Photovoltaic systems distributed monitoring for performance optimization

DOCTORAL THESIS: PHOTOVOLTAIC SYSTEMS DISTRIBUTED MONITORING FOR PERFORMANCE OPTIMIZATION AUTHOR: FRANCISCO JOSÉ SÁNCHEZ PACHECO DIRECTORS: PROF. DR. JUAN RAMÓN HEREDIA LARRUBIA PROF. DR. MARIANO SIDRACH DE CARDONA ORTÍN PROF. DR. FRANCISCO PÉREZ HIDALGO MÁLAGA, 2015 AUTOR: Francisco José Sánchez Pacheco EDITA: Publicaciones y Divulgación Científica. Universidad de Málaga Esta obra está sujeta a una licencia Creative Commons: Reconocimiento - No comercial - SinObraDerivada (cc-by-nc-nd): Http://creativecommons.org/licences/by-nc-nd/3.0/es Cualquier parte de esta obra se puede reproducir sin autorización pero con el reconocimiento y atribución de los autores. No se puede hacer uso comercial de la obra y no se puede alterar, transformar o hacer obras derivadas. Esta Tesis Doctoral está depositada en el Repositorio Institucional de la Universidad de Málaga (RIUMA): riuma.uma.es THESIS INDEX i - DOCTORAL THESIS PHOTOVOLTAIC SYSTEMS DISTRIBUTED MONITORING FOR PERFORMANCE OPTIMIZATION WRITTEN BY: FRANCISCO JOSÉ SÁNCHEZ PACHECO DIRECTED BY: PROF. DR. JUAN RAMÓN HEREDIA LARRUBIA PROF. DR. MARIANO SIDRACH DE CARDONA ORTÍN PROF. DR. FRANCISCO PÉREZ HIDALGO TO APPLY TO THE DEGREE OF DOCTOR FROM THE UNIVERSITY OF MÁLAGA THESIS INDEX ii THESIS INDEX iii - INDEX OF CONTENTS Abstract Chapter 1: INTRODUCTION 1.1 THE ROLE OF THE PHOTOVOLTAIC ENERGY NOWADAYS 1-2 1.2 THE CONVENIENCE OF MONITORING 1-3 1.3 THE PV MODULES EFFECTIVE PERFORMANCE RATIO ESTIMATION 1-4 1.4 THE CHALLENGE OF MONITORING AT LOW COST 1-4 1.5 THE OBJECTIVES OF THE THESIS 1-5 1.6 THE STRUCTURE OF THE THESIS 1-6 1.7 REFERENCES 1-8 Chapter 2: THE PV CELL AND MODULE MODELS AND MONITORING 2.1 OPERATIONAL PARAMETERS OF THE ILLUMINATED PHOTOCELL 2.1.1 Short Circuit Current (ISC) 2-6 2.1.2 Open Circuit Voltage of the photocell (VOC). 2-6 2.1.3 Diode reverse saturation current (Io) 2-7 2.1.4 Maximum Power Point 2-7 2.1.5 Fill Factor. Parasitic and characteristic resistances 2-8 2.2 PV MODULE MATHEMATICAL MODELS 2-11 2.3 PV MODULES PARAMETERS. REFERENCE TO STC AND NOCT 2-14 2.4 PV CELL AND MODULE TEMPERATURE ESTIMATION 2-17 2.5 TRANSLATION CRITERIA TO REAL OPERATING CONDITIONS 2-18 2.6 PV PLANTS MONITORING SYSTEMS 2-24 2.6.1 Requirements of a PV Plant Monitoring System 2-28 2.6.2 Commercial specific PV plants monitoring systems 2-29 2.6.3 General purpose industrial Data Acquisition Systems 2-29 2.6.4 Custom Design Monitoring Systems 2-31 2.7 REFERENCES 2-33 CHAPTER 3: SMART MONITORING AND PLC COMMUNICATIONS MODULE - SMCM 3.1 MONITORING SYSTEM ARCHITECTURE 3-2 3.2 SMCM DESIGN CONSIDERATIONS 3-6 3.3 SMART MONITORING AND COMMUNICATIONS MODULE STRUCTURE 3-8 3.3.1 MSP 430 Microcontroller 3-9 3.3.2 Analog Front End for PV module Voltage (VPV) signal conditioning and measurement. 3-12 3.3.3 Analog Front End for IPV signal conditioning and measurement 3-14 3.3.4 AFE for TPVBp signal conditioning and measurement 3-15 3.3.5 Analog Front End for Data Transmission (Tx) and Reception (Rx). 3-16 3.3.6 SMCM-CCS configuration 3-18 3.4 PLC LINE AND TX AND RX DATA ANALYSIS 3-20 THESIS INDEX iv 3.4.1 DC lines PLC data transmission simulation and validation 3-20 3.4.2 Bypass capacitors for PV modules low impedance data path implementation 3-26 3.4.3 EPS Renewable Energies LAB validation of simulated model and experimental data 3-27 3.5 SMCM GRAPHIC USER INTERFACE 3-29 3.6 CONCLUSIONS 3-31 3.7 REFERENCES 3-32 Chapter 4: SMCM MEASURING CHAINS UNCERTAINTY ESTIMATION 4.1 INTRODUCTION TO THE UNCERTAINTY IN THE MEASUREMENTS 4-2 4.2 MEASURING CHAINS STANDARD UNCERTAINTY ESTIMATION 4-2 4.3 MEASUREMENT UNCERTAINTY ESTIMATION DUE TO AMBIENT CONDITIONS 4-5 4.4 MEASURING CHAINS ELECTRONIC COMPONENTS DRIFT UNCERTAINTY ESTIMATION 4-7 4.5 CONCLUSIONS 4-10 4.6 REFERENCES 4-11 Chapter 5: MODEL VALIDATION WITH EXPERIMENTAL DATA 5.1 INTRODUCTION 5-2 5.2 PROPOSED MATHEMATICAL PV MODULE MODEL VALIDATION 5-4 5.2.1 Modeled values vs field experimental data 5-4 5.3 CONCLUSIONS 5-12 Chapter 6: MODEL APPLIED TO SMCM MONITORED DATA 6.1 INTRODUCTION 6-2 6.2 EPS PHOTOVOLTAIC LABORATORY TEST ASSEMBLY DESCRIPTION 6-2 6.2.1 Elements and instruments that compose the laboratory test bench 6-6 6.2.2 Isofoton ISF-245 PV module main electrical characteristics 6-7 6.2.3 Atersa calibrated cell electrical characteristics 6-8 6.3 MEASUREMENT PROCEDURE 6-9 6.4 EXPERIMENTAL DATA OBTAINED WITH ISOFOTON PV MODULES 6-11 6.5 CONCLUSIONS 6-20 Chapter 7: PV MODULES REAL-TIME PERFORMANCE RATING 7.1 INTRODUCTION 7-2 7.2 PHOTOVOLTAIC MODULES INSTANTANEOUS PERFORMANCE RATIO ANALYSIS 7-5 7.3 EXPERIMENTAL RESULTS 7-10 7.3.1 Measurements on UMA RREE Lab Modules 7-10 7.3.2 Measurements on EPS PV LAB Modules 7-16 THESIS INDEX v - 7.4 CONCLUSIONS 7-21 7.5 REFERENCES 7-22 Chapter 8: CONCLUSIONS, FUTURE RESEARCH LINES & SCIENTIFIC PRODUCTION 8.1 GENERAL CONCLUSIONS 8-2 8.2 FUTURE RESEARCH LINES 8-5 8.3 SCIENTIFIC PRODUCTION 8-5 8.3.1 Articles in Indexed Journals 8-5 8.3.2 Articles in International Conferences 8-6 Summary in Spanish (Resumen en español) THESIS INDEX xii Figure 7.11 Thermophotography of Isofoton ISF-245 PV modules showing a defective cell on the upper right hand corner (courtesy of Intermal) 7-19 Figure 7.12 STC, translated and real I-V curves for Tm=40,4 ºC and G=650 W/m2 7-20 Figure 7.13 STC, translated and real P-V curves for Tm=40,4 ºC and G=650 W/m2 7-20 Chapter 8: CONCLUSIONS, FUTURE RESEARCH LINES & SCIENTIFIC PRODUCTION THESIS INDEX xiii - INDEX OF TABLES Chapter 1: INTRODUCTION Chapter 2: THE PV CELL AND MODULE MODELS Table 2.1 Isofotón ISF-245 PV module data-sheet parameters at STC 2-18 Table 2.2 Isofotón ISF-245 PV module data-sheet parameters at NOCT 2-19 Table 2.3 PV Plants parameters to be monitored 2-24 Chapter 3: SMART MONITORING AND PLC COMMUNICATIONS MODULE Chapter 4: SMCM MEASURING CHAINS UNCERTAINTY ESTIMATION Table 4.1 Measuring chain calibrated constants calibration 4-4 Table 4.2 Uncertainty estimation of measuring chains calibration constants 4-5 Table 4.3 Stable conditions consecutive readings (G=226,63 W/m2 , Ta= 24,06 ºC) 4-6 Table 4.4 Uncertainty estimation of electronic components drift 4-8 Table 4.5 Uncertainty estimation of voltage, current and temperature measuring chains 4-9 THESIS INDEX xiv Chapter 5: MODEL VALIDATION WITH EXPERIMENTAL DATA Table 5.1 Yocasol PCB-195 module electrical characteristics at STC 5-3 Table 5.2 One clear sky day long proposed model theoretical and experimentally measured PV modules operating parameters. Relative Error of most significant electrical parameters 5-8 Chapter 6: MODEL APPLIED TO SMCM MONITORED DATA Table 6.1 Isofoton ISF-245 PV module electrical parameters at STC 6-7 Table 6.2 Isofotón ISF-245 PV module data-sheet operational characteristics 6-7 Table 6.3 Atersa calibrated cell electrical characteristics 6-8 Table 6.4 Mean values, STD deviation and outdoor conditions of readings 6-11 Table 6.5 Results of run #5 (maximum output at G=217 W/m2 and Tm=33.1 ºC) 6-13 Chapter 7: PV MODULES REAL-TIME PERFORMANCE RATING Table 7.1 PV plant Estimated Energy Losses Factors Values – LFj 7-3 Table 7.2 Yocasol PV modules measurements performed at UMA RREE Lab 7-11 Chapter 8: CONCLUSIONS, FUTURE RESEARCH LINES & SCIENTIFIC PRODUCTION THESIS INDEX xv - LIST OF SYMBOLS Symbol Name or description a Photovoltaic module temperature coefficient b Photovoltaic module temperature coefficient Drift(VAFE) Derate of the VAFE Drift(IAFE) Derate of the IAFE Drift(TAFE) Derate of the TAFE Eg Energy gap of the semiconductor material atom FF Fill Factor FFO Ideal Fill Factor FFOt Translated ideal Fill Factor FSOA Operational Amplifier Full Scale G Incident irradiance Ge Effective irradiance in the plane of the photovoltaic module Gg Global incident irradiance G0 Reference irradiance at STC I Current at the output of the photovoltaic cell/module ID Diode current Imp Maximum power point current ISC Photovoltaic module short circuit current ISC0 Photovoltaic module short circuit current at STC ISCt Photovoltaic module translated short circuit current I0 Diode saturation current Iph Cell/module photogenerated current kB Boltzmann constant kISENS Hall effect current sensor sensitivity kVC VAFE measuring chain constant kVN VAFE measuring chain nominal constant kIC IAFE measuring chain constant kIN IAFE measuring chain nominal constant kTC TAFE measuring chain constant THESIS INDEX xvi kTN TAFE measuring chain nominal constant LF Loss Factor LeDC Effective DC Losses LiDC Instant read PV Module DC overall Losses LtDC ROC translated DC overall Losses NIADC ADC digital counts of photovoltaic module current NOCT Normal Operating Cell Temperature Ns Number of series connected photocells NTADC ADC digital counts of photovoltaic module back-plane temperature NVADC ADC digital counts of photovoltaic module voltage P PV module output power PDC Photovoltaic module DC generated power Pi Photovoltaic module instant power PmLF Photovoltaic module maximum power with loss factor Pmp Maximum power point Pmpt, Pmt Translated maximum power point Pmp0 Maximum power point at STC PR Traditional Performance Factor PRDC DC side Performance Ratio PRiDC Photovoltaic module instant DC performance ratio PRtDC Photovoltaic module translated DC performance ratio q Charge of the electron RCh Characteristic resistance Rp Parallel parasitic resistance Rs Series parasitic resistance Rst Translated series parasitic resistance Sci Uncertainty sensitivity coefficient SR Spectral Radiation STC Standard Test Conditions SW Wind speed T Absolute temperature (K) Ta Ambient temperature THESIS INDEX xvii - Tc Photovoltaic cell temperature Tct Translated photovoltaic cell temperature Tm Photovoltaic module temperature Tmt Translated photovoltaic module temperature T0 Reference module temperature at STC U Expanded uncertainty u(C) Uncertainty contribution of measuring chains calibrated constants u(R) Uncertainty contribution of reading unstable conditions u(D) Uncertainty contribution of components drift Uc(y) Combined uncertainty uD(Xi) Components drift uncertainty terms uD(VAFE) Uncertainty of VAFE drift uD(IAFE) Uncertainty of IAFE drift uD(TAFE) Uncertainty of TAFE drift uR(Xi) Reading uncertainty terms V Voltage at the output of the photovoltaic cell/module Vq ADC quantization voltage VD Diode voltage Vmp Maximum power point voltage VOC Photovoltaic module open circuit voltage VOC0 Photovoltaic module open circuit voltage at STC VOCc Photocell open circuit voltage VOCct Translated photocell open circuit voltage VOCt Photovoltaic module translated open circuit voltage vOC Normalized VOC voltage vOCt Translated normalized VOC voltage VT Thermal voltage VTt Translated thermal voltage ∆T Temperature difference α Photovoltaic module current temperature coefficient αVIO Opamp input offset voltage thermal drift β Photovoltaic module voltage temperature coefficient THESIS INDEX xviii δ Photovoltaic module irradiance temperature coefficient γ Photovoltaic module power temperature coefficient THESIS INDEX xix - A mis Padres A mi mujer Pilar y a mis hijos Carlos y Alvaro A mi hermano Carlos THESIS INDEX xx ACKNOWLEDGEMENTS AGRADECIMIENTOS Esta Tésis Doctoral recoge los resultados del trabajo desarrollado durante más de 5 años ya. Ha sido una labor árdua, que ha requerido mucho tiempo, gran parte del cual se lo he quitado a mi Familia. Sin embargo, ha constituido para mi una oportunidad única de profundizar en un campo apasionante, al tiempo que me ha permitido compartir muchos fructíferos momentos con grandes profesionales que me han transmitido su buen saber hacer. Espero y deseo que esto sea un punto de continuidad habida cuenta de las perspectivas que este trabajo ha permitido vislumbrar. Quiero expresar mi mayor agradecimiento a mis Directores de Tésis, los Dres. Juan Ramón Heredia Larrubia, Mariano Sidrach de Cardona y Francisco Pérez Hidalgo, por su constante apoyo y asesoramiento, durante todos estos años. Sus aportaciones en cuanto a los contenidos, estructura y valoración de los resultados experimentales han sido inestimables. Ha sido para mi un privilegio haber podido contar con ellos como Directores. También quiero agradecer a los profesores del DIEM de la Universidad de Salerno, que tan amablemente me acogieron durante mi estancia en dicha Universidad. Tuve la oportunidad de contactar con los profesores Giovanni Spagnuolo y Giovanni Petrone, renombrados expertos en el área de la Tecnología Fotovoltaica cuyos conocimientos y sugerencias me ayudaron mucho en el desarrollo de la Tésis. Mis compañeros de Departamento han constituido el elemento de apoyo necesario en esta andadura. Sus constantes palabras de ánimo me han servido para no cejar en el esfuerzo. Las sugerencias hechas por la profesora Ana Pozo y por el profesor David Trujillo han sido de gran ayuda. Muy particularmente, quiero agradecer al Profesor Pedro Sotorrío su inestimable ayuda y asesoramiento en varios apartados de la Tésis. También quiero agradecer el asesoramiento que me facilitó el profesor Lorenzo Sevilla. THESIS INDEX xxi - Quiero agradecer asimismo el trabajo de revisión que han hecho los Dres. Damien Picault y Sonia Leva, que me han permitido acceder a la mención de Doctorado Internacional. El tiempo robado a la Familia espero que no haya sido en vano. Han sido muchas las horas dedicadas los fines de semana y los días cogidos en períodos de vacaciones. El apoyo incondicional que en todo momento me ha transmitido mi mujer Pilar, así como mis Padres ha sido providencial para haber podido consolidar este trabajo. Una vez echada la vista atrás, sin la participación de todas estas personas, uno no se imagina haber podido llegar hasta aquí y no puede menos que estar enormemente agradecido. No es posible compensar dichos esfuerzos, salvo con el compromiso personal de intentar transmitir sus conocimientos y sabios consejos cuando se me dé la oportunidad de poder hacerlo en la labor cotidiana que uno intenta llevar a cabo de la mejor manera posible. CHAPTER 1 1-2 … And it follows [1]: “Photovoltaic (PV) energy is one of the most promising emerging technologies. The cost of PV modules has been divided by five in the last six years; the cost of full PV systems has been divided by almost three. The levelised cost of electricity of decentralised solar PV systems is approaching or falling below the variable portion of retail electricity prices that system owners pay in some markets, across residential and commercial segments. For bulk power on grid, PV electricity can already be competitive at times of peak demand, especially in areas where peak electricity is provided by burning oil products. And there remains ample room for improvements…” 1.1 THE ROLE OF THE PHOTOVOLTAIC ENERGY NOWADAYS It is a matter of fact that Renewable Energies play an important role in the Green House Gases (GHG) emissions reduction [1]. Figure 1.1 is a graph of the potential contribution of the different technologies in emissions reduction. The solar PV has a potential contribution of 20 %, but, furthermore, the electricity efficiency improvements can contribute a 23 % more. Figure 1.1. Cumulative technology contributions to power sector emission reductions in ETP 2014 hi-Ren scenario (source: IEA) INTRODUCTION 1-3 Figure 1.2 shows the cumulative PV based energy production by countries, as per 2013 [1]. 