Design, feasibility study and environmental analysis of a ...
102
Design, feasibility study and environmental analysis of a grid connected PV system in Cambodia Edoardo Gionta Thesis to obtain the Master of Science Degree in Energy Engineering and Management Supervisor: Prof. Rui Manuel Gameiro de Castro Examination Committee Chairperson: Prof. Luís Filipe Moreira Mendes Supervisor: Prof. Rui Manuel Gameiro de Castro Member of the Committee: Prof. Nuno Alexandre Soares Domingues November 2018
Transcript of Design, feasibility study and environmental analysis of a ...
Design, feasibility study and environmental analysis of
a grid connected PV system in Cambodia
Edoardo Gionta
Thesis to obtain the Master of Science Degree in
Energy Engineering and Management
Supervisor: Prof. Rui Manuel Gameiro de Castro
Examination Committee
Chairperson: Prof. Luís Filipe Moreira Mendes
Supervisor: Prof. Rui Manuel Gameiro de Castro
Member of the Committee: Prof. Nuno Alexandre Soares Domingues
November 2018
II
III
Abstract
Cambodia’s economy is growing rapidly as its CO2 emissions. The low quality of the electricity from the grid, its
high price and the country high insolation make Cambodia favorable for the development of solar energy. The
thesis aims to analyze performance indicators of the expected production of a medium-scale PV system
connected to a MV grid. A garment factory in Cambodia has been analyzed, with PVsyst software, as case study.
The influence of the recently approved new solar regulation was added to the traditional energetic,
technological, economic and environmental performance indicators. The project design together with the
shading analysis of the system for the garment factory have been developed. With PVsyst software, the solution
that maximizes the performance ratio has been found to be 707 kWp, with JKM340PP-72 polycrystalline
modules from Jinko Solar coupled with SUN2000-36KTL Huawei inverters. The result has been guided by the
regulation restrictions of no grid export and PV system size limited to less than half the contracted size. From
the economic assessment of the project has emerged how the investments are affected by the regulation. The
pay-back time of the system goes from around 4 years with the old regulation to more than 65 years with the
new regulation. Both the NPV and net cash-flow, which were positive with the old regulation, are now resulting
in extremely negative values. Nevertheless, from the carbon balance of the project has emerged that the
installation can save up to 11200 tCO2 during its lifetime.
Keywords: Photovoltaic System, Solar Energy, Cambodia, Project Design, PVsyst
IV
Resumo
A economia do Camboja está a crescer rapidamente, assim como as suas emissões de CO2. A qualidade baixa
da rede elétrica, o seu alto preço e a sua alta insolaçao tornam o Camboja favorável para o desenvolvimento da
energia solar. A tese tem como principal objetivo analisar minuciosamente a produção esperada de um sistema
fotovoltaico de média escala conectado a uma rede de MT. Uma fábrica de roupas no Camboja foi analisada,
com o software PVsyst, como estudo. A influência da nova regulação solar recentemente aprovada foi
adicionada aos tradicionais indicadores de desempenho energético, tecnológico, econômico e ambiental. O
design e análise de sombreamentodo do sistema foram desenvolvidos para uma fábrica de vestuário. De acordo
com o software PVsyst, a solução que optimiza o performance ratio é de 707kWp com módulos policristalinos
JKM340PP-72-V da Jinko Solar com inversores SUN2000-36KTL da Huawei. O resultado é restricto pela regulação
que não só não permite a injecção de energia na rede, como também limita a dimensão máxima do sistema a
metade da potência contratada. O tempo de retorno do investimento sobe de 4 para 65 anos com a mudança
de regulamentação. Tanto o Valor Actual Líquido como o fluxo de caixa liquido, que eram positivos com a antiga
regulamentação, apresentam-se agora negativos. Ainda assim, olhando para o balanço do carbono, a instalação
do projecto pode levar à poupança de emissões na ordem dos 11200 tCO2 durante o tempo de vida do projecto.
Palavras chave: Sistema Fotovoltaico, Energia Solar, Camboja, Design de Projeto, PVsyst
V
Table of Contents
Abstract .......................................................................................................................... III
Resumo ........................................................................................................................... IV
Table of Contents ............................................................................................................. V
List of Figures ................................................................................................................. VII
List of Tables ................................................................................................................... IX
Abbreviations ................................................................................................................... X
Chapter One – Introduction .............................................................................................. 1
1.1. Cambodia Country Summary ............................................................................................................1
1.2. Cambodia’s Environment ..................................................................................................................1
1.3. Energy and Business in Cambodia .....................................................................................................1
1.4. Purpose of the Work.........................................................................................................................2
1.5. Thesis Structure ................................................................................................................................2
Chapter Two – Cambodia State of Art ............................................................................... 4
2.1. Major Institutions of Electricity Industry in Cambodia ......................................................................4
2.2. Electricity Sector in Cambodia ..........................................................................................................5
2.3. Environmental aspects .....................................................................................................................8
2.4. Solar Power in Cambodia ..................................................................................................................9
2.4.1. The New Solar Regulation in Cambodia ........................................................................................ 10
2.4.2. Energy Challenges and Solar Generation ...................................................................................... 13
2.5. Economic sector in Cambodia ......................................................................................................... 14
Chapter Three – Case Study Definition and Preliminary Assessment ................................ 16
3.1. Garment Factory Case Study: Overview, Goals and Scopes ............................................................. 16
3.2. Site Visit and Load Profile ............................................................................................................... 16
3.3. Preliminary Design.......................................................................................................................... 20
3.3.1. Shading Losses .............................................................................................................................. 20
3.3.2. 3D modelling ................................................................................................................................. 22
Chapter Four - System Design with PVsyst ...................................................................... 27
4.1. Geographical Location and Meteorology ........................................................................................ 27
4.2. Orientation ..................................................................................................................................... 35
VI
4.3. Matching PV arrays and Inverters ................................................................................................... 37
4.4. Detailed Losses ............................................................................................................................... 40
4.5. Module Layout ............................................................................................................................... 48
4.6. Simulation Results and Methodology ............................................................................................. 49
Chapter Five – Economics and Environment .................................................................... 54
5.1. Economic Assessment of the Project .............................................................................................. 54
5.2. Environmental Analysis .................................................................................................................. 60
Further Work .................................................................................................................. 64
Conclusion ...................................................................................................................... 65
References ..................................................................................................................... 66
Appendices .................................................................................................................... 69
Appendix 1 ................................................................................................................................................. 69
Appendix 2 ................................................................................................................................................. 72
Appendix 3 ................................................................................................................................................. 75
Appendix 4 ................................................................................................................................................. 76
Appendix 5 ................................................................................................................................................. 79
Appendix 6 ................................................................................................................................................. 83
Appendix 7 ................................................................................................................................................. 86
Appendix 8 ................................................................................................................................................. 90
VII
List of Figures
Figure 1 - Global Horizontal Irradiation and Photovoltaic Power Potential in Cambodia [6]. ................................ 2
Figure 2 - Access to electricity in Cambodia [9]. ..................................................................................................... 5
Figure 3 - Goals of Cambodia National Grid Development for 2020 [7]. ................................................................ 7
Figure 4 – Average electricity prices for MV customers in Southeast Asian countries in USD/kWh [8].................. 8
Figure 5 - Comparison of Cambodian CO2 emissions and GDP trend considering a baseline of 100 for 2010 [13].9
Figure 6 - 3P4W wiring connection diagram for TES3600 Power Analyzer. ......................................................... 17
Figure 7 - Weekly load profile of Transformer 1 in the garment factory. ............................................................. 19
Figure 8 - Weekly load profile of Transformer 2 in the garment factory. ............................................................. 19
