A Revision of IEC nd Edition Data Correction Procedures 1 and 2: PV Module Performance at Murdoch University

Transcription

1 School of Engineering and Information Technology ENG470 Engineering Honours Thesis A Revision of IEC nd Edition Data Correction Procedures 1 and 2: PV Module Performance at Murdoch University Tim Blagojevic 2016 A thesis submitted to the School of Engineering and Energy, Murdoch University in patrial fulfilment of the requirements for the degree of Bachelor of Engineering Unit Coordinator: Dr Gareth Lee Supervisor: Dr David Parlevliet 1

2 Author s Declaration Except where I have indicated, the work I am submitting in this report is my own and has not been submitted for assessment in another course. Signed: Date: ii

3 ENG460 Engineering Thesis Academic Supervisor endorsement pro forma I am satisfied with the progress of this thesis project and that the attached report is an accurate reflection of the work undertaken. Signed: Date: iii

4 Abstract The focus of this project is to review and effectively assess the first two photovoltaic module electrical performance data correction procedures contained in the international engineering standard IEC 60891: Photovoltaic Devices- Procedures for temperature and irradiance corrections to measured I-V characteristics. The formulated workings of the project were used to assess the effectiveness of the correction methods in translating electrical performance data for determining the degradation or performance of photovoltaic modules. A preliminary literature review of concepts involved in the implementation of project procedures was conducted, so that appropriate experimental testing conditions could be formulated. This project covers information regarding factors that may affect photovoltaic module performance variation and degradation. Over a period of months in autumn/winter, outdoor field electrical performance data for different PV module technologies at the Murdoch University location was recorded and processed. The data collected was obtained under varying atmospheric conditions, with the tilts and orientations of the modules altered to change the total amount and nature of solar irradiation reaching the modules. The algebraic equations of the first and second standard correction procedures utilised parameters with values that could be measured directly from the outdoor testing of modules, or deduced from electrical performance data obtained from testing modules indoors at known values of irradiance, temperature and atmospheric spectra. Indoor performance data simulated with solar irradiance levels and cell temperatures recognised as those matching international standard test conditions, was obtained for use in effectively implementing the correction procedures. The data was also independently analysed and compared. Outdoor module test performance data was corrected with both correction procedures and collated for analysis. The results highlighted the effects of and correlations between factors that influence module I-V curve dynamics. When implemented for data translation, correction procedure one was found to produce a range of maximum power mismatch accuracy levels from 0.09 to 22.97% with an average accuracy mismatch level of 9.54%. Correction procedure two was found to produce a range of accuracy maximum power mismatch levels of 0.19 to 28.64%, with an average accuracy mismatch level of 8.58% An assessment of the correction procedures showed that they could be effectively used to gauge module degradation or for comparison of module performance against factory specifications. Both methods showed similar variations in accuracy, with correction procedure 2 being better suited to situations where the irradiance level difference between two data sets is more than 20%. Correction procedure 2 has more working parameters and takes more time to establish for correct implementation. iv

5 Acknowledgments I would like to wholeheartedly thank my supervisor Dr David Parlevliet for his consistent and effective guidance, direction and patience while assisting me in the undertaking of this thesis project. I would like to express my pleasure in dealing with the facilities and staff at Murdoch University. Without the user-friendly facilities and great efforts of the academic staff and technical assistants, this project would not have been possible. Finally, I would like to thank my family and friends for their support and help that they have contributed over the duration of this project and my university studies. v

6 Glossary Abbreviation Voc Isc FF AM Poly-Si Mono-si Amorphous-si STC PV I-V MPP Pmax Definition Open Circuit Voltage Short Circuit Current Fill Factor Air Mass Coefficient Polycrystalline Silicon Monocrystalline Silicon Amorphous Silicon Standard Test Conditions Photovoltaic Current-Voltage Maximum Power Point Maximum Power Voltage at the maximum power point Current at the maximum power point STC W Δ Standard Test Conditions Watt as a unit of electrical power measurement The symbol delta, representing a change in a parameter vi

7 Contents Author s Declaration... ii ENG460 Engineering Thesis... iii Abstract... iv Acknowledgments... v Glossary... vi Contents... vii List of Figures... x List of Tables... xii Chapter 1: Introduction Background Aims and Objectives... 3 Chapter 2: Literary Review Solar Radiation and the Solar Spectrum PV Module Performance, Measurement and IV Curves Standard Test Conditions PV Performance Data Mapping Methodologies IEC Standards and IEC Edition PV Module Degradation Factors Affecting IV Curves and Performance Solar Irradiance Module Temperature Module Soiling Module Shunt Resistance Module Shading Module Cell Cracking Environmental Reflection Angle of Incidence / Tilt Orientation PV Module Technologies Mono-Crystalline Silicon Modules Poly-Crystalline Silicon Modules Amorphous Silicon Modules vii

8 2.8 Module Spectral Response Chapter 3: Method PV Module Selection Outdoor Testing PV Module Electrical Performance Measurements Irradiance Data Measurements PV Module Temperature Measurements PV Module Tilt and Orientation Measurements Solar Spectrum Measurements Indoor Testing Procedure Correction Factors Experimental Device Limitations Chapter 4: Results Indoor Testing at STC Correction Procedure 1 and 2 Parameters Outdoor Testing Correction Procedure Correction for Outdoor Tested Modules to Non-STC Correction for Outdoor Tested Modules to STC Correction Procedure Correction for Outdoor Tested Modules to Non-STC Correction for Outdoor Tested Modules to STC Correction Procedure 1: ΔPmax and Module Orientation Correction Procedure 2: ΔPmax and Module Orientation Averaged ΔPmax Variation for Influencing Parameters ΔPmax vs Irradiance and Temperature Spectral Distribution and AM Chapter 5: Analysis and Discussion of Correction Procedure Performance Orientation, Tilt and Correction Procedure Accuracy Irradiance and Correction Procedure Accuracy Temperature and Correction Procedure Accuracy Module Technology and Correction Procedure Accuracy Correction Procedure Comparison Measurement Device Uncertainty viii

9 Chapter 6: Future Works Chapter 7: Conclusion References Appendices Appendix A: Data taking Procedures for the PROVA 210 Solar Curve Tracer Appendix B: Graphs for Procedure 1 and 2 Parameters ix