1.2 THE CONVENIENCE OF MONITORING Electricity production PV power plants of medium to large dimensions are very sensitive to optimal performance from operational and financial points of view [2][3][4]. Projects shareholders expect to have the maximum benefit of their investment and have a prompt pay-back. PV plants monitoring is becoming a real need, since any malfunction can create important revenue losses [5]. In this sense, the International Energy Agency (IEA) states clearly the need to know exactly the PV performance data [2][6]. Different adverse circumstances may lead to a significant decrease in the conversion of solar radiation into electricity production [7]. Most relevant adverse conditions are due to:  Atmospheric and ambient variables like temperature, humidity, wind speed and air mass index  Permanent or temporary shadowing, dust, soiling  Devices degradation / failures Figure 1.2. Cumulative PV energy production capacity (source:IEA) CHAPTER 1 1-4 Several of these power losses might apparently be due to incident radiation or occasional shadowing conditions [8]. This being the case, it is somewhat difficult to determine the root cause of any potential failure. The fact to have reliable real time data of every single PV module functional parameters, namely voltage, current and back-plane temperature can help to determine their optimal performance and evaluate any problem root cause. In case of system failure, since proposed solution is able to identify the defective module, immediate corrective actions can be undertaken, which improves maintenance activities. In case of non-monitored plants, a defective module might be masked into the string with which it is associated and no major evidence would arise. Nowadays, there are different possible solutions available in the market and referred in the literature. These solutions are based on general purpose industrial monitoring and control systems (SCADA), on application specific systems designed for PV plants monitoring, or in some cases this is performed by µinverters endorsed to the PV modules. All these options represent a significant cost of the system itself and of the installation in the plant. These solutions are tailored mainly for domestic or small scale facilities and not always suitable to monitor at PV module level. Regarding the transmission of the monitored data, here again the possible solutions are based on standard industrial communications configurations, which require additional wiring installation; another option is based on radio communications topologies, expensive and highly sensitive to noise and communications problems. Finally, if the case of module level monitoring, in the literature, most of the solutions are based on costly magnetic coupling. Furthermore, the data mining and processing of the monitored parameters can help to improve the PV plant performance. For this purpose, it is proposed a concept of Performance Ratio that helps to real-time identify a PV module that is underperforming according to the applying real outdoor conditions. INTRODUCTION 1-5 1.3 THE PV MODULES EFFECTIVE PERFORMANCE RATIO ESTIMATION Once the PV module functional parameters have been monitored, the main advantage is that this information can help to quantify the PV module real time Effective Performance Ratio (EPR). This EPR by difference with the traditional one, gives a more accurate information of the module operation, taking into account the applying outdoor ambient conditions. This being done, the resulting values help to identify in real time any defective module, working below the expected values. 1.4 THE CHALLENGE OF MONITORING AT LOW COST Arrived at this point, the convenience of monitoring at PV module level is clear and justified to be convenient. Now the challenge is to define a monitoring structure which were suitable for complying with requested objectives. Mainly, these are related to tight the costs as low as possible of:  System electronic components  Data transmission and reception (Tx/Rx)  System deployment The main challenges were to develop a monitoring device, embedded in a whole monitoring system that might be cost competitive. If the monitoring module cost were excessive, it might not be justified in terms of investment in the monitoring system by the PV plant owners. On the other hand, the overall monitoring system, including the data Tx/Rx communications layers is also an important issue in terms of installation costs. Furthermore, the system deployment cost-effectiveness would be questioned if finally its usefulness were not demonstrated. Maintenance costs due to non-identified defective modules may have an important impact. The identification of a defective PV module in a string requires a skilled technician that would have to analyze each single module in a time consuming activity. CHAPTER 1 1-6 As an orientation example, in a 100 kW PV which might invoice 50.000 €/year, an accumulated continuous 10 % of not detected losses could represent a lack of incomes of around 5.000 €/year. Considering that the proposed monitoring system has an estimated materials cost of around 10 €/module (which is a 2 % of a 250 Wp PV module cost), a 550 PV panels plant monitoring system deployment would represent an investment of roughly 5.500 €. The return of the investment is done in one year of PV plant operation. 1.5 THE OBJECTIVES OF THE THESIS The work performed and the results obtained in order to comply with above referred challenges have been detailed in the present Thesis. In essence, the objectives have been:  To demonstrate the possibility to monitor the PV modules operational parameters at module granularity level  The possibility to receive and transmit monitored data using the already existing direct current (DC) power line, applying the Power Lines Carriers technology  To implement an overall monitoring system at the lowest possible cost  To demonstrate the Effective Performance Ratio proposed model advantages 1.6 THE STRUCTURE OF THE THESIS On the basis of these premises, this thesis describes all the tasks that have been undertaken to comply with them and which have helped to demonstrate the viability of the idea. It includes:  The models definition  Modules and Tx/Rx lines simulation  Laboratory prototypes and test benches  The field experimental results INTRODUCTION 1-7 The present Thesis is structured as follows: After this introductory chapter, in chapter 2 an overall review is made of the PV cells and modules functional parameters, as well as their dependence on the applying outdoor ambient conditions. The identification of these parameters behavior is essential to define the monitoring system. Additionally, the translation criteria of referred parameters to real outdoor operating conditions have been depicted which will be necessary to further estimate the proposed EPR and subsequent resulting Losses Factors (LF). The different mathematical models; of the cell, the module, the translation criteria and the resulting Performance Ratios and Losses coefficients are deployed. Chapter 3 is devoted to the PV modules monitoring concept. A description is made of the proposed Smart Monitoring and Communications Module (SMCM), together with the Power Lines Carrier application to the monitoring system as physical communication layer. The electronic design, simulation and laboratory prototypes are full depicted. Afterwards, in chapter 4, the uncertainty related to above referred SMCM in terms of measuring device is described and quantified. In chapter 5, the models proposed in chapter 2 are validated by means of consolidated field experimental data. This has allowed to confirm the viability of the proposed models. In chapter 6, the same procedure has been applied to the monitoring data resulting from the experimental measurements made at the Escuela Politécnica Superior (EPS) Photovoltaic Lab, where two Isofotón ISF-245 PV modules have been installed together with a laboratory instrumentation test bench for experimental purposes. In chapter 7, a detailed analysis is made of monitored data obtained by means of the SMCM, with the purpose of quantifying the PR and subsequent LF´s. Finally, chapter 8 is devoted to the conclusions and future research lines that can result from present research work. CHAPTER 1 1-8 1.7 REFERENCES [1] IEA. International Energy Agency. Technology Roadmap. Solar Photovoltaic Energy. 2014 Edition. [2] L. Cristaldi, M. Faifer, A. Ferrero and A. Nechifor, “On line monitoring of the efficiency of photo-voltaic panels for optimizing maintenance schedule”. I2MTC 2010 - International Instrumentation and Measurement Technology Conference, Austin, TX, 3-6 may 2010. [3] L. Dorobantu, M.O. Popescu and C.L. Popescu, “Yield loss of photovoltaic panels caused by depositions”. The 7th International Symposium on Advanced Topics in Electrical Engineering, Bucharest, may 2011. [4] G. Petrone, G. Spagnuolo, R. Teodorescu, M. Veerachary and M. Vitelli, “Reliability Issues in Photovoltaic Power Processing Systems”. IEEE Transactions on Industrial Electronics, Vol. 55, No. 7, july 2008. [5] B. Marion, K. Adelstein, K. Boyle, H. Hayden, B. Hammond, T. Fletcher, B. Canada, D. Narang, A. Kimber, A. Kimber, L. Mithcell, G. Rich and T. Townsend, “Performance parameters for grid-connected PV systems”. Conference Record of the Thirty-first IEEE Photovoltaic Specialists Conference, 2005. [6] IEA. International Energy Agency. Performance Prediction of Grid-Connected Photovoltaic Systems using Remote Sensing. Report IEA PVPS T2-7: 2008. [7] A. Detrick, A. Kimber and L. Mitchel, “Performance Evaluation Standards for Photovoltaic Modules and Systems”. Conference Record of the Thirty-first IEEE Photovoltaic Specialists Conference, 2005. [8] S. Vergura, G. Acciani, V. Amoruso and G. Patrono, “Inferential Statistics for Monitoring and Fault Forecasting of PV Plants”. IEEE International Symposium on Industrial Electronics, ISIE 2008. Chapter 2 PV CELL AND MODULE MODELS AND MONITORING In chapter 1, it has been justified the convenience of PV modules monitoring, in order to have reliable information related to their behavior and by extension to know how the PV plant is performing. A PV modules monitoring system is a very important resource which makes possible to empower different aspects once it is implemented. On one hand, it allows the user to have real time information of the behavior of the Photovoltaic plant; it can store data for further retrieving. Furthermore, according to the data processing algorithms, it may issue the corrective actions to be generated. For this purpose, a detailed description is made of the main operating parameters of the photocell and by extension of the PV module, which will be further monitored. CHAPTER 2 2-2 2.1 OPERATIONAL PARAMETERS OF THE ILLUMINATED PHOTOCELL A solar cell is in essence a p-n semiconductor based device capable of generating electricity from the incident sunlight, by means of the PV process [1]. Since 1954, when the first efficient photovoltaic cells were manufactured, up to date, the technology has made possible to reach efficiencies in the surrounding of the 45 % [2]. In the dark, a PV cell behaves like a p-n junction diode and requires an external voltage source to allow a given amount of current through the diode (Figure 2.1). The current through the cell (ID) would be defined by the voltage of the external power supply (VD) when in direct polarization and when in reverse polarization, the current obtained is the dark saturation current (IO) [1]. The expression that characterizes both parameters is the typical characteristic curve of the diode: (2.1) Where VD is the voltage across the diode; VT is the thermal voltage, given by equation (2.2) and which value at 25 ºC is considered to be around 25 mV for Silicon. (2.2) Figure 2.1. The photocell in the dark equivalent diode and characteristic curve         1 0T D V V DeII p n VD VD ID ID VD I0 q Tk VCB T   PV CELL AND MODULE MODELS AND MONITORING 2-3 Where kB is the Boltzmann constant (kB=1,38 · 10-23 J/K); TC is the absolute temperature of the photocell under ambient conditions and q is the charge of the electron (q=1,60 · 10-19 C). On the other hand, the diode dark saturation current (IO) is given by equation (2.3), according to [1]: (2.3) Where Eg is the energy gap of the semiconductor material used for the photocell manufacturing (1,2 eV in case of Silicon). This makes the photocell energy conversion efficiency dependent upon the wavelength of the incident photons (Figure 2.2.) [4]. When the photocell is being submitted to the sun