Figure 9 - Average load profile of T1 (green) and T2 (blue). ................................................................................. 19
Figure 10 - Shading situation where two to eight modules in a string are shaded, where the black curves = no
shading, blue curves = 2 modules shaded, green curves = 4 modules shaded, orange curves = 6 modules shaded
and the red curves = 8 modules shaded [27]. ....................................................................................................... 21
Figure 11 - Shading situation of an array where two of the strings are shaded from two to eight modules, where
the black curves = no shading, blue curves = 2 modules shaded, green curves = 4 modules shaded, orange
curves = 6 modules shaded and the red curves = 8 modules shaded [27]. ........................................................... 21
Figure 12 - Shading situation of an array where two modules on one to four strings are shaded, where the black
curves = no shading, blue curves = 2 modules shaded, green curves = 4 modules shaded, orange curves = 6
modules shaded and the red curves = 8 modules shaded [27]. ............................................................................ 22
Figure 13 - Aerial image of the garment factory (left) and its location in Cambodia (right). ............................... 23
Figure 14 - View of the first building, where the solar system connected to T1 should be installed, and the
corresponding electrical room. ............................................................................................................................. 25
Figure 15 - Example of different flight modes for drone mapping in Pix4D app. .................................................. 26
Figure 16 - Beam, diffuse and ground reflected radiation [29]. ........................................................................... 29
Figure 17 - Synthetic table with monthly average values for the site location defined in PVsyst. ........................ 30
Figure 18 - Cell/module I-V characteristic curves depending on the irradiance (left) and on the temperature
(right). ................................................................................................................................................................... 31
Figure 19 - Schematic of the distribution of diffuse radiation over the sky dome showing the circumsolar and
horizon brightening components added to the isotropic component [29]. .......................................................... 33
Figure 20 - Possible losses in respect to the optimum situation computed with PVsyst for the garment factory
case study. ............................................................................................................................................................ 36
Figure 21 - Inverter/array sizing in PVsyst with 63 strings of 21 modules model JKM 340PP-72-V by Jinko Solar
and 11 inverters SUN2000 36-KTL by Huawei. ..................................................................................................... 39
Figure 22 - Incident Angle Modifier for different parametrizations...................................................................... 43
Figure 23 - Graph of the increasing mismatch of the PV system simulated in PVsyst for the Jinko Solar panels
model JKM 340PP-72-V. The green dot represents the losses simulation at the 10th year of the system
considering the increasing mismatch. .................................................................................................................. 46
VIII
Figure 24 - Parallel strings (left) and group of parallel strings (right) wiring connection schemes in PVsyst. ...... 47
Figure 25 - Different modules connection methods: conventional method (1a) and leapfrog wiring (1b) [44]. .. 49
Figure 26 - Roof layout for T1 solution drawn in Sketchup. .................................................................................. 50
Figure 27 - Roof layout of T2 drawn in SketchUp. ................................................................................................ 50
Figure 28 - Cumulative net cash flow of the old regulation scenario (green), of the new regulation scenario (red)
and of the new regulation with T2 re-sizing (yellow) ........................................................................................... 58
Figure 29 - Scheme of the business strategy of Kamworks according to the market sector. ............................... 71
Figure 30 - T1 (left) and T2 (right) electrical cabinet measurements. .................................................................. 72
Figure 31 - Transformer T1 (left), power analyzer (center) and electrical cabinet T2 (right). .............................. 72
Figure 32 - Panoramic view of the first building roof of the garment factory. ..................................................... 73
Figure 33 - T1 building roof side top view (left), metal roof cover (center) and connection buildings cover (right).
.............................................................................................................................................................................. 73
Figure 34 - Garment factory electricity bill. .......................................................................................................... 74
Figure 35 - 3D design (up) and 2D map (down) of a residence of orphans in Phnom Penh, Cambodia. .............. 76
Figure 36 - Rice milling factory 3D design (up) and 2D design (down) situated in Phnom Penh, Cambodia. ....... 77
Figure 37 - 3D elaboration (up) and 2D map (down) of a factory manufacturing and distributing home decor,
gifts, and holiday products in Sihanoukville, Cambodia. ...................................................................................... 78
Figure 38 - SUN2000 36KTL inverter from Huawei specifications. ........................................................................ 81
Figure 39 - SunnyBoy Tripower 25000TL JP-30 inverter from SMA specifications. ............................................... 82
Figure 40 - Carbon balance settings for T1 system (first image) and the overview for T1 (second image). ......... 90
Figure 41 - Carbon balance settings for T2 system (first image) and the overview for T2 (last image). .............. 91
Figure 42 - The grid energy mix of Cambodia (including imports) common at T1 and T2. ................................... 92
IX
List of Tables
Table 1 - Overview of the electricity industry with definitions. ............................................................................... 5
Table 2 - Generation and imports of electricity in Cambodia [11]. ......................................................................... 6
Table 3 - The new tariff for consumers with the Solar PV system [17]. ................................................................ 11
Table 4 - Example on how the new solar tariff will affect savings in installing solar both for buyout and PPA
contract................................................................................................................................................................. 12
Table 5 - Summary of the challenges for Cambodia energy sector and the related opportunities of PV. ............ 13
Table 6 - Data about the consumption of Grace Glory Garment Factory. ............................................................ 18
Table 7 - Main roof characteristics of the garment factory. ................................................................................. 24
Table 8 - Reference temperature for the garment factory site. ............................................................................ 35
Table 9 - Summary of the module layout characteristic of the garment factory case study in Phnom Penh area
for both installation regarding T1 and T2. ............................................................................................................ 51
Table 10 - Summary of the outputs of the PVsyst simulations. P_nom is the nominal power of the PV system
under STC, E_load is the energy needs of the user, E_user is the solar energy supplied to the user, E_grid is the
excess energy exported to the grid, E_total is the total energy produced with the PV system, Grid Export
Percentage represents the percentage of solar energy, in respect to the total solar energy produced, that is lost
because exported. All the losses are computed as explained in section 3.4.4. ..................................................... 52
Table 11 - Summary of T2 simulations in PVsyst. ................................................................................................. 53
Table 12 - Investment cost and maintenance cost for the installation at the garment factory (707 kWp). ......... 55
Table 13 - Characteristics of the garment factory. The EDC consumption charge and contracted load charge are
referred to the connection from a MV (22kV) sub-transmission or distribution network of the national grid as
referred in the solar regulation. ............................................................................................................................ 56
Table 14 - Scenarios comparison summary. All the data about PV production have been taken from the
simulations of the optimal systems in PVsyst. ...................................................................................................... 60
Table 15 - Energy generation mix and energy import mix from 2017 [11]. .......................................................... 62
Table 16 - Carbon balance results for the whole system of the garment factory both with PVsyst default values
and with values insert manually taking into account the energy mix of Cambodia for 2017 and the specific PV
installation at the garment factory. ..................................................................................................................... 63
Table 17 - T1 and T2 load measurements at the garment factory. ...................................................................... 75
Table 18 - Main technical specifications of all modules under STC and I-V curves. .............................................. 79
X
Abbreviations
AC – Alternate Current
BoS – Balance of System
DC – Direct current
EAC – Electricity Authority of Cambodia
EDC – Electricité du Cambodge
EPC – Engineering Procurement and Construction
EPI – Environmental Performance Index
GDP – Gross Domestic Product
HV – High Voltage
IPP – independent Power Producer
JRC – Joint Research Centre
LCE – Life Cycle Emissions
LCOE – Levelized Cost of Energy
LID – Light Induced Degradation
LV – Low Voltage
M2M – Machine to Machine
MEF – Ministry of Economy and Finance
MME – Ministry of Mines and Energy
MPP – Maximum Power Point
MPPT – Maximum Power Point Tracker
MV – Medium Voltage
NOCT – Normal Operating Cell Temperature
NPV – Net Present Value
NREL – National Renewable Energy Laboratory
O&M – Operations and Maintenance
PBT – Pay Back Time
PPA – Power Purchase Agreement
PR – Performance Ratio
PV – Photovoltaic
REF – Rural Electrification Fund
ROI – Return on Investment
SHS – Solar Home System
STC – Standard Test Condition
T1 – Transformer 1 of the garment factory case study
T2 – Transformer 2 of the garment factory case study
VAT – Value Added Tax
1
Chapter One – Introduction
1.1. Cambodia Country Summary
Cambodia, officially the Kingdom of Cambodia, is located in the tropical area of Southeast Asia in the Lower
Mekong region, between latitudes 10° and 15°N, and longitudes 102° and 108°E. It is 181,035 Km2 in area,
bordered by Thailand to the northwest, Laos to the northeast and Vietnam to the east. It has a population of
over 15 million people spread all around the country. Phnom Penh, the capital, is the biggest city as well as the
political, economic and cultural center of Cambodia. Geographically, Cambodia’s region is characterized by low-
lying central plain that is surrounded by low mountains and Tonle Sap Lake area.
