Technical and Economic Assessment of a 500W Autonomous Photovoltaic System with LiFePO 4 Battery Storage João Filipe Esteves Carriço Thesis to obtain the Master of Science Degree in Electrical and Computer Engineering Supervisor: Prof. José Paulo da Costa Branco Examination Committee Chairperson: Prof. Rui Manuel Gameiro de Castro Supervisor: Prof. José Paulo da Costa Branco Member of Committee: Prof. João José Esteves Santana Prof. Carlos Alberto Ferreira Fernandes November 2015
Transcript
Technical and Economic Assessment of a 500W
Autonomous Photovoltaic System with LiFePO4 Battery
Storage
João Filipe Esteves Carriço
Thesis to obtain the Master of Science Degree in
Electrical and Computer Engineering
Supervisor: Prof. José Paulo da Costa Branco
Examination Committee Chairperson: Prof. Rui Manuel Gameiro de Castro
Supervisor: Prof. José Paulo da Costa Branco
Member of Committee: Prof. João José Esteves Santana
Prof. Carlos Alberto Ferreira Fernandes
November 2015
II
III
Acknowledgments
I would like to express my gratitude and appreciation to the supervisor of this work, Professor Paulo
Branco for giving me this opportunity, his guidance, persistency and encouragement. It would not be
possible without his contribute. I would like to thank all my family, specially my mother, father and sister
for all the patience and support and a very special thanks to my girlfriend Inês Freire who gave me
strength. Finally, to all my colleagues and friends who listened, advised and made their questions mine
contributing to a better research.
IV
V
Abstract
The access to electricity in some African countries is still very poor. For isolated villages where the
distance to the major cities is higher than hundreds of kilometres and the number of populations is very
low, it is not economically viable to build a connection to the electrical grid. Autonomous photovoltaic
(PV) installations are the key for future full autonomy of the household at a low infrastructure cost. More
particularly, this thesis studies and tests the behaviour and efficiency of a low-cost isolated
(autonomous, off-grid or stand-alone) photovoltaic (PV) system with the novel Lithium Iron Phosphate
(LiFePO4) battery storage, for the rural area near Luena in Angola. The system (solar panel, batteries,
controller and inverter) is designed having in mind the required household load and energy available
from the sun. These determined the sizing of the PV panels’ nominal power, LiFePO4 battery pack
storage capacity, the energy monitoring system as well as the power of the inverter.
In this context, the autonomous solar energy production system was developed for a nominal power of
500W. It was configured considering the load diagram of a typical rural house in order to achieve a high
lifespan of LiFePO4 batteries with less maintenance as possible. The experimental tests were made
under real conditions with the relevant electrical parameters being measured and logged along the day.
These tests have shown that, although the system ensures the energy supply effectively between the
months of February to November, during the rainy season the system should be complemented with a
second source of alternative energy.
Keywords
Autonomous/Off-grid PV system, Angola, system efficiency, LiFePO4
VI
VII
Resumo
O acesso à eletricidade em alguns países africanos é ainda muito precário. Em vilas isoladas onde a
distância às grandes metrópoles é superior a algumas centenas de quilómetros e o número de
habitantes é pequeno, não é economicamente viável construir uma interligação à rede elétrica. Os
sistemas fotovoltaicos autónomos são a chave para pequenas habitações autossuficientes em termos
energéticos a um preço de infraestruturas reduzido. Em particular esta tese estuda e testa o
comportamento e eficiência de um sistema isolado de custo reduzido com as recentes baterias de Lítio-
Ferro-Fosfato (LiFePO4) para a zona rural de Luena em Angola. O sistema (painéis solares, baterias,
controlador e inversor) é dimensionado tendo em vista a energia necessária ao abastecimento dos
consumidores e a energia solar disponível. Estes irão determinar os valores de potência nominal dos
painéis, a capacidade do “pack” de baterias de lítio, do sistema de gestão de energia assim como a
potência nominal do inversor.
Neste contexto, o sistema de produção elétrica autónomo foi desenvolvido para uma potência nominal
de cerca de 500W. O sistema foi configurado tendo em consideração o diagrama de carga típico destas
pequenas habitações rurais, de forma alcançar o maior tempo de vida útil das baterias de LiFePO4 com
a mínima manutenção possível. Os ensaios experimentais foram realizados em condições reais com
todos os sinais elétricos relevantes medidos e gravados ao longo do dia. Estes mostraram que, apesar
de o sistema assegurar o abastecimento energético de forma eficaz entre os meses de Fevereiro a
Novembro, durante a época das chuvas o sistema deverá ser complementado com uma segunda fonte
de energia alternativa.
Palavras-chave:
Sistema fotovoltaico autónomo, Angola, rendimento, LiFePO4
VIII
IX
Contents
Acknowledgments .................................................................................................................................. III
Abstract .................................................................................................................................................... V
Resumo ................................................................................................................................................. VII
Contents ................................................................................................................................................. IX
List of Figures ....................................................................................................................................... XIII
List of Tables .........................................................................................................................................XV
List of Symbols ................................................................................................................................... XVII
List of abbreviations ............................................................................................................................. XIX
8.15.1 Net Present Value (NPV) ............................................................................................. 115
8.15.2 Internal Rate of Return (IRR) ....................................................................................... 116
8.15.3 Payback Period ............................................................................................................ 116
XII
XIII
List of Figures
Figure 2.1 - Global cumulative growth of PV Capacity (Source IEA) ...................................................... 5 Figure 2.2 - Grid-tied PV system block diagram ...................................................................................... 7 Figure 2.3 - Block diagram of an isolated PV system – Inverter connected to the regulator .................. 9 Figure 2.4 - Block diagram of an isolated PV system – Inverter connected to the battery ..................... 9 Figure 2.5 - Block diagram of a hybrid isolated PV-Wind system with energy storage ........................ 10 Figure 2.6 - Incident radiation on a tilted surface (source: adapted from The Irradiation Data, Andres Cuevas) ................................................................................................................................................. 10 Figure 2.7 - Angles of a tilted surface (source: adapted from ITACA website) ..................................... 12 Figure 2.8 - Solar rays on a) geographical and b) panels referential .................................................... 13 Figure 2.9 - Solar irradiance measuring instruments – a) pyranometer and b) pyrheliometer (source: NREL and IBPSA-USA website) ........................................................................................................... 15 Figure 2.10 - Campbell-Stokes sunshine recorder (source: WeatherBug blog) ................................... 16 Figure 2.11 - Africa and Middle East Global Horizontal Irradiation average annual sum between 04/2004 and 03/2010 [10] .................................................................................................................................... 17 Figure 2.12 - Angola Global Horizontal Irradiation (GHI) map [10] ....................................................... 18 Figure 2.13 - Daily mean solar radiation averages in Luena, Angola (source: PVGIS Climate-SAF database 2001-2012) ............................................................................................................................ 19 Figure 2.14 - Average temperatures in Luena, Angola [30] .................................................................. 20 Figure 2.15 - Angola electrical sector intervenients [33] ....................................................................... 20
Figure 3.1 - Autonomous PV System electric block diagram ................................................................ 25 Figure 3.2 - EU energy efficiency labels (source: which.co.uk website) ............................................... 27 Figure 3.3 - Fluorescent Lamp Lighting block diagram (source: next electronics website) .................. 27 Figure 3.4 - Sony TV operating current ................................................................................................. 28 Figure 3.5 - Multi-junction cell (source: solar cell central) ..................................................................... 29 Figure 3.6 - Exothermic reaction evolution with temperature [45] ......................................................... 32 Figure 3.7 - LiFePO4 simplified cross section cell [46] .......................................................................... 33 Figure 3.8 - LiFePO4 typical cell charging curve [49] ............................................................................ 34 Figure 3.9 - Parallel cell charging with power supply ............................................................................ 35 Figure 3.10 - LiFePO4 balancing time [48] ............................................................................................ 37 Figure 3.11 - BMS block diagram (adapted from “The Electropaedia” website) ................................... 37 Figure 3.12 - Types of balancing [51] .................................................................................................... 39 Figure 3.13 - Passive vs active Cell Balancing (Source: Texas Instruments) ....................................... 39 Figure 3.14 - Passive switched shunting resistor balancing (based on [51]) ........................................ 41 Figure 3.15 - Example of a passive cell balancing circuit TI BQ77PL900 [52] ..................................... 41 Figure 3.16 - a) Bottom balancing and respective b) after charging ..................................................... 42 Figure 3.17 - a) Top balancing and b) after discharging ....................................................................... 42 Figure 3.18 - Passive cell balancing based on voltage [52] .................................................................. 43 Figure 3.19 - Block diagram of a sine wave inverter (adapted from Texas Instruments) ..................... 45 Figure 3.20 - Inverter output waveform and corresponding harmonics ................................................ 45 Figure 3.21 - Inverter efficiency characteristic a) Xantrex Prosine off-grid inverter efficiency curve for 1000 and 1800W 24V system (source: Xantrex) and b) 1500W Livre Inverter ..................................... 47 Figure 3.22 - Cable losses ..................................................................................................................... 48 Figure 3.23 - Off-grid PV system connection diagram .......................................................................... 49
Figure 4.1 - KENT 201E refrigerator load diagram (1h) ........................................................................ 52 Figure 4.2 - Total daily load diagram ..................................................................................................... 53 Figure 4.3 - Refrigerator a) starting and b) operating current ............................................................... 57
Figure 5.1 - a) Autonomous PV system testing bench and b) system components ............................. 64 Figure 5.2 - Battery pack and BMS monitoring boards ......................................................................... 64 Figure 5.3 - Battery pack voltage during discharge on a constant resistive load of 6Ω a) without and b) with initial balancing ............................................................................................................................... 65
XIV