irradiance, the PV process begins and it starts generating electrons in movement to create the electric current. An ideal photocell can be represented by an ideal photo-generated current source (Iph) and a diode in parallel, as per Figure 2.3, [3]: 25 0/105,1 cmAeI Tk E B g           Figure 2.2. Spectral distribution of sunlight of the sun as a 6000 K black body, with AM0 and AM1,5 CHAPTER 2 2-10 In order to establish a relationship between FF and FF0, which would give an idea of the cell ideality, it can be considered that [1]: (2.14) Equation (2.14) gives a maximum value for FF when Rs=0 Ω, resulting in FF=FF0. Solving Rs from equation (2.14), it can be deducted that: (2.15) Substituing equation (2.13) in (2.15), it results equation (2.16), which is the expression that empirically expresses the value of Rs, as a function of the characteristic parameters of the photocell. (2.16) Figure 2.10. PV cell characteristic resistance Rch                 SC OC I V FF FF Rs 0 1 V I ISC VOC Imp Vmp Pmp Rch=VOC/ISC       Rch Rs FFFF 1 0         0 1FF FF RchRs PV CELL AND MODULE MODELS AND MONITORING 2-11 The effect of the parallel resistance Rp represents the leakages due to the solar cell crystal defects and impurities. The value of Rp is significant, and for ease of calculation, considered ∞ and hence removed from the electrical model. This parasitic resistor affects mainly the slope of the I-V in the region of ISC. All these considerations are depicted in Figure 2.11. The values of Rp and Rs that finally affect the PV cell, will define the coordinates of the MPP. 2.2 PV MODULE MATHEMATICAL MODELS In order to compare measured PV module operating parameters with the values that theoretically should be obtained, under the applying atmospheric conditions, a mathematical model of the PV module operational parameters (Figure 2.12) is required. Figure 2.11. Effect of the parasitic resistances on the I-V PV cell characteristic curve V I ISC VOC Effect of Rp Effect of Rs Slope=1/Rs Slope=1/Rp Ideal curve Figure 2.12. PV module operational parameters. V I Back-plane T CHAPTER 2 2-12 There are several options that have already been described in the literature. The most detailed and extensive one is to consider the two diodes model, as described in [10], by Adamo et al. Nevertheless, the single diode option has been considered (Figure 2.13a) and for ease of operations, the option without parallel loss resistance (Rp) has been selected, which, being accurate, significantly simplifies the mathematical model [11] and [12] (Figures 2.13b). It is suitable in order to obtain the I-V fingerprint, which involves the PV module main electrical operating parameters: (Figure 2.14). - ISC : short circuit current (V=0 V) - VOC : open circuit voltage (I=0 A) - Vmp: Voltage at MPP - Imp: Current at MPP - Pmp : Power at Maximum Power Point V I ISC VOC Pmp Imp Vmp Pmp Figure 2.14. PV module electrical operational parameters Figure 2.13a. PV cell one diode equivalent model with Rp and Rs I Iph IO V Rs Rp I Iph IO V Rs Rp Figure 2.13b. PV cell one diode equivalent model without Rp PV CELL AND MODULE MODELS AND MONITORING 2-13 The expression that characterizes the one diode based single exponential model of a PV module (Figure 2.13a) is given by equation (2.17). (2.17) Where V and I are respectively the voltage and current supplied by the PV module (Figure 2.3); Iph stands for the photo-generated current; n is the diode ideality factor; Ns is the number of series connected cells of the PV module; VT is the equivalent diode thermal voltage and Rp and Rs represent the ohmic losses that affect the final energy yield. One of the constraints of equation (2.17) is that it is an implicit one, which requires significant computational resources to solve it. Beside traditional computational environments, other possible solutions have been described in the literature. Petrone et al. in [13] describe a solution based on the use of the Lambert W-function. Furthermore, Merino et al. in [14] propose the so-called Reverse Decomposition Method to solve the PV cell electrical model. In order to simplify one step ahead the analysis process, in this case and since in comparison, Rp>>Rs, it is generally accepted to neglect Rp (Figure 2,13b), and hence, equation (2.17) can be written as follows: (2.18) Above equation represents the non-implicit simplified mathematical one diode model of the illuminated PV module. This equation is based on six variables, namely I, V, Iph, I0, VT and Rs. Once solved equation (2.18), the PV module fingerprint I-V curve can be obtained. This can be done any time the values of Iph, Io and VT are known. It has been considered the option based on PV module           1 0Ts s VNn IRV ph eIII Rp IRsV eIII Ts s VNn IRV ph             1 0 CHAPTER 2 2-14 manufacturer´s data sheets, as in [15]. This can apply to the photo-generated current (Iph) and consequently to the PV module output voltage in absence of load, that is, the open circuit PV module voltage (Voc). Given all these data, finally, the PV module fingerprint I-V characteristic curve can be drawn. For this purpose, starting from equation (2.18) and solving for V, we obtain equation (2.19) is obtained: (2.19) Which can be solved for Iph=ISC. This equation will be used in order to determine the theoretical value of the PV module output voltage, for the range of current (I) between 0 A and ISC for given outdoor conditions. Calculated values will be compared with real measurements and the result will help us to determine the PV module instant Performance Ratio. 2.3 PV MODULES PARAMETERS. REFERENCE TO STC AND NOCT A PV Module is an array structure of PV cells combined in series and parallel connections. The array is assembled into an aluminum frame, with the front covered with a tempered glass and the back protected with a tedlar cover (Figure 2.15). This affects the response of the cells and of the module, to the ambient temperature and incident irradiance and, hence, this must be taken into account for further estimation of their performance RsI I II NnVV ph sT           0 1ln PV CELL AND MODULE MODELS AND MONITORING 2-15 . The nominal operational parameters depend upon the final series-parallel cells array configuration and, in essence, come from the parameters of the cells that conform the array (Figure 2.16). These parameters are obtained and certified in laboratory environments and detailed in the manufacturer´s data sheets (Tables 2.1 and 2.2). In order to establish common criteria that might make possible the comparison and analysis of these parameters, they are referred to two different test conditions, according to [6]: Figure 2.15. PV module array and mechanical structure Figure 2.16. PV module I-V (red line) and P-V (blue line) characteristic curve V I ISC VOC Pmp Imp Vmp Pmp CHAPTER 2 2-16 a) Standard Test Conditions – STC: Most of the PV cells and modules manufacturers and qualification laboratories perform the qualification tests at the so-called Standard Test Conditions (STC). According to [6], STC corresponds to: – Irradiance on the plane of the PV module: 1000 W/m2 – PV cell temperature: 25°C – Solar spectral irradiance (Air mass): AM 1,5 Table 2.1 reports referred parameters of Isofotón ISF-245 PV module. b) Normal Operating Cell Temperature – NOCT: Alternatively, it is also common to define the operating parameters to different applying conditions, the so named as Normal Operating Cell Temperature (NOCT) [6], considering an open-rack mounted module in the following standard reference environment: – Tilt angle: 45° from the horizontal – Total irradiance: 800 W/m2 – Ambient temperature: 20 °C – Wind speed: 1 m/s Table 2.1. Isofotón ISF-245 PV module data-sheet electrical parameters at STC Electrical parameter Value Rated Power (Pmax) 245 W Open-circuit Voltage (VOC) 37,3 V Short-circuit Current (ISC) 8,70 A Maximum power point Voltage (Vmax) 30,2 V Maximum power point Current (Imax) 8,12 A Efficiency 14,8 % Power tolerance (% Pmax) +/-3% PV CELL AND MODULE MODELS AND MONITORING 2-17 Table 2.2 gives Isofotón ISF-245 PV module parameters, referred to NOCT. 2.4 PV CELL AND MODULE TEMPERATURE ESTIMATION One of the critical parameters that will affect the overall performance of the PV cell and of the PV module is the cell temperature (Tc) that it reaches during operation. The temperature has different effects on the PV cell operating parameters. It is more relevant on the VOC , and consequently on the Pmp, while its effect on the ISC is less significant [1]. This effect can be quantified by means of corresponding temperature coefficients, supplied by the PV cell manufacturer. Since it is technically somewhat difficult to measure the cell temperature when this one is assembled in a module frame, it can be calculated in two different ways: a) Procedure based on NOCT value This procedure is based on the Nominal Operating Cell Temperature (NOCT), which is a parameter supplied by the PV cell manufacturer. This parameter refers to the temperature that the cell would reach under operation at specific ambient conditions, namely an air temperature of 20 ºC, an incident irradiance of 800 W/m2 and a wind speed of 1 m/s [4]. According to this, the cell temperature can be given by equation (2.20). (2.20) Table 2.2. Isofotón ISF-245 PV module data-sheet electrical parameters at NOTC Electrical parameter Value Maximum Power (Pmax) 176 W Open-circuit Voltage (VOC) 34,2 V Short-circuit Current (ISC) 7,02 A Maximum power point Voltage (Vmax) 26,8 V Maximum power point Current (Imax) 6,56 A G mW NOCT TT ac    2 /800 20 CHAPTER 2 2-18 Where Ta is the ambient temperature (ºC), NOCT is the manufacturer reported cell temperature (ºC) and G is the incident irradiance (W/m2). b) Procedure based on the module temperature Since above procedure is not exempt of a significant uncertainty, and technical difficulty as well, in case of cells assembled in module frame, the cell temperature can be estimated based on the module measured temperature, as per the thermal model indicated in equation (2.21) [5]. (2.21) Where Tm is the back-plane temperature module (ºC), G is the incident irradiance on the plane of the module (W/m2), G0 is the reference irradiance (1000 W/m2) and ∆T is the difference temperature between the cell and back-plane temperature (2 to 3 ºC for c-Si and tedlar). In case that the back-plane temperature could not be measured, but the wind speed (SW) were known or estimated, the module temperature can be given by equation (2.22) [5]. (2.22) Where G is the incident irradiance on the module surface, Ta is the ambient temperature (ºC), Sw is the wind speed (m/s) measured at a height of 10 meters and a and b are empirical coefficients that depend on the PV module technology. 2.5 TRANSLATION CRITERIA TO REAL OPERATING CONDITIONS Although previously referred test conditions references, STC and NOCT as well are generally accepted, one of the problem that arises is that these conditions T G G TT mc  0   a Sba mTeGT W  PV CELL AND MODULE MODELS AND MONITORING 2-19 are difficult to comply and probably will never be met. Furthermore, if the aim is to analyze the PV module performance under real operating conditions (ROC), a procedure must be applied that can help to predict the values of the PV module operating parameters. Once these parameters have been estimated, they can be compared with the field measured PV modules operating parameters under applying ambient conditions in order to quantify the resulting performance ratio. The Real Operation Conditions (ROC) refer to the effective incident irradiance (G), the ambient temperature (Ta) and the wind speed (SW) as the most relevant ones. Additionally, for a more accurate estimation of the PV module performance ration, the module temperature (Tm) is required to be also considered. Other parameters affecting the performance ratio will be detailed in chapter #7 referred to the losses of the PV module performance. Figure 2.17 (Isofotón) depicts clearly the effect of the incident solar irradiance (G) on the short-circuit current (ISC) of the PV module. As it can be seen, ISC decreases as G decreases. The irradiance also has an effect on VOC, although it is not so relevant. The effect of the cell temperature (TC) on the module performance can be seen in Figure 2.18. An increase on the cell temperature has a negative impact on the value of the open circuit voltage (VOC). As the cell temperature increases, the result is a significant decrease of VOC. Figure 2.17. Variation of I-V curve depending on incident solar irradiation at constant cell temperature (source: Isofoton) Effect of G on ISC Effect of G on VOC CHAPTER 2 2-26 A PV monitoring system allows Data logging, storing, user representation, retrieval and transmission as well. This will ensure following benefits:  The correct operation of the plant  To determine the performance of the different PV plant components  To identify devices malfunction which may subsequently generate corrective and preventive maintenance actions In order to define the system configuration, layout and integration, following points have to be considered:  Hardware and Software for PV modules parameters data monitoring and transmission  Hardware for atmospheric ambient conditions monitoring  Hardware, Firmware and Software for data processing, user information and corrective actions to be taken The monitoring system configuration will depend on the PV system layout and dimensions, as well as on the working conditions under which it will be submitted. Since the PV systems can be tailored for many type of applications, from low scale up to huge power plant, working under different conditions, several scenarios can be considered, each with different requirements and constraints. Figure 2.19. Different Monitoring Levels PV CELL AND MODULE MODELS AND MONITORING 2-27 Different solutions can be adopted with scaled upgrades of Hardware and Software (Hw/Sw) resources and cost:  Systems configured around already existing industrial data acquisition platforms  Dedicated systems developed specifically for PV plants applications  System based on the PV modules low cost Smart Monitoring and Communications Module (SMCM), which is described in this thesis Additionally, a specific chapter must be dedicated to the data communications in the monitoring system. In case of high dimensions PV plants, connectivity and communications topologies are a critical issue for real time monitoring and control. Different options can be considered :  Wired technology: additional cabling network must be implemented. Cost impact.  