10 List of Figures Figure 1: Cumulative Household and Commercial Solar PV Installation for 2007 to 2014 in Australia [3]... 1 Figure 2: Solar Zenith Angle, AM1.5 and AM Figure 3: Standard Solar Spectra with AM0, AM1.5 Direct and AM1.5 Global... 6 Figure 4: I-V Curve Illustrating Voc, Isc, MPP, Vmp, Imp and Determinants of Fill Factor... 8 Figure 5: I-V Curves for a Change in Global Irradiance Levels Figure 6: I-V Curves for a Change in Module/cell Temperature Figure 7: Azimuth Angle as Seen in the Southern Hemisphere Figure 8: Spectral Response Plots of Different Silicon Solar Cell Materials Figure 9: Amorphous Silicon PV Module Used for Testing Figure 10: Amorphous Silicon PV Module ID: PN Figure 11: Poly-Crystalline Silicon PV Module Used for Testing Figure 12: Poly-Crystalline Silicon PV Module Used for Testing Figure 13: Mono-Crystalline Silicon PV Module Used for Testing Figure 14: Solar E Mono-Crystalline Silicon PV Module Model: SE-150M Figure 15: PROVA 210 Make and Model Solar Analyser Figure 16: PROVA 210 Make and Model Solar Analyse 4-Wire Connections Figure 17: Kipp and Zonen Irradiance Meter Measuring Irradiance on the Plane of the module Figure 18: Kipp and Zonen Irradiance Meter Figure 19: PROTEK 506 Multimeter Temperature Reader Figure 20: Data Recording Location Monument Facing True North Figure 21: StellarNet Spectrometer Sensor used for Solar Spectrum Data Recording Figure 22: StellarNet Spectrometer used for Solar Spectrum Data Recording Figure 23: The Spire 5600SLP Solar Simulator Located at Murdoch University Figure 24: The Spire 5600SLP Solar Simulator Control Monitor Figure 25: Indoor Measured Current vs Voltage Output for Mono-Crystalline Silicon at STC Figure 26: Indoor Measured Current vs Voltage Output for Poly-Crystalline Silicon at STC Figure 27: Indoor Measured Current vs Voltage Output for Amorphous Silicon at STC Figure 28: Corrected Difference in Pmax for -75 Degree Module Orientation Figure 29: Corrected Difference in Pmax for Degree Module Orientation Figure 30: Corrected Difference in Pmax for 0 Degree Module Orientation Figure 31: Corrected Difference in Pmax for 37.5 Degree Module Orientation Figure 32: Corrected Difference in Pmax for 75 Degree Module Orientation Figure 33: Corrected Difference in Pmax for -75 Degree Module Orientation Figure 34: Corrected Difference in Pmax for Degree Module Orientation Figure 35: Corrected Difference in Pmax for 0 Degree Module Orientation Figure 36: Corrected Difference in Pmax for 37.5 Degree Module Orientation Figure 37: Corrected Difference in Pmax for 75 Degree Module Orientation Figure 38: Average Percentage Difference in Pmax after Correction vs Module Technology and Correction Procedure Figure 39: Average Percentage Difference in Pmax after Correction vs Module Orientation and Correction Procedure x

11 Figure 40: Average Percentage Difference in Pmax after Correction vs Module Tilt and Correction Procedure Figure 41: Correction Procedure 1- Percentage Difference of Pmax (%) vs Irradiance Difference from Reference Module (W/m^2) Figure 42: Correction Procedure 2- Percentage Difference of Pmax (%) vs Irradiance Difference from Reference Module (W/m^2) Figure 43: Correction Procedure 1- Percentage Difference of Pmax (%) vs Temperature Difference from Reference Module (W/m^2) Figure 44: Correction Procedure 2- Percentage Difference of Pmax (%) vs Temperature Difference from Reference Module (W/m^2) Figure 45: Outdoor Tested Spectral Irradiance Distribution and AM xi

12 List of Tables Table 1: Parameters and Values for Standard Test Conditions (STC)... 9 Table 2: Uncertainties of Measurement Parameters by Data Recording Apparatus Table 3: Critical Electrical Parameters at STC for Different Module Technologies Table 4: Correction Procedure 1 and 2 Parameters for Different Module Technologies Table 5: Mono-Si: Outdoor Testing Electrical Performance Data Table 6: Poly-Si: Outdoor Testing Electrical Performance Data Table 7: Amorphous-Si: Outdoor Testing Electrical Performance Data Table 8: Mono-Si: Difference(%)in Pmax(W) After Correction to 1007 Using Method 1 39 Table 9: Poly-Si: Difference(%)in Pmax(W) After Correction to 968 Using Method Table 10: Amorphous-Si: Difference(%)in Pmax(W) After Correction to 1001 Using Method Table 11: Mono-Si: Difference(%) in Pmax(W) After Correction to STC Using Method Table 12: Poly-Si: Difference(%) in Pmax(W) After Correction to STC Using Method Table 13: Amorphous-Si: Difference(%) in Pmax(W) After Correction to STC Using Method Table 14: Mono-Si: Difference(%)in Pmax(W) After Correction to 1007 Using Method Table 15: Amorphous-Si: Difference(%)in Pmax(W) After Correction to 1001 Using Method Table 16: Poly-Si: Difference(%)in Pmax(W) After Correction to 968 Using Method Table 17: Mono-Si: Difference(%) in Pmax(W) After Correction to STC Using Method Table 18: Poly-Si: Difference(%) in Pmax(W) After Correction to STC Using Method Table 19: Amorphous-Si: Difference(%) in Pmax(W) After Correction to STC Using Method xii

13 Solar PV Installed Capacity (MW) Chapter 1: Introduction 1.1 Background In recent times, there has been an increased focus on progressing energy generation away from systems that cause damage to the environment and towards more cost effective and sustainable sources, that cause less damage to the environment. Energy generation systems that use renewable energy generation technology represent a way to generate energy with low emissions. The increased generation of energy from renewable sources means that there is less reliance on fossil fuels. Current substitutes for fossil fuel generation include systems that utilise wind, biomass, tidal/ocean, geothermal and solar energy mechanisms and unlike fossil fuels, the energy sources are essentially unlimited [1]. Photovoltaic, PV or solar panels use semi-conductors to generate electrical energy from photons, sourced from the sun s radiation [2]. This energy can be utilised as electricity that can be injected into established power electricity grids or used independently as a power source at the site of generation. PV panels are used increasingly often as a source of electricity for family homes [3]. Recent data analysis illustrates that renewable energy technology provided % of the total electricity generated in Australia in the year 2014, with 15.47% of this electricity coming from solar generation [3]. Solar energy generation continues to be an important facet of total electricity production in Australia. With continuous improvements in energy storage and panel technologies [3], there is further room for growth and expansion in the solar sector Cumulative Household and Commercial Solar PV Installation Capacity for 2007 to 2014 in Australia (MW) Figure 1: Cumulative Household and Commercial Solar PV Installation for 2007 to 2014 in Australia [3] Year 1