Figure 5.4 - Cell’s voltage during discharge on a resistive load of 6Ω a) without and b) with initial individual charging ................................................................................................................................. 66 Figure 5.5 - Discharge at approximate 0.06C current rate or 8A (3.3Ω resistor) .................................. 67 Figure 5.6 - Discharge on a resistive load of 6Ω a) bottom and b) top balancing ................................. 67 Figure 5.7 - Charge at constant current I=6A with a) bottom and b) top balancing .............................. 68 Figure 5.8 - PV installation site and sun trajectory in March 2015 ........................................................ 69 Figure 5.9 - Total solar irradiance incident on a horizontal (red) and 35º tilted plane (black) on 14th March 2015 with clouds .................................................................................................................................... 70 Figure 5.10 - PV (blue) and battery pack voltage (black) on 14th March 2015 ...................................... 71 Figure 5.11 - Battery pack voltage on 14th March 2015 ........................................................................ 71 Figure 5.12 - PV panel output power on 14th March 2015 .................................................................... 73 Figure 5.13 - PV module temperature and efficiency on 14th March 2015 ............................................ 74 Figure 5.14 - Global horizontal irradiance measured at IST meteorological station (red) and PVGIS global clear-sky irradiance (black) ......................................................................................................... 75 Figure 5.15 - Global irradiance incident on the PV panel plane calculated from experimental GHI (red) and PVGIS database (black) ................................................................................................................. 76 Figure 5.16 - Monthly global irradiation average and temperature on a 35º tilted plane [26] ............... 77 Figure 5.17 - Global horizontal irradiation in a) Angola and b) Portugal [10] ........................................ 80 Figure 5.18 - Monthly average global irradiation on a β=35º plane in Lisbon, Portugal (blue) and β=19º in Luena, Angola (black) [26] ................................................................................................................. 81 Figure 5.19 - Monthly average temperatures in Lisbon, Portugal (blue) and Luena, Angola (black) [26, 30] .......................................................................................................................................................... 81
Figure 8.1 - LA 25-NP current transducer characteristic ....................................................................... 99 Figure 8.2 - Connecting a Differential Voltage Signal [NI USB-6008/6009 User Guide] .................... 100 Figure 8.3 - LabVIEW Signal Express Monitor / Record interface ...................................................... 100 Figure 8.4 - LabVIEW Signal Express Playback interface .................................................................. 101 Figure 8.5 - LEM LA25-NP equivalent circuit according to datasheet parameters [58] ...................... 102 Figure 8.6 - Experimental setup to obtain LA-25 NP characteristic .................................................... 103 Figure 8.7 - Experimental characteristic test of the LEM LA25-NP ..................................................... 103 Figure 8.8 - Equivalent electrical model of a solar cell a) three parameters and b) five parameters [59] ............................................................................................................................................................. 106 Figure 8.9 - Stationary characteristic I(V,G) of a photo-diode exposed to solar light [59] .................. 107 Figure 8.10 – BMS 123 a) dashboard and b) system settings ............................................................ 110 Figure 8.11 - Inverter characteristic with a resistive load .................................................................... 111 Figure 8.12 - Capacitor-start motor a) connections and b) phasor diagram at starting [60] ............... 112 Figure 8.13 - Single-phase equivalent circuit with core losses [61] .................................................... 113
XV
List of Tables
Table 2.1 - PV module prices in 2012 in European countries [5] ............................................................ 6 Table 2.2 - PV systems combinations ..................................................................................................... 7 Table 2.3 - Average annual GHI, GTIopt and D/G in some Angola cities [10, 26] ................................ 19 Table 2.4 - Annual sunshine hours and climate conditions for different Angola regions [30, 31] ......... 20 Table 2.5 - Normal low voltage electricity tariff in Angola (single phase) [33] ....................................... 22
Table 3.1 - Battery technologies comparison [42, 43, 44] ..................................................................... 32 Table 3.2 - Balancing topologies comparison [51] ................................................................................ 40 Table 3.3 - Comparison of Balancing Algorithms [48] ........................................................................... 43 Table 3.4 - Typical actuation voltage levels of off-grid pure sine wave inverters and PWM regulators 47
Table 4.1 - Loads electrical experimental parameters .......................................................................... 53 Table 4.2 - Experimental dairy loads energy ......................................................................................... 53 Table 4.3 - Irradiation on Luena, Angola [26] ........................................................................................ 54 Table 4.4 - Pure sine wave inverter price and specifications ................................................................ 60 Table 4.5 - Total system cost ................................................................................................................ 60
Table 5.1 - Energy of the discharge tests in Figure 5.3(a) - unbalanced and Figure 5.3(b) - balanced cells ........................................................................................................................................................ 65 Table 5.2 - Energy and mean daily efficiency of the different components .......................................... 72 Table 5.3 - Statistical test results .......................................................................................................... 76 Table 5.4 - Autonomous PV system average technical data ................................................................ 77 Table 5.5 - Monthly average vs 14th March irradiation ......................................................................... 78 Table 5.6 - Estimated energy available to the loads ............................................................................. 78 Table 5.7 - Percentage of days of full and empty battery [26] ............................................................... 79 Table 5.8 - Radiation comparison between Angola and Portugal (source: PVGIS climate-SAF Europe and Africa maps database 2001-2010) [10, 57, 26] .............................................................................. 80 Table 5.9 - Global irradiation in Luena, Angola and Lisbon, Portugal on the horizontal and optimal inclined plane [10, 26, 30]..................................................................................................................... 81 Table 5.10 - Estimated energy available to the loads in Luena, Angola ............................................... 82 Table 5.11 - Percentage of days of full and empty battery [26] ............................................................. 83
XVI
XVII
List of Symbols
A – Amp
𝐴 – PV surface azimuth angle
𝐴𝑆 – Solar azimuth angle
𝐶 – Cell Capacity
D – Daylight saving time
EOT – Equation of Time (min)
𝐸𝑑 ≡ 𝑊𝐷 – Daily energy required by the loads
𝐸𝐼𝑁 – Energy received (in)
𝐸𝐿𝑜𝑎𝑑𝑠 – Energy of the loads
𝐸𝑂𝑈𝑇 – Energy delivered (out)
𝐸𝑃𝑉 – Energy produced by the PV panels
𝑓𝑠 – Sampling rate
𝐺 – Solar Irradiance (W/m²)
𝐺𝑏 – Beam radiation
𝐺𝑏_ℎ𝑜𝑟𝑖𝑧 – Incident (beam) radiation component perpendicular to the ground horizontal plane
𝐺𝑏_𝑖𝑛𝑐𝑖𝑑𝑒𝑛𝑡 – Maximum incident (beam) radiation
𝐺𝑏_𝑚𝑜𝑑𝑢𝑙𝑒 – Module (beam) radiation or radiation incident on the tilted plane
𝐺𝑑 – Diffuse radiation
𝐺𝑔𝑙𝑜𝑏𝑎𝑙 – Global radiation
𝐺𝑟 – Reflected radiation
𝐻ℎ – Irradiation on horizontal plane (Wh/m²/day)
𝐻𝑖 – Solar Irradiation (kWh/m²)
𝐻𝑜𝑝𝑡 – Irradiation on optimally inclined plane (Wh/m²/day)
𝐼𝑛 𝐷𝐶 𝑆𝑤𝑖𝑡𝑐ℎ – Nominal current of the DC switch
𝐼𝑅𝑀𝑆 – Root mean square current
𝐼𝑏𝑝1, 𝐼𝑏𝑝2, . . , 𝐼𝑏𝑝𝑁 – Cell’s by-pass current
𝐼𝑝𝑎𝑐𝑘 – Current circulating in the battery pack
𝐼𝑟𝑒𝑔_𝑚𝑎𝑥 – Maximum regulator current
𝐼𝑧 – Admissible current on the cables
LCT – Local Clock Time (h)
LiCoO2 – Lithium Cobalt Oxide
LiCoPO4 – Lithium Cobalt Phosphate
LiFePO4 - Lithium Iron Phosphate
LL – Local Longitude (°)
LSTZ – .Longitude of Local Standard Time Meridian (°)
𝑚 – the diode ideality factor
XVIII
mmf – Magnetomotive force
𝑁 – Day of the year
NiCd – Nickel-Cadmium
NiMH – Nickel-metal hydride
𝑃𝐽𝑜𝑢𝑙𝑒 – Joule losses on the cables
𝑃𝑃𝑉 - Power produced by the panel(s)
𝑃𝑅 – Power dissipated in a resistor (Joule Power)
𝑃𝑖𝑛𝑗𝑒𝑐𝑡𝑒𝑑 - Power at the end of the cables terminals, injected in the batteries
𝑃𝑖𝑛𝑣_𝑚𝑖𝑛 – Minimum inverter power
𝑃𝑙𝑜𝑎𝑑𝑠,𝑝𝑒𝑎𝑘 – Peak power of the loads
𝑅 – Resistance
𝑅𝐷𝑆𝑂𝑁 – Transistor resistance
𝑅𝐵𝑎𝑙 – On-Board resistor in parallel with the cell at passive balancing
𝑆 – Incident radiation or sun ray vector
𝑡𝑎𝑟𝑖𝑓𝑓 – Electrical tariff (€/kWh)
𝑡𝑠 – solar hour (h)
𝑈𝐵𝐴𝑇,𝑇𝑜𝑡 ≡ 𝑈𝐷𝐶 – Total battery DC voltage (V)
𝑈𝑅𝑀𝑆 – Root mean square voltage (V)
V – Volt
W – Watt
Wh – Watt-hour
𝑊𝐵𝑎𝑡,𝑇𝑜𝑡 – Total energy of the batteries
𝑊𝐷 ≡ 𝐸𝑑 – Daily energy required by the loads (Wh)
𝛼 – Solar altitude or solar elevation angle (º)
𝛽 – Surface inclination angle (°)
𝛽𝑜𝑝𝑡 – Optimal inclination angle (°)
𝜃𝑖 – Incidence angle (°)
𝛿 – Declination angle (°)
∆𝑇𝑠 – Sampling time
∆𝑈 – Cables Voltage drop
𝜂𝐵𝐴𝑇 – Battery efficiency (%)
𝜂𝐵𝑀𝑆 – BMS efficiency (%)
𝜂𝐵𝑀𝑆+𝐵𝐴𝑇+𝐼𝑛𝑣 – BMS, batteries and inverter efficiency
𝜂𝑃𝑉 – PV panels efficiency
𝜂𝑖𝑛𝑣 – Inverter efficiency
𝜂𝑙𝑜𝑠𝑠𝑒𝑠 – Cable losses factor
𝜂𝑠𝑦𝑠𝑡,𝑎𝑢𝑡 – Autonomous/Off-grid system efficiency
𝜙 – latitude (°)
𝜔 – hour angle (°)
XIX
List of abbreviations
AC – Alternating Current
AM1.5 –Spectral Distribution of Solar Radiation
BMS – Battery Management System
CC – Constant Current
CFL – Compact Fluorescent Lamps
CPV – Concentrated Photovoltaics
CRT – Cathode Ray Tube
CSP – Concentrated Solar Power
CV – Constant Voltage
DAQ – Data Acquisition
DC – Direct Current
DEEC – Department of Electrical Energy and Computers
D/G – Beam vs Diffuse Radiation Ratio
DIF – Diffuse Horizontal Irradiance/Irradiation
DOD – Depth of Discharge
DNI – Direct Normal Irradiance/Irradiation
EDEL – Empresa de Distribuição de Electricidade de Luanda
ENDE – Empresa Nacional de Distribuiçao de Electricidade
ENE – Empresa Nacional de Eletricidade
EOT – Equation of Time
EU – European Union
GAMEK – Gabinete de Aproveitamento do Medio Kwanza
GER – Gabinete de Energias Renováveis
GHI – Global Horizontal Irradiance/Irradiation
IEA PVPS – International Energy Agency Photovoltaic Power Systems
IRR – Internal Rate of Return
IRSE - Instituto Regulador do Sector Energético (Energy Sector Institute Regulator)
IST – Instituto Superior Técnico
LCT – Local Clock Time
LFP - Lithium Iron Phosphate (LiFePO4)
LL – Local Longitude
LSTZ – .Longitude of Local Standard Time Meridian
MBE – Mean Bias Error
MINEA – Ministério da Energia e Águas
MPPT – Maximum Power Point Tracking (or Tracker)
NOCT - Normal Operating Cell Temperature
NPV – Net Present Value
OCV – Open Circuit Voltage
XX
Op Amp – Operational Amplifier
OV – Overvoltage
PFC – Power Factor Correction
PV – Photovoltaic
PVGIS – Photovoltaic Geographical Information System
PRODEL – Empresa Nacional de Produção de Electricidade
RELOP – Associação de Reguladores de Energia dos Países de Língua Oficial Portuguesa (Portuguese
Official Language Energy Regulators Association)
RMSE – Root Mean Square Error
RNT – Rede Nacional de Transporte (National Transport Grid)
SOC – State of Charge
SR – Switched Shunt Resistor
STC – Standard Test Conditions
THD – Total Harmonic Distortion
UPS – Uninterruptible Power Supply
USD – Unites States Dollar
Wp – Watt-peak
1
1 Introduction
In the rural areas of developing countries, the access to electricity is still very deficient. This happens
once the distances involved are very large and do not justify the investment for supplying small
populations. The focus on rural electrification aims to give people better living conditions at low cost,
with the certainty that there will be no short or medium term return, but those conditions will make all the
difference on the development of that country.