Wireless technology: radio systems must be deployed around the whole PV plant (zigbee, bluetooth, etc…). Communications problems.  PLC technology: takes profit of already existing power wiring between PV modules and connection points. This option means important savings in the installation costs. The communications hierarchy levels are described in Figure 2.20:  Level 1: between primary sensors and signal conditioning / monitoring modules  Level 2: between monitoring modules and Central Computer System (CCS)  Level 3: Between CCS and end users interfaces CHAPTER 2 2-28 2.6.1 Requirements of a PV Plant Monitoring System In order to define a PV plants Monitoring System (PVPMS) architecture, most important requirements to be taken into consideriation are:  PV plant size and dimensions  Geographical location  Data monitoring and logging technology  Data Tx and Rx communications layer into system integration  Data transmission requirements (baud rate, distance, line specifications)  Data processing and Graphic User Interface (GUI) format for end user interface and information generation  Final control elements requirements  Hardware, firmware and software integration  Total system cost For the deployment and implementation of the monitoring system, several possible solutions can be considered, as described in following paragraph. Figure 2.20. Data communications hierarchy levels Level 1 connectivity PV plant - SMCM Level 2 connectivity SMCM– CCS SMCM Level 3 connectivity CCS end user ≈ = PV CELL AND MODULE MODELS AND MONITORING 2-29 2.6.2 Commercial specific PV plants monitoring systems Systems based on specific PV plants monitoring systems are currently available in the market. They are based on PC architecture or dedicated programmable logic controllers. They process data at string level and use inverters generated parameters for PV plant performance evaluation. Some manufaturers supply specific data acquisition modules. These options require manufacturer´s proprietary application software for system supervision. They are mainly intended to comply with following requirements:  Real Time Data Acquisition  Operation supervision (energy metering, alarms, data transmission and storing, meteorological parameters)  Web access for remote supervision and control capabilities. Wired / wireless communications protocol  Allow MPPT plant operation  Grid monitoring (voltage stability, power generation, phase angle supervision) But, they have also some constraints:  PC based + DAS + Management and propietary user interface Sw  Not fully end user configurable  Closed systems from programming point of view 2.6.3 General purpose industrial Data Acquisition Systems Another possibility is based on general purpose industrial Data Acquisition Systems (DAS). This option has the advantage to use broadly tested industrial systems and to give to the end user the possibility to develop its own software application, according to PV plant management requirements. They are usually based on PC platforms and use RS232/RS485 (for digital) or 4-20 mA industrial standard current loop (for analog) wired data transmission. Referred options require additional wiring for both monitoring and data transmission. Some of them use radio modules for wireless communications. This CHAPTER 2 2-30 is a constraint if monitoring system deployment is large. On the other hand, radio systems are not exempt of operation problems, due to noise, antennas mistuning, etc… In any case, this means an important additional installation cost. The main advantages of this option are:  Modularity, adaptability and expansion capabilities  All type of primary parameters sensors (G, Ta, SW)  Communications between sensors and CCS wired or wireless  Web access is also granted in most cases  Allow data transmission, storing and display  Software development tools allow customer self-development according to the plant operation requirements Meanwhile, main constraints are:  These systems are already available in the market from different makers at considerable costs  Not properly suitable for monitoring at PV module level Several monitoring systems of this type have already been presented [18] and more recently [19], which are considered essential in order to improve the PV facility efficiency. In Figure 2.21, the layout of the industrial type monitoring system installed in the EPS PV Lab is depicted. PV CELL AND MODULE MODELS AND MONITORING 2-31 2.6.4 Custom Design Monitoring Systems The third option is based on PV custom design application specific monitoring system. It is specifically designed to comply with PV plants requirements, from monitoring, communications as well as management and operation point of view. These systems are scalable and can be plant deployed according to monitoring needs and financial available budget. Supervisory, operations and management activities can be further defined and implemented since plant management application Sw is end user developed. In our case, taking into consideration all these advantages, this is the option adopted. The main advantages of monitoring at PV module level are:  The use of low cost devices  Real Time Data availability (V, I and T)  Real Time defective PV module detection  No delay on maintenance/corrective actions to be taken  End user programmable and reconfigurable Figure 2.21. Industrial type DAS for PV modules monitoring at EPS PV Lab CHAPTER 2 2-32  PLC based communications layer (uses already existing DC power wiring for data Tx –reduces communications cost)  PC based + SMCM + Management and propietary user interface SW  Not affected by radiation noise Nevertheless, they have some constraints:  Must be installed in each PV module  Data Tx / Rx sensible to line conditions In the literature, several options have been described. In [20], Román et al. present a monitoring module with a magnetic coupling to the DC line. In [21], Carullo et al. describe an in-situ PV modules calibration system; meanwhile, Guerriero et al. in [22] write about a wireless communications based PV modules monitoring device. In 2011, Sánchez et al. in [23] presented in the IEEE PowerEng Conference, the first approach to the SMCM proposal, based on low cost concept based on the PLC technology for data transmission. More recently, in [24], corresponding experimental results and uncertainty quantification were published. This is the model that has allowed to implement the field experiments in order to validate the proposed concept of low cost Real Time monitoring for PV plant efficiency improvement. PV CELL AND MODULE MODELS AND MONITORING 2-33 2.7 REFERENCES [1] M. Green, Solar Cells, Operating Principles, Technology and System Applications. University of New South Wales. 1992. [2] M. Green, K. Emery, Y. Hishikawa, W. Warta and E. Dunlop, Solar cell efficiency tables (version 43), Progress in Photovoltaics, Volume 22, Issue 1, pages 1–9, January 2014. [3] A. Mc Evoy, T. Markvart and L. Castañer, Solar Cells. Materials, Manufacture and Operation. Elsevier. 2013. [4] A. Luque, S. Hegedus, Handbook of Photovoltaics Science and Engineering. Wiley, 2003. [5] D. King, W. Boyson and J. Kratchovil, Photovoltaic Array Performance Model. Sandia Report SAND2004-3535, Sandia National Laboratories, 2004. [6] EN 61646:208, Thin-film terrestrial photovoltaic (PV) modules. Design qualification and type approval. AENOR, 2008. [7] IEC 60891:2009, Photovoltaic devices – Procedures for temperature and irradiance corrections to measured I-V characteristics. International Electrotechnical commission, 2009. [8] D. King, J. Kratochvil, W. Boyson and W. Bower, Field experience with a new performance characterization procedure for photovoltaic arrays. Sandia National Laboratories, 1998. [9] B. Marion, A method for modeling the current-voltage curve of a PV module for outdoor conditions. Progress in Photovoltaics, Research and Applications, 2002. [10] F. Adamo, F. Attivisimo, A. Di Nisio and M. Spadavecchia, Characetrization and Testing of a Tool for Photovoltaic Panel Modeling, IEEE Transactions on Instrumentation and Measurement, Vol. 60, N. 5, May 2011. [11] L. Cristaldi, M. Faifer, M. Rossi, and S. Toscani, A Simplified Model of Photovoltaic Panel, IEEE International Instrumentation and Measurement Technology Conference (I2MTC), 2012. [12] A. Chouder, S. Silvestre, N. Sadaoui, and L. Rahmani, Modeling and simulation of a grid connected PV system based on the evaluation of main PV module parameters, Simulation Modelling Practice and Theory, Elsevier, 2012. CHAPTER 2 2-34 [13] G. Petrone, G. Spagnuolo and M. Vitelli, Analytical model of mismatched photovoltaic fields by means of Lambert W-function, Solar Energy Materials and Solar Cells 91, Elsevier, 2007. [14] S. Merino, F.J. Sánchez-Pacheco, P. Rodríguez, and C. Sánchez, Photovoltaic Cells electrical model resolution by applying Reverse Decomposition Method, Proceedings of the Applications of Computer Algebra (ACA) Congress, 2013. [15] D. Sera, R. Teodorescu and P. Rodríguez, PV panel model based on data sheet values, IEEE Int´l. Symposium on Industrial Electronics (ISIE), 2007. [16] EN 61724, Photovoltaic system performance monitoring. Guidelines for measurement, data exchange and analysis, april 2000. CENELEC. [17] S. Vergura, G. Acciani, V. Amoruso, G.E. Patrono, and Francesco Vacca, Descriptive and Inferential Statistics for Supervising and Monitoring the Operation of PV plants. IEEE Transactions on Industrial Electronics, (Vol. 56, No. 11), 2009. [18] L. Cristaldi, M. Faifer, A. Ferrero and A. Nechifor, On-line monitoring of the efficiency of Photo-Voltaic panels for optimizing maintenance scheduling. IEEE International Instrumentation and Measurement Technology Conference, 2010. [19] A. Carullo and A. Vallan, Outdoor Experimental Laboratory for Long-Term Estimation of Photovoltaic-Plant Performance, IEEE Transactions on Instrumentation and Measurement, 2012. [20] E. Román, R. Alonso, P. Ibáéz, S. Elorduizapatarietxe and D. Goitia, Intelligent PV Module for Grid-Connected PV Systems, IEEE Transactions on Industrial Electronics, 2006. [21] A. Carullo, S.Corbellini, A. Luoni, and A. Neri, In Situ Calibration of Heterogeneous Acquisition Systems: The monitoring System of a Photovoltaic Plant, IEEE Transactions on Instrumentation and measurement, (Vol. 59, No. 5), may 2010. [22] P. Guerriero, V. d’Alessandro, L. Petrazzuoli, G. Vallone, and S. Daliento, Effective Real-Time Performance Monitoring and Diagnostics of Individual Panels in PV Plants, IEEE International conference on Clean Electrical Power (ICCEP), 2013. [23] F.J. Sánchez-Pacheco, P.J. Sotorrío-Ruiz, J.R. Heredia-Larrubia, F. PérezHidalgo, M. Sidrach-de-Cardona, Low cost DC lines PLC based Photovoltaic PV CELL AND MODULE MODELS AND MONITORING 2-35 plants parameters smart monitoring communications and control module. IEEE III International Conference on Power Engineering (POWERENG 2011), Energy and Electrical Drives, 2011. [24] F.J. Sánchez-Pacheco, P.J. Sotorrío-Ruiz, J.R. Heredia-Larrubia, F. PérezHidalgo, M. Sidrach-de-Cardona, PLC-Based PV Plants Smart Monitoring System: Field Measurements and Uncertainty Estimation. IEEE Transactions on Instrumentation and Measurement (TIM), 2014. [25] B. Marion, J. del Cueto and B. Sekulic, Modeling Current-Voltage Curves Using Bilinear Interpolation. NREL/CP-520-36232, 2004. [26] C. R. Osterwald, Translation of Device Performance Measurements to Reference Conditions / Device Performance, Solar cells, 18,3-4 Pages 269-279, Elsevier, 1986. CHAPTER 3 3-6 3.2 SMCM DESIGN CONSIDERATIONS From the electronic design point of view, several considerations have been taken into account. They have conditioned the electronic design, board dimensions, operating characteristics, power consumption and cost. A. Components cost: The aim has been to develop a low cost device which might be inserted in the PV module connection box. This requirement is mandatory, since the additional cost of the installation of a monitoring system must tight as much as possible, so that to minimize the financial impact. This has led to following considerations:  Selection of low cost electronic devices, including the μController.  The coupling to the DC line has been made by capacitances, instead of magnetic devices, since these are more expensive.  The current sensor is not invasive and requires no interruption of the DC line, so no additional hand-work is required for device wiring.  The SMCM final printed circuit board dimensions must fit into the PV module connection box, so that it uses the already existing ones. Final version will be designed using Surface Mounted Technology (SMT).  For data Tx and Rx no commercial transceiver has been used, which are quiete expensive. Instead of it, a general purpose driver has been used with corresponding signal conditioning circuitry. B. Supply voltage: The SMCM is directly connected to the associated PV module which acts as its own power supply. This means that:  The SMCM cannot suppose an excessive load to the PV module in terms of consumed power.  