14 Determining the output power production of a PV panel over a period of time is a very important characteristic in its working dynamics. The output power slowly reduces in magnitude from its original amount due to a number of environmental factors, such as the sun, high temperature and moisture. This decline in power is known as one of the factors in the degradation rate of a panel. PV degradation rate is of a particular importance to any stakeholders which make use of the technology, including utility companies, investors, and researchers. Any changes in power output to grid connected PV systems can cause possible interruptions to power quality and can cause other problems in an electricity grid [4]. It is also important to understand PV degradation from a financial viewpoint, because higher degradation rates of energy system modules means earlier replacement costs meaning lower future cash flows, due to the loss of output power. Other financial concerns involve end user situations where panel array space is an energy system design constraint. Having higher panel efficiency from the panels used in a system array on a roof top with limited space for example, means more power per given area. This can mean that the end user has more electricity from their limited roof space [5]. Comparing the degradation rates and performances of PV panels can lead to better financial and technical choices for energy systems [6,7]. Measuring the electrical performance of a particular photovoltaic or PV panel enables an analysis of any performance degradation to be conducted. The current-voltage characteristic or I-V curve is a graphical representation of the relationship between the current and the voltage of a panel under operation [8]. For a correct assessment of the operating characteristics of a specific panel, the performance data must be standardised to allow for a direct comparison with data from another panel. The data must be translated or mapped to produce results that would have been obtained had the panel been operating under agreed standardised testing conditions [9] when tested. International and other specific engineering standards provide instruction and guidelines for the use of algebraic translational methods. The mapping methods contain parameters that account for some environmental and performance factors that can affect module output power and therefore the measured data and I-V curves. Some factors that affect the output of a panel are not accounted for by data translation methods alone. PV Modules are commonly made of different derivatives of silicon and are also made of other materials. Radiation from the sun comes in varying wavelengths of light and therefore has a certain spectral distribution. Particular module material types respond to different wavelengths in the sunlight spectrum at varying rates when compared to other module types. Power output levels can therefore vary with spectral differences. The varying module spectral response must be taken into account when mapping panel data, as the outdoor sunlight spectral characteristics may not match the spectrum of the light under standard test conditions [10]. This project reviews and illustrates algebraic translational data mapping methods, with a specific focus on methods that are outlined in the standard IEC edition 2. The mapping methods were applied to electrical performance data collected in the field at Murdoch University, for three different solar module technologies. Variations in factors that affect electrical performance were exploited to allow for a wider data set for analysis. Alternative performance data was obtained from 2

15 testing involving a sun simulator, or an indoor energy source that mimics conditions experienced when panels receive sunlight under ideal standard test conditions [10]. The mapped data and ideal performance data collected indoors was then compared to examine the effectiveness of the mapping methods implemented. Any discrepancies in the data were investigated. 1.2 Aims and Objectives This thesis aims to examine methods that assess the degradation and performance of different PV modules individually and comparatively. Performance data at the Murdoch University location was to be comparatively assessed using the first two data translation methods outlined in the international engineering standard: IEC nd Edition ( ). For the proper assessment of the data using the translation methods contained in this standard, a number of preliminary steps were necessary for completion in the achievement of these objectives. An initial literary review was necessary for development as to understand all physical concepts and phenomena involved in the processes of PV modules producing electrical energy and the recording and mapping of performance data. One necessary objective was to effectively test and record valid electrical performance of different panel technologies outdoors. The practical methodology of this project was implemented and carried out to achieve this. Performance data was selectively set to be affected by changing irradiance and temperature. The orientation of the incident solar radiation to testing PV modules and the tilt of the modules were to be used as factors to selectively vary irradiance levels. The specific spectral wavelength distribution of the sunlight was to be noted with each variation in data recording conditions. There was an aim to produce valid electrical performance data under ideal standard test conditions, which would be obtained indoors using solar simulation equipment. Further practical methodology in this project was formulated to achieve this. Following on from the outdoor and indoor performance data collection, all valid data was mapped using the three different methods outlined in the standard. Once the data mapping was complete, the different data translation methods could be assessed. The validity, effectiveness and selectivity of the different methods could be determined. 3

16 Chapter 2: Literary Review A broad range of literary research was necessary to effectively achieve the outcome aims and objectives set out in this project. An understanding and establishment of concepts involved within the physical phenomena, instruments, international engineering standards and practical methodology involved in this project was necessary to examine the validity of the data translation methods implemented. 2.1 Solar Radiation and the Solar Spectrum This research project involved the processes of capturing solar radiation and the analysis of electrical power obtained from this captured energy. It is necessary to establish how these processes occur and note factors that may affect the nature of the solar radiation. Energy travels from the sun in an electromagnetic form, which reaches the earth s atmosphere as sunlight. This light that reaches earth contains infrared, ultraviolet and visible light as part of its spectrum. The particles representative of the light are termed photons. The photonic energy can be represented by a function of its wavelength as represented by the following equation: Where represents the photon energy, h is Planck s constant and c is the speed of light [11]. The position of the sun will change compared to the surface of the earth and consequently, the atmospheric distance that the photons travel through will change also. Air Mass (AM) is descriptive of the measured distance or thickness of the atmosphere that solar flux must travel through, when the sun is above the horizon, to reach the earth s surface. The air mass will represent the shortest path length that is possible for the sunlight to travel through, when finally reaching ground level [12]. The actual spectrum of sun that reaches the earth s surface is called the global radiation. Global radiation has multiple components. Radiation that comes directly from the sun unmodified by atmospheric processes is termed direct radiation. Solar radiation that reaches the earth s surface after atmospheric interaction is described as diffuse radiation [13]. The global irradiation level is equal to the sum of the diffuse and direct irradiation components multiplied by the cosine of the solar zenith angle as seen by the formula: 4

17 Example solar zenith angles for AM 2.0 and AM 1.5 are illustrated in figure 2 below: Figure 2: Solar Zenith Angle, AM1.5 and AM2.0 Upon entering the earth s atmosphere, the nature of sunlight will change. The main influences of the change to the light coming into the atmosphere are due to particular chemical components like ozone, water vapour, oxygen and carbon dioxide. Atmospheric composition exhibits dissimilarity, changing with different geographical locations on the planet. The solar radiation intensity differs according to wavelength and the relationship between measured solar intensities and solar radiation wavelengths is called the solar spectral distribution [13]. When environmental factors change the physical nature of sunlight, PV system electrical output performance can fluctuate. Some major atmospheric effects can be seen to cause a change in the total amount of energy and also the particular colour component wavelength distribution of the solar radiation that finally reaches the land. These changes can result in a reduction of power levels from a solar panel. The specific effects that cause changes to solar radiation reaching the land include atmospheric reflections, absorption, and the atmospheric scattering of light. Locality dependent effects such as water vapour, clouds and pollution can also affect the power, spectrum and direction of the light incident to a PV panel [12]. For sunlight, a longer path length through the atmosphere will mean that the aforementioned atmospheric affects can most often have a more pronounced influence on the spectrum and intensity of the sunlight. The spectral distribution of the radiation can change due to scattering and absorption of some wavelengths. Normally, the proportional amounts of radiative energy compared to the total solar energy available show that roughly around 6% of terrestrial solar energy contains wavelengths in the UV region of the spectrum, around 50% in the visible light range and around 44% of wavelengths in the infrared light portion of the spectrum [13]. 5