This work aims to create a final solution that is actually possible to use immediately, add value and solve
these issues improving the living standard of people. The fully autonomy depends on the energy storage,
its efficient management and control is the key to ensure a long operation period.
Photovoltaic (PV) systems are a convenient solution for having a clean and reliable source of energy
and also provide electricity in locations where it is not possible to connect to the conventional electricity
grid or the connection has a high cost. The decrease of photovoltaic panels manufacturing cost make
autonomous photovoltaic systems more beneficial over the years. Consequently, PV installations have
been growing significantly in many countries over the past years. The use of PV installations meets then
economic, environmental and social causes. It is imperative to provide access to energy to ensure
socioeconomic development in the world’s poorest and developing countries. Although its relatively high
cost, autonomous/isolated or off-grid PV systems gradually expands with government economic support
and private investments.
Energy production is the basis for the working time capacity of an autonomous system which additionally
requires energy storing, in other words, batteries. Within the range of available chemical types, lithium
type has high electrochemical potential which makes it one of the most reactive of metals. These
properties give lithium the potential to achieve very high energy and power densities in many
applications such as automotive and photovoltaic. Lithium Iron Phosphate (LiFePO4) have some
advantages compared with other lithium battery types, having a very stable thermal behaviour. For this
reason they are object of study and also the best ways of using this type of cells.
This is the context of this study, more particularly PV isolated systems with LiFePO4 storage to supply
residential buildings in Angola and all the related elements necessary for its operation.
1.1 Motivation and problem definition
Electrical energy is a fundamental good for the development and well-being of a society. Africa is a
continent with a low rate of access to electricity, so it is important to develop solutions to improve life
conditions of the populations without harming the environment. Despite its huge natural resources
(mostly hydric potential), Angola is a country inserted in this context, with a per-capita consumption
2
below the average in Africa [1]. It is estimated that in Angola less than 20% of the population has access
to energy [2]. The power grid is still very limited supplying only the big cities. There are groups of diesel
generators that run on local networks to supply industries located in remote places, with problems of
continuous electricity supply. The prolonged drought affects river flows which in turn require electricity
production to stop. Also the dependence of the price of diesel fuel and its transportation costs, which
despite being relatively low compared with European countries, are continually growing due to the
decrease of natural resource reserves.
LiFePO4 have some advantages compared with other lithium battery types but some working constrains.
Lithium cells tend to have different state of charge (SOC) when used in groups (connected in series or
parallel) and submitted to several charge/discharge cycles. Enumerating some of the reasons, this could
happen due to external temperature differences, cell’s impedance differences or capacity manufacturing
differences. The SOC of each cell individually is not detected by any SOC measuring instruments. The
strategy to have the SOCs as close as possible is to do what is called balancing which should always
depend on system application and is generally done by a Battery Management System (BMS).
Protecting a single cell has a certain complexity, protecting a battery (a series string) is harder: cell
voltages do not divide equally, temperatures vary, etc. The higher the protection the higher the system
cost, for this reason the project is a balance between costs and capabilities as usual.
The electrical needs in developing countries is constantly growing which sets new horizons for energy
isolated systems.
1.2 Objectives
This thesis aims on the use of autonomous photovoltaic systems with LiFePO4 batteries to supply
individual or collective residential buildings in rural areas, more particularly at Luena zone in Angola.
The main objectives are the following:
1. Definition of the energy pathways in an autonomous photovoltaic home system (recognition the
involved elements);
2. Identification of the typical consumers profile on a small residential building in Angola. Based
on this, project the nominal values of each element that constitutes the system regarding the
typical consumers profile;
3. Cost–benefit analysis of the system in study by verification and analysis of total experimental
system efficiency results.
4. Study and test the behaviour of the novel LiFePO4 batteries under load when used in group as
a battery pack; Identify the best balancing typology for LiFePO4 batteries using the available
measuring equipment;
3
1.3 Thesis structure
This thesis is constituted by six chapters organized in the following sequence:
In Chapter 1 is done a general overview of the work;
In Chapter 2 the different types of autonomous photovoltaic systems are introduced, basic solar
radiation concepts and the contextualization of Angola and the introduction of economic
indicators;
In Chapter 3 is introduced the photovoltaic system and the various features of the components
necessary for its implementation;
In Chapter 4 is made the project of the components is done accordingly to the objectives;
In Chapter 5 is done the experimental study system elements behavior and its efficiency;
In Chapter 6 is made a summary of the conclusions drawn from work and some proposals for
future developments;
4
5
2 Autonomous PV Systems – Actual Panorama
The thermic conversion of the solar radiation can be used in several applications from water heating
(concentrated solar thermal, also known as concentrated solar power - CSP) to electricity production
(electrical) since the discovery in 1954 by Bell Labs who found that silicon doped with certain impurities
was very sensitive to light. As a result emerged the first solar modules that allowed a practical application
and with an energy conversion efficiency around 6% in laboratory. Photovoltaic (PV) technology is an
active solar system type based on the photoelectric effect, converting light into electrical energy through
photovoltaic cells that in turn form a photovoltaic panel. Nowadays, the most established solar PV
22.9% at Standard Test Conditions1 (STC) in laboratory [3]. Each cell generates a Direct Current (DC)
power of about 1.5W (0.5V and a 3A current) and the number of cells depends on the panel technology
[4]. A PV System consists of several PV panels connected in series and parallel forming what is called
a string. Photovoltaic modules power is usually between 50W and 315W. This systems may be
connected to the grid (grid-tied) or be isolated (off-grid) and are basically constituted by controllers,
batteries and inverters so they may feed the consumers. With the correct maintenance the durability is
higher than 20 years, making this a very reliable source of energy.
The global PV capacity grew 49% a year on average since 2003 as illustrate in Figure 2.1. Decentralised
systems represent approximately 60% of the global market, while centralised systems are close to 40%.
The share of off-grid installations is very small compared to grid-tied, once grid-tied systems expanded
in large scale from 2000 to 2013. However, like PV technologies in general off-grid systems are
expanding, especially in developing countries with high solar exposure. [5] [6]
Figure 2.1 - Global cumulative growth of PV Capacity (Source IEA)
1 Cell temperature, 𝜃𝑟 = 25°𝐶 ≡ 𝑇𝑟 = 298.16 K; Incident radiation, 𝐺𝑟 = 1000 𝑊/𝑚2; Spectral Distribution of Solar
Radiation AM1.5 - Total solar spectral irradiance distribution (direct and diffuse) at sea level, clear sky day, at an inclined plane at 37° tilt toward equator facing the sun, an absolute air mass (AM) of 1.5 (solar zenith angle 48.19°S)
6
As with other investments, prices tend to be smaller as the investments in photovoltaic systems expand.
In autonomous systems, batteries represent the largest share of investment usually greater than 40%
and can raise the cost of Watt-peak (Wp) to twice of a grid-tied system. The photovoltaic panels
constitute about 15-20% of the investment and the inverter another 20%. The remaining stands for the
control system and other costs. Additionally, the price is difficult to define as it depends on location, size,
and component specifications regarding the consumer’s needs. With data from 2013, a survey of system
prices in International Energy Agency Photovoltaic Power Systems (IEA PVPS) European reporting
countries2 showed the system prices in the off-grid sector (<1 kW), irrespective of the type of application,
typically ranged from about 2.7 to 20 USD/W. This means that a 500W system may cost between 1250
and 10.000 USD [5].
Data from selected IEA PVPS reporting countries showed that the price evolution of PV modules and
small-scale systems (<1 kW) has been decreasing specially since 2007. Table 2.1 specifies that different
PV technologies module prices varies from 0.5 to 1.54 USD/W in European countries. This represents
a price between 112.5 to 346.5 USD for a 225W PV panel. In 2012 some manufacturers have reported
losses which resulted in a price increase. The future trend is a price stabilisation of about 1 USD/W for
PV modules and 2.5 USD/W3 for residential systems [5].
Table 2.1 - PV module prices in 2012 in European countries [5]
Country USD/W
Austria 0.5 – 0.72
Denmark 0.72 – 1
France 0.96
Germany 0.92
Italy 0.67 – 0.87
Netherlands 1.39
Spain 0.73
Sweden 1.54
Switzerland 0.86 – 1.08
2.1 Photovoltaic Systems Types
In terms of PV systems network connectivity, there are currently two types: grid-tied systems and
isolated or autonomous systems with and without battery energy storage. In some literature [4] [7] [8]
[9], the hybrid systems with several energy sources (PV, diesel generators, wind, etc.) are considered
a third type; however due to the hybrid definition is more correct to portray it as a category. Table 2.2
shows different possible combinations of PV systems that are further explained in more detail. The
electrical loads may be Direct Current (DC) or Alternating Current (AC) power depending on application,
however only AC loads go in line with the context of this work. There may also be found grid-tied systems
with energy storage, the batteries work as a backup of energy in case the grid fails, Uninterruptible
2 Austria, Denmark, France, Italy, Spain, Sweden and Switzerland 3 Battery costs are not included
7
Power Supply (UPS), however this solution makes the installation more expensive and therefore used
only in very specific cases.
Table 2.2 - PV systems combinations
Types of PV systems Energy storage No energy storage Hybrid
Grid-tied X X X
Autonomous/Off-grid X X X
2.1.1 Grid-tied Systems
In grid-tied systems, the connection to the electric grid allows the sale of electricity to the power
distribution companies and the feeding of the loads. The generated energy is injected directly in the grid.
Therefore, batteries and regulators are not necessary, what makes the system simpler and less
maintenance.
Figure 2.2 represents the main equipment of a grid-tied system with AC loads (the protection elements
are not represented). Purchase and sell back energy meters may be, in some cases, substituted by bi-
directional meters. A power inverter makes possible the connection to the grid and the connection of the
consumers is done at the grid side node.
Medium power grid-tied inverters use a Maximum Power Point Tracker (MPPT) to optimize PV panels
performance. This device constantly adjusts the input of a DC/DC converter (accordingly to the actual
irradiance and temperature) which selects the voltage output based on a simulated model. Selecting an
output voltage the current automatically adjusts and its value depends on the IV curve of the panel
(Figure 8.9). In many countries there are several incentive policies for small home systems sell your
energy to the grid with advantageous rates.
PV
Consumers
Energy meter:
Sell back
Energy meter:
Purchase
Grid
PDC
PACInverter Isolation Transformer
DC/DC Converter
MPPT
Figure 2.2 - Grid-tied PV system block diagram
Another recent solution that is being implemented is the connection of the inverter directly to consumers’
houses circuit for immediate consumption. This is done using an inverter for each panel – Microinverter4.
A photovoltaic system typically includes several panels wired together in series/parallel, with their total
4 Inverter used for a single panel that automatically synchronizes with the grid
8
DC output going to a central “string” inverter. This design has a few significant weaknesses. First, if the
inverter goes down, the whole system is down. Second, at any given moment, each solar panel may be
producing different amounts of power depending on shading, age, wear, and other factors. In the central
configuration, the overall system performance is dragged down by the weakest link. A string inverters
has to go with the average of the array rather than optimizing the outputs of each panel.
Microinverters are dedicated to a single PV panel and use the MPPT algorithm for each panel
individually, optimizing the overall system output regardless of minor shading issues and other variations
among individual panels. The connection of this type of inverters to consumers’ houses circuit allows
power reduction required from the network and consequent energy cost reduction for consumers without
the need of a producer contract, thus being autonomous (off-grid system).