The DC supply voltage of the SMCM must be from 0 to +12 VDC, since the voltage regulators used for internal supply can support a considerable range of input power variation (large line regulation). SMART MONITORING AND PLC COMMUNICATIONS MODULE - SMCM 3-7  Since the SMCM supply voltage is from 0 to + 12 VDC, rail-to-rail single supply operational amplifiers must be used. C. Functional configuration In order to optimize the SMCM functionality/cost, both the SMCM and the SMCM-CCS version are based on the same design. The only difference are the external elements that can be connected; that is, the PV module voltage/current for the first ones and the calibrated cell and ambient temperature for the CCS version. This philosophy allows us to use the same design and a specific components configuration for each option. D. PV back-plane temperature measurement As previously described, one of SMCM functions is the measurement of the PV back-plane temperature. Although in the prototype, a PT-100 RTD temperature sensor endorsed to the PV module has been used, in the original design, in order to reduce the components cost. The aim was to use the μController internal temperature sensor capability. For this purpose, an estimation of the temperature gradient and impedance has to be calculated from the back-plane to the SMCM μController (Figure 3.4). μController internal temperature sensor SMCM PCB PV module connection box Figure 3.4. Back-plane temperature measurement using the MSP 430 μController internal temperature sensor and thermal resistivity CHAPTER 3 3-8 E. SMCM-CCS version USB connection to the PC The CCS version of the SMCM is based on the standard SMCM and configured to perform two different functions: on one hand, it has the functionality to operate as a data node concentrator which routes all the received data to the CCS PC. Additionally, this device monitors the incident radiation given by the calibrated cell output; the ambient temperature and all this information is routed to the CCS PC for further processing and performance estimation. 3.3 SMART MONITORING AND COMMUNICATIONS MODULE STRUCTURE As it has previously described, the SMCM is the element around which, all the monitoring system has been implemented. It is endorsed to each single PV module and performs the two basic functions of monitoring the PV module operating parameters, namely the voltage and current photogenerated as well as the temperature of the back-plane and then the transmission through the DC power line, of all the data resulting from the monitoring. The SMCM is a smart device in the sense that it can support bidirectional communications. It receives in a polling sequence from the CCS, the SMCM identification code which wakes it up and in response, it delivers the monitored data according to a given data format. The detail of the SMCM structure is given in Figure 3.5. The SMCM incorporates a low cost MSP 430 Controller, specific Analog Front Ends signal conditioning for V, I and T measurement as well as an Analog Front End for Data Transmission (Tx)/Reception (Rx) through capacitive coupling. SMART MONITORING AND PLC COMMUNICATIONS MODULE - SMCM 3-9 3.3.1 MSP 430 µController. The core element of the SMCM is Texas Instruments® MSP430 Controller [5]. This µController has been selected among other ones basically because its cost (< 1 €/piece), low power consumption in standby, has an internal Analog to Digital Converter (ADC) and temperature sensor. The data bus is composed of an 8 bits general purpose I/O bus that can be user configured as digital inputs or outputs and analog inputs for further conversion to digital by the ADC. One of the I/O lines is used for Tx/Rx of data and identification codes. Its main characteristics are:  Low supply-voltage range: 1,8 to 3,6 V  Ultralow power consumption  Active mode: 220 μA at 1 MHz, 2,2V  Standby mode: 0,5 μA  Off mode (RAM retention): 0,1 μA  Five power-saving modes  Ultrafast wake-up from standby mode in less than 1 μs  16-Bit RISC architecture, 62,5 ns instruction cycle time  Clock module Figure 3.5. SMCM block diagram for V-I-T PV module parameters monitoring and Data Transmission (Tx) and Reception (Rx) PV panel (V) PV panel (I) DC POWER LINE SMCM V-I-T AFE PLC Tx/Rx AFE Data I/O 0 0 0 0 0 MSP 430 uCONTROLLER Data Tx Data Rx Data in Data out 0 0 0 0 0I/O 1 I/O 2 Cc PV panel (T) 0 0 0 0 0 I/O 3 CHAPTER 3 3-10  Internal frequencies up to 16 MHz  External digital clock source  16-Bit Timer with two capture/compare registers  Universal Serial Interface (USI) supporting SPI and I2C  Brownout detector  10-Bit 200-ksps A/D converter with internal reference, sample-and-hold and autoscan  Serial onboard programming. No external programming voltage needed  On-chip emulation logic with spy-by-wire interface All these characteristics make possible the compliance of “smart” concept. In Figure 3.6, it can be seen its internal functional block diagram. Figure 3.6. MSP 430 μController functional block diagram. SMART MONITORING AND PLC COMMUNICATIONS MODULE - SMCM 3-11 In order to speed-up the first prototype, it has been used the TI MSP 430 launch-pad development kit. The main characteristics of this kit are:  USB debugging and programming interface featuring a driverless installation and application UART  serial communication with up to 9600 Baud  Supports all MSP430G2xx and MSP430F20xx devices in PDIP14 or PDIP20 packages  Two general-purpose digital I/O pins connected to green and red LEDs for visual feedback  Two push buttons for user feedback and device reset  Easily accessible device pins for debugging purposes or as socket for adding customized extension boards  High-quality 20-pin DIP socket for an easy plug-in or removal of the target device Another of the advantages is that all the μController pins are user available and can be connected to an expansion board. Figure 3.7 shows the functional block diagram of the MSP 430 launch-pad. Figure 3.7. MSP 430 launch-pad CHAPTER 3 3-12 3.3.2 Analog Front End for PV module Voltage (VPV) signal conditioning and measurement. One of the parameters that the SMCM is due to monitor is the PV module output voltage. This parameter can range from 0 V up to VOC, according to the manufacturer data sheet with maximum values in the surroundings of 40 VDC. Since the µController ADC input can be maximum 3,6 VDC, a first conditioning must be done to convert the signal to acceptable values. The PV module output voltage (VPV) monitoring Analog Front End (AFE) block diagram is depicted in Figure 3.8. It is configured of a by 10 resistors based voltage divider. The signal is then applied to a high input impedance, unity gain voltage follower (TLC271 Opamp) and further routed to one of MSP430 analog inputs for Analog to Digital Conversion. This solution is simple, efficient and very stable from the temperature point of view. The TLC 271 Opamp has low offset voltage drift and high input impedance and manufactured with a technology that provides stable offset voltage. Figure 3.9. shows the schematic diagram of referred signal conditioning configuration. Figure 3.8. SMCM AFE for PV module Voltage monitoring block diagram monitoring. ÷10 PV module voltage divider X 1 voltage driver μC D I/O SMART MONITORING AND PLC COMMUNICATIONS MODULE - SMCM 3-13 Figure 3.10. shows the simulated circuit, where it can be seen the signal conditioning and the resulting 3 V output of the Opamp, for a PV module generating 30 V connected to the equivalent circuit of the DC power line. V3 12 V R28 100kΩ U2 TLC271ACD 3 2 6 715 4 8 R3 90kΩ R1 10kΩ IO1 8 IO1 0 2 0 0 VPVPOS VPVNEG PV panel Voltage monitoring Analog Front End VPVPOS VPVNEG Probe1,Probe1 V: 2.72 V I: 27.2 uA X3 PFV72C06V5A IO1 IO1 IO2 IO2 Cb 15uF L1 22uH 0 Rlinea 5mΩ Cl1 1pF L2 50uH Cq1 1pF Rq 6Ω 00 Cq2 1uF 0 Cl2 100pF 0 19 3 V3 12 V R28 100kΩ U2 TLC271ACD 3 2 6 7 1 5 4 8 R3 90kΩ R1 10kΩ IO1 0 2 0 0 PV panel Voltage monitoring Analog Front End 0 4 8 IO1 Probe10,Probe2 V: 30.0 V I: 5.00 A Probe1,Probe1 V: 3.00 V I: 30.0 uA Probe11,Probe3 V: 30.0 V I: 5.00 A Probe2,Probe4 V: 30.0 V I: 300 uA Probe3,Probe5 V: 3.00 V I: 6.76 pA Figure 3.9. SMCM AFE for PV module Voltage monitoring schematic diagram Figure 3.10. SMCM AFE for PV module Voltage monitoring simulated model CHAPTER 3 3-14 3.3.3 Analog Front End for IPV signal conditioning and measurement. Current (IPV) measurement requires a different solution. A non-invasive halleffect sensor configuration has been used. The output current of the PV module is sensed by this current transducer and its output is fed into a unity gain voltage follower for further connection to the µController input. Figure 3.11 shows the functional block diagram of this module. The hall-effect current sensor used is the Premo HCT06DSR5 model [6]. This is a hall-effect bidirectional current sensor, which supplies 0,625 V for a 6 A span, with an offset of 2,5 V. Since µController input voltage must be maximum 3,6 V, a voltage divider has been implemented, which multiplies voltage by 0,9. Under these conditions, in case a PV module were supplying 6 A, maximum current sensor output voltage generated would be 3,125 V; after voltage divider, the signal obtained is 2,81 V. This signal is then applied to a unity gain voltage follower amplifier based on TLC271 Opamp, which output is applied to MSP430 µController input for AD conversion. One of the characteristics of the HCT06DSR5 current sensor is that it has a 2,5 V direct voltage output reference. This signal is supplied by an internal Zener diode and hence very stable from temperature variations point of view. This output is directed to the µController input and is used as a reference for system integrity checking and even calibration. ÷1/9 Current transducer signal conditioner μC D I/O X 1 voltage driver Hall effect current transducer Figure 3.11. SMCM AFE for PV module Current monitoring block diagram SMART MONITORING AND PLC COMMUNICATIONS MODULE - SMCM 3-15 Figure 3.12 shows the schematic functional diagram of referred configuration. The schematics shows also the results of the electrical simulation. 3.3.4 Analog Front End for TPVBp signal conditioning and measurement. PV module backplane temperature (TPVBp) measurement is made using a platinum PT 100 type Resistance Temperature Detector (RTD) endorsed to the PV module back-plane. The AFE has been implemented around a standard Wheatstone bridge RTD signal conditioning and amplifier module (Figure 3.13). The configuration is designed to supply an amplifier output value of 3 V, for a PV module back-plane temperature of 100 ºC. In this case, the RTD resistor value should be 138,5 Ω. For a temperature of 0 ºC, the RTD resistance would be 100 Ω and the amplifier would deliver 0 V. V3 12 V R28 100kΩ U2 TLC271ACD 3 2 6 7 1 5 4 8 V1 0.104 Ω V2 2.5 V R3 10kΩ R1 90kΩ HCT06DSR5 Premo Hall Effect current sensor IO1 8 IO1 7 0 5 0 2 0 0 IO2 IO3 IO2 IO3 PV panel current measurement Analog Front End Probe1,Probe1 V: 2.72 V I: 27.2 uA Probe3,Probe5 V: 3.02 V I: 30.2 uA Figure 3.12. SMCM AFE for PV module Current monitoring functional diagram CHAPTER 3 3-22 The design of the transmission line has been implemented to simulate real scenarios, considering a 1000 m (≈ 3000 feet) DC cable length, with its resistive (Rline) and capacitive (Cline) equivalent impedance values. Input impedance of typical EMC filters at the input of DC/AC inverters has been considered as well. SMCM-PLC line driver has been simulated, performing the transmission of a 100 kHz square signal. The transmitter coupling to the DC line is made by C1 capacitor. DC line is coupled for this transmission rate, using two 22 µH terminal suppression coils, one at the PV panels end (L1) and the other one just before the inverters entry (L2). Data recovery is made by OP1. Its output is able to show received data at the end of the line (Figure 3.19). Following simulation results have been obtained. Figure 3.20 shows waveforms at the transmitter level (VF1), modulated signal on the DC line (VF5) as well as received signal at the end of the DC line (VF6). As it can be seen, the received simulated signal is clearly defined. Figure 3.20. Data transmitted waveforms at transmitter level (VF1), DC line modulated signal (VF5) and received signal at the end of DC line (FV6) SMART MONITORING AND PLC COMMUNICATIONS MODULE - SMCM 3-23 Figure 3.21 shows DC levels at the PV generator side (VFV), at the DC line – PLC (VF1), as well as the signal at the inverter side (VF3). Effect of L1 and L2 suppression coils can be clearly seen. AC transmitted signal components are duly filtered at the purely DC segment of the transmission line. In figure 3.22, the graphic of the Fourier analysis at the inverter level (VF3) is shown. It can be seen that magnitude of present harmonics are not significant, at the end of the PLC line (Figure 3.22). No AC component is transferred to the inverter. Figure 3.21. PLC line voltage waveforms at the DC line Figure 3.22. Fourier analysis of signal at the inverter level (VF3); Spectrum analysis Figure 3.21. Fourier analysis of signal at the inverter level (VF3). Spectrum analysis. CHAPTER 3 3-24 In Figure 3.23, THD calculation results are presented, indicating main carrier presence and no consecutive harmonics. Figure 3.24 shows the characteristic impedance at the head of the PLC transmission line. The major impedance appears at the proximity of the 100 kHz (signal carrier base frequency). Figure 3.23. Fourier analysis of signal at the inverter level (VF3). THD calculation Figure 3.24. AC transfer characteristic and impedance at the transmission point (VF1) SMART MONITORING AND PLC COMMUNICATIONS MODULE - SMCM 3-25 Figure 3.25 shows the laboratory test bench mounted to implement the DC PLC line. It simulates the line impedance to evaluate empirical data for simulation comparison. This test bench includes all components specified in electric diagram of figure 3.19. Power resistors have been necessary to dissipate all the heat generated by transferred current. As it can be seen in figure 3.26, the experimental laboratory test bench obtained results are quiete similar to the results of simulation. In the lab, spikes and noise are much more significant, although with no major effect on the receiver side, since