18 Spectral Irradiance (Wm^(-2)nm^(-1) Filtering of wavelengths less than around 300nm occur as a result of light absorption by gas molecules like ozone, nitrogen and oxygen. Specifically, the infrared portion of the spectrum has a loss of energy and dips because of the effects of atmospheric water and carbon dioxide [13]. For solar radiation outside the atmosphere, the spectral distribution is sometimes known as the extra-terrestrial or air mass zero (AM 0) spectrum. The graphical apex of this spectrum coincides with a wavelength value of around 500 nm [14].The measured amount of irradiance is called known as the solar constant. If a surface was set to face the sun outside the atmosphere at a right angle to it, it would receive energy at approximately 1360, kept to an average [15]. This value of extraterrestrial total irradiance represents the total integrated energy of the whole spectrum. The spectrum for direct radiation alone, when the sun directly overhead and has a zenith angle of zero degrees can be known as Air Mass Direct, or AM1D. When the sun is in the same position and global radiation is accounted for, the spectrum is termed AM1G, which is abbreviated from Air Mass Global. A standard spectrum that is recognized globally is Air mass 1.5 or AM1.5, corresponding to a solar zenith angle of approximately 48.2 degrees. For PV data analysis AM1.5 Global is usually used as a spectral reference. An example plot of the spectrum can be seen in Figure 3 below: 2.50 AM AM1.5 Global 1.50 AM1.5 Direct Wavelength (nm) Figure 3: Standard Solar Spectra with AM0, AM1.5 Direct and AM1.5 Global 6

19 2.2 PV Module Performance, Measurement and IV Curves As the analysis of data for this project dealt with the electrical performance of PV modules, it is of importance to establish how modules capture and harness electrical energy and how this energy is measured, presented and interpreted. In a PV cell, the production of electricity relies on a number of crucial occurrences and physical characteristics. The unique semi-conductive characteristics of silicon and other materials used in solar cells allow electrical conductivity, while some insulating properties are still present. For Silicon based cells, the internal structure of the semiconductor is purposely interrupted by the addition of other elements, creating two distinctive zones. The n-type zone is negatively charged with excess electrons, while the p-type zone is positively charged, with holes [16]. Sunlight travelling from the sun as photons can either reflect off the surface of silicon material, or cause silicon to emit electrons in a process described as the photoelectric effect. This will occur only if the energy of the photon is higher than the band-gap energy of the semi-conductor. If the energy is less than the band-gap energy, the photon will not be absorbed and any excess energy over the band gap amount will be dissipated as heat [5]. Excess electrons from the n-type zone will cross over to the p-type zone to fill the electron holes. Electric current will flow in the depletion region, where the free electrons have merged with the holes. When conductive material is connected to the n-type cathode and the p-type anode, electrons can flow. The electrons balance total system electrical neutrality by recombining with the p-type zone holes near the back electrical contact in the solar cell [17]. Different atomic materials and molecular configurations of silicon allow for different solar cell technologies, which can have different electrical performance and efficiency. A single PV cell will typically produce between 1 and watts of power, at a voltage of between 0.5 to 0.6 volts, under standard test conditions. Multiple single solar cells are connected in series so that the overall voltage and power levels of the series circuit are much higher in value. Multiple cells encapsulated and manufactured to perform in a series circuit are called solar modules. Often, the voltage potential is manipulated to match that of a 12V battery for practical purposes. Very often, solar modules will contain 36 cells in series to account for typical load operating voltage fluctuations and other excess energy requirements [18]. The power output and efficiency exists as the main contributing factor in discerning between different PV modules. Generated electrical power and electrical performance of modules can be recorded and is expressed graphically in the form of an I-V curve or PV characteristic curve. Three critical points of interest in an analysis of these plotted curves are the short circuit current ( ), the open circuit voltage ( ) and the maximum power point (MPP) [19]. The short circuit current can be described physically as the maximum current level that will flow through the output terminals of a PV module, meaning that the resistance at the output terminals is very small in magnitude. This can usually be measured by having a conductive wire or cable connected to the output terminals, which has a very low resistance. This connection is often the wiring of an appropriate measuring device. 7

20 The open circuit voltage is physically described as the maximum measured voltage potential difference across the output terminals of a PV module, with the negative terminal being grounded. This is usually measured by a connection being made across the output terminals with an appropriate measuring device [19]. Another variable fill factor is also a key indicator for module performance. The fill factor can act as a gauge for the squareness or how close the I-V curve actually resembles that of a perfect situation where it would be a perfect rectangle. Graphically illustrated, it can be seen that the FF represents the proportionality between the output power level that is determined from the product of and compared to the power calculated, using actual observed MPP voltage and current levels and [20]. As an equation, this can be stated as follows: Figure 4: I-V Curve Illustrating Voc, Isc, MPP, Vmp, Imp and Determinants of Fill Factor 2.3 Standard Test Conditions The expansion of the photovoltaic industry and manufacturing levels of PV modules in the 1980s warranted a need for established standards for performance referencing. Standard reporting conditions for PV modules were developed as a benchmark by the ASTM or the American Society of 8