1.2. Cambodia’s Environment
Regarding the environment, Cambodia has an extremely bad Environmental Performance Index (EPI) with a rank
of 150 out of 180 countries, one of the worst in Southeast Asia region, only ahead of Laos and Myanmar. The
environmental areas where Cambodia performs worst (i.e. highest ranking) are the air quality (164), the water
resources (143), the forests (134) and the climate and energy (103) [1]. Cambodia’s climate, likewise the other
Southeast Asia countries, is tropical. Dominated by monsoons, it has two main seasons: the wet season, from
May to October, with temperatures that drop down to 20-22 °C and a high humidity level and the dry season,
from November to April, where the climate is particularly arid with peak temperatures of more than 40 °C.
According to The United Nations, due to an overreliance on fishing and rice production for livelihoods, Cambodia
has been identified as the most vulnerable country among Southeast Asia to the effects of climate change [2].
Shortages of clean water, extreme flooding, mudslides, higher sea levels and potentially destructive storms are
of particular concern according to the Cambodia Department of Climate Change.
1.3. Energy and Business in Cambodia
Several years of war (1979-91) have critically damaged Cambodia’s power and communication sectors putting
an already fragile economy in crisis. After the restoration of order in the country, the Government has followed
a program focused on rehabilitation and development of the basic infrastructure, with the aim of improving the
socio-economic conditions. Without energy, promoting economic growth, overcoming poverty, and supporting
human development are impossible [3]. Indeed, sustainable energy is the 7th of the 17 UN Sustainable
Development Goals to ensure access to affordable, reliable, sustainable, and modern energy for all by 2030.
Despite Cambodia now is one of the fastest growing economies in Asia with an average GDP growth rate around
6 percent over the last decade [4], there are many factors that prevent the agriculture in favor of garment,
textiles, and tourism sectors. These sectors are expanding, bringing increased foreign investment and
international trade but the development is facing many difficulties. Cambodia ranked 183 out of 190 countries
in the 2017 World Bank Doing Business report [5] to start a business and 137 to get electricity. Solar power has
2
a critical role to play to meet the United Nations Sustainable Development Goals and to facilitate the business
development in the country. Despite solar photovoltaic (PV) high potential in the region, it is currently a very
small industry. The country has great solar resources with a global horizontal solar irradiation that can reach
2000 kWh/m2 and a daily yield up to 1600 kWh/kWp as can be seen from Figure 1.
Solar energy can be the solution for both access to electricity and business development in Cambodia. PV has
the possibility to help to meet the 2030 targets, delivering clean and affordable energy to rural area via off-grid
systems and helping the business cutting down the costs of electricity and providing a better, cleaner and more
reliable service.
Figure 1 - Global Horizontal Irradiation and Photovoltaic Power Potential in Cambodia [6].
1.4. Purpose of the Work
The purpose of this thesis is to show the energy situation in a developing country with a growing economy such
as the one of Cambodia, analyze the process of design of a solar system for a big garment factory, from the first
site visit till the final design of the whole solar system, to investigate the new solar regulation recently put in
place and to shows how this will affect both the economic and private investments in Cambodia.
1.5. Thesis Structure
After the introduction, the work has been divided into four main sections.
Chapter two analyzes the major institutions of the electricity industry in Cambodia and their specific tasks, the
electricity sector of the country with the recent developments and the environmental aspects related to the
energy sector. The solar power sector has been then described with a deep analysis of the new regulation about
3
solar recently approved. Finally, the energy challenges related to solar power have been outlined and an
overview about the economic sector of Cambodia has been presented.
Chapter three presents the case study of a garment factory rooftop PV installation that have been examined in
this work. An overview of the factory with the goals of the project is followed by the first site visit description as
well as the load profile measurement analysis. The preliminary design of the project is then presented. with a
focus on the shading losses and the 3D modelling.
Chapter four deeply analyzes the project design of the case study with PVsyst software. All the simulations
settings will be examined and discussed in this section. The chapter follows the process of the design of a grid-
connected PV system with PVsyst. The simulations results and the optimal solution for the PV installation are
presented at the end of this section.
Chapter five analyzes the profitability of the project through an economic assessment. A comparison between
three different scenarios is presented to show the impact of the new solar regulation on PV projects in
Cambodia. Finally, the carbon balance of the PV installation is investigated in order to evaluate the CO2 emissions
reductions that can be achieved thanks to the proposed solution.
The conclusions summarize the findings of all the work done.
4
Chapter Two – Cambodia State of Art
This section presents an overview of the country situation about electricity sector. The major institutions are
presented and the electricity situation is outlined together with the environmental aspects related to the energy
mix of the country and the import situation. Solar power in Cambodia is then examined with a focus on the new
solar regulation recently approved and the energy challenges of the country that solar energy can help to solve.
2.1. Major Institutions of Electricity Industry in Cambodia
The primary policy making and planning entity for the electricity industry in Cambodia is the Minister of Mines
and Energy (MME). The responsibilities of the MME include [7]:
• Setting and administrating the government policies, strategies and planning in the power sector.
• Setting the technical, safety, and environmental standards for the power sector.
• Promoting energy conservation.
• Preparing and implementing energy sector development plans.
The second authority of the country regarding energy industry is the Electricity Authority of Cambodia (EAC). It
is an autonomous agency established under Cambodia’s Electricity Law that has its own budget funded from
license fees [8]. The responsibilities of EAC include [7]:
• Issue regulations.
• Issue, revise, suspend, revoke or deny the licenses for the provision of electric power services.
• Approve tariff rates and charges and terms and conditions of electric power services of licensees.