HF filtering is made. No significant AC harmonic is detected at the DC side of the PLC line (inverter proximity). So, no negative effect is expected from inverter point of view. Figure 3.25. Laboratory test bench for PLC line testing Figure 3.26. Test bench signals. Square signal by µController (CH1). Received signal at the end of the PLC line (CH2). DC signal at the inverters input (AC filtered)(CH3) CHAPTER 3 3-26 3.4.2 Bypass capacitors for PV modules low impedance data path implementation. Once the concept of DC PLC data transmission has been seen to be viable, the following step was to validate the previous concept in the scenario of the EPS PV Lab test bench. One of the problems that arised was related to the impedance seen by the Data Tx SMCM AFE, since from the PV module up to the CCS input, a complex signal path was figured-out. The aim of ensuring the data reception required to consider alternative data paths that voided all the PV modules internal parasitic parameters influence (Rs, Rp, parasitic capacitances, etc …). The proposed solution has been to implement a bypass capacitor (Cb) in parallel with each PV module. This ensured a controlled low impedance path for the data transmitted from the SMCM, as depicted in Figure 3.27. Adjustable Resistive Load SMCM DI/O V I T Cb Cb Cc CCS L2 coil Line Trap L1 coil Line Trap PV #1 PV #2 SMCM-CCS DI/O V I T CALIBRA TED CELL USB A DC V DC Z PT-100 temp. sensor Figure 3.27. PV modules parallel low impedance data path bypass capacitors SMART MONITORING AND PLC COMMUNICATIONS MODULE - SMCM 3-27 3.4.3 EPS Renewable Energies LAB validation of simulated model and experimental data. The model depicted in Figure 3.28 shows the Multisim (National Instruments) based simulation model. This model comprises a PV module, as well as the transmission line simulation components, including the inverter input. The proposed model incorporates the 15 μF bypass capacitor that simulates the low impedance capacitive data path for the transmitted signal for a x 12 PV modules string. Figure 3.29 shows the simulated waveforms of the transmitted data (CH1), the waveform of the data on the PLC line (CH2) and finally, the recovered signal (CH4) applied to the µController. X3 PFV72C06V5A IO1 IO1 IO2 IO2 Cb 15uF L1 22uH 0 Rlinea 5mΩ Cl1 1pF L2 50uH Cq1 1pF Rq 5Ω 00 C3 1uF R7 100kΩ 0 Cq2 1uF 0 Cl2 100pF 0 V3 12 V V4 BIPOLAR_VOLTAGE R28 100kΩ U2 TLC271ACD 3 2 6 7 1 5 4 8 0 8 7 0 0 0 10 V1 12 V R2 100kΩ U1 TLC271ACD 3 2 6 7 1 5 4 8 0 50 0 XSC1 Tektronix 1 2 3 4 T G P C1 10uF R1 1kΩ 0 6 11 4 9 1 3 Figure 3.28. PLC data Tx EPS PV Lab est bench signals simulation model CHAPTER 3 3-28 For validation purpose, the configuration implemented simulates two series connected 24 V, 5 A PV modules, with its corresponding bypass capacitors (Cb). It simulates a string with maximum 1000 m (3000 feet) DC cable length, with its resistive (Rline) and capacitive (Cl1 and Cl2) equivalent impedance values. Equivalent input impedance of typical EMC filters at the input of DC/AC inverters has been considered as well. Line trap coils (L1 and L2) represent a high impedance at the data signal frequency, while for the DC power signal, this impedance is insignificant. Field measurements of generated, PLC transmitted and received data are shown in Figure 3.30. Oscilloscopes picture shows in CH1 the data generated in the PV module SMCM. CH2 shows the distorted data waveform on the PLC line. Finally, CH3 shows the data that arrives to the µController input, after corresponding conditioning. Data Tx Data waveform on PLC line Data to uController Figure 3.29. PLC data Tx simulated transmitted and received signals SMART MONITORING AND PLC COMMUNICATIONS MODULE - SMCM 3-29 3.5 SMCM GRAPHIC USER INTERFACE All single module parameters are real time processed by CCS. A Graphic User Interface (GUI) application has been developed to allow the representation of all the information generated by the SMCM. This application has been developed under National Instruments Labview® environment. Figure 3.31 shows the GUI screen shot, where each single PV modules parameters can be visualized. In the “Temperature Map” option, PV modules temperature is displayed. Meanwhile, in the “Power Map” option, the actual Maximum Power Point at which every PV module is working under present ambient conditions is displayed. This tool is useful to detect any malfunction in real time. A scaled color based display allows operator to sightsee any operation fault, alarm situations being displayed in red. The color scale is related to the percentage to the Maximum Peak Power (MPP) that under actual conditions, according to the information supplied by the calibrated cell (Gg and Ta), each single PV module should be generating. Data Tx Data waveform on PLC line Data to uController Figure 3.30. PLC data Tx EPS PV Lab test bench transmitted and received signals CHAPTER 3 3-30 If an adverse condition is applying, that affects to the PV module MPP or its temperature, the label color would be changing. In case of extreme failure, the PV module alarm label would be displayed blinking in red. This criteria applies both to the PV modules back-plane temperature and to the power generated as well. The option is operator selectable. Figure 3.32 shows the operator interface to introduce PV module parameters manufacturer´s nominal values, which will be used to determine the different values in percentage for the display options Figure 3.31. Graphic User Interface for power and temperature visualization Figure 3.32. GUI operator interface for PV module parameters insertion SMART MONITORING AND PLC COMMUNICATIONS MODULE - SMCM 3-31 3.6 CONCLUSIONS The possibility to real-time monitor the operating parameters of PV modules requires the adequate electronics to both perform this task and transmit the monitored data to the CCS. In this chapter, a detailed description is made of the SMCM hardware design and simulation of the different AFE. The AFE described are the necessary to monitor the PV modules operating parameters, namely voltage, current and module temperature and data transmission as well. For the current measurement, a non-invasive hall-effect current sensor has been used. Additionally, a customized version of the SMCM is in charge of the monitored data reception and further routing to the CCS. This data concentrator has required a specific signal conditioning electronics, due to the signals level of noise. All the components used are standard industrial devices, which cost allows to tight the monitoring system overall cost. The PLC based data transmission hardware AFE design and simulation is described and the viability demonstrated experimentally. The specifically designed low impedance data path by means of the PV modules bypass capacitors (Cb) has been demonstrated to be effective. This low cost solution also contributes to the initial objective to design a low cost system. The line trap coils are required to adapt the data transmission line impedance (in this case, the DC power line), to the digital data transmission electronics. They are only required at the head and trail of each modules string. CHAPTER 4 4-4 In order to quantify actual calibrated values, a set of experimental data have been processed. Regarding kTC, a reference voltage of 0,962 V to simulate a TPVBp of 24 ºC has been applied. For the kIC evaluation, a reference voltage of 2,494 V has been applied, which is equivalent to a IPV current of -0,057 A. Finally, for kVC estimation, a 24 V reference voltage has been applied to simulate the PV module output voltage. Results are summarized in Table 4.1. According to above results calibrated constants actual mean values result to be: kTC=24,828 ºC/V, kIC=1,10 A/V and kVC=11,885 V/V. These constants values will be further considered when measuring real data. Furthermore, once this potential systematic error has been corrected, the different contributions to the uncertainty in the measuring chains constants estimation have been evaluated. After kTC, kIC and kVC have been calculated, their contribution to the Combined Uncertainty has been quantified considering them as type A. Accordingly, a total of 10 series of 30 consecutive readings have been made under given PV module working conditions. These results obtained are summarized in Table 4.2. Table 4.1. Measuring chain calibrated constants calibration kTC (ºC/V) kIC (A/V) kVC (V/V) Nominal values: 33,33 1,11 10,00 Applied reference voltages 0,962 V @ 24 ºC 2,494 V @ -0,057 A 24,003 V @ 24 V 1 24,955 1,100 11,893 2 24,864 1,100 11,914 3 24,864 1,100 11,893 4 24,774 1,100 11,914 5 24,774 1,100 11,872 6 24,774 1,100 11,852 7 24,864 1,098 11,831 8 24,774 1,100 11,872 9 24,774 1,100 11,934 10 24,864 1,100 11,872 µ kxic 24,828 1,100 11,885 SMCM MEASURING CHAINS UNCERTAINTY ESTIMATION 4-5 4.3 MEASUREMENT UNCERTAINTY ESTIMATION DUE TO AMBIENT CONDITIONS The second contribution to the Combined Uncertainty is related to the unstable ambient conditions impacting on the PV module which might affect the readings in an unknown way. In this sense, a run of measurements has been made under known ambient conditions (G=226,63 W/m2, Ta=24,06 ºC). A total of 30 successive readings have been processed by the SMCM, at a rate of 16 ksa/s (kilosamples/second); that is, the 30 readings have been made in 1,87 ms. This fulfills the criteria that in a less than 2 ms of time frame, no ambient conditions changes are expected that might affect the PV module operation [7]. In Table 4.3 columns NTADC, NIADC and NVADC show the output of the SMCM µController AD converter (digital counts from 0 to 1023). Columns TPVBp, VPV and IPV show corresponding duly converted engineering values. Applying procedures described in [1], which stipulate that: (4.4) Table 4.2. Uncertainty estimation of measuring chains calibration constants # VAFE kVM IAFE kIM TAFE kTM 1 11,95 0,994 24,801 2 11,96 0,994 24,795 3 11,95 0,997 24,815 4 11,98 1,036 24,822 5 12,14 1,002 24,830 6 12,05 1,001 24,827 7 11,94 1,015 24,800 8 11,95 0,999 24,832 9 11,96 1,002 24,829 10 11,98 1,012 24,835 µ 11,98 1,005 24,819 Var 0,003453 0,000149 0,000188 Std. Dev 0,062 0,013 0,015 Sens. Coeff 1,000 1,000 1,000 divisor 1,000 1,000 1,000 uC (kXiM) ±0,062 V ±0,013 A ±0,015 ºC     iii XSdevScXuR  CHAPTER 4 4-6 and considering that Sci is the sensitivity coefficient (Sci=1) and Sdev(xi) is the Standard deviation of measured data, resulting uncertainty values are as follows: uR(VPV)=±0,442 V, uR(IPV)=±0,013 A and uR(TPVBp)=±0,095 ºC. Table 4.3. Stable conditions consecutive readings (G=226,63 W/m2 , Ta= 24,06 ºC ) # NTADC NIADC NVADC TPVBp (ºC) VPV (V) IPV (A) 1 257 663 799 30,07 33,871 0,485 2 256 662 795 29,95 33,701 0,466 3 257 662 789 30,07 33,447 0,466 4 256 662 777 29,95 32,938 0,466 5 257 663 774 30,07 32,811 0,485 6 257 663 772 30,07 32,726 0,485 7 260 662 771 30,42 32,684 0,466 8 257 663 774 30,07 32,811 0,485 9 257 662 782 30,07 33,150 0,466 10 256 662 790 29,95 33,489 0,466 11 256 662 799 29,95 33,871 0,466 12 257 663 800 30,07 33,913 0,485 13 257 663 796 30,07 33,744 0,485 14 256 662 788 29,95 33,404 0,466 15 257 664 780 30,07 33,065 0,504 16 256 663 774 29,95 32,811 0,485 17 257 662 771 30,07 32,684 0,466 18 257 663 772 30,07 32,726 0,485 19 256 663 775 29,95 32,853 0,485 20 256 663 782 29,95 33,150 0,485 21 256 662 790 29,95 33,489 0,466 22 257 664 798 30,07 33,828 0,504 23 256 662 798 29,95 33,828 0,466 24 256 663 795 29,95 33,701 0,485 25 256 663 787 29,95 33,362 0,485 26 256 664 777 29,95 32,938 0,504 27 257 663 774 30,07 32,811 0,485 28 257 664 773 30,07 32,769 0,504 29 256 662 772 29,95 32,726 0,466 30 256 663 776 29,95 32,896 0,485 µ 30,022 33,204 0,480 Var 0,003285 0,19684 0,000174 Std. Dev 0,095 0,442 0,013 Sens.Coeff CCoef 1,000 1,000 1,000 divisor 1,000 1,000 1,000 uR(xi) ±0,095ºC ±0,442V ±0,013A SMCM MEASURING CHAINS UNCERTAINTY ESTIMATION 4-7 4.4 MEASURING CHAINS ELECTRONIC COMPONENTS DRIFT UNCERTAINTY ESTIMATION The third contribution to the Combined Uncertainty is due to the measuring chains AFE electronic components drift. In this case, the standard uncertainty related to the electronic components drift u(D) contribution to Combined SMCM measurement Uncertainty is quantified considering [1] as follows: (4.5) where Drift(xi) is each measuring chain quantified drift, considering they have a uniform distribution and hence the divisor is √3 (= 1,73) and Sci are corresponding sensitivity coefficient factors. This approach is made separately for the three measuring chains. In all of them, one of the contributions is due to the µController internal Analog to Digital Converter (ADC). The term that globally quantifies the ADC overall uncertainty contribution is the so defined Total Unadjusted Error (TUE) and is related to the ADC Vq quantization voltage. According to the Controller manufacturer´s [6] data sheets, this TUE is quantified to be ±5LSB. In this case, referred TUE is ±17,6 mV, which represents an uncertainty of ±0,49 % of the 3,6 V µController maximum input voltage. The TLC 271 Opamp generates an uncertainty due to the input offset voltage thermal drift (  VIO), quantified to be 2 V/ºC, which represents a maximum drift of 0,1 mV in an operating temperature range of 50 ºC. This gives a ±0,01 % of drift in the considered full scale (FSOA) input voltage of 3,6 V. Total derate in the VAFE is Drift(VAFE)=±0,4983 %, hence, the resulting uncertainty is uD(VAFE)=Drift(VAFE)/√3=±0,29 % or ±0,107 V related to the 37,3 V PV module open circuit voltage (Voc). The hall effect current sensor has an offset voltage temperature drift of ±0,5 mV/ºC. This represents an uncertainty of ±0,75 % of full scale. Considering also       3/ iii xDriftScxuD  CHAPTER 4 4-8 the ±0,01 % of the Opamp, and ±0,488 % of the ADC, this results in Drift(IAFE)=1,2983 % that represents an estimated components drift uncertainty uD(IAFE)=Drift(IAFE)/√3=±0,75 %, or ±0,065 A of the 8,7 A PV module short circuit current (Isc). Regarding the PV