2.1.2 Off-Grid Systems
The term isolated, autonomous or off-grid system owes its name due to the lack of any connection to
the electrical grid. In an isolated system without energy storage and DC or AC loads, consumers utilize
immediately the electric energy produced by the photovoltaic module. The investment in this type of
system is lower compared to systems with energy storage elements incorporated. However, the supply
to consumers is only possible during energy production hours. For example, a very common application
is water pumping systems.
Autonomous PV systems with energy storage are constituted by a string of PV panels, a solar
regulator/controller to monitor the batteries’ voltage levels and a battery pack. The energy produced by
the PV panels is delivered to the loads, and the surplus energy is stored in the batteries. The stored
energy may be used during the night when there is no solar radiation. Consequently, the batteries must
have enough capacity to feed the load during the night and/or in low solar radiation days. The regulator
has the ability to cut-off the PV and also cut-off the consumers supply accordingly to the battery voltage
levels. The use of DC loads only may avoid the use of an inverter. Inverter efficiency and consumption
in stand-by reduces the amount of available power to the loads. However, the use of DC loads is not
very usual and this equipment is always more expensive than AC equipment. Using only AC loads is a
cheaper solution and has more interest in terms of project.
Two configurations are possible – connect the inverter to the solar regulator (Figure 2.3) or directly to
the batteries (Figure 2.4). The use of an off-grid inverter is required when using AC loads. Solar
regulators are usually projected for currents up to 50A. In case of loads without high current peaks, the
inverter may be connected to the regulator as shown in Figure 2.3.
9
PV
DC Loads
Battery
Pack
Solar Reg.
PDC
PAC
PDC
AC Loads
Inverter
Figure 2.3 - Block diagram of an isolated PV system – Inverter connected to the regulator
Figure 2.4 - Block diagram of an isolated PV system – Inverter connected to the battery
AC loads are usually constituted by electrical motors, lights, TVs and other equipment with high starting
currents that produce prohibitive currents on the DC side. These peaks may reach 100A in the battery
pack and going through the regulator. To avoid such currents in the solar regulator, a direct connection
of the inverter to the batteries is a possible solution, as shown in Figure 2.4. With this configuration, the
regulator may connect/disconnect the PV panels and the DC loads. The inverter acts like a regulator
being the under-voltage protection for the batteries thus disconnecting the AC Loads. In this case, the
efficiency of the inverter has to be taken into account once it might be working at 20% of its nominal
power once an oversizing is necessary due to power peaks. The two different typologies choice depend
on the cut off parameters of the inverter and regulator nominal current.
2.1.3 Hybrid Systems
Hybrid systems consist of a combination of several energy sources in photovoltaic applications.
Diesel/gas generators, wind power and other sources may be used besides PV panels. This systems
need more complex control and protection circuits and are generally used for medium to high power
applications. For example in PV/Diesel medium power systems, the generator should work only when
the batteries reach the low voltage level and shuts down when the batteries are charged.
Figure 2.5 shows a hybrid isolated system with PV and wind energy sources. Different from configuration
in Figure 2.4, this configuration allows the charge of the batteries during the day and night. Wind
generators are always AC power, so the wind regulator makes possible the connection to the batteries
in DC current.
PV DC Loads
Battery Pack
Solar Reg.
PDCPDC PAC
AC Loads
Inverter
10
PV
Consumers
PAC
Battery Pack
Wind Generator
Solar Reg.
Wind Reg.
PDC
PDC
PAC
Inverter
Figure 2.5 - Block diagram of a hybrid isolated PV-Wind system with energy storage
For autonomous PV systems, there are no regulations in many countries, giving some freedom to their
implementation. Autonomous PV systems with electric energy storage based on LiFePO4 battery pack
is the focus of this work, being studied in more detail. Particular care is the requirement to ensure the
batteries and the power plant to have the longest possible useful lifetime.
2.2 Solar Radiation
Before analysing some data is important to make a brief introduction to the concepts of solar radiation.
Sunlight is part of the electromagnetic radiation given off by the Sun. The visible spectrum is constituted
by infrared, visible and ultraviolet light. The incident solar power per unit area, 𝐺, is the solar irradiance
and is measured in 𝑊 𝑚2⁄ . The solar energy incident per unit area, 𝐻𝑖 , is designated irradiation and is
measured in 𝑘𝑊ℎ 𝑚2⁄ . Only a part of the solar radiation hits the earth surface and this radiation is
classified in two different types regarding its way of incidence: direct and diffuse [4].
2.2.1 Direct and diffuse radiation on a tilted plane
When the surface under study is tilted with respect to the horizontal, the total irradiance on the tilted
surface is the direct (beam) normal radiation projected onto the tilted surface, plus the diffuse plus the
reflected radiation on the tilted surface. Figure 2.6 illustrates the types of radiation for a tilted surface
being the reflected radiation a particular case of diffuse radiation. Note that if the surface is not tilted the
reflected radiation component does not exist as explained below. [4, 10, 11]
Figure 2.6 - Incident radiation on a tilted surface (source: adapted from The Irradiation Data, Andres Cuevas)
The second more common lamps are compact fluorescent lamps (CFL) or compact fluorescent lights
using three to five times less energy and lasting eight to fifteen times more than incandescent light bulbs
(Figure 3.3). They also produce harmonics, if possible CFLs with low THD (below 30 percent) and power
factors greater than 0.9 should be selected.
28
Traditional (iron-core ballast and starter) fluorescent lamps also draw a higher current during the switch-
on cycle. During the start-up process, there are filaments at each end of the tube that are heated, and
this draws more current than normal operation. This surge current is typically between 1.25 and 1.5
times the normal current. Power Factor Correction (PFC) capacitors are used in parallel with many
fluorescent lamp ballasts, especially those designed for industrial use. These are necessary to minimise
the excess current drawn by a relatively linear but reactive load. When power is turned on, the surge
current may be very high - typically up to 30 Amps or more depending on the exact point in the main
cycle when power is applied. This is many times the rated current of the PFC capacitor (as determined
by the capacitance, voltage and frequency). The ballast is basically an inductor that is in series with the
gas tube to limit the current through it, which would otherwise rise to destructive levels, due to the tube's
negative resistance characteristic.
3.1.3 CRT TV
Comparing with Plasma, LCD (Liquid-Crystal Display), LED LCD and other recent TV technologies, CRT
TV’s are the least energy-efficient but also the cheaper and most common in Angola.
CRT TVs have also a high surge current when heating the metal plate at the back of the electron gun.
They are non-linear loads – a switchmode power supply. Next Figure 3.4 shows the typical operating
current waveform (blue curve) of a CRT TV experimentally obtained for the Sony TV (appendix 8.12)
with high harmonic content. The orange curve is the voltage at the terminals of the TV.
Figure 3.4 - Sony TV operating current
Most of the electronic equipment has a single-phase rectifier with a capacitive filter producing currents
with impulsive character and high THD, approximately centered in the voltage wave peak (orange). The
power factor does not apply in this case, once the current wave is not sinusoidal; however, we might
say that non-linear circuits have a poor power factor because the current waveform is distorted.
3.2 PV Panels
There are several types of active (conversion of sunlight into other forms of thermic or electric energy)
and passive (heating buildings through constructive strategies) solar systems. PV panels are active
29
solar systems constituted by solar cells associated in series and parallel to create a PV module with
appreciate power. The first generation technology is the crystalline silicon with a market share of 87%
in three mainly types: monocrystalline, polycrystalline and silicon tapes. The polycrystalline cells are
less efficiency but represent 49% of the market against 35% of monocrystalline cells. They have also a
lower manufacturing cost (about 20%) being the right choice to project the cheapest system possible
[4].
3.2.1 Working Principle
The working principle of solar cells is the photovoltaic effect discovered by Alexandre Edmond Becquerel
in 1839. In photovoltaic effect the electrons-hole pairs generated are transferred between different bands
(valence bands to conduction bands) within the material itself, resulting in the development of electrical
voltage between two electrodes. Solar cells are made of the same kinds of semiconductor materials,
such as silicon, used in the microelectronics industry. For solar cells, a thin semiconductor wafer is
specially treated to form an electric field, positive on one side and negative on the other. When photons
strike the solar cell, electrons are knocked loose from the atoms in the semiconductor material. If
electrical loads are attached to the positive and negative sides, forming an electrical circuit, the electrons
can be captured in the form of an electric current.
Figure 3.5 - Multi-junction cell (source: solar cell central)
The use of several layers revealed higher efficiencies in the conversion process with less thermal energy
lost. Since sunlight will only react strongly with band gaps roughly the same width as their wavelength,
the top layers are made very thin so they are almost transparent to longer wavelengths. This allows the
junctions to be stacked, with the layers capturing the shortest wavelengths on top, and the longer
wavelength photons passing through them to the lower layers. The example of a multi-junction cell in
Figure 3.5 has a top cell of gallium indium phosphide, then a "tunnel diode junction", and a bottom cell
of gallium arsenide. The tunnel junction allows the electrons to flow between the cells and keeps the
electric fields of the two cells separate.
A number of solar cells electrically connected to each other and mounted in a support structure or frame
is called a photovoltaic module. Modules are designed to supply electricity at a certain voltage, such as
30
a common 12 volts system. The current produced is directly dependent on how much light strikes the
module.
3.2.2 Electrical parameters
To uniform the measurements of the characteristic parameters the manufactures accepted to use the
Standard Test Conditions (STC) represent by the index r [4]:
Cell temperature, 𝜃𝑟 = 25°𝐶 ≡ 𝑇𝑟 = 298.16 𝐾;
Incident radiation, 𝐺𝑟 = 1000 𝑊/𝑚2;
Spectral Distribution of Solar Radiation AM1.514.
The thermal potential at the reference conditions:
𝑉𝑇𝑟 =
𝐾𝐵𝑇𝑟
𝑞= 0.0257 V (3.1)
The maximum output power at STC is the peak-power:
𝑃𝑃 = 𝑃𝐷𝐶𝑟 = 𝑉𝑀𝑃
𝑟 𝐼𝑀𝑃𝑟 (3.2)
The efficiency at STC:
𝜂𝑟 =𝑃𝐷𝐶
𝑟
𝐴𝐺𝑟=𝑃𝑝𝐴𝐺𝑟
(3.3)
where 𝐴 is the area of the cell/panels. With other working conditions:
𝜂 =𝑃𝐷𝐶𝐴𝐺
(3.4)
where 𝐺 is the total (beam and diffuse) radiance incident on the PV surface. The fill factor (FF) is:
𝐹𝐹 =𝑃𝐷𝐶
𝑟
𝑉𝑐𝑎𝑟 ∙ 𝐼𝑐𝑐
𝑟 (3.5)
It should as high has possible to take the most of the panel power.
A simplified model to estimate the temperature of the module uses a relation proportional to the incident
irradiance. Manufacturers provide the Normal Operation Cell temperature (NOCT15)
𝑇𝑐 = 𝑇𝑎 +𝐺(𝑁𝑂𝐶𝑇 − 20)
800 (3.6)
The efficiency of the panel may be estimated knowing the temperature of the cells, decreasing with
increasing temperature due to the higher dark current [40]. It will produce better results with experimental
data using the following expression [14]:
14 Total solar spectral irradiance distribution (direct and diffuse) at sea level, clear sky day, at an inclined plane at
37° tilt toward equator facing the sun, an absolute air mass (AM) of 1.5 (solar zenith angle 48.19°S) 15 NOCT conditions: ambient temperature of 20°C, G=800W/m². The typical value is 45°C
31
𝜂(𝑇) = 𝜂𝑆𝑇𝐶[1 + 𝛾(𝑇𝑐 − 𝑇𝑆𝑇𝐶)] (3.7)
where 𝑇𝑐 is the temperature of the cells, 𝑇𝑆𝑇𝐶 = 25°C is the temperature of the cells at STC
conditions, 𝜂𝑆𝑇𝐶 is the efficiency of the module at STC conditions and γ is the empirically estimated
relative efficiency temperature coefficient, approximately equal to -0.004/K for polycrystalline silicon
cells [41].