• Resolving disputes between licensees and between customers and licensees.
• Monitoring and enforcing compliance by licensees with technical and performance standards.
• Prepare and publish reports of power sector and relevant information received from licensees for the
benefit of the Government and the public.
The other major player in the Cambodian electricity industry is Electricité du Cambodge (EDC). It is the state-
owned electricity supplier with responsibility for the generation, purchase, transmission and distribution of
electricity throughout the country. EDC is jointly-owned by the MME and Ministry of Economy and Finance
(MEF). EDC is the only entity licensed to conduct activities across the full electricity supply chain. Nevertheless,
Cambodia has always been heavily dependent on the private sector to generate electricity and sell it to EDC due
to financial constraints. Those private third parties are called Independent Power Producers (IPPs). IPPs were
formalized following the creation of EAC which licensed them and brought them under regulatory control. In
more recent years, third party entities have also entered the transmission and sub-transmission sectors. An
overview of the Electricity industry in Cambodia with functions and definitions has been summarized in Table 1.
5
Table 1 - Overview of the electricity industry with definitions.
The Electricity Industry
Generation Generation of electricity.
Transmission Long distance transmission of electricity using HV lines from generators to distributors.
Sub-transmission (Cambodia)
Transmission of electricity using MV lines from HV transmission networks to LV distribution networks in rural areas.
Distribution Distribution of electricity to final users through MV and LV lines.
Retail Supply Financial activity of buying electricity from generators and selling it to customers
High Voltage (HV) Refers to 230 kV and 115 kV
Medium Voltage (MV) Refers to 35 kV and 22 kV
Low Voltage (LV) Refers to less than 22 kV
2.2. Electricity Sector in Cambodia
The electric grid has expanded rapidly, with the electrification rate greatly increasing, from the initial percentage
of 9.3% in 1997 to 34.4% in 2009 and to 56.1% in 2014 as shown in Figure 2 [9]. At the present time in Cambodia,
100% of the urban population have access to electricity while only 36.5% of rural population have it [10]. In the
rural areas 60% of the connection are through the grid while the other 30% (of the households with electricity
access) uses off-grid solutions and the majority of the aforementioned uses Solar Home Systems (SHS) [3].
The first problem of Cambodia’s electricity system is the generation capacity that is lower than the total
consumption. Until 2007, the supply was almost entirely from oil and coal.
0.0%
20.0%
40.0%
60.0%
80.0%
100.0%
19
97
19
98
19
99
20
00
20
01
20
02
20
03
20
04
20
05
20
06
20
07
20
08
20
09
20
10
20
11
20
12
20
13
20
14
20
15
20
16
Acc
ess
to
Ele
ctri
city
(%
)
Year
Access to Electricity in Cambodia
Access to Electricity (% of total population)
Access to electricity, urban (% of population)
Access to electricity, rural (% of population)
Figure 2 - Access to electricity in Cambodia [9].
6
With the establishment of the connection with Thailand in 2007 and with Vietnam in 2009, the imports have
replaced the internal generation, reaching in 2011 about two third of the country needs. Big step forwards have
been made after 2013 with the growth of hydropower and coal to make the electricity system more independent
from the neighboring countries [8]. As can be seen from Table 2, Cambodia still imports a significant amount of
energy from the neighboring countries. Predominantly from Vietnam followed by Thailand and Laos [11]. This
put the country’s economy in a high dependence from other countries that can influence the electricity price.
Table 2 - Generation and imports of electricity in Cambodia [11].
production from: Electricity 2016 [GWh]
% Electricity 2017 [GWh]
% Expected Electricity 2018 [GWh]
%
Coal 2551.17 35.6% 2829.12 35.5% 3020.58 33.3%
Oil 362.13 5.0% 381.14 4.8% 260.43 2.9%
Hydro (also pumped) 2619.11 36.5% 3217.79 40.4% 4209.25 46.4%
Renewable Energies 43.35 0.6% 48.61 0.6% 60.06 0.7%
Captive Generation by Industries and Licensees
16.44 0.2% 9.59 0.1% 8.68 0.1%
Total Generation 5592.2 77.9% 6486.25 81.4% 7559.00 83.3%
Imports:
- Thailand 346.19 4.8% 269.56 3.4% 266.37 2.9%
- Vietnam 1201.78 16.7% 1153.85 14.5% 1178.19 13.0%
- Lao 34.88 0.5% 56.52 0.7% 75.62 0.8%
Total Import 1582.85 22.1% 1479.93 18.6% 1520.18 16.7%
Final consumption 7175.05 7966.18 9079.18
Another obstacle is the energy generation mix of the country. Power generation in Cambodia comes from hydro
and coal that accounts respectively for 40.4% and 35.5% (in 2017). Almost 5% of the production is from oil and
only 0.6% comes from other renewable energies. This provides another significant challenge for the energy
sector that results significantly dependent also on coal and oil imports since the country does not have coal
mines or significant oil reservoirs.
Heavy investment in the extension of the transmission grid has significantly increased the proportion of people
with access to electricity and aims to have a wider coverage of the territory has shown in Figure 3. In fact,
multiple generation projects are in the pipeline in order to increase the energy generation to become more
energy independent and cover energy demands when access to electricity increases. These projects include [7]:
• 135 MW Coal Fired Power Plant in Sihanoukville. Commissioned in 2017 (partially commissioned).
• 400 MW Se San Hydroelectric Project. The plant is proposed to have 8 units of 50 MW each. PPA has
been signed between EDC and Hydro Power Lower Sesan 2 Co., Ltd on 26 November 2012. The project
is expected to be commissioned in 2017 (still not commissioned).
• 135 MW (Net) coal fired project in Sihanoukville by Cambodia Energy 2 Co., Limited. EDC is negotiating
the PPA (still no date).
7
• 10 MW Solar Farm at Bavet in Svay Rieng Province is being implemented by Sunseap Asset (Cambodia)
Co., Ltd. The PPA has been signed between the company and EDC. The solar plant is expected to be
commissioned in 2017 (still not complete).
• 8 MW Biomass project in Koh Kong Province is being implemented by Kohkong Sugar Industry Limited.
The company has signed PPA with EDC.
Another challenge to the electrification of the country and in particular to the development of businesses such
as big garment or textiles factories is the high price of electricity. The average price of electricity in Cambodia
for a LV consumer is between 800 and 1000 Riels/kWh (0.20-0.25 USD/kWh). It does not seem so elevated but
is incredibly high for Cambodians, especially compared to the GDP of the country. As can be seen from Figure 4,
Cambodia has the highest price of electricity compared all other Southeast Asian countries. What makes the
price even more inadequate is the quality of the service. Based on personal experience, in the urban context of
Phnom Penh, the power cut offs are rare and isolated (few times a month, no more than one hour) but outside
the city center, in rural areas far away in the provinces in particular, the energy quality is very poor. Many energy
audits in factories around Phnom Penh have shown that power cut offs occur almost every day and in certain
occasions last for more than 6 hours. Moreover, Cambodia continues to have mismatches between supply and
demand during different times of the year and the day. Due to the high seasonality of hydropower generation
for instance, during the wet season the production is higher but the consumption is lower and the surplus energy
is not yet sold to other countries such as Vietnam, both for institutional problems and for the similarity in the
production/consumption schemes.
Figure 3 - Goals of Cambodia National Grid Development for 2020 [7].