back-plane temperature measurement AFE, the contribution of the PT100 RTD sensor is due to its temperature drift to be of ±0,8 % in the range of 100 ºC. This, added to the OA amplifier drift (±0,01 %) and ADC drift (±0,488%), gives a total of Drift (TAFE)=±1,6783 % which represents an uncertainty uD(TAFE)=Drift (TAFE)/√3=0,97 % that corresponds to ±0,242 ºC referred to STC Tm=25 ºC. These data are summarized in Table 4.4. Taking into account previous data, combined and expanded standard uncertainty can be quantified for the three measuring chains. Following the GUM recommendations [1], this can be expressed as: Table 4.4. Uncertainty estimation of electronic components drift VAFE components drifts uncertainty Total drift Voltage AFE (±V) 0,1858 Sensitiviy coefficient (Sc) 1 Divisor (√3) 1,7321 Voltage AFE uD(VPV) ±0,107 V IAFE components drifts uncertainty Total drift Current AFE (±A) 0,1651 Sensitiviy coefficient (Sc) 1 Divisor (√3) 1,7321 Current AFE uD(IPV) ±0,065 A TAFE components drifts uncertainty Total drift RTD temp sensor (±ºC) 0,324 Sensitiviy coefficient (Sc) 1 Divisor (√3) 1,7321 Temp. AFE uD(TPVBp) ±0,24 ºC SMCM MEASURING CHAINS UNCERTAINTY ESTIMATION 4-9 (4.6) Furthermore, it follows that the expanded uncertainty is given as: (4.7) where k is the coverage factor. In this case, with k=2, a confidence level of 95 % can be considered. Table 4.5 summarizes obtained results of combined and expanded uncertainty values. Table 4.5. Uncertainty estimation of voltage, current and temperature measuring chains Source Type Value Divis. Sens Coeff Standard Uncertainty uR(VPV) A 0,442 1 1 0,442 uC(VPV) A 0,062 1 1 0,062 uD(VPV) B 0,107 1,73 1 0,062 combined uncertainty coverage factor Voltage AFE expanded uncertainty (±V) 0,451 2,000 0,901 uR(IPV) A 0,013 1 1 0,013 uC(IPV) A 0,013 1 1 0,013 uD(IPV) B 0,065 1,73 1 0,038 combined uncertainty coverage factor Current AFE expanded uncertainty (±A) 0,042 2,000 0,084 uR(TVPVBp) A 0,095 1 1 0,095 uC(TPVBp) A 0,015 1 1 0,015 uD(TVPVBp) B 0,242 1,73 1 0,140 combined uncertainty coverage factor PV BP Temp. expanded uncertainty (±ºC) 0,170 2,000 0,340           2 1 222     ii M ic xuDxuRkxuCyu   yukU c  CHAPTER 4 4-10 4.5 CONCLUSIONS As an electronic measurement device, the SMCM is not exempt of uncertainty in the results delivered. In this sense, applying the recommendations of the GUM of the BIPM, it has quantified its uncertainty of both A type and B type, due to 3 different contributions. On one hand, the contribution of the calibrated constants has been quantified as well as the related to the unstability of the applying ambient conditions; being both A type. Additionally, the contribution to the electronic components drift has also been quantified, being this B type. It has resulted that the different contributions give an uncertainty of ±0.9 V in the case of the voltage measurement AFE; ±0.08 A for the current measurement AFE and of ±0.34 ºC in the case of the PV module temperature measurement AFE. It can be seen that the higher uncertainty is related to the voltage measurement, mainly due to the stability of the ambient conditions, together with the noise present in the PV module voltage signal. SMCM MEASURING CHAINS UNCERTAINTY ESTIMATION 4-11 4.6 REFERENCES [1] BIPM. Bureau International des Poids et Mesures, Evaluation of measurement Data – Guide to the expression of uncertainty in measurement JCGM 100:2008, corrected version 2010. [2] A. Giordani and L. Mari, Measurement, Models and Uncertainty. IEEE Transactions on Instrumentation and Measurement, Vol. 6, NO. 8, August 2012. [3] A. Ferrero and S. Salicone, Measurement Uncertainty. Part 8 in a series of tutorials in instrumentation and measurement. IEEE Instrumentation & Measurement Magazine, June 2006. [4] BIPM. Bureau International des Poids et Mesures, International Vocabulary of metrology – Basic and general concepts and associated terms (VIM), 3rd edition, JCGM 200:2012. [5] W. Bich, From Errors to Probability Density Functions. Evolution of the Concept of Measurement Uncertainty. IEEE Transactions on Instrumentation and Measurement, Vol. 61, No.8, August 2012. [6] A. Ferrero, M. Lazzaroni and S. Salicone, A Calibration Procedure for a Digital Instrument for Electric Power Quality Measurement. IEEE Transactions on Instrumentation and Measurement, 2002. [7] M. Seapan, C. Limsakul, T. Chayavanich, K. Kirtikara, N. Chayavanich and D. Chenvidhya. Effects of dynamic parameters on measurements of IV curve. IEEE 33rd PVSC Photovoltaic Specialists Conference, 2008. CHAPTER 4 4-12 Chapter 5 MODEL VALIDATION WITH EXPERIMENTAL DATA In this chapter, the validation of the proposed mathematical model, together with the translation criteria is made on the basis of consolidated field experimental data obtained in the UMA RREE Laboratory. Both are deployed and the comparison has been made between the expected values obtained with the proposed model and the real ones. The results have allowed to quantify the maximum Relative Error that can be generated and the conditions in which it happens. CHAPTER 5 5-8 Table 5.2. One clear sky day long proposed model theoretical and experimentally measured PV modules operating parameters. Relative Error of most significant electrical parameters M1 M2 M3 M4 M5 M6 M7 M8 M9 M10 Time 9:11 9:51 11:06 12:26 13:36 14:36 15:31 16:41 17:46 18:31 Tm (ºC) 14,5 19,7 24,2 31,4 33 35,4 35,3 28,8 23,9 18,5 G (W/m2) 308 502 793 985 1039 992 878 635 344 118 ISC (A) translated 2,5 4,1 6,6 8,2 8,6 8,3 7,3 5,3 2,8 1,0 measured 2,3 4,0 6,4 8,0 8,5 8,1 7,1 5,1 2,7 0,8 RE (%) 8,5 3,8 2,0 1,8 1,9 2,4 2,6 2,9 6,8 19,2 VOC (V) translated 33,4 32,8 32,3 31,5 31,4 31,1 31,1 31,8 32,4 33,1 measured 32,2 32,4 32,4 32,2 32,1 32,1 32,1 32,0 32,0 30,1 RE (%) 3,9 1,2 0,4 2,0 2,4 3,1 2,9 0,6 1,3 9,9 Pmp (W) translated 60,7 97,6 152,2 185,5 194,9 184,9 163,8 120,5 66,2 23,1 measured 57,5 97,6 153,2 182,8 190,5 180,1 160,4 120,5 64,0 18,5 RE (%) 5,6 0,0 0,6 1,5 2,3 2,6 2,1 0,0 3,5 24,8 Vmp (V) translated 26,7 26,3 25,8 25,2 25,1 24,9 24,9 25,5 25,9 26,5 measured 26,8 26,3 25,7 24,9 24,6 24,3 24,4 25,5 25,9 25,1 RE (%) 0,3 0,3 0,6 1,3 1,9 2,2 1,9 0,1 0,0 5,5 Imp (A) translated 2,3 3,8 6,0 7,4 7,9 7,5 6,6 4,8 2,6 0,9 measured 2,1 3,7 6,0 7,3 7,7 7,4 6,6 4,7 2,5 0,7 RE (%) 7,1 1,5 0,1 1,3 1,5 1,5 1,3 1,2 4,7 19,8 MODEL VALIDATION WITH EXPERIMENTAL DATA 5-9 From Table 5.2, it can be seen that most significant Relative Errors values appear with low values of incident irradiance for ISC, VOC and mainly for Pmp and associated Vmp and Imp. Figure 5.7 gives an overview of the RE for all the operating parameters vs the evolution of the incident irradiance on the PV module. As it can be seen, the RE has bigger values at low irradiance levels, while in the mid ranges, the RE becomes much smaller. Regarding the maximum power point (Pmp) and its coordinates, the maximum power point current (Imp) and the maximum power point voltage (Vmp), following figures depict in detail the evolution of referred parameters vs. G in the range of values of the outdoor measurements made. Regarding the maximum power parameter (Pmp), the RE (clear gray line) is kept < 3,5 % for irradiances > 340 W/m2, as seen in Figure 5.8. Figure. 5.7. General overview of the parameters Relative Errors, vs G 0 200 400 600 800 1000 0,0 2,0 4,0 6,0 8,0 10,0 12,0 14,0 16,0 18,0 20,0 9:11 9:51 11:06 12:26 13:36 14:36 15:31 16:41 17:46 18:31 Irradiance G (W/m2) Relative Error (%) Measuremnt time hh:mm Translated vs Measured parameters Relative Error (%) ISC RE VOC RE Pmp RE Vmp RE Imp RE G (W/m2) CHAPTER 5 5-10 Figure 5.9 represents the evolution of the Vmp RE vs the irradiance. It can be seen a small increase of the RE in low range values, but in any case, this RE is kept in the boundaries of the 2 %. Figure 5.8. Evolution of Pm RE vs G Figure 5.9. Evolution of the Vmp RE vs G values 0 200 400 600 800 1000 0,0 2,0 4,0 6,0 8,0 10,0 12,0 14,0 16,0 18,0 20,0 9:11 9:51 11:06 12:26 13:36 14:36 15:31 16:41 17:46 18:31 Irradiance G(W/m2) Relative Error (%) Measuremnt time hh:mm Translated vs Measured parameters Relative Error (%) Pmp RE G (W/m2) 0 200 400 600 800 1000 0,0 2,0 4,0 6,0 8,0 10,0 12,0 14,0 16,0 18,0 20,0 9:11 9:51 11:06 12:26 13:36 14:36 15:31 16:41 17:46 18:31 Irradiance G(W/m2) Relative Error (%) Measuremnt time hh:mm Translated vs Measured parameters Relative Error (%) Vmp RE G (W/m2) MODEL VALIDATION WITH EXPERIMENTAL DATA 5-11 Regarding Imp, Figure 5.10 shows the evolution of the complementary parameter, namely the Imp vs G. In the high range values, the RE is in the surrounding of the 1,5 %. In order to complete the detailed analysis, the graphs corresponding to the evolution of VOC and ISC operating parameters RE vs G are depicted in figures 5.11 and 5.12 respectively. Figure 5.10. Evolution of the Imp RE vs G values Figure 5.11. Evolution of the VOC RE vs G values 0 200 400 600 800 1000 0,0 5,0 10,0 15,0 20,0 25,0 9:11 9:51 11:06 12:26 13:36 14:36 15:31 16:41 17:46 18:31 Irradiance G(W/m2) Relative Error (%) Measuremnt time hh:mm Translated vs Measured parameters Relative Error (%) Imp RE G (W/m2) 0 200 400 600 800 1000 0,0 5,0 10,0 15,0 20,0 9:11 9:51 11:06 12:26 13:36 14:36 15:31 16:41 17:46 18:31 Irradiance G(W/m2) Relative Error (%) Measuremnt time hh:mm Translated vs Measured parameters Relative Error (%) VOC RE G (W/m2) CHAPTER 5 5-12 Putting the focus on the incident irradiance boundaries < 200 W/m2, and specifically on Pmp and associated Vmp and Imp associated RE, in figure 5.13, it can be seen that the major contribution to the Pmp RE comes from the Imp RE. For incident irradiances values > 200 W/m2, the RE is kept < 2,2 % and 1,5 % respectively for Vmp and Imp. Figure 5.12. Evolution of the ISC RE vs G values 0 200 400 600 800 1000 0,0 5,0 10,0 15,0 20,0 9:11 9:51 11:06 12:26 13:36 14:36 15:31 16:41 17:46 18:31 Irradiance G(W/m2) Relative Error (%) Measuremnt time hh:mm Translated vs Measured parameters Relative Error (%) ISC RE G (W/m2) 0 200 400 600 800 1000 0,0 2,0 4,0 6,0 8,0 10,0 12,0 14,0 16,0 18,0 20,0 9:11 9:51 11:06 12:26 13:36 14:36 15:31 16:41 17:46 18:31 Irradiance G(W/m2) Relative Error (%) Measuremnt time hh:mm Translated vs Measured parameters Relative Error (%) Pmp RE Vmp RE Imp RE G (W/m2) Figure 5.13. Evolution of the Pmp, Vmp and Imp RE vs G values MODEL VALIDATION WITH EXPERIMENTAL DATA 5-13 5.3 CONCLUSIONS In this chapter, the proposed translation model has been validated with consolidated experimental data obtained from the UMA RREE Lab monitored modules. These results highlight the magnitude of the proposed model generated Relative Error that acquires significant values in the low irradiation frames, roughly by < 200 W/m2, which happens mainly at the beginning and the end of the day. Nevertheless, these errors are reduced significantly in the center day-time measurements, corresponding to higher values of irradiance (> 200 W/m2). From a practical point of view, this is the irradiance value frame that in fact has to be considered, since the central hours of the day are the ones where the PV modules are generating the maximum power. The results allow us to consider the model as reliable for the purpose of identifying a defective module performing under below the predicted values as per the proposed model. Regarding the PV module maximum power generated Pmp, in comparison with the estimated modeled one, the maximum RE generated in the center hours is of 3,5 %. This means that all the measured values with an error above this one might be interpreted as an abnormal operation of corresponding PV module. Similar conclusions can be extracted from the other parameters considered. CHAPTER 5 5-14 Chapter 6 MODEL APPLIED TO SMCM MONITORED DATA In this chapter, the proposed PV module model, after its validation, as described previously, is applied to the SMCM experimentally monitored PV modules parameters. The results and analysis are reported. This demonstrates the functionality of the SMCM as a PV module parameters monitoring and data transmission device. The results allow an evaluation of the real behavior of the PV modules under applying outdoor conditions and its comparison with the optimal results that should be obtained. CHAPTER 6 6-2 6.1 INTRODUCTION The PV modules operational parameters based on mathematical model described and validated in previous chapters are now compared with the real-time data of referred PV modules and applying outdoor conditions, obtained by means of the SMCM based monitoring system. This procedure has been applied to the PV modules test assembly installed in the Málaga University Escuela Politécnica Superior Photovoltaic Laboratory (EPS PV Lab). The aim is to establish the comparison between the modelled parameters and the SMCM monitored data and according to the obtained results, evaluate the effective performance of the PV modules working under referred conditions. This procedure has been applied to different measurements in order to validate it, which are described in this chapter. It will be demonstrated the viability of the application to quantify the PV modules operation. 