21 Testing Materials committee [21]. PV modules respond differently to different atmospheric spectra. Taking into account that a majority of the world s major population locations with solar PV installations exist at equatorial locations, the standard spectra AM1.5 was developed as illustrated in Figure 4. This serves as the common reference for international standard, which was developed directly from the ASTM standards. Standard references for module cell temperature are required as module temperature can vary greatly. The STC reference cell temperature is 25 degrees Celsius. The standard reference level of irradiance is representing the normalised surface irradiance at sea level on a clear day [22]. A summary of STC can be seen in Table 1 below: Table 1: Parameters and Values for Standard Test Conditions (STC) Testing Conditions Parameter Value Irradiance ( 1000 Module Cell Temperature (degrees Celsius) 25 Air Mass Coefficient (AM) AM PV Performance Data Mapping Methodologies In order to compare different PV Modules for performance and degradation analysis, different mapping methodologies are used. These methods translate I-V curve data to desired performance conditions. For mapping to STC, different algebraic and numerical methods can be used. IEC 60891and ASTM E [23] are commonly used standards that contain guidelines for and methods of translation. Other examples methods of that are used with similar procedures to that of ASTM E , are the Blaesser method [24] and the Anderson method [25]. Numerical methods can also be used to translate data. An example of a numerical method is mentioned by Hermann and Weisner [26], which relies on a model of the electrical circuit of solar cells and requires certain cell parameters such as the series resistance, shunt resistance, diode ideality factors, and generated photo-current and dark saturation currents. This project focuses specifically on the implementation and assessment of the methods used in the second edition of IEC IEC Standards and IEC Edition 2 The International Electrotechnical Commission or IEC are a non-government, non-profit organisation. The IEC is the world s leading producer and publisher of international standards for engineering in the electrical and electronic fields. The role of the IEC is to provide documental publications that contain content which focuses on instruction and guidelines that may apply when providing certain services, or dealing with certain products and systems. There are numerous standards that apply to the application of PV technology [27]. The standard IEC provides procedures that can make adjustments to measured I-V curves that account for the irradiance and temperature differences from a desired norm, usually 9

22 standard test conditions. Applying methodology outlined in the standard, two I-V curves that were produced under different temperature and irradiance levels can be compared to each other at a new standard level for both formally different variables. The standard requires certain prerequisites when applying the methodologies. These pre-requisites actually involve other standard practises themselves. When measuring temperature of the PV device, IEC stipulates that the measurement sensor must have an accuracy of 1% [28].When taking measurements for global irradiance, the appropriate device must be calibrated in accordance with the requirements listed in IEC and be within ±2 degrees of the testing module [29]. When comparing one PV module to another, the reference module technology type must be the same as the comparative technology or it has to be spectrally matched according to IEC [30]. In addition, the reference module must adhere to the guidelines of IEC stating that the region must have appropriate linearity. For conversions to STC, the reference device measurements must not fluctuate by more than ±1 % and the global irradiance level must be at least [28]. When recording PV module output I-V curve data using an I-V curve tracer, certain operating requirements are to be adhered to. The current and voltage values that are measured from module operation are to be obtained by an instrument with a 4 wire connection to the output terminals of the module and with a ± 0.2% accuracy level for values of and. The standard IEC contains different algebraic translation methods. All methods are used to map an initial set of module output data to a set of data that would reflect different temperature and irradiance conditions. The methods can be used for any PV module technology type, but the output test data must show linear behaviour when factoring in changes to temperature and irradiance [31] Correction Procedure 1 The first method in IEC uses an empirical approach and is based on the work of Sandstrom [32]. Two temperature coefficients alpha (α) and beta (β) are used. Both coefficients illustrate how electrical output parameters behave with changing temperature, with α representing the behaviour of and β representing the behaviour of. Other parameters are necessary in calculations, such as the series resistance ( ) and the curve correction factor (κ). These parameters actually reflect any changes in the graphical form of an I-V curve when temperature or solar irradiance levels change [33]. Two equations are used in the first method: ( ) (1) (2) [33] In these equations, and represent a pair of points on a measured I-V curve. and are the new corrected current and voltage that is desired. and are the measured solar irradiance for the testing conditions and the desired solar irradiance for the new conversion, respectively. is the 10

23 measured temperature of the module that has data to be translated. is the temperature that is desired when performing data mapping. is the short circuit current of the testing module when there is an irradiance level corresponding with parameters and. The coefficients and β are implemented as the current and voltage temperature coefficients for the test specimen, which refer to the standard or target irradiance for correction and also within the temperature range of interest. is the internal series resistance of the testing sample and κ is the curve correction factor [23] Correction Procedure 2 For correction procedure 2, there are 2 equations to use in deducing the new corrected values for current and voltage: (3) ( ( )) ( - (4) [33] This particular method is developed from the one diode model circuit for PV modules. The translation equations that are used are semi-empirical in nature. The equations contain 5 different correction parameters that are able to be deduced by measurement of I-V curves at differing temperature and irradiance conditions. Instead of α and which are from method 1, an initial coefficient κ is employed for use to allow for a representation of changes to the fill factor and internal series resistance with temperature. is the open circuit voltage at test conditions. and in the equations are the relative temperature coefficients for current and voltage respectively, of the test module. The coefficients are measured at 1000 and are related to STC. The parameter a represents the irradiance correction factor for the open circuit voltage which is related to the diode thermal voltage of the p-n junction in the solar module cell material, and the number of solar module cells ( ) which are connected in series within the module in use. is the internal series resistance of the testing module, while κ is the temperature coefficient of the internal series resistance [33] Thermal Coefficients Thermal coefficients α,, and from correction procedure 1 and 2 can be determined using methods outlined in section 4 of IEC [33]. To determine the coefficients, a constant irradiance level is selected. I-V characteristics are then calculated at different operating temperatures. The selected performance parameter (, or ) is then plotted with module operating temperature. A least squares regression line is then calculated from the performance parameters temperature plot and added to the same plot, with the gradient of the regression line being the thermal coefficient. The standard recommends that the thermal coefficients be applied to irradiances which are within 30% of the irradiance level which they were determined at. 11

24 There are slight differences in the thermal coefficients between correction method 1 and 2, with the second procedure having coefficients that are normalised by the performance parameter (, or ) at STC, such that the parameters are dimensionless. There are options to determine the coefficients indoors with a solar simulator, or also outdoors using natural solar resources. If determining the temperature coefficients outdoors, the temperature range must be within at least ± 2 degrees. If determining the coefficient indoors, one can rely on an irradiance source being provided by a solar simulator and temperature of the testing module being controlled by temperature control apparatus, such as air-conditioners of close contact heat radiation devices. The indoor method provides less uncertainty in when finding coefficient values [33] Correction Factors Correction factors used in the procedures outlined in IEC must be determined experimentally, as they are not usually found in PV module specifications. The standard states that the processes involved in determining and from (1) and (2) are different but the processes for determining κ and κ are the same. The difference in the procedure for finding and is that the parameter a must be found for. Finding from correction procedure 1 relies on three different I-V curves being obtained from a particular module in question. The three curves must all be traced at the same temperature, but at different irradiance levels. The equations from correction procedure 1 take on a new form as = : + (7) (8) [33] Equations 7 and 8 show that to obtain, the temperature coefficients are not required. For correction procedure 2, as = the equations take on a new form: ) (9) ) - (10) [33] Similarly, with correction procedure 2, the temperature coefficient parameters are not required to obtain the relevant series resistance of the module. Applying these methods to obtain series resistances for both correction procedure 1 and 2 relies on the I-V curves that are of lower irradiance levels being translated to the level of the highest one. While doing this, one must start with in equation 8 being set to 0 and and a in equation 10 being set to 0. The next step is to increase in steps of 10mΩ and when the value of the two translated lower irradiance value I-V curves are within 0.5% of each other in value, the value for the series resistance in the equations at that time is the correct one. 12