8
2.3. Environmental aspects
Energy development and climate change have important effects on water resources in all Southeast Asia,
especially in the countries of the Lower Mekong River Basin such as Cambodia and Vietnam. Hydropower
resources on Mekong River are significant and upstream dams combined with El Niño weather patterns caused
water levels at their historical minimum. In 2015 and 2016, Cambodia has been hit by strong drought that
affected almost 100,000 households.
Rapid economic and population growth coupled with urbanization in the last 15 years has led to a large increase
in energy consumption in all Southeast Asian region, met principally by fossil fuel. It is estimated that combustion
of coal for power production and industrial processes is the largest source of energy-related sulfur dioxide (SO2)
emissions in Southeast Asia today, accounting for more than half of the total 2.3 Mt of SO2 emissions [12].
A study from the General Department of Energy and the General Department of Petroleum of Cambodia shows
that CO2 emissions increase at an average rate of 10.4% per year in Cambodia from 2010 to 2015 reaching 7,367
kt [13]. Most of the emissions are due to transport sector, especially the burning of gasoline and diesel to use
motor vehicles in Phnom Penh, but un important fraction (around 28%) is connected to the solid fossil fuel (coal)
burning.
Garments, constructions and agriculture have driven Cambodia’s growth, resulting in a growth of GDP where oil
and coal play an important role, resulting in high CO2 emissions. In Figure 5 is shown the trend of growth in GDP
and CO2 emissions for Cambodia. It is not surprising the strong link between these two factors. Sunder those
entities is hard, especially in a developing country with an incredibly rapid growth as Cambodia at the present
time. On the contrary, if no major measures will be taken, is foreseen that climate change will cause a reduction
in the absolute GDP of 2.5% in 2030 and reaches a reduction of 9.8% in 2050 [14]. This only means that GDP will
be 3 times higher in 2030 instead of 3.08 times. However, this signify that the grow rate for 2050 will fall from
4.5% to 4.2% and these will lead to more serious impacts over productivity sector resulting in high income losses
Figure 4 – Average electricity prices for MV customers in Southeast Asian countries in USD/kWh [8].
17.716.6
13.112.3 11.7 11.4
10.0 9.7
7.96.5
9
and labor productivity losses. Public policies should focus on that, paying more attention on measures to protect
supply chain and workers and doing more specific research on the impacts and on the possible paths to mitigate
the climate change. The first way of better, is to switch the fossil fuel-based economy to a more renewables-
oriented society, exploiting the available internal resources such as solar, decreasing the dependency from other
country and the high emissions related to the generation sector.
2.4. Solar Power in Cambodia
Solar power in Cambodia is currently a very small industry that is trying to emerge with different results
depending on the context. The country, as already shown in the introduction, has a great solar potential with
high values of solar horizontal irradiation that can lead to high values of yearly yields up to 1600 kWh/kWp for
big systems, making the solar generation a very profitable market.
Thanks to the Rural Electrification Fund (REF) set up in 2004 to disburse grants from the World Bank and
managed by EDC since 2012, many solar off-grid programs including mini-grids have been realized. The REF is
responsible for disbursing subsidies and providing technical assistance. The REF has three programs relevant to
mini-grids that are [15]:
• Power to the Poor. This program provides an interest-free loan to households (through the licensee) to
finance their connection. This loan is intended to cover: “1) costs of connection fees of the electricity
supplier; 2) costs of deposit to be deposited to the electricity supplier; 3) costs for purchase of materials
and labor for the installation of wires from the connection point to the house; and 4) costs for purchase
of material and labor for the installation of in-house wiring.” The REF lends to the licensees, who in turn
lend at no interest to households.
• Program for Providing Assistance to Develop Electricity Infrastructure in Rural Areas (PPADEIRA). This
program provides grants, interest-free loans, and guarantees to private licensees to develop the
distribution infrastructure and cover the entire license area.
• The tariff subsidy. This subsidy corresponds to the money given to licensees buying power from EDC to
bridge the gap between their calculated cost-recovery tariff, and the standardized tariff.
Figure 5 - Comparison of Cambodian CO2 emissions and GDP trend considering a baseline of 100 for 2010 [13].
10
On the contrary, for what concerns on-grid solar projects, the regulation has always been almost absent until
the beginning of 2018, with the only statement from EDC of mandatory net zero export from the inverters,
meaning that no energy from the solar system could have never been injected into the grid if not consumed by
the customer. The 26TH of January 2018 a new regulation concerning solar has been approved by EAC. The new
regulation will be discussed separately in section 2.4.1 due to its importance.
Without any financial help from the government for on-grid PV systems, Power Purchase Agreement (PPA) are
becoming more popular in Cambodia. PPAs are used, normally for big projects (>100kWp), when an investor is
willing to finance a solar installation for a customer that either does not have the capital to invest for the PV
system or is not willing to take the risk of installing solar. The second case is very common in Cambodia where
the trust on PV technology is not yet totally built. The investor pays the EPC (Engineering, Procurement and
Construction company) to build the solar system at the client’s site and then charges the client for the electricity
consumption at a defined cost per kWh for a defined period of time (normally around 20-25 years). In this period
the client commit himself to pay the electricity according to his consumption and the investor is responsible for
the PV system (and owns it). After the established period the ownership of the solar system can be either
transferred from the investor to the client or the terms can be renegotiated depending on the PPA contract. The
current rates for PPAs ranges from 0.08 to 0.15 $/kWh depending on the installation size. Usually the bigger the
system the lower the electricity rate. The huge convenience of the PPAs is that they require no down payment
or only a small one and represents low risk option for the companies that want to save money on their electricity
bill. The customers that can profit more from solar installation and from this type of contract are the factories
running 7 days per week. Their load is spread in all the 7 days of the week, allowing the investors to be payed
also for the Sunday consumption that normally is wasted.
2.4.1. The New Solar Regulation in Cambodia
On the 26th of January 2018, EAC approved a new regulation, “On General Conditions For Connecting Solar
Generation Sources To The Electricity Supply System Of National Grid Or To Electrical System Of A Consumer
Connected To The Electricity Supply System Of National Grid” [16].
Up until this time, no official regulation was set to regulate the installations of solar in Cambodia. The only
statement already made from EAC was that no solar power could have been exported to the grid.
After the release of the initial regulation, another document has been released, called “Announcement by the
Electricity Authority of Cambodia On Public Discussion on Tariff Plans for Bulk and Large Purchase Consumers
Installing Solar Power Connecting to the Distribution Network of the National Grid” [17] setting the new tariff
for the new regulation.
The key points of the new solar regulation have been summarized:
• The regulation affects only grid connected solar system. Off-grid solar installations are not affected by
the new regulation.
• No electricity can be injected into the grid, only EDC has the right.
• The maximum capacity of solar cannot exceed the 50% of contracted power.
11
• Only Big Consumers (MV) and Bulk Consumers (HV) are allowed to connect and synchronize the PV
systems with the grid.
• MV and HV consumers that want to install solar and synchronize to the grid will have to apply to EDC
for a license to install solar. EDC will decide within a month after receipt of all information. If permission
is granted, then the customer has one year to commission project. Failure to commission after one year
will result in having to complete process again.