The voltage level is also very important as it determines the currents flowing in the circuit given a certain
power load. The higher the voltage, the lower the currents, and therefore less losses, which leads to a
cheaper system (lower regulator, fuses, cables cross section).
PV nominal voltage is used to make sure the module is compatible with a given system and refers to
the voltage of the battery that the module is expected to charge. The real PV output voltage changes
with environmental conditions, so there is never one specific voltage at which the module operates, but
a voltage rang, instead.
3.3 Energy Storage - LiFePO4 Batteries
An electric battery is a device that converts stored chemical energy into electrical energy in accordance
to the chemical reactions – redox equations. Rechargeable batteries have largely replaced primary cells,
as they save resource and reduce pollution. The most commonly used batteries for storing applications
are lead-acid batteries [7] type, but they are being substituted by Lithium-ion batteries over time due to
several advantages. Li-ion batteries have high capacity, high electrochemical potential, superior energy
density, durability, as well as the flexibility in design. All the above outstanding properties accelerate the
substitution of conventional secondary batteries.
The batteries used in autonomous PV systems must have the following characteristics:
Reduced maintenance requirements;
Long service time;
Reduced self-discharge and high energy efficiency;
High storage capacity and power density;
Good performance/price relation;
Protection against the occurrence of hazards to the environment and health.
The four most promising cell chemistries considered for energy storage applications are
LiMn2O4/graphite cell chemistry, which uses low-cost materials that are naturally abundant; LiNi1-X-
Y2CoXAlYO2/graphite cell has high specific energy and long life; Li4Ti5O12 is used as the negative
electrode material in Li-ion batteries with long life and good safety features and LiFePO4/graphite (or
carbon) cell chemistry is a type of Lithium that has a very good thermal and chemical stability, leading
to safety characteristics and very low risk of fire.
32
Table 3.1 represents a lithium ion cathode chemistry comparison (using carbon anodes). Lithium-ion
batteries advantages are evident. Less weight, more life cycles and less cycle discharge.
Table 3.3 - Comparison of Balancing Algorithms [48]
Voltage Based Final Voltage Based SOC History Based
Principle
of
operation
Balances whenever charging,
regardless of SOC. Strives to match
cell voltages.
Balances at high SOC. Strives to match cell voltages
Balances all the time. Strives to match cell DOD,
based on previous history of cells
Pros Very simple method
At high SOC the cell voltage changes rapidly, it gives
better data on the true SOC. Charging current can be reduced so errors due to
internal resistance variations are minimized
The BMS balancing current can be lower, and balancing can be
done in fewer cycles, as balancing can occur all the time. Cell resistance has little effect.
Cons
Cell voltage as an indications of SOC
is not effective. OCV vs SOC curve is quite flat at mid
SOC levels. Strongly affected by
cell’s resistance.
Balancing only at the top means there is less time to balance with high currents
Requires more computing power and more memory to store the history of each cell
Table 3.3 resumes the balancing typologies most frequent. Once each method has its pros and cons,
the choice depends on the application. Voltage based method produces poor results so the best choice
is to combine the final voltage based method with computing and memory capabilities. Although 123
electric BMS™ (appendix 8.9) uses the final voltage based method only, this results in the best
performance vs. price. Figure 3.18 is an example of voltage evolution in two different cells when
charging.
Figure 3.18 - Passive cell balancing based on voltage [52]
The charging process ends properly by reaching the by-pass voltage. In each cycle, the BMS shortens
the balancing time once the voltages get closer each cycle. The bigger the cells the higher the needed
time to balance.
44
3.4.3 SOC estimation
As seen before in 3.3.2, the usual method for estimating the SOC is the coulomb counting method done
with a “fuel gauge” that consists of a DC current meter. The basic calibration in this type of BMS consists
in the following steps:
1. At the beginning the algorithm considers the SOC given by the user.
2. If the first cycle is a discharging cycle, then Amperes are deducted until 2.5/V per cell is reached.
At the moment when 2.5V (or different number in Vmin value) is reached, 0% is shown at SOC
gauge.
3. After charging starts and once the voltage reaches 3.6V (or different Vmax value), the gauge is
set to 100%. Anytime when Vmin or Vmax value is reached by any cell
4. The 𝑆𝑂𝐶0 is recalibrated for the newer calculated value.
3.5 Autonomous or Off-Grid Inverters
For any photovoltaic (PV) system, the off-grid inverter is the essential electronic device that converts
low voltage DC electricity from a battery or other power source to 100V-120V or 220V-240V AC signal.
Off-grid Inverters produce a voltage wave, with an independent frequency from the grid. This is the point
where the off-grid inverters differ from grid-tied once the last need the grid voltage wave to “couple” and
inject power. Not only does the inverter convert DC to AC power but it may also regulates the PV system
if correctly dimensioned according to the battery voltage levels.
Some requirements are indispensable for an off-grid inverter [7]:
Auto starting and adequate protection warning signs;
Peak power capacity – it should support more than two times its nominal power;
Low THD;
Low stand-by power;
High efficiency;
Voltage stability – between the range of 230V±10%;
Possibility to connect other inverters in parallel.
The waveform should be selected accordingly to the type of application. They are classified in three
classes:
1. Square wave – The simplest and cheapest DC-AC converter. Usually constituted by thyristors
controlled by the grid clock, not appropriated if the loads are not purely resistive (not suitable
for off grid systems) and high THD (≈45%);
2. Modified Sine or semi-sine wave – Provide rectangular pulses with an approximate 42% THD;
3. Sine Wave – The inverter with the lowest THD (<3%). Appropriate for motor loads such as
medical equipment, refrigerators, laser printers, etc. It is also used in grid-tied applications. It is
the most expensive inverter of the three types.
45
Pure sine wave inverters are the best choice to the correct working of motor or compressor loads once
it is the inverter with less harmonics which is seen in more detail.
Pure sine wave inverters
These inverters are controlled by a Pulse Width Modulation which allows a wide amplitude of the output
sine wave and reduce the amplitude of the low order harmonics. The general block diagram of a sine
wave inverter is represented in Figure 3.19.
Bridge
Isolation Transformer
L1
N
V
V -
+
DSP Control
Filter Switch
Figure 3.19 - Block diagram of a sine wave inverter (adapted from Texas Instruments)
The battery bank DC voltage (12, 24 or 28V) is the input of the bridge which may be full bridge or half
bridge depending on the requirement. The power transformer is generally used to isolate the battery
from the load and to step up the input voltage in grid-tied inverters or just isolate the output in grid-tied
inverters. The IGBT or MOSFET transistors are modulated by a square wave switching frequency in the
range of 20 kHz (PWM – class D circuit) to create an AC voltage. A sine wave is compared with a
triangular wave to generate a PWM output which drives the transistors bridge – triangulation scheme.
This square wave is controlled by the digital signal processor (DSP Control) which also controls the
output switch accordingly to certain alarms (low input voltage, overload, short circuit, over temperature
protection, etc). The output of the insulation transformer is filtered (low pass filter) to obtain a correct
sine wave. The typical waveform of this type of transformers it shown in advance (the inverter was
chosen and test in chapter 4) in Figure 3.20:
a)
b)
Figure 3.20 - Inverter output waveform and corresponding harmonics
0 20 40 60 80 100-400
-300
-200
-100
0
100
200
300
400
Time [ms]
Vo
lta
ge
[V
]
0 100 200 300 400 500
0
50
100
150
200
Frequency [Hz]
Vo
lta
ge
[V
]
46
The experimental output voltage of the Livre Inverter (Appendix 8.10) with a RMS=200,9V confirmed a
good quality waveform due to its low harmonic content in Figure 3.20 a). The mainly harmonic is the 1st
at 50Hz with 195.68V (45.83 dB). The 3rd, 5th, 7th and 9th harmonics (odd) have a much lower amplitude
than the 1st. The THD is 1.19% (2nd to 9th harmonics) THD lower than 3% for this inverters (Figure 3.20
b).
The use of an insulation transformer means protection against indirect contacts (no need of TN-C
differential protection) and also reduces the electromagnetic interferences. However, these inverter may
not have it and transformerless inverters are gaining popularity in solar systems. The idea behind
transformer-less switching has existed long before the PV market was even developed. A pair of field-
effect transistors operates most efficiently in a complete ON or OFF state, when no current flows through
them, and they dissipate no power. Thus, amplifying an ideal square wave would theoretically be 100%
efficient. For this this reason, transformer-less inverters have a lower stand-by consume.
Efficiency
The efficiency rating indicates what percentage of DC power is converted to usable AC power. This will
never be 100% once the inverter uses some of the input DC power itself, generally around 10-30 W. In
general, off-grid inverters have an extremely flat efficiency curve for higher than 30% of the nominal
power so less battery power is wasted. This means that the drive must be sized to operate as near as
possible of inverter’s rated power.
The efficiency of the inverter is given by [4]:
𝜂𝑖𝑛𝑣 =𝑃𝑂𝑈𝑇𝑃𝐼𝑁
=𝑃𝐴𝐶𝑃𝐷𝐶
(3.12)
An inverter's efficiency is shown in Figure 3.21 efficiency curve. The inverter's efficiency increase sharply
until it reaches its peak efficiency point. It will then remain close to level, decreasing slightly as it
approaches its rated power output. Ideally, the power used should be at or above inverter's peak
efficiency point.
The manufacturer efficiency curve of the Xantrex Prosine inverter may be seen in Figure 3.21 a) for the
model of 1000W and 1800W showing the peak after 30% of the nominal power. Figure 3.21 b) shows
in advance17 the experimental characteristic of the 1500W Livre Inverter chosen to the off-grid system.
The procedure of the test is described in more detail in Appendix 8.10 plotted with 10 points until the
maximum power supply was reached (the dashed part of the curve should does not have enough points)
. The efficiency ranged from 84-90% between 200-300W with the highest value at 800W (50% of the
nominal power like in Xantrex Inverter).
17 the inverter was selected and tested after the project in chapter 4
47
a)
b)
Figure 3.21 - Inverter efficiency characteristic a) Xantrex Prosine off-grid inverter efficiency curve for 1000 and 1800W 24V system (source: Xantrex18) and b) 1500W Livre Inverter
The actual inverters available on the market (mostly PWM regulators) are designed for lead-acid, NiCd
or NiMH batteries typically ranging in the following voltage levels:
Table 3.4 - Typical actuation voltage levels of off-grid pure sine wave inverters and PWM regulators
Nominal Voltage Minimum voltage cut-off Maximum voltage cut-off
12 V 10.5V±5% 15V±5%
24 V 20 V±5% 31±5%
48 V 41V±5% 60±5%
For example, to have a 24V nominal voltage, 8 LiFePO4 cells are necessary. The minimum voltage of
this cells is 2.5V. The inverter minimum voltage of 20V divided by 8 cells will make each cell have 2.5V
which is too close to the limit and will damage the cells. For this reason the minimum and maximum
voltage levels are not in the correct range for LiFePO4 batteries requiring the use of a BMS or an inverter
with controllable voltage levels.