25 For correction procedure 2, the irradiance correction factor for or a, must be found before finding the value of To achieve this objective, is kept to zero and a is increased from zero in steps of until the of the translated curves are within 0.5% of each other in value. The value of a at this point will be correct. Using this value of a, is increased in increments of 10mΩ until the values of from the two lower irradiance value translated I-V curves are within 0.5% of each other. When deducing the curve correction factors κ and κ, three or more different I-V curves are required. The different I-V curves must all have the same tested solar irradiance level but also must have different module operating temperatures. As the solar irradiance levels do not change for any particular data set, the equations for correction procedure 1 with parameters, can also be expressed alternately: (11) (12) For correction procedure 2, the original equations will become: (13) (14) Equations illustrate that to obtain the curve correction factors κ and κ, the temperature coefficients and series resistance for both correction procedure 1 and correction procedure 2 are all required. When deducing the correction factors, the selected I-V curve with the lowest temperature is selected as the reference curve with the two lower curves then translated to match the I-V curve with the lower temperature. For correction procedure 1, parameter is set initially to zero and for correction procedure 2, parameter κ is also set to zero. The values of each curve correction factor are then increased by increments of 1 and the desired value is reached when the two translated I-V curves values of are within ±0.5% of each other [33]. 2.5 PV Module Degradation As a PV module ages, the power performance can drop. Other environmental factors can induce immediate performance loss. From an end user point of view, it is desirable to prevent and avoid as much degradation of a module as possible, to get more useable power. 13

26 One study showed that over a ten year period of a PV module can lose 1-2% of its original factory specified output power performance capabilities due to degradation factors [34]. Another study found that degradation contributed to a performance loss of around 0.5% per year in a polycrystalline module over an eight year period [35]. There are numerous different causes of module degradation. They can be grouped together in different descriptive categories. Degradation due to cell failure can include problems with panel hot spots and cells cracking. Package material degradation can occur as a result of encapsulate material degradation, glass breakage or delamination. Module failure can be induced as a result of soiling and shading. Changes to the shunt and series resistances can all contribute to power degradation. Panel hot spots occur when a short circuit occurs in a series connection between cells and the cell overheats [36]. Cell cracks can occur when an external force or thermal stress is applied. The packaging material of a module always degrades over time but can hot spot heating of the module and water intrusion can greatly increase the rate at which this degradation occurs [24]. The glass encapsulation packaging on the module may break, leading to lowered performance, possibly due to leakage current. If the series resistance of a solar cell increases, the short circuit current of a solar module can decrease. If the shunt resistance decreases, the open circuit voltage of a module can reduce. The series resistance in a solar cell will normally increase as a result of water vapour inducing corrosion or delamination of contacts. The shunt resistance can decrease as a result of partial shading, thermal stress, hot spot occurrence or ohmic shorts [37]. Recombination in a solar cell occurs when an electron hole in the cell material disappears. Electrons can fall back into the valence band, recombining with the holes. This process will reduce voltage and current and lower power [8]. This process can often occur at the contacts of the module and on the surface. Different types of soiling on the outside surface of a PV module can lead to reduced electrical current. Dust, animal faecal matter, mud, frost, snow or soot can all accumulate on the surface of the panel and block or reflect solar radiation. Module shading can be caused by external environmental factors or also other localised factors affecting the amount of sunlight reaching the surface of the module. Obstacles such as rooftops, trees and walls can cause localised shading and horizon shading can be caused by very large objects such as hills at a distance. The amount of direct radiation reaching a solar module can be greatly affected by shading. 2.6 Factors Affecting IV Curves and Performance A number of factors affect the voltage and current and therefore power output levels of a PV module. Of these factors, some are a product of degradation. Irradiance, temperature, series 14

27 resistance, shunt resistance, cell cracking, soiling, reflections contributing additional diffuse irradiance and shading all have an effect on the shape of an I-V curve Solar Irradiance Increasing global irradiance has a positive effect on module power. As seen in Figure 5, increasing solar irradiance levels will in turn increase both the values of and. Many other environmental factors directly affect I-V curve shape by changing the amount of irradiance available to a solar module. Figure 5: I-V Curves for a Change in Global Irradiance Levels Module Temperature Temperature increases in a module can be caused by ambient environmental temperature changes, cloud patterns and wind speed. Temperature increases in a PV module have an effect in the graphical representation of the power output. Isc increases slightly, while Voc decreases with rising temperature. PV module performance is less sensitive to temperature than irradiance changes, but temperature changes are still significant [19]. 15

28 Figure 6: I-V Curves for a Change in Module/cell Temperature Module Soiling The soiling of a module can lower the levels of sunlight that penetrate the absorbing surface of a panel and consequently, each current reading for every different voltage level is reduced. The I-V curve has a similar shape, but the height is affected. This phenomenon can occur with uniform and non-uniform soiling [20] Module Shunt Resistance If the shunt resistance of a module changes, the slope of the module performance output I-V curve near the region can change. Any shunt resistance reductions may cause the slope to be steeper than normal and appear less flat. The changes to the resistance can be due to shunt paths existing in the PV cells Module Shading Shading of a module in any form will cause a reduction in module output current. If a particular cell is shaded, the other cells connected in series will have a reduction in the maximum current that they may otherwise have produced. Graphically, shading will be shown by notches in an I-V curve [20]. The graphical value of can be reduced. 16

29 2.6.6 Module Cell Cracking If a module cell has cracks in it, some physical parts of the cell may become electrically isolated. Cell cracks can have the same effect on module performance I-V curves as seen by shading Environmental Reflection Any reflecting solar radiation received by a module from close foreign objects can actually increase the power output. Any module performance I-V curve will be different graphically and behave similarly as if receiving a larger amount of irradiance [20]. The effect of reflection and increased irradiance is more pronounced in the early day and late afternoon Angle of Incidence / Tilt The angle of incidence a panel in relation to the sun overhead will have an effect on the total solar radiation that hits the surface of the module [38]. Other factors can come into play when examining how much irradiance will change if a panel is tilted, such as latitude, albedo and clearness index. Taking into account the sum of direct, diffuse and any reflected light, the yearly average optimal maximum solar radiation levels available to a module occur around an angle of incidence that coincides with the latitude angle of the module location [39]. The power output of a panel and therefore graphically, the of a module performance output I-V curve, can be affected by the geometrical positioning of the sun (involving the tilt angle of the panel) and also by any optical effects that depend on the module design. Optical effects on captured irradiance levels are caused by certain optical characteristics of module materials that are in the path between the solar radiation and the cells. An example of this is the reflection of radiation from glass front surfaces on flat plate modules. Reflecting factors become more significant when the tilt angle of a module exceeds approximately 50 degrees [40] Orientation PV Module output power production is close to being linearly proportional to the amount of solar radiation that reaches the surface of the module. The orientation of a solar panel can refer to the azimuth angle of the sun. When the azimuth angle is zero degrees, it is solar noon and the sun will be directly south in the northern hemisphere and directly north in the southern hemisphere. Most of 17