• MV and HV customers that install solar will be moved to a new tariff structure called “the tariff for
consumers with the solar PV system”
The new tariff is defined in Table 3. It can be seen that the new tariff presents a fixed price for the contracted
power and a consumption charge based on the effective consumption. It is also important to notice that the
new consumption charge is much lower than the preview one. Before the new regulation, the price range for
MV and HV consumers was between 0.195 USD/kWh and 0.165 USD/kWh while now the price can be as low as
0.093 USD/kWh, meaning a reduction of more than 50% for the consumption and in contrast a high contract
load charge, independent on the consumption. This will affect largely solar installations, especially PPA that have
now to compete with very low price of 0.093-0.12 USD/kWh that is now the lowest PPA rate for a large PV
system installation.
Table 3 - The new tariff for consumers with the Solar PV system [17].
No Conditions Contracted load charge Consumption charge
1 Connection at HV substation (115/230kV) 7.00 USD/kW/month 0.093 USD/kWh
2 Connection at MV (22kV) substation 7.10 USD/kW/month 0.094 USD/kWh
3 Connection at MV (22kV) substation in Phnom Penh
9.10 USD/kW/month 0.105 USD/kWh
4 Connection from a MV (22kV) distribution network in Phnom Penh and Ta Khmao City
10 USD/kW/month 0.120 USD/kWh
5 Connection from a MV (22kV) sub-transmission or distribution network of the national grid
10 USD/kW/month 0.120 USD/kWh
There are still unknows and vague points in the new solar regulation. There is nothing stating if this new tariff
structure will be applied to customers who have already installed solar or if it will only effect new solar
instillations. It is unknown if a client can reduce their contracted load charge to effectively meet their new
maximum power demand if they install solar. This is a fundamental point since EDC tends to oversize the
transformers and has always stipulated contract with higher expected capacity in a “future expansion”
perspective. However, also decreasing the contracted load charge the amount of solar that can be installed will
decrease too due to the maximum 50% solar restriction. They have not stated what is required for the
application to install solar and there is no information about a “minimum solar capacity” to switch to the new
regulation. This means that hypothetically a customer can install one solar panel, having almost no investment
cost, and then can switch to the new regulation that is much more profitable for what concern the consumption
tariff. In such case this will affect very badly all EPC companies and in general the solar market, with solar energy
that will be seen only as an easy shortcut to reduce the electricity bill.
12
The new regulation was seen as a quick response from the Cambodian government to the missing legislation
about solar on-grid. However, all solar companies in Cambodia and all solar investors were disappointed about
the regulation that has been put in place. The EAC gave no explanation on why solar power has been targeted
but in the solar business environment have been speculated why it has occurred. The main suspected reason is
that EDC is the only authorized generator, transmission and distributor of electricity in Cambodia. As they have
heavily invested in the transmission and distribution lines in the past years and as they rely on the tariffs they
charge to recover from these costs, EDC is worried about decrease in their profit due to a possible decrease in
the energy demand due to the solar generation. It is also important to notice that the chairman of the board of
EDC is a relative of the current prime minster of Cambodia, Hun Sen, as well as three of other relatives of him
that are listed as owners of Cambodia Electricity Private [18]. It looks clear why the regulation aims to keep the
authority and the money on the country side more than on the customers and investors side.
This regulation of solar appears as a short-sighted view and is one of the reasons why the solar companies and
investors complained. Looking at Table 4 is clear why this regulation will affect badly the solar market. The quick
simulation has been done to show briefly the impact of the new regulation and have been analyzed more deeply
in Section 5.1 about the economic assessment of the case study. The simulation has been done with the
assumption of a yearly yield of 1591 kWh/kWp (as suggested by PVsyst software for Phnom Penh area in
Cambodia), a contracted electricity price of 0.165 $/kwh (Big consumer) a tariff of type 4 of the new regulation
(10 $/kW/month of contracted load charge and 0.12 $/kWh of consumption charge) and that all the electricity
produced by the solar system is used by the customer (not conservative hypothesis). The contracted size and
the monthly bill have been adapted from the real case study of a garment factory in the suburb of Phnom Penh.
Table 4 - Example on how the new solar tariff will affect savings in installing solar both for buyout and PPA contract.
Old regulation New regulation
Contracted size 1000 kW 1000 kW
Current monthly bill $20,000 $20,000 Monthly consumption 121212 kWh 121212 kWh
PV size 500 kWp 500 kWp
PV monthly production [kWh] 66292kWh 66292 kWh
New bill (buyout) $9,061.9 $16,590.5
Monthly savings (buyout) $10,938 $3,409.5
New bill with PPA at 0.10 $/kWh $15,691.0 $23,219.62
Monthly savings with PPA $4,308.96 -$3,219.62
From this rough estimation of the savings, we can understand the important points that will influence mostly
the solar market. First, the high risk from the investor side, since for example in this case a PPA will not be
profitable at all at a rate of 0.10 $/kWh, having a monthly price higher than the bill without solar. Lower charge
rate should be considered but only with very big system they could be applied. Second is the oversizing of the
transformers for the contracted size that involves high costs for the contracted load charge. The only solution
would be to re-size the transformer (or the contracted load charge in case of no transformer) meaning a lower
13
monthly fixed charge and a lower possible size of the solar system. Because of the vague regulation, missing a
lot of details, solar industry has been affected by uncertainty both from the investors and the clients side. Most
investors have put their projects on hold until further information are released or until someone else “try and
see what happen”.
2.4.2. Energy Challenges and Solar Generation
Solar power has the ability to assist Cambodia in expanding the energy access by 2020 to meet the goals set
from the government. The intention is to electrify 100% off the Cambodian villages and at least 70% of the
households. It is an ambitious plan, that is not easily reachable, where solar energy plays a key role. Off-grid
solar solutions can provide energy even in the most remote areas of Cambodia, reducing the reliance on
generators and thus fossil fuels, improving the living conditions of the families and decreasing the consumption
of charcoal for cooking and heating.
PV can also facilitate the growth of businesses such as garment factories, agricultural facilities and food
industries. Utility on-grid solar projects, with or without PPA, can lower down the high price of electricity
affecting Cambodian market. On-grid solar can be built to be profitable at around 0.10 $/kWh (with PPA)
compared to 19.5 $/kWh of most of the consumers.
Combined with batteries or coupled with a backup generator, PV systems can be used to provide a backup
electricity source where the grid is particularly intermittent and unreliable.
Eventually, solar generation will be able to provide a substantial amount of clean energy at an affordable price,
cutting off the energy dependence from other countries and lowering down the traditional fossil fuel
consumption.
The energy sector challenges can thus become opportunities in Cambodia thanks to solar PV. In Table 5 have
been summarized the challenges/opportunities.
Table 5 - Summary of the challenges for Cambodia energy sector and the related opportunities of PV.
Challenge Opportunity
Access to electricity Solar off-grid solutions including mini-grids to ease the access also in the most remote areas.
High electricity prices On-grid solar and PPA can help to cut the electricity prices for factories and industries.
Unreliable grid Solar with batteries or Generator as a backup option for unreliable and unstable grid.
Energy imports dependence Solar farms and solar installations in general can help to lower down the dependency from the neighboring countries.
Fossil fuel dependence Solar PV is a renewable source of electricity that is not dependent on import of fossil fuels.
14
2.5. Economic sector in Cambodia
Cambodia has experienced strong economic growth over the last decade. GDP grew at an average annual rate
of over 8% between 2000 and 2010 and about 7% since 2011. The tourism, garment, construction and real
estate, and agriculture sectors accounted for the bulk of growth [19].