18 Xantrex prosine 1000/1800 Owner’s Manual
0 200 400 600 8000
20
40
60
80
100
Power [W]
Eff
icie
ncy [
%]
48
3.6 Cables
Figure 3.22 represents the losses on the DC cables, the voltage drops and power losses associated to
it. The losses on the cables connecting the batteries to the loads are not represented for simplification19.
The length of these cables is small when compared with the length of the PV cables and might be
neglected.
PV
R
R
I
PPV BatteryPinjected
ΔV
Figure 3.22 - Cable losses
where 𝑅 is the resistance on the cable PV-battery in Ohms, 𝑃𝑃𝑉 is the power at the panel’s terminals,
𝑃𝑖𝑛𝑗𝑒𝑐𝑡𝑒𝑑 is the power at the battery's terminals and ∆𝑉 is the voltage drop on the cables. R is generally
calculated according to the manufacturer resistance and length of the cable. As an example, the
specifications of the TUV cable may be consulted in Appendix 8.8.
Joule losses on the cables:
𝑃𝐽𝑜𝑢𝑙𝑒 = 𝜌𝑙
𝑆∙ 𝐼2 = 𝑅 ∙ 𝐼2 (3.13)
The power produced by the panel 𝑃𝑃𝑉, knowing the power at the end of the cables terminals 𝑃𝑖𝑛𝑗𝑒𝑐𝑡𝑒𝑑, is
given by:
𝑃𝑃𝑉 = 𝑃𝑖𝑛𝑗𝑒𝑐𝑡𝑒𝑑 + 2𝑃𝐽𝑜𝑢𝑙𝑒 (3.14)
Voltage drop, for both DC and AC cases, is given by the following formula:
∆𝑈 = 𝐾𝑐 (𝜌1𝑆cos𝜑 + 𝜆 sin𝜑) × 𝐿 × 𝐼𝐵 (3.15)
where 𝐾𝑐 is a coefficient equal to 1 for three phase and equal to 2 for single-phase circuits; 𝜌1 is the
resistivity of the conductors at the usual working temperature (1.25 times the resistivity at 20ºC – 0.0225
Ω.mm²/m for copper and 0.036 Ω.mm²/m for aluminium); S is the section of the conductors in square
millimetres; cos𝜑 is the power factor (cos𝜑 = 1 for pure resistive loads making sin𝜑 = 0); λ is the linear
reactance per length unit of the conductors (default value 0.08 mΩ/m); L is the length of the cable (m)
and 𝐼𝐵 is the current (A).
19 accordingly to the experiment made in chapter 5
49
The voltage drop in percentage is given by:
∆𝑈 [%] =∆𝑈
𝑈0× 100 (3.16)
where 𝑈0 is the voltage between the phase and the neutral.
3.7 PV System Efficiency
PVAC
Consumers
PAC
Battery Pack
BMS
PDCPDC, injectedInverter
ƞinv
ƞPV
ƞBAT
ƞBMS
Figure 3.23 - Off-grid PV system connection diagram
The total efficiency of the module interface system with the loads accordingly to the diagram of Figure
3.23 may be obtained by:
𝜂𝑠𝑦𝑠𝑡,𝑎𝑢𝑡 = 𝜂𝐵𝑀𝑆 ∙ 𝜂𝐵𝐴𝑇 ∙ 𝜂𝑖𝑛𝑣 ∙ 𝜂𝑙𝑜𝑠𝑠𝑒𝑠 (3.17)
where 𝜂𝑙𝑜𝑠𝑠𝑒𝑠 = 1 − 𝑃𝑙𝑜𝑠𝑠𝑒𝑠𝑐𝑎𝑏𝑙𝑒𝑠[%] is a factor that considers the losses on the cables.
In case the energy “injected” in the battery pack and the energy consumed by the loads is known, the
respective efficiency may be calculated by:
𝜂𝐵𝑀𝑆+𝐵𝐴𝑇+𝐼𝑛𝑣 =𝐸𝑂𝑈𝑇𝐸𝐼𝑁
=𝐸𝐿𝑜𝑎𝑑𝑠𝐸𝑃𝑉
(3.18)
However, as the batteries store energy, this efficiency needs to be estimated during a time period where
the batteries start at a certain SOC return to that same SOC point. This may be more accurately done
by a battery voltage reference point. In other words, the battery should start the test charged with a
voltage 𝑈𝑜, discharge during a certain time and charge again until 𝑈𝑜. Knowing the energy in and out
during this period, the efficiency may be calculated.
50
3.8 Conclusions
The purpose of this chapter is to introduce the basic principles of operation of all the system components
– consume, photovoltaic production, storage, control and energy conversion.
To supply the loads of the system, polycrystalline panels were chosen, once the price is more
competitive consequence of the highest presence in market. Besides, this was the highest power panel
gently provided by Resul – Energy Equipments, S.A on 2014.
Comparing weight versus available energy storage, LFP batteries are about one-third the weight and
about half the volume of a lead-acid (LA) battery with equivalent energy storage. Additionally, the
storage capacity of LAs drops by 50% at -20°C, compared to 8% with LFP. Keeping lead-acid batteries
warm so that they maintain reasonable capacity in cold climates can be challenging, giving advantage
to LFPs. LiFePO4 have the best thermal stability of all seen lithium ion types being then appropriate to
photovoltaic applications.
The characteristic curve of these batteries requires special care and project of the controllers. A battery
management system (BMS) to manage the “health” of the batteries is the best option to ensure reduced
maintenance and long service time. Some solar regulators, like PWM regulators, are not yet design for
this type of batteries. This process is done better with top balancing, once the batteries are expected to
be at full status as much as possible. This way the balancing will be done more often. At high SOC, the
cell voltage changes rapidly, which also gives better data on the true SOC.
To convert DC to AC power, pure sine wave inverters are the best inverters type to feed motor loads
and have less than 40% THD voltage compared with modified sine and square wave. Inverters should
have no screen, as this always increases stand-by power. Besides, its standby consume may be
reduced with no insulation transformer on the AC side (requires differential circuit breaker).
51
4 Autonomous PV System – Project
This chapter aims to project the photovoltaic system for Luena zone in Angola. The project is carried
out in accordance with the equipment kindly provided by Resul – Energy Equipments, S.A.20, constituted
by two 190W STP190S monocrystalline and two 225W STP225 Suntech PV Panels. The LiFePO4
batteries were also previously obtained by Electrical Department (DEEC) with the objective of studying
its performance in this type of systems.
There are several possible processes to project an autonomous PV system. Processes differ in the
initial condition that will affect the design of the remaining components [7]:
Maximum daily energy calculation;
Maximum power consumption calculation;
Number of PV modules calculation;
Battery capacity calculation.
To scale a PV system is necessary to know the average radiation values at the installation site. To
determine these values radiation PVGIS software was used [26]. This software allows to estimate online
the photovoltaic solar energy for isolated or grid-tied systems within the European and African continent
and uses Google maps ® to choose the location. This is a reliable and free software that justified the
choice for the project.
The maximum daily energy calculation process considers the following steps which will be seen in detail
[7]:
1) Determine the dairy energy consumed by the loads;
2) Calculate the energy production of the PV modules at a given tilt, β, and azimuth angle, γ;
3) Project the peak power of the PV modules;
4) Calculation of the batteries parameters;
5) Project of the solar regulator;
6) Project of the autonomous or off-grid inverter.
4.1 Loads
The more accurate the consumption data the better the project, for this reason the loads consumptions
and peak currents were analyzed and are presented next. In this chapter some assumptions were made
to match the loads profile with the average consumption of a small family house in Angola.
20 Resul website may be consulted at http://www.resul.pt/
52
4.1.1.1 Refrigerator Load Diagram
The refrigerator is the load with the load diagram more complex. The refrigerator used in the experiments
is the Kentt 201E which technical data can be seen in Appendix 8.11. This refrigerator has 7 refrigeration
levels consuming more energy for higher levels. The load diagram was obtain experimentally for a more
precise project. For a reasonable temperature vs energy use, level 3 is the appropriate one. However,
once it was not possible to simulate a diary use of the refrigerator, the level 4 was the chosen one to
compensate the extra consume caused by the opening of the door on level 3. Figure 4.1 represents the
refrigerator active power diagram for one hour at level 4 with no opening of the door measured using
Fluke 1735 power logger (Appendix 8.3). The refrigerator compressor is on for periods of approximately
4 min at an average power of 127W four times per hour.
Figure 4.1 - KENT 201E refrigerator load diagram (1h)
4.1.1.2 Lights and TV Load Diagram
For lighting purposes is estimated a power of 60W during 5 hours per day, from 18:30 to 22:30 and from
6:00 to 7:00 according to the working hours in Angola. This power may be achieved with a single lamp
or several with less power. This is the second more important load of a family’s house.
For a regular house entertaining purposes is estimated a 40W TV which is on for 2 hours per day from
20:00 to 22:00 or it can be substituted by a more important electronic device if consuming the same
equivalent energy.
4.1.1.3 Total Load Diagram
The total load diagram is the sum of the three previous loads’ consume profile represented in Figure 4.2
logged with Fluke 1735 (Appendix 8.3).
00:10 00:20 00:30 00:40 00:50 01:000
50
100
150
200
250
300
350
Time [h:min]
Po
we
r [W
]
53
Figure 4.2 - Total daily load diagram
The refrigerator switches on and off in regular periods along the day and the lights and TV working
periods are those in 4.1.1.2. Unlike the peaks of the inductive loads (refrigerator and TV) are not visible
once the integration period of the logging is higher. The maximum power of the system is 250W between
19:00 and 22:00.
4.1.2 Experimental loads power
Table 4.1 shows the experimental electrical parameters of Kentt 201E refrigerator (Appendix 8.11),
Philips 60W incandescent light and Sony KV-14LT1E 13’’ TV (appendix 8.12) measured with Fluke 1735
power logger (Appendix 8.3). The energy values represent the refrigerator daily consumption at level 4.
The autonomous PV system essays were performed in March 2015 with two 225W panels (Appendix
8.7) installed on top of the third floor of North tower building at IST (38º44’15.9’’N and 9º08’17.6’’W).
The approximate sun's trajectory is shown in Figure 5.8. The solar maximum angle is 48°at 13:00.25 The
panels are oriented to the geographical south (7° deviation from south to the west) with an inclination
angle of 𝛽 = 35° (optimal angle generally used in Portugal)26 [7]. For the calculations, it was used the
convention of Figure 2.7 (𝐴𝑧𝑠 = −173°). Due to its location, the PV panels are shaded between 7:00
and 8:30 caused by the building height and after 14:30 caused by the north tower building (located right
on the image) which reduces the solar radiation at least 3 hours/day resulting in a test constraint.
Figure 5.8 - PV installation site and sun trajectory in March 2015
This test was done for a load constituted by the refrigerator, TV and lights projected in 4.1 having a daily
consume of 1300 Wh and aims to prove the viability of the system.