30 the global irradiance comes from the direct irradiance component and more of this direct irradiance can be captured from the sun by a PV module if it directly faces the sun. The more a module is orientated away from the solar noon azimuth angle, the less average solar irradiance it will receive over the course of a day. These principles are at the origin of the reasoning for the existence of solar trackers, which will physically change the orientation of solar panels throughout the day to maximise the amount of absorbed solar radiation [41]. Graphically, the of an I-V curve will be generally effected in the same way by changing module orientation as a change in tilt angle, as both factors result in a change in the total global irradiance received by a PV module. Figure 7: Azimuth Angle as Seen in the Southern Hemisphere 2.7 PV Module Technologies Silicon is the second most abundant element on the earth s surface. Silicon is refined from its oxidised form of silicon dioxide (Si ) through great heat and a combination with carbon. Ninety eight percent pure silicon is produced that is further purified through the use of trichlorosilane (SiH ) to produce silicon that is pure enough to use in producing solar cells and PV modules [42]. Numerous different cell technologies exist and are employed in the use of functional PV modules. In achieving the aims of this project, three different more common module technologies were used for analysis to give different physical material performance responses and data sets for contrast and comparison. The three technologies of focus were amorphous silicon, mono-crystalline and poly crystalline, with one module representing each technology used as a testing specimen. 18

31 2.7.1 Mono-Crystalline Silicon Modules Modules made from this material use a form of crystalline silicon that consists of a crystal lattice that is continuous throughout the solid material and doesn t have any grain boundaries. Very large ingots of silicon mono-crystals are grown and thinly cut into wafers that are ready for more processing to reach the final product stage. The modules are usually black in colour. Lab efficiencies for modules of this technology currently rank at 20 percent as of One square metre of monocrystalline cells will potentially generate around 190W of power. The modules are more efficient than polycrystalline or amorphous silicon modules and are used commonly when space saving is of a concern. [45] Poly-Crystalline Silicon Modules PV Modules made of this form of silicon are made of small crystals that are commonly known as crystallites, which can contain grain boundaries or 2D defects that can decrease the electrical and thermal conductivity of the solar cells made of this material. Large rods of the material are made into ingots and cut into wafers to make cells. Poly-crystalline silicon cells exhibit a metal flake effect causing the outward appearance to have random internal patterns. Modules using this technology are usually blue in colour, although some newer types are darker blue. A square metre of polycrystalline cells will generate around 180W of power. Modules have a lower temperature coefficient than mono-crystalline silicon modules and over a period of time can generate more useable electrical power than mono-crystalline modules of the same power rating [19] Amorphous Silicon Modules Amorphous silicon is known as a second generation cell technology. The material is an alloy of hydrogen and silicon. The formation of silicon atoms in amorphous silicon resembles a continuous random network. Amorphous silicon modules typically have performance efficiency percentage levels in the 6-8% range. The efficiency of this material drops when exposed to light. The efficiency levels also drop in winter but are better in summer due to annealing. Many modules now use a hydrogenated dilution to increase module operation quality. The modules are relatively low in cost to produce, being cheaper than mono or poly crystalline variants. PV applications of amorphous silicon cells are more typically used indoors [19]. Graphically, performance output I-V curves are affected by light induced module degradation. The fill factor is less and short circuit current changes, but the open circuit voltage remains relatively unchanged. [42]. 19

32 2.8 Module Spectral Response The ability to respond to sunlight is influenced by the band gap of a solar cell material, which is the amount of energy needed to release electrons into the conduction band from the valence band. A larger band gap corresponds with higher energy, so solar cell materials with a higher band gap respond to higher frequency and more energetic portions of the solar spectrum. More precisely, the spectral response limiters of a cell material are the band gap at long wavelengths and the material absorption at shorter wavelengths. For spectral response other influencing factors may include: the independent device design; the material system and the electrical contacts [42,43]. The band gap of amorphous silicon, which is around 1.7 ev, is higher than crystalline silicon (at around 1.1 ev). As a consequence of the difference in band gap levels, amorphous silicon responds to the visible part of the solar spectrum more than the higher power density areas of the spectrum like the infrared frequency area. In summer, the average light path is shorter than in the winter months, as the sun is more overhead. This means that the higher energy blue light portion becomes larger than the AM1.5 standard reference spectra. In winter, as the average path length for the sunlight is longer, the spectrum will contain more red or higher wavelength light. As a result of these occurrences, amorphous silicon modules will have less response in winter months because for this module technology the spectral response upper limits are at about nm in wavelength [44]. The spectral response region for poly-crystalline silicon and mono-crystalline silicon is different in the range of wavelengths when compared to the typical response for amorphous silicon. Both poly and mono-crystalline silicon are more responsive to the portion of sunlight that has a higher wavelength. When the air mass decreases, generally there is more blue light in the solar spectrum and this will mean that poly or mono-crystalline modules will be more sensitive to performance change [45] The typical differences in the spectral response between amorphous silicon, mono-crystalline and poly-crystalline silicon cell materials can be seen graphically in figure 8 below: 20

33 Figure 8: Spectral Response Plots of Different Silicon Solar Cell Materials 21

34 Chapter 3: Method 3.1 PV Module Selection Three different PV module technologies were selected for an electrical performance analysis from those available at Murdoch University. An amorphous silicon module was selected for testing and was identifiable by the number string of PN The module can be seen in figure 8 below: Figure 9: Amorphous Silicon PV Module Used for Testing 22

35 Figure 10: Amorphous Silicon PV Module ID: PN-7-02 A poly-crystalline silicon module was selected for testing and was identifiable by the make and model: Sharp PN The module can be seen in figure 10 below: Figure 11: Poly-Crystalline Silicon PV Module Used for Testing 23

36 Figure 12: Poly-Crystalline Silicon PV Module Used for Testing A mono-crystalline silicon module was selected for testing and was identifiable by the make, model and ID number: Solar E SE-150M. The module can be seen in figure 12 below: Figure 13: Mono-Crystalline Silicon PV Module Used for Testing 24