The Royal Government of Cambodia has set directions to transform Cambodia into a middle-income economy
by 2030 and high-income country by 2050, as mandated in the National Strategic Development Plan for 2014–
2018. The directions emphasize the role of industry and small and medium enterprises as a key driver of future
growth [20].
The manufacturing sector accounted for 31 percent of Cambodia’s economy in 2016 [21]. The Asian
Development Bank has forecast the country’s manufacturing to continue to grow by around 9.6%, “with a
slowdown in garments and footwear off set by stronger growth in emerging industries: electrical parts,
automobile components, bicycles, milled rice, and rubber.” In 2017, Cambodia became the European Union’s
number one supplier of bicycles [22]. After garments and footwear, bicycles are the third largest export category
in Cambodia.
There are now signs of moderation in the export of textile and apparel articles and in construction activity. The
first six months of 2017 have registered a 5.4% growth in the export of clothing and other textile products, down
from the 8.4% of the same period of 2016. However, the export related this sector reached US$3.3 billion.
Exports are facing increasing competition from other countries while improvements in the productivity remains
low. Moreover, minimum wages are rising. The minimum salary in 2017 was 153$ (excluding allowances) for the
garment and footwear sector, it reaches 170 US$ in 2018 and is expected to grow more. This trend is ongoing
since 2013, with an average increase of 17.6% per year [23].
In addition, improved weather condition in the last two years have helped agriculture sector expansion,
especially rice production. This sector provides 44% of employment, accounting for the 53% of employment in
rural areas [23].
As a Least Developed Country Cambodia benefits from the most favorable regime available under the EU's
Generalized Scheme of Preferences (GSP), namely the “Everything But Arms” scheme which grants full duty free
and quota free access to the EU Single Market for all products. This helped and continue to help a lot the export
market. Recently, EU sent a delegation of the European Commission and the European External Action Service
to Cambodia to evaluate the situation on the ground following recent worrying human rights and labor rights
developments in the country [24]. If a suspension of the economic scheme would happen it will cause serious
damage to the export sector in Cambodia. To sustain the garment industry and maintain its competitiveness,
several initiatives and measures are being introduced. Also because of that, an effort to upgrade the skills of
textile factory workers through on-the-job training with the establishment of Cambodian Garment Training
Institute located in the Phnom Penh Social Economic Zone has been made. Moreover, initiatives that are helping
to contain living costs of factory workers including access to lower utilities tariffs and provision of free health
insurance, pensions, and transportation have been recently introduced.
15
Tourism has also grown in the last year, with new direct flights from china that have increased the Chinese
tourists in the country, overcoming Vietnamese, Thai, Lao PDR and Korean people. Also, arrivals from Europe
and USA have increased despite they are still the minority in comparison with the Asian countries tourists.
While new emerging manufacturing industries, such as electrical appliances and components, and auto parts
are trying to emerge aiming to attract foreign investments, and older manufacturing sector are trying to remain
competitive, key bottlenecks including high electricity and logistic costs, and growing skills constrain are
preventing the success.
Solar power can assist Cambodia in the development of these new industries and in the subsistence of the
garment, food and agricultural sector. Even with the recent reductions, Cambodia’s electricity price for big
consumers remains the highest in the region as shown in chapter 2.2. Industrial customers can act to reduce
their electricity costs by improving energy efficiency and thus reducing the direct consumption of electricity for
their processes. Adding a self-generation method such as solar PV or solar water pumping (for the agricultural
sector) is another way to lower down the grid consumption, reducing on their side the high cost of electricity
bills and on the grid side lowering down the peak demand.
16
Chapter Three – Case Study Definition and Preliminary
Assessment
This section defines the case study that is analyzed in this work: the scopes of the installation are defined and
the first site visit is briefly described together with the load profile measurement. The first step of the project,
the preliminary design, is presented and it discuss about the first two important consideration for a PV project:
the evaluation of the shading and the 3D design of the installation. This analysis has been carried out during the
internship realized with Kamworks Ltd. The company profile and business strategy can be found in Appendix 2.
3.1. Garment Factory Case Study: Overview, Goals and Scopes
The garment factory analyzed is a big factory that produce garment and textiles for the export. Its business is
growing as it is growing the Cambodian economy. The company has a high electricity consumption that results
in an incredibly high electricity bill. The garment sector in Cambodia represents the core of exports of the country
and it is highly impacted from the cost of electricity. This impact reduces the margin on the profit and thus also
the willing of entrepreneurs and investors to remain in Cambodia instead of moving the business elsewhere.
The garment factory of the study is located in the suburb of Phnom Penh, the capital city. It is connected to the
main distribution grid through two different transformers that give electricity to two different sections of the
factory. Nevertheless, the electricity bill is one and comprehensive of both transformers consumption. The
factory is a MV consumer, connected from a MV (22kV) distribution network and it has a consumption tariff of
0.165 USD/kWh.
Due to the high cost of electricity in Cambodia, the impact of the consumption onto the business of the company
is relevant. The PV system installation can ease the business of the garment factory as well as affects positively
the peak demand from the grid. The main objective of the project design of a roof mounted solar PV system for
the garment factory is to evaluate the possible savings in the electricity consumption from the grid. The optimal
solution needs to be defined respecting the Cambodia regulation about solar and the system will be sized
according to that. Moreover, the impacts of new solar tariff on the economics of the project have been
investigated to show the consequences that such measure can have on a private customer, on his business and
on renewable investment in general in the country. Finally, all the energy produced with the solar system will
be able to reduce the national CO2 emissions, especially if a wider range of factories will choose solar PV as a
self-generation method.
3.2. Site Visit and Load Profile
The first step before planning a grid-connected PV system is a site visit. This enables an assessment of the basic
conditions for the PV system. First, it is important to establish whether the building is suitable for installing a PV
system. A precise initial investigation avoids planning errors. During the site visit, the installation work for the
PV arrays, the installation site for the panels and the inverters, the wiring routes and for expanding or modifying
17
the meter cupboard can be better estimated. Moreover, the most important step for the first evaluation of the
feasibility of the project is the measure of the load profile. This step is fundamental to understand the
consumption pattern of the factory, evaluate the possible savings through a solar system and design the system
to avoid most of the energy exports in order to comply with the regulation limits.
The garment factory that has been investigated in this study has two different transformers of 1000 kVA situated
in two different electrical rooms outside the buildings of the factory. The electrical rooms are of recent
construction and are well designed with a lot of free space around the transformers and the electrical cabinet,
they are ventilated and would not any problems for the inverters installation. The two transformers supply
power to two different section of the factory: the first (from now on referred as T1) is used to power the garment
factory with all the sewing processes and the administrative buildings while the second one (from now on
referred as T2) is used to power a section of the factory dedicated to the printing processes. The two
transformers are under the same contract and the monthly electricity bill is comprehensive of both transformers
consumption, but since the consumption patterns are different and the future PV system has to be split into two
different parts both for technical reasons and to comply with the regulation, they will be analyzed separately.
The load profile analysis presented in this study has been done with two TES 3600 Power Analyzer. The
configuration 3P4W meaning 3 phases and 4 wires connection with neutral has been used to connect the power
analyzers. The connection of the clamps to measure the currents on the 3 phases and of the “alligators” to the
lines to measure the voltages have been done following the manual to avoid errors. The scheme used to connect
the two power analyzers to both transformers is shown in Figure 6 while the pictures of the real installation of
the power analyzers are shown in Figure 6. Some site visit pictures and one month of electricity bill of the
garment factory can be found in Appendix 2.