6.2 EPS PHOTOVOLTAIC LABORATORY TEST ASSEMBLY DESCRIPTION In order to validate the proposed idea of monitoring and data comparison with corresponding model, a test bench assembly has been deployed in the EPS PV Lab, comprising the PV modules, the SMCM monitoring devices, together with the additional electronic devices required for the proper operation, like the PC Central Computer System (CCS) an electronic load and laboratory measuring instruments. The test assembly layout is depicted in Figure 6.1, where it can be seen the layout of the PV modules, calibrated cell, temperature sensors, power wiring and data path. One SMCM module has been used for data monitoring and transmission of one of the PV modules and a SMCM-CCS for data reception and further routing to the PC CCS via USB. MODEL APPLIED TO SMCM MONITORED DATA 6-3 The EPS PV Lab prototype assembled is comprised of two Isofotón ISF-245 PV modules, with a PT100 RTD temperature sensor for back-plane PV module temperature monitoring (Figures 6.2 and 6.3). The effective incident irradiance is monitored by means of an Atersa calibrated photo cell and an additional Pt100 RTD is used for ambient temperature measurement. The cell has been assembled in the plane of the PV modules, so that to obtain the effective irradiance incident on their plane. This allows to have incident irradiance data compensated with tilt and orientation angles, ambient temperature and air mass index. The SMCM and SMCM-CCS are connected to the PV modules, the DC power line and the CCS computer. The whole PV modules set has been assembled on a metallic structure, south oriented in order to comply with the optimal tilt angle, corresponding to Málaga latitude, that is of 36º as it can be seen in Figure 6.2, so as to ensure that the PV modules are able to deliver the maximum power depending on applying ambient conditions. Adjustable Resistive Load SMCM DI/O V I T Cb Cb Cc CCS L2 coil Line Trap L1 coil Line Trap PV #1 PV #2 SMCM-CCS DI/O V I T CALIBRA TED CELL USB A DC V DC Z PT-100 temp. sensor Figure 6.1. EPS PV Lab PV test bench layout schematic CHAPTER 6 6-10 current AFE (IAFE) measuring chain calibrated constant (kIC=1,1 A/A). Furthermore, kISENS is the current sensor calibrated constant (kISENS =9,6 A/V); in this case, divided by 2, since the current sensor has been connected with two loops of the PV module wire, so that to have more reliable values, as indicated in Figure 6.8. According to the measured module operational parameters (V and I), the resulting power output of the PV module has consequently been obtained by equation (6.3): (6.3) Finally, the module temperature is obtained by following equation: (6.4) where NTADC are the digital counts resulting from the ADC of the module temperature sensor; Vq is the ADC quantization voltage (Vq=3,52 mV) and kTC is the Temperature Analog Front End (TAFE) measuring chain calibrated constant (kTC=24,828 V/V). Figure 6.8. PREMO HCT06DSR5 Hall effect current set-up for non-invasive current measurement. IVP  TCqADCm kVNTT MODEL APPLIED TO SMCM MONITORED DATA 6-11 6.4 EXPERIMENTAL DATA OBTAINED WITH ISOFOTON PV MODULES The first validation approach has been performed based on field measurements obtained in the EPS PV Lab experimental test bench, where the SMCM has been connected to one of the PV modules for V, I and Tm measurement. Additionally, a SMCM-CCS has been installed which has been in charge of the ambient incident effective irradiance monitoring, G (from the Atersa calibrated cell), the ambient temperature, Ta (by means of a Pt100 RTD) and to the CCS as a data concentrator. In a first attempt, a total of 11 runs of 30 consecutive measurements each have been performed. This has allowed to have field data of the PV module operational parameters, under the applying outdoor conditions. For each of the runs, the 30 measurements have been tabulated and the mean and standard deviation values have been calculated. The resulting mean values and outdoor conditions are summarized in Table 6.4 and depicted in the graph of Figure 6.9 which represents the results in a similar way to the typical I-V and P-V curves, but constructed with the mean values of the 11 measurements. This justifies the lack of continuity of the curves. Despite the resulting graph, it can be seen that the procedure is capable of generating reliable data of the PV module operation. Table 6.4. Mean values, STD deviation and outdoor conditions of readings # G (W/m2) Ta (ºC) Tm (ºC) V (V) I (A) P (W) Tm STD V STD I STD P STD 1 197,3 23,4 28,0 6,8 0,92 6,32 0,06 0,42 0,02 0,40 2 201,4 23,4 31,3 12,6 0,89 11,17 0,07 0,61 0,02 0,64 3 223,2 23,8 31,7 12,8 0,92 11,78 0,07 0,64 0,01 0,58 4 199,3 23,5 33,2 20,0 0,84 16,91 0,06 0,59 0,01 0,50 5 217,4 23,6 33,1 27,8 0,85 23,56 0,06 0,47 0,02 0,68 6 274,4 24,1 33,1 31,5 0,69 21,74 0,06 0,45 0,02 0,52 7 229,0 23,1 33,2 32,1 0,55 17,77 0,07 0,42 0,01 0,45 8 203,4 24,1 33,1 32,2 0,46 14,90 0,06 0,44 0,02 0,54 9 200,4 24,0 33,0 32,4 0,39 12,54 0,08 0,44 0,02 0,63 10 204,1 24,0 33,0 32,6 0,35 11,28 0,10 0,43 0,01 0,47 11 226,6 24,1 33,0 32,7 0,32 10,40 0,07 0,44 0,01 0,42 CHAPTER 6 6-12 On the basis of previously indicated equations and applying corresponding current and voltage measurement chains coefficients and constants, the resulting values of the PV module monitored voltage, current and temperature are described in Table 6.5. This table corresponds to the 30 measurements of the run #5, which is the one that has allowed to obtain the bigger value of output power, under applying outdoor conditions, namely a sun irradiance of 217 W/m2 and a module temperature of 33,1 ºC. Column NIADC shows the digital counts resulting from the PV module current measurement ADC process; NVDC correspond to the PV module voltage AD conversion; T refers to the module temperature; V column represents the converted resulting values of the PV module voltage; I column corresponds to the measured currents, while PPV is the resulting power obtained by the product of voltage and current as specified by equation (6.3). MODEL APPLIED TO SMCM MONITORED DATA 6-13 Table 6.5. Results of run #5 (maximum output at G=217 W/m2 and Tm=33,1ºC) meas # NTADC NIADC NVADC Tm (ºC) V (V) I (A) P (W) 1 379 691 667 33,1 27,9 0,84 23,5 2 379 691 652 33,1 27,3 0,84 23,0 3 379 691 652 33,1 27,3 0,84 23,0 4 379 691 662 33,1 27,7 0,84 23,3 5 380 692 677 33,2 28,3 0,86 24,4 6 379 692 682 33,1 28,5 0,86 24,6 7 379 692 660 33,1 27,6 0,86 23,8 8 378 691 652 33,0 27,3 0,84 23,0 9 380 692 655 33,2 27,4 0,86 23,6 10 380 692 668 33,2 27,9 0,86 24,1 11 380 693 682 33,2 28,5 0,88 25,1 12 379 692 672 33,1 28,1 0,86 24,2 13 379 691 652 33,1 27,3 0,84 23,0 14 379 691 650 33,1 27,2 0,84 22,9 15 379 691 660 33,1 27,6 0,84 23,3 16 379 692 675 33,1 28,2 0,86 24,3 17 380 692 683 33,2 28,6 0,86 24,6 18 379 692 666 33,1 27,9 0,86 24,0 19 380 691 651 33,2 27,2 0,84 22,9 20 379 692 653 33,1 27,3 0,86 23,5 21 380 692 664 33,2 27,8 0,86 23,9 22 380 691 679 33,2 28,4 0,84 23,9 23 379 691 678 33,1 28,4 0,84 23,9 24 380 692 659 33,2 27,6 0,86 23,7 25 379 691 652 33,1 27,3 0,84 23,0 26 379 691 654 33,1 27,4 0,84 23,1 27 377 688 667 32,9 27,9 0,79 22,0 28 379 691 678 33,1 28,4 0,84 23,9 29 378 690 672 33,0 28,1 0,82 23,2 30 379 690 652 33,1 27,3 0,82 22,5 Mean values 33,1 27,8 0,85 23,6 Standard Deviation 0,06 0,47 0,02 0,68 CHAPTER 6 6-14 Additionally, several different cases are described. The I-V and P-V curves have been obtained in a single run under referred outdoor conditions. In the corresponding graphs, for comparison purposes, the STC curve, the ROC translated one and the corresponding to the instant readings are represented. The first case analyzed corresponds to a measurement made at G=810 W/m2 and Tm=43,9 ºC. Figure 6.9 shows the I-V at STC (blue solid line), translated to ROC (red dashed line) and read (green dotted line). Figure 6.10 depicts corresponding P-V curve. It can be seen that the read results follow the translated model. Figure 6.9. STC, translated and real measured I-V curves of VI_2 reading at G=810 W/m2 and Tm = 43,9 ºC 0,00 1,00 2,00 3,00 4,00 5,00 6,00 7,00 8,00 9,00 0,00 5,00 10,00 15,00 20,00 25,00 30,00 35,00 40,00 PV module I(A) PV module V(V) I-V VI_2 stc,translated & real I-V real 1 I-V tras1 I-V stc 0,00 50,00 100,00 150,00 200,00 250,00 0,00 5,00 10,00 15,00 20,00 25,00 30,00 35,00 40,00 PV module P(W) PV module V(V) P-V VI_2 stc,translated &real P STC P tras Preal Figure 6.10. STC, translated and real measured P-V curves of VI_2 reading at G=810 W/m2 and Tm = 43,9 ºC MODEL APPLIED TO SMCM MONITORED DATA 6-15 In the second case analyzed (VI_3), which corresponds to a measurement made at G=820 W/m2 and Tm=46,6 ºC, it will be seen the effect of partial shadowing at the boundaries of the 25 V. Figure 6.11 shows the I-V at STC (blue solid line), translated to ROC (red dashed line) and read (green dotted line). This shadowing seems to have reduced the Pmp and hence the module temperature at that point, which has made the open circuit voltage (VOC) to result bigger than expected according to the translated curve for given outdoor conditions. In Figure 6.12 corresponding P-V STC, translated and real curves are depicted. 0,00 1,00 2,00 3,00 4,00 5,00 6,00 7,00 8,00 9,00 0,00 5,00 10,00 15,00 20,00 25,00 30,00 35,00 40,00 PV module I(A) PV module V(V) I-V stc,translated &real I-V real 1 I-V tras1 I-V stc Figure 6.11. STC, translated and real I-V curves of VI_3 reading at G=820 W/m2 and Tm= 46,6 ºC Figure 6.12. STC, translated and real P-V curves of VI_3 reading at G=820 W/m2 and Tm = 46,6 ºC 0,00 50,00 100,00 150,00 200,00 250,00 0,00 5,00 10,00 15,00 20,00 25,00 30,00 35,00 40,00 PV module P(W) PV module V(V) P-V stc,translated &real P STC P tras Preal CHAPTER 6 6-16 In the third case analyzed (VI_4), the ambient conditions were G=750 W/m2 and Tm=48,2 ºC. The curves show a behavior of the module, subject to a partial shadowing and the negative effect of the ambient temperature, which results in a significant reduction in the VOC read, vs the estimated translated one. This can be seen in Figures 6.13 (I-V curves) and 6.14 (P-V curves). 0,00 1,00 2,00 3,00 4,00 5,00 6,00 7,00 8,00 9,00 0,00 5,00 10,00 15,00 20,00 25,00 30,00 35,00 40,00 PV module I(A) PV module V(V) I-V stc,translated &real I-V real 1 I-V tras1 I-V stc Figure 6.13. STC, translated and real time measured I-V curves of VI_4 reading at G=750 W/m2 and Tm= 48,2 ºC 0,00 50,00 100,00 150,00 200,00 250,00 0,00 5,00 10,00 15,00 20,00 25,00 30,00 35,00 40,00 PV module P(W) PV module V(V) P-V stc,translated &real P STC P tras Preal Figure 6.14. STC, translated and real time measured P-V curve of VI_4 reading at G=750 W/m2 and Tm = 48,2 ºC MODEL APPLIED TO SMCM MONITORED DATA 6-17 Finally, the fourth case analyzed corresponds to a low irradiance G= 720W/m2 and a module temperature of 42,3 ºC. These conditions result in a value of ISCt=6.32 A) and a value of VOCt=34.9 V and consequently, the maximum power point results to be Pmpt=166,6 W. In Figure 6.15, both the STC, translated and real measured I-V curves are depicted. There it can be seen that the STC and translated curves are correlated. In the boundaries of VOC=7 V, the graph shows the effect of a partial shadowing of the module, which significantly reduces the resulting current. Corresponding P-V curves are drawn in figure 6.16. Figure 6.15. VI_1 STC, translated and measured data I-V curves at G=720 W/m2 and a module temperature Tm=42,3 ºC. 0,00 1,00 2,00 3,00 4,00 5,00 6,00 7,00 8,00 9,00 0,00 5,00 10,00 15,00 20,00 25,00 30,00 35,00 40,00 PV module I(A) PV module V(V) I-V stc,translated &real I-V real 1 I-V tras1 I-V stc Figure 6.16. VI_1 STC, translated and measured data P-V curves at G=720 W/m2 and a module temperature Tm=42,3 ºC. 0,00 50,00 100,00 150,00 200,00 250,00 0,00 5,00 10,00 15,00 20,00 25,00 30,00 35,00 40,00 PV module P(W) PV module V(V) P-V stc,translated &real P STC P tras Preal CHAPTER 6 6-18 6.5 CONCLUSIONS The aim of described proposed SMCM and SMCM-CCS as a PV modules level monitoring system was to have the capability of real time parameters monitoring and further processing. In this chapter, the SMCM functionality has been demonstrated for data monitoring, which has made possible to get the PV module real time I-V and P-V fingerprint. This evidence has been additionally confirmed in a controlled laboratory environment with laboratory instrumentation and electronic loads and a specifically developed monitoring and data processing software. The former challenge of proposed device is its capability to measure real time I, V and Tm values of the PV module under applying ROC, in large dimensions PV fields. The scenario in this case is significantly different. The maximum power point in which the PV module is working will depend, as previously stated on the incident irradiance and on the ambient temperature on a first step and subsequently on other factors and ambient conditions. The specific coordinates of the PV module maximum power point power will finally depend on the load faced by the PV modules assembly. Since the system is intended to be working in grid connected large PV facilities, it is considered to be working in the MPP. Hence, the initial key factors that define this point are the ambient conditions related previously. If these conditions are met, the PV module should be working properly and the maximum power expected to be obtained. It is obvious the necessity of SMCM monitored PV modules operating parameters. Its comparison with the translated values, by means of proposed translation model helps to identify if it is working properly. Under these premises, the SMCM system is a resourceful tool for failing PV modules real-time identification. If the measured generated power is abnormally smaller than the expected one, the corrective actions must be undertaken. This technology grants the financial operation of the whole plant, since abnormally operating modules can be immediately identified and maintenance activities can be significantly improved. This is a required way to ensure the optimal financial performance of the whole PV plant. Chapter 7 PV MODULES PERFORMANCE RATIO ANALYSIS Once it has been demonstrated the possibility to monitor the operational parameters at PV module level, the main advantage of proposed technology is the possibility to have real-time information of their performance, under applying operating conditions. The Performance Ratio of a PV plant, so far, has been related to the energy it generates and the maximum energy it might supply if it were working under STC conditions. In this chapter, a specific instant Performance Ratio related to the power generated by a single PV module will be defined. This parameter gives us the power losses of the PV module, in relation with the power they might deliver under the so called standard conditions.