37 Figure 14: Solar E Mono-Crystalline Silicon PV Module Model: SE-150M This project relied on the sustained use of 3 different PV module technologies. Performance data was taken as a means for analysis and to test data mapping methodologies. The 3 modules used in the project had some noted signs of degradation, such as cell cracking and soiling. If any further panel degradation occurred, whether a change in performance was permanent or semi-permanent, the integrity of any recorded data could be compromised. A visual inspection of the modules was performed to check for any changes in appearance due to cracking or extra physical stress. Any controllable factors that could lead to degradation were avoided. Any shading obstacles were avoided when data capture was done. The modules were very seldom exposed to sunlight as they were kept indoors when not in use. The storage area was a dry, clean environment with normal ranges for room temperatures. The modules were also cleaned when any soiling was noted. No abnormal physical stress was applied to the module surfaces. As a result of these practices, the condition of the three modules from the first instance of analytical data capture until the last occurrence was kept as unchanged as possible, meaning that any panel degradation was very low or negligible and data integrity could be maintained. 25

38 3.2 Outdoor Testing As established, this project relied on the sustained use of 3 different PV module technologies for outdoor testing. Outdoor electrical performance data was taken as a means for analysis and to test data mapping methodologies. For a given module orientation and tilt, simultaneous measurements of module tilt irradiance and module temperature were made using PV modules. Solar spectral data was also taken at the time of the other measurements. All outdoor testing was conducted on the campus of Murdoch University, Perth, Western Australia. The testing location was at a latitude and longitude of degrees south and degrees east PV Module Electrical Performance Measurements Outdoor electrical performance data was recorded outdoors using a 4 wire connection PROVA 210 I-V curve tracer. I-V Curve data is displayed on the screen for immediate inspection and saved for further use. The device meets the standard requirements for data recording. Connections are made directly to the PV module output terminal wires by clip. Appendix A gives more detailed instructions as to how data was taken using the curve tracer. Figure 15: PROVA 210 Make and Model Solar Analyser 26

39 Figure 16: PROVA 210 Make and Model Solar Analyse 4-Wire Connections Irradiance Data Measurements Irradiance data was taken with the aid of a Kipp and Zonen irradiance meter. The Instrument displayed the irradiance level directly in. The instrument specifications are contained in appendix C. Different measurements were taken for each individual orientation and tilt combination of the 3 modules. The measuring device and technique complied with standard requirements as described in the international standard. Figure 17: Kipp and Zonen Irradiance Meter Measuring Irradiance on the Plane of the module 27

40 Figure 18: Kipp and Zonen Irradiance Meter PV Module Temperature Measurements Temperature measurements were made with a Protek 506 Multimeter, switch to temperature data reading mode. A light gauge wire was used as a back sheet material temperature sensor. This wire was attached directly to the module via the multimeter. The positioning of the wire contact with the back sheet was kept away from the cooler perimeter of the module. The positioning was selected to obtain an average temperature of the module. Figure 19: PROTEK 506 Multimeter Temperature Reader 28

41 3.2.5 PV Module Tilt and Orientation Measurements The principle of a change in global irradiance levels as a result of changing module tilt angle and orientation is used to get different levels of irradiance from a given set of environmental conditions over a very short period of time. This is an alternative to recoding module performance data at different times of the day where the atmospheric spectrum distribution could be different and therefore add more uncertainty and additional scope for data discrepancies. Tilt angles of 30,35,40,45 and 50 degrees were implemented for each module. Orientations of -75, -37.5, 0, 37.5 and 75 degrees were implemented, with 0 degrees as a reference taken from true north. Figure 20: Data Recording Location Monument Facing True North Solar Spectrum Measurements Solar spectrum data measurements were taken for each new module tilt/orientation combination. The instrument in use was a StellarNet Spectrometer[46]. The apparatus was placed next to test modules and matched to the tilt angle. Data was transferred via the spectrometer straight to a laptop, which could be obtained for analysis. 29

42 Figure 21: StellarNet Spectrometer Sensor used for Solar Spectrum Data Recording Figure 22: StellarNet Spectrometer used for Solar Spectrum Data Recording 30

43 3.3 Indoor Testing Indoor testing of the selected modules was implemented so that a reference to data measured at STC was possible. The instrument used for simulation was the Spire 5600SLP solar simulator. The use of this simulator allows the user to determine parameters of PV modules such as module efficiency and fill factor. Modules are placed face down between two fastener guides as seen in figure 23. The computer control monitor shown in figure 24 is used to control the use of the simulator by appropriate software. Users can change the irradiance, temperature and air mass coefficient settings that testing panels will be subject to via the control monitor. The measuring performance specifications of the simulator quote a repeatability of 0.15% Pmax, Isc, Voc and FF [47]. Data was taken for each of the poly and mono-crystalline silicon modules as well as the amorphous silicon module at three different irradiance levels of 1000, 700 and 400. For these different irradiance levels, all data was taken with a module temperature of 25 degrees and an air mass coefficient of AM1.5. Data was also taken with the irradiance level held constant at 1000 at and air mass coefficient held constant at AM1.5 but with a range of different temperatures. The data was simulated under these circumstances to allow the temperature coefficients to be evaluated. Figure 23: The Spire 5600SLP Solar Simulator Located at Murdoch University 31

44 Figure 24: The Spire 5600SLP Solar Simulator Control Monitor 3.4 Procedure Correction Factors All the parameters described in section and were determined by the use of the indoor testing data, as described in section 2.4. The graphical results for the determination of the correction factors can be seen in appendix B. 3.5 Experimental Device Limitations Accuracy of experimentally recorded data can be greatly affected by the uncertainty and consistency of a particular recording apparatus. Table 2 lists the uncertainties of the devices used in this project when recording data. These uncertainties can become cumulative and become greater when device data is combined in relation with another set of data from a different device. Table 2: Uncertainties of Measurement Parameters by Data Recording Apparatus Measuring Device Measured Parameter Measurement Uncertainty PROTEK 506 Multimeter Module Temperature ±3% Rdg + 5D Kipp and Zonen Irradiance Global Irradiance ± 0.1% Meter PROVA 210 Curve Tracer DC Voltage ± 1% ± (1% of Voc ± 0.1V) DC Current ± 1% ± (1% of Isc ± 9mA) StellarNet Spectrometer Spectral Irradiance Stray Light: 0.02% at 435nm 0.2% at 200nm SPIRE SLP5600 Solar Simulator I-V Characteristics Measurement Repeatability 0.15% Pmax, Isc, Voc, FF 32