Figure 5.9 shows in red, the measured solar radiation on a horizontal surface for an almost clear sky
working day on 14th March 2015 at the IST meteorological station which is installed on the top of the
South Tower where the shading effect after 14:30 does not happen (Appendix 8.13). It also shows in
black, the radiation incident on the PV tilted plane (35°) calculated accordingly to the expression (2.11)
in section 2.2.3. To simplify the calculations the direct irradiance on the horizontal plane 𝐺𝑏_ℎ𝑜𝑟𝑖𝑧 was
substituted by the total horizontal irradiance measured at the IST meteorological station once the diffuse
radiation represents less than 10% of the total radiation as seen in 2.2.3. The irradiance is available
from 7:00 until 18:30 at the meteorological station. At the PV location the irradiance is similar but is not
available from 7:00 to 8:30 and after 14:30 due to the location limitation described before - represented
by a grey shadow area on the graph in Figure 5.9. This day particularly had some clouds in the morning
which is visible the irradiance from 9:00 to 12:00 and clear sky from 12:00 until the end of the day. The
25 According to the webtool: http://www.sunearthtools.com/dp/tools/pos_sun.php 26 Accordingly to a report by Morse & Czarnecki (1958) the optimally fixed tilt angle is a value 0.9 times the
latitude. In Lisbon that is 38º*0.9=34.2º [25]
70
maximum irradiance on the module was 1056 W/m2 at 13:00, an average wind speed of 3.33 m/s, an
average ambient temperature of 12ºC and a cells’ temperature of 15ºC [54].
Figure 5.9 - Total solar irradiance incident on a horizontal (red) and 35º tilted plane (black) on 14th March 2015 with clouds
The solar power is only injected in the batteries when the irradiance is available and voltage of the
panels is higher than the voltage of the batteries. The figure below shows the evolution of both voltages
during the day. The voltage of the panels depends on the total irradiance incident on the module on 14th
March plotted in Figure 5.9. The PV voltage (blue curve) is approximately one volt higher than the battery
pack voltage from 8:30 until 14:00 when the sun is hitting the panels. Between 9:00 and 11:00 is visible
a decay of the voltage caused by the clouds in this period which consequently reduces the power
injected as seen further in Figure 5.12. The peaks seen on the graph after 14:00 o’clock represent the
opening of the charging relay when the charging is done and one of the cells voltage is above 3.5 V,
remaining open for approximately 10 minutes. Then, the relay closes again connecting the panels when
the voltage of the cells goes under 3.35 V. Therefore, the voltage observed is the PV open circuit voltage
(𝑉𝑂𝐶) of 34V for an approximate 1000W/m² irradiance. After 14:30 the PV voltage is lower than the
<10mA). The control system had some limitations in reading the output currents of the batteries. Since
the inverter voltage input signal is a square wave the current is not a constant DC value. This introduces
small errors on the real SOC Coulomb counting estimation method which may lead the controller to
open the circuit before the batteries are actually discharged. This may be fixed by calibrating the current
sensor offset or ignoring the estimation of SOC that could make the system simpler and cheaper to
88
implement. Nevertheless, the cells’ protection is ensured once the relays open the circuit based on
voltage signals which work properly.
On site, the polycrystalline solar panels used showed a daily average efficiency of 10.8% and the total
system a 75% efficiency which may be improved about 5%, if the panels are installed at a distance lower
than 10m of the battery pack.
Actually, PV modules present a relatively high cost, however its cost have been reducing across the
years. In the future, it is expected to observe a growing investment in these systems by developing
countries such as Angola. This work was an approach to autonomous PV isolated systems with the
objective of clarify and deepen some aspects in the management and control of LiFePO4 battery storage.
It is expected to open new opportunities for further investigations which may improve with more detail
methods of operation of these systems.
89
6.2 Future Work
Due to the need for access to energy, environmental concerns and the growing demand for renewable
energy, isolated photovoltaic systems should also feel an increase in demand. It is expected that "plug
and play" (easy installation) solutions emerge. The controller used in this system has very favorable
characteristics to its comprehension and modification that may in the future be improved and compacted
in order to integrate and manufacture a system ready easy to install.
For end users, this BMS has features that are not of the most importance to isolated systems. This
happens since this BMS was firstly designed to electric vehicles and then integrated in isolated systems.
Among this features is the SOC estimation. For isolated systems in rural places, the consumer will not
have a computer access the sensors’ readings, so he will not have information on the batteries’ SOC.
Thus, a controller of this type may be based only on voltage levels, both switch on/off as well as
reconnection voltage. This way, it may be set that the relays are to be reconnected when the
predetermined voltage is reach, not the predetermined SOC as it happens actually. The system will be
more economical if the electronic boards are pre-programmed to the voltage levels, in LiFePO4 this
levels are between 3 to 3.6V.
It would be also interesting to study the batteries’ manufacturing energy cost versus the energy delivered
during its lifetime in a more environmental approach. The temperature management is also an important
aspect to improve, however this is more relevant with high current applications.
As concluded before, to improve significantly its autonomous capacity, these systems should be
integrated with alternative energy sources that allow the supply of the loads not only during the hours of
sun but during all day. Wind sources are the ideal solution, this will increase the cost of the installation
but decrease the effort of batteries as the loads can be supplied almost exclusively by generation. It
would be interesting to analyse the cost benefit of implementing a wind generator, in opposition to
increase the solar power production and storage capacity.
The development and continuous improvement of energy sources shows a promising future for these
systems making possible to provide electric energy for all.
90
91
7 Bibliography
[1] “Relatório de Energia em Angola,” Centro de Estudos e Investigação Científica da Universidade Católica de Angola, Angola, 2011.
[2] “Emissions Reduction Profile - Angola,” UNEP RISØ Centre - United Nations, Roskilde, Denmark, 2013.
[3] M. A. Green, K. Emery, Y. Hishikawa, W. Warta and E. D. Dunlop, “Solar cell efficiency tables (Version 45),” Progress in photovoltaics: Research and applications, vol. 23, no. 1, pp. 1-9, 2015.
[4] R. Castro, Uma Introdução às energias Renováveis: Eólica, Fotovoltaica e Mini-hídrica, 1st ed., IST Press, 2011.
[5] IEA-PVPS, “Trends in Photovoltaic Applications - Survey Report of Selected IEA Countries between 1992 and 2013,” IEA-PVPS T1-25:2014, 2014.
[6] “Technology Roadmap - Solar Photovoltaic Energy,” International Energy Agency, 2014.
[7] F. A. d. Sousa Pereira e M. Â. Sarmento Oliveira, Curso Técnico Instalador de Energia Solar Fotovoltaica, Publindústria, Edições Técnicas, 2011.
[8] F. M. Emanuel, “Módulo fotovoltaico com seguimento da posição solar,” Faculdade de Ciências e Tecnologia - Universidade Nova de Lisboa, Dissertação para obtenção do grau de Mestre em Engenharia Electrotécnica e de Computadores, Lisboa, 2009.
[9] A. Luísa, “Validação Experimental de Modelos Matemáticos de Componentes de Sistemas Fotovoltaicos,” Instituto Superior Técnico, Universidade de Lisboa, Dissertação para obtenção do grau de Mestre em Engenharia Mecânica, Lisboa, 2007.
[11] Frank Vignola, Joseph Michalsky and Thomas Stoffel, “Solar Resource Definitions and Terminology,” in Solar and Infrared Radiation Measurements, CRC Press, 2012, pp. 7-28.
[12] Abdul Qayoom Jakhrani, Al-Khalid Othman, Andrew R.H. Rigit, Saleem Raza Samo and Shakeel Ahmed, “Estimation of Incident Solar Radiation on Tilted Surface by Different Empirical Models,” International Journal of Scientific and Research Publications,, vol. 2, no. 12, 2012.
[13] Ali Mohammad Noorian, Isaac Moradi and Gholam Ali Kamali, “Evaluation of 12 models to estimate hourly diffuse irradiation on inclined surfaces,” Renewable Energy, vol. 33, pp. 1406-1412, 2008.
[14] Joze Rakovec, Klemen Zaksek, Kristijan Brecl, Damijana Kastelec and Marko Topic, “Orientation and Tilt Dependence of a Fixed PV Array Energy Yield Based on Measurements of Solar Energy and Ground Albedo - a Case Study of Slovenia,” in Energy Management Systems, Dr Giridhar Kini , Ed., InTech, 2011, Available from: http://cdn.intechopen.com/pdfs-wm/17056.pdf, pp. 145-160.
[15] Abdul Majeed Muzathik, Wan Mohd Norsani Wan Nik, Khalid Samo and Mohd. Zamri Ibrahim, “Hourly Global Solar Radiation Estimates on a Horizontal Plane,” Journal of Physical Science, vol. 21, no. 2, pp. 51-66, 2010.
[16] B.Y.H Liu and R.C Jordan, “The interrrelationship and characteristic distribution of direct, diffuse and total solar radiation,” Solar Energy, vol. 4, no. 3, pp. 1-9, 1960.
92
[17] S. A. Klein, “Calculation of Monthly Average Insolation on Tilted Surfaces,” Solar Energy, no. 19, p. 325, 1977.
[19] R. W. Harrigan and W. B. Stine, Solar Energy Systems Design, John Wiley and Sons, 1986.
[20] M. Iqbal, An Introduction to Solar Radiation, Academic Press, Inc, 1983.
[21] P. I. Cooper, “The Absorption of Solar Radiation in Solar Stills,” Solar Energy, vol. 12, no. 3, pp. 333-346, 1969.
[22] A. George and R. Anto, “Analytical and Experimental Analysis of Optimal Tilt angle of Solar Photovoltaic Systems,” in International COnference on Green Technologies (ICGT), Trivandrum, 2012.
[23] J. A. Duffie and W. A. Beckman, Solar Engineering of Thermal Processes, 2nd ed., New York (USA): Wiley-Interscience, 1991.
[24] M. M. El-Kassaby and M. H. Hassab, “Investigation of a variable tilt angle Australian type solar collector,” Renewable,, vol. 4, no. 3, pp. 327-332, 1994.
[25] R. N. Morse and J. T. Czarnecki,, “Flat heat solar absorbers: The effect on incident radiation of inclination and orientation,” in Rep E.E. 6 Commonwealth CSIRO, Melbourne, Australia, 1958.
[26] “PVGIS Climate-SAF database,” Institute for Energy and Transport (IET), 2001-2012. [Online]. Available: http://re.jrc.ec.europa.eu/pvgis/imaps/index.htm.
[27] D. R. Myers, I. M. Reda, S. M. Wilcox and T. L. Stoffel, “Uncertainty Analysis for Broadband Solar Radiometric Instrumentation Calibrations and Measurements: An Update,” in Preprint NREL/CP-560-3620, to be presented at the World Renewable Energy Congress VIII, Denver, Colorado, 2004.
[28] A. Spena and C. Cornaro, “Global, direct and diffuse radiation measurements at ground by the new environmental station of university of Rome Tor Vergatta,” Rome - Italy, 1990.
[29] S. Cristóvão, “Fontes de Energias renováveis no Contexto angolano - Seminário sobre Energia Eléctrica,” Ministério das Energias e das Águas, Luanda, 2011.
[32] M. Shahin, Hydrology and Water Resources of Africa, Dordrecht: Luwer Academic Publishers, 2003.
[33] “Empresa Nacional de Electricidade, E.P.,” [Online]. Available: http://194.79.83.194:9091/Pages/Home.aspx. [Accessed 9 June 2015].
[34] “African Energy,” [Online]. Available: http://www.africa-energy.com/angola?type=articles. [Accessed 15 July 2015].
[35] B. C. d. C. Balão, “Dimensionamento de Sistemas Autónomos Modulares Eólicos/PV com armazenamento,” Faculdade de Engenharia da Universidade do Porto, Dissertação para obtenção do grau de Mestre em Engenharia Electrotécnica e de Computadores, Porto, 2012.
93
[36] “Programa de Transformação do Sector Eléctrico,” in Conferência, Luanda, 2013.
[37] V. Azevedo, “A lâmpada de Edison vai apagar-se de vez,” Expresso.pt, 25 August 2012. [Online]. Available: http://expresso.sapo.pt/a-lampada-de-edison-vai-apagar-se-de-vez=f748822. [Accessed March 2015].
