC (Cell temperature 25 ºC, Irradiance W/m 2, AM.5 G SPECTRUM) Open circuit voltage V oc 89.8 V Short circuit current I sc.62 A Rated maximum power P MPP 85 W Voltage at maximum power V MPP 66.3 V Current at maximum power I MPP.33 A Module efficiency ɳ 9 % The table 5 includes the temperature coefficients at Irradiance W/m 2 and AM.5 G SPECTRUM. Table 5 Temperature coefficients of the Thin Film CIGS 85W modules. Temperature Coefficient of I SC α Temperature Coefficient of V OC β Temperature Coefficient of P MPP γ. ±.4 %/ C.3 ±.4 %/ C.38 ±.4 %/ C 7

32 The IV curves of the thin film modules at different temperatures and irradiances are represented in the figure 7. Figure 7 IV curves for the Thin Film modules at different temperatures and irradiances Inverter Each of the systems of the Hall-45 has a string inverter to convert the DC to AC. The model of the inverters is SUNNY BOY 2. Figure 8 SUNNY BOY 2 inverter. [27] The table 6 gathers the main properties of this inverter. Table 6 Characteristics of the SUNNY BOY 2 inverter. Maximum DC voltage MPP voltage range AC nominal power 4 V V 32 V 2 W 8

33 3..4. TIGO optimizer The model of the TIGO optimizers used in the corresponding system is a Dual Maximizer MM- 2ES 5. The table 7 shows the electrical characteristics of this type of optimizer: Table 7 Characteristics of the TIGO optimizer. Nominal Input Power 375W Absolute Maximum Input Voltage V OC 52Vd MPPT Voltage Range V MP 6-48Vdc Maximum Input Current I SC A Maximum Input Current I MP 9.5A Maximum Output Power 75W Maximum Output Current 9.5A Operating Output Voltage -4V 3.2. Measurements To carry out the project, real measurements of the systems have been collected. Below, the devices used are described. The calibration of the new logger was successfully concluded on April 25 th, so from then on, the data has been collected daily Power measurements: new egauge logger The logger used to measure the power generated by each system in the Laboratory is the egauge EG3 logger, acquired by the University this year. Figure 9 egauge EG3 logger. [28] This device measures the electric AC power on up to 2 circuits. In this case, it has been installed so that it collects the generation data of the four studied systems: Si modules with diodes, Si modules with Tigo optimizer, Si modules with module micro-inverter and the thin film system. The resolution for this device is of seconds, allowing to collect data from every second or longer periods of time (minutes, hours, daily values). The data from the logger can be directly obtained from an internet connection. 9

34 Other data acquisition The laboratory had already installed another older logger with lower resolution (minute resolution). This logger also records the power output (in kwh) for the four systems installed in the Laboratory. Moreover, there are two pyranometers installed in the Laboratory roof, figure 2. One of them is positioned horizontally, so that it measures the total horizontal irradiation. The other pyranometer is placed with the same inclination as the modules (45 ), so that it measures the total irradiation in this tilted plane. Figure 2 Pyranometers installed in the Laboratory. The values of the irradiation are recorded by the logger with minutes resolution. In addition to the measures of the total irradiation in both planes, a third value is given, corresponding to the diffuse irradiation in the horizontal plane. Finally, the ambient temperature is also recorded at the Laboratory every minutes. 2

35 3.3. PROCEDURE AND OPERATION Calibration of the egauge logger First of all, to ensure that the installation of the new logger was correct, this has been calibrated. For this purpose, the data from both loggers for a whole day has been compared. More precisely, from 28 th April General procedure In the procedure two different parts can be distinguished: building up a theoretical model and treating the real data from the logger. For the theoretical model, some parameters need to be calculated. First, the angle of incidence θ is obtained by applying the following equation: Cos(θ)=cos(δ)sin(ω)sin(β)sin(γ)+cos(δ)cos(ω)sin(λ)sin(β)cos(γ)- sin(δ)cos(λ)sin(β)cos(γ)+cos(δ)cos(ω)cos(λ)cos(β)+sin(δ)sin(λ)cos(β) (4) For this calculation, a prepared Excel sheet has been used, where the following input data needs to be introduced: Table 8 Data from the installation for the calculation of the angle of incidence. Latitude (Gävle) 6.7 Day number (from the start of the year) e.g. st May, n=2 Surface tilt (from horizontal plane) 45 Surface azimuth (S= E=-9 W=9) -2 Standard meridian -5 Local meridian -7. As a result, the angle of incidence during a whole day is given, with a value provided every ten minutes. Once the angle of incidence is known, the correction factor of the direct radiation can be calculated by the equation: Where b o=, (approximation for the system). K b(θ)= -b ( ) (5) cos (θ) Moreover, the data collected by the pyranometer is required in order to know the irradiation. The pyranometer takes three different measurements: the total irradiation in the horizontal plane, the total irradiation in the modules plane (tilted 45 ) and the horizontal diffuse irradiation. The required values are the direct irradiation in the modules plane I b 45 and the diffuse irradiation in this plane I d 45. The last one is obtained from the diffuse irradiation in the horizontal plane I d. Where β=45 (the tilt of the modules) 45 I d = ( +cos (β) ) I d (6) 2 2

36 From this value and the total irradiation in the tilted plane, the direct irradiation in the modules plane is calculated: I b 45 =I 45 - I d 45 (7) Once the measures from the pyranometers were taken, it was appreciated lower values than the expected ones. Because of that, the pyranometers were calibrated again and to correct the data already acquired, this was multiplied by a correction factor of.. Another data that is necessary for the calculations is the ambient temperature. The temperature of the cell is approximated using this value by the expression: T cell=t amb+h G (8) where G is the total irradiance reaching the modules (I 45 ) and h=2 C m 2 /Kw Once all these values are obtained, the theoretical power can be calculated by the following expression: P(T, θ) = P(25,) [K b(θ) I b + K d I d ] [+(T cell 25) α] (9) Where P(25,) is the power of each system at standard conditions. These values are shown in Table. K b(θ), I b 45, I d 45 and T cell are obtained as it has been explained above. K d is the correction factor for the diffuse irradiation. This has in average a relatively high angle of incidence, so the following value is assumed: K d=.9 α is the temperature coefficient for the power of the module and it takes the value α -.4%/ C for Silicon. As for the irradiations and temperatures the data is recorded every minutes, values for the theoretical model will be obtained in the same interval. Regarding the real values, measurements are taken with both loggers. The data from the new logger will be collected also in mean values every ten minutes so that the data from both loggers can be compared directly. The egauge logger provides the generated power in kw. The old logger gives this production in kwh, and as the data is from every minutes, the power in kw can be obtained just my multiplying by 6. Moreover, when starting the analysis, it is also important to set the time period for the comparison. For that, the irradiance and power generated will be plotted for days with different weather conditions and the possible shadow problems will be studied. To compare the theoretical model and the real values, the calculations have been done for 8 different days. From all these values, four days will be taken as a reference to perform the study, including sunny days and cloudy days to see how the model behaves in different weather conditions. 22