[38] “Angola reduz emissão de gases na atmosfera,” Jornal de Angola, p. 40, 28 March 2015.
[39] C. Barthel and T. Götz, “Test procedures, measurements and standards for refrigerators and freezers,” bigEE, 2012.
[40] M. A. Green, Solar cells: operating principles, technology, and system applications, Englewood Cliffs NJ: Prentice-Hall, 1982.
[41] D. E. Carlson, G. Ling and G. Ganguly, “Temperature dependence of amorphous silicon solar cell PV parameters,” IEE Phot Spec C, vol. 28, pp. 707-712, 2000.
[42] “Battery+4Life Research,” Battery Plus For Life, [Online]. Available: http://batteryplusforlife.com/research.html. [Accessed 03 02 2015].
[43] F. P. Tredeau and Z. M. Salameh, “Evaluation of Lithium Iron Phosphate Batteries for Electric Vehicles Application,” IEEE, pp. 1266-1270, 2009.
[44] C. M. d. C. Menaia, “Desenvolvimento do sistema de protecção,monitorização e gestão de baterias de iões de lítio do tipo LiFePO4 para aplicação em veículos eléctricos,” Instituto Superior de Engenharia de Lisboa, Dissertação para obtenção do grau de Mestre em Engenharia Electrotécnica, ramo de Energia, 2013.
[45] G. Arnold, J. Garche, R. Hemmer, S. Strobele, C. Vogler and M. Wohlfahrt-Mehrens, “Fine-particle lithium iron phosphate LiFePO4 synthesized by a new low-cost aqueous precipitation technique,” Journal of Power Sources 119-121, pp. 247-251, 2003.
[46] M. J. Almeida, “Analysis of the Novel LiFePO4 Batteries for Electrical Traction Batteries,” Instituto Superior Técnico - Universidade de Lisboa. Dissertação para a obtenção de grau de Mestre em Engenharia Electrotécnica e de Computadores, Lisboa, Setembro 2015.
[47] M. F. C. d. Reis, “State-of-Charge (SOC) Prediction of Lithium Iron Phosphate (LiFePO4) Batteries for Automotive Application Based on Intelligent Systems,” Instituto Superior Técnico, Universidade de Lisboa, Dissertação para obtenção do grau de Mestre em Engenharia Electrotécnica, 2014.
[48] D. Andrea, Battery Management Systems for Large Lithium-Ion Battery Packs, Artech House, 2010.
[49] Y. Xing, W. He, M. Pecht and K. L. Tsui, “State of charge estimation of lithium-ion batteries using the open-circuit voltage at various ambient temperatures,” Applied Energy, vol. 113, pp. 106-115, 2014.
[50] M. R. Khan, G. Mulder and J. V. Mierlo, “An online framework for state of charge determination of battery systems using combined system identification approach:,” Journal of Power Sources, vol. 246, pp. 629-641, 2014.
[51] M. Daowd, N. Omar, P. Van Den Bossche and J. V. Mierlo, “A Review of Passive and Active Battery Balancing based on MATLAB/Simulink,” International Review of Electrical Engineering, 2011.
[52] S. Wen, “Cell balancing buys extra run time and battery life,” Texas Instruments Incorporated.
94
[53] B. Herteleer, J. Cappelle and J. Driesen, “An autonomous photovoltaic system sizing program for office applications in Africa,” in International Conference on Renewable Energies and Power Quality, Santiago de Compostela - Spain, 2012.
[55] M. Šúri, T. Huld, E. D. Dunlop and H. A. Ossenbrink, “Potential of solar electricity generation in the European Union member states and candidate countries,” Solar Energy, 81, 1295–1305, http://re.jrc.ec.europa.eu/pvgis/, 2007.
[56] T. Huld, R. Müller and A. Gambardella, “ A new solar radiation database for estimating PV performance in Europe and Africa,” , 86, 1803-1815, 2012.
[57] M. G. Cardoso Pitra, “Renewable Energies in Angola - Actual situation and perspectives,” in 5th Anual RELOP Conference, 2012.
[58] M. GIMÉNEZ, V. GUZMÁN, J. RESTREPO, J. ALLER, A. BUENO, J. VIOLA, A. MILLÁN and A. CABELLO, “Plataforma: Development of an integrated dynamic test System to determine power electronics systems performance,” Revista de la Facultad de Ingeniería Universidad Central de Venezuela, Vol. 23, N° 3, pp. 91-102, May 2008.
[59] S. K. Nashih, “Numerical Approach for the Analysis of Shading Effects on Photovoltaic Panels,” Universidade de Lisboa - Instituto Superior Técnico, Dissertação para obtenção do grau de Mestre em Engenharia Electrónica, Lisboa, 2015.
[60] A. E. Fitzgerald, J. C. Kingsley and S. D. Umans, Electric Machinery, 6th ed., New York: McGraw-Hill, 2003.
[61] T. Wildi, Electrical Machines, Drives, and Power Systems, New Jersey: Prentice Hall, 2002.
[62] W. Borong, R. Yonghuan and L. Ning, “LiFePO4 Cathode Material,” in Electric Vehicles - The Benefits and Barriers, D. S. Soylu, Ed., InTech, 2011.
[63] Y. Barsukov, “Battery Cell Balancing: What to Balance and How,” Texas Instruments.
[64] R. Lieberwirth, S. Lilwitz, J. Reinbach, D. G. Reiss and N. Reidel, “Virtual Energy Cells,” dSpace Magazine, pp. 8-9, 2010.
[65] 123electric.nl, “123electric.nl,” December 2012. [Online]. Available: http://www.123electric.nl/uploads/file/BMS_EV_Manual_v1_3.pdf. [Accessed October 2014].
[66] P. Umbelino, Instituto Superior Técnico, Universidade de Lisboa, Dissertação para obtenção do grau de Mestre em Engenharia Electrotécnica e de Computadores, 2015. [Online].
[67] M. Lefèvre, M. Albuisson and L. Wald, “Description of the software Heliosat-II for the conversion of images acquired by meteosat satellites in the visible band into maps of solar radiation available at ground level,” Groupe Télédétection & Modélisation - Centre d'Energétique – Armines / Ecole des Mines de Paris, France, 2002.
[68] D. W. Medugu, A. B. Adisa, F. W. Burari and M. A. Abdul' Azeez, “Solar radiation - Correlation between measured and predicted values in mubi, Nigeria,” International Journal of Science and Technology Education Research, vol. 4, no. 1, pp. 11-17, January 2013.
[69] N. Omar, M. A. Monem, Y. Firouz, J. Salminen, J. Smekens, O. Hegazy and J. V. Mierlo, “Lithium iron phosphate based battery – Assessment of the aging,” Applied Energy, vol. 113, pp. 1575-1585, 2014.
95
8 Appendix
8.1 National Instruments NI-USB6008/9 Data Acquisition Device
96
97
8.2 Current Transducer LA-25-NP Datasheet
98
99
Figure 8.1 - LA 25-NP current transducer characteristic
8.3 Power Logger Fluke 1735
8.4 Tektronix TDS 2001/2012C Oscilloscope
y = 5,6026x - 0,0833
0
1
2
3
4
5
6
7
8
9
0 0,2 0,4 0,6 0,8 1 1,2 1,4 1,6
Prim
ary
cu
rre
nt [A
]
Secondary voltage [V]
Model Fluke 1735 Three-Phase Power Logger
Memory 3.5MB for measuring data
Sample Rate 10.24 kHz
V-RMS wye resolution 0.1V
Operating Error ±0.5% of measured value +10 digit
A-RMS resolution 0.01A
Operating Error ±1% of measured value +10 digit
Brand Tektronix
Model TDS 2001/2012C
Analog Bandwidth 100MHz
Sample Rate 2GS/s
Record Length 2.5k Point
Analog Channels 2
100
8.5 Data logging
Data logging is the process of using a computer to collect data through sensors, analyse, save and
output the results. Data logging also implies the control of how the computer collects and analyses the
data. Data logging is commonly used in scientific experiments and in monitoring systems where there
is the need to collect information faster than a human can and in cases where accuracy is essential.
8.5.1 Voltage log
The voltage logging in this work was done using the four National Instruments NI USB-6009 DAQ® input
channels and LabVIEW Signal Express software for data acquisition.
The maximum input voltage of the DAQ is 20V (Appendix 8.1). All the signals lower than 20V like the
cell’s voltage may be directly connected to the DAQ. The voltage higher than 20V as the PV voltage or
the battery bank voltage are read through a voltage divider to decrease the voltage by a 2 times factor.
Figure 8.2 - Connecting a Differential Voltage Signal [NI USB-6008/6009 User Guide]
After the DAQ have been connected the LabVIEW Signal Express or LabVIEW software may be used
to log the data. LabView software offers the capabilities of LabView Signal Express but in at a lower and
more complex level of interface. For logging purposes Signal Express is more intuitive and has an easy
to use interface.
Figure 8.3 - LabVIEW Signal Express Monitor / Record interface
For a primary nominal current, 𝐼𝑃𝑁 = 25 𝐴 and primary maximum current, 𝐼𝑃 = 36 𝐴, the turns ratio is
(according to datasheet 8.2):
𝐾𝑁 = 𝑚 =𝑈1𝑈2⁄ = 1/1000 (8.2)
Substituting (8.2) in (8.1):
𝐼2 = 𝑚 × 𝐼1 (8.3)
With a peak current no higher than 36𝐴, the sensor works in the linear zone which means that the
current and the voltage have a linear relation.
This secondary voltage 𝑈2 = 𝑈𝑆, is the voltage at the terminals of the resistor 𝑅𝑀 and can be theoretically
calculated using expression (8.4) knowing the secondary current.
𝑈𝑆 = 𝑅𝑀 × 𝐼2 (8.4)
The secondary voltage 𝑈𝑆 is logged using the NI-6009 and the primary current may be obtained by the
following voltage/current relation substituting (8.3) in (8.4) :
𝐼1 =𝑈𝑆
𝑅𝑀 ×𝑚 (8.5)
Figure 8.5 - LEM LA25-NP equivalent circuit according to datasheet parameters [58]
To achieve the best relation, the sensor’s experimental characteristic may was obtained measuring
several points with a current source connected to the primary of LA-25 NP starting from 0A and reading
the secondary’s voltage using the following circuit:
103
DCV
1 A
0 V
M
OUT
IN
UpUs
Figure 8.6 - Experimental setup to obtain LA-25 NP characteristic
This leads to the relation between voltage and current:
𝐼 [𝐴] =𝑉𝑠
0.1785− 0.0833 (8.6)
Figure 8.7 - Experimental characteristic test of the LEM LA25-NP
The voltage measured is 0.927 𝑉 which corresponds to a current of 5.224 A according to the expression
(8.1). The current read by the current probe is 5.157 A, giving a 47 mA or a 0.9% current error.
104
8.6 CALB SE130AHA Battery Cell Datasheet
Model name SE130AHA Alternative product marking CSE130AH Nominal voltage 3,3 V Operating voltage under load is 3,0 V Capacity 130 Ah +/- 5% Internal impenetrableness <1 mOhm 1kHz AC Operating voltage min 2.6V - max 3,6 V At 80% DOD Discharging cut-off voltage 2.5 V The cells is damaged if voltage drops
below this level Charging cut-off voltage 3,65 V The cells is damaged if voltage exceeds this level Recommended charging -
discharging Current 39 A 0,3C
Maximum short-time discharging current 1000 A 10C period = 10s