To obtain reliable data for the simulation, the power analyzers must remain connected for the longer period
possible. Ideally one-year analysis will be able to give the optimal yearly consumption pattern of a factory or a
company.
Figure 6 - 3P4W wiring connection diagram for TES3600 Power Analyzer.
18
For this analysis, due to practical limitations, the power analyzers have been connected slightly more than one
week, measuring currents, voltages and calculating the power every two minutes for each transformer. The
data, once downloaded on the computer, have been then elaborated via Excel to extrapolate the typical weekly
profile of the factory.
From the two minutes step data, the average consumption of each hour of the day have been computed. The
load profile of each day of the week have been plotted over the hours of the day to show the trend. Figure 7
and Figure 8 show the load profile of T1 and T2. As can be seen from the graphs, both transformers have similar
trends with different peaks values. The hourly power consumption and the averages for the week days can be
found in the Appendix 3.
The garment factory is operating normally from Monday to Saturday while is not operating on Sunday.
Considering that, the typical working day profile and the typical holiday day profile have been calculated and
plotted as can be seen in Figure 9. To compute the working day load, the average for each hour have been
calculated in the range Monday to Friday. The holiday day load profile has been considered to be as the Sunday
load. The values have been then used to create a yearly consumption pattern through another Excel sheet. The
workbook needs the hourly data of a working day, of a Saturday, of a Sunday, of a holiday day and the number
of holidays in that year as inputs. It produces one column with 8760 hourly data representing the entire year of
consumption thanks to a macro function. This file has been then imported into PVsyst to simulate the user needs
in order to compare the behavior of the solar system production and the internal consumption to optimize the
system. Since the load analysis has been conducted on a restricted period of time, to understand the reliability
of the measures, the load profile has been compared with the electricity bill. The energy consumed has been
computed from the hourly power measurements using the trapezoidal rule to approximate the finite integral
below the curves. The daily averages have been then added together, taking into account the number of holidays
of the month of the electricity bill to have a more accurate estimation. The resulting energy consumption has
been found to be 0.57% lower of real consumption from the electricity bill, meaning the measurements and the
calculations are consistent with the real consumption and can be used for the PVsyst simulation. The data about
the factory and the estimation done for the consumption have been summarized in Table 6.
Table 6 - Data about the consumption of Grace Glory Garment Factory.
Garment Factory Consumption Data
Location Phnom Penh, Cambodia Transformer size (T1 + T2) 2000 kVA (1000 kVA + 1000 kVA)
Type of connection for the regulation 4 (Connection from a MV (22kV) distribution network in Phnom Penh and Ta Khmao City)
EDC electricity price 0.165 $/kWh Monthly bill 24,686.00 USD
Monthly Consumption from bills 149,612.12 kWh Monthly Consumption estimated from measurements
(3 days of holiday) 148,755.49 kWh
Error of the estimation 0.57%
19
Figure 7 - Weekly load profile of Transformer 1 in the garment factory.
Figure 8 - Weekly load profile of Transformer 2 in the garment factory.
Figure 9 - Average load profile of T1 (green) and T2 (blue).
20
3.3. Preliminary Design
The first important consideration that has to be made in the project design of a PV solar system is the location
for the installation. After the first site visit, a preliminary design helps to fast check the possible size of the
installation, the limitations and the customer willing. The location is important because it can affect the energy
production for many reasons. For rooftop installation, as in the case study analyzed, is necessary to measure the
available space on the roof to understand the maximum PV system size and to evaluate which technology and
which type of panel is better. The roof type and quality (material, profile, load resistance) must be considered
as well as its slope, that is important matches with the optimal tilt angle. In Cambodia the optimal tilt angle
range is between 10° and 15° that is also the range of the common factories roofs inclinations. Furthermore, is
important to analyze the roof surrounding area to compute the possible shading losses. Shadows on a PV system
has a much greater effect on the solar yield than, for example, in the case of solar thermal systems.
3.3.1. Shading Losses
The amount of effect produced by a PV module relies on how much radiation it receives at its precise location.
Shades can reduce significantly the amount of energy that reaches the panels.
Shading can be classified as temporary, resulting from the location, the building or caused by the system itself
(self-shading). Direct shadows can have a critical effect on PV-array output reducing the incident energy on the
cells by approximately 60% to 80% [25]. The impact of a single shaded cell can be critical especially if all the
panels cells are connected in series. In that case, the short circuit current of the shaded cell will affect the total
current of the panel. If the defected cells are forced to pass a current higher than their generation capabilities,
they become reverse biased, even enter the breakdown regime, and sink power instead of sourcing it. Intense
local heating can produce very high temperatures (hot spots) that can damage the module [26]. The common
approach to solve the hot spots problem and to mitigate the shading effect that causes reverse currents is to
put a diode. This diode, called bypass diode, is connected in parallel, but with reverse polarity, with a group of
cells. When one or several cells are shaded, they are reverse biased only to the point where the diode across
the group starts forward conduction. The diode carries away the necessary current to keep the group near short-
circuit. It is clear that the smaller number of cells per bypass diode, the lower the efficiency losses. Bypass diodes
are nowadays of common use by the solar panels manufacturers that insert them inside the junction box, but
due to the higher production cost, the commercial solar panels normally present two or three bypass diodes,
meaning in the best case one over 20 cells for 60-cells panels or one over 23 cells for 72-cells panels.
Due to its impacts, shading should be eliminated as much as possible. Even small obstacles like chimneys,
telephone poles etc. shouldn't be neglected in the design phase. While is quite easy to compute the power losses
due to the complete or partial shading of one or more cells in a module, compute the impact of nearby objects
to the yearly production of a PV array or a complete PV system is much more complex. The effect of the shades
depends strongly on the modules layout. Modules in series and strings in parallels have different behaviors
depending on the number of modules shaded and their position in respect to the array.
21
Figure 10, Figure 11 and Figure 12 show how the I-V and power characteristic curves are affected differently as
multiple modules or sub-modules are shaded either in a series or a parallel connection.
From the comparison of Figure 11 and Figure 12 it can be seen that a shading situation covering fewer modules
in several strings is preferable to several modules in few strings. The power loss remains almost constant for an
array with parallel connections when shaded modules (or sub modules) on the same string increases. The power
loss for parallel connections increases when the numbers of shaded strings increase [27].
Figure 10 - Shading situation where two to eight modules in a string are shaded, where the black curves = no shading, blue curves = 2 modules shaded, green curves = 4 modules shaded, orange curves = 6 modules shaded and the red
curves = 8 modules shaded [27].
Figure 11 - Shading situation of an array where two of the strings are shaded from two to eight modules, where the black curves = no shading, blue curves = 2 modules shaded, green curves = 4 modules shaded, orange curves = 6
modules shaded and the red curves = 8 modules shaded [27].
22
While the evaluation of the shading is complex and needs a proper software to be evaluated precisely, for a
preliminary design some useful tools can be used to address the criticism of the area and give an idea of the best
location for the installation before to deeply investigate the problem with PVsyst in the design phase of the
project.
3.3.2. 3D modelling
3D design is fundamental part of the preliminary design. It helps in several ways the definition of the project.
First, the definition of a 3D scene helps to have a better overall idea about the buildings that, especially when
particularly big, are not as easy to visualize as after a 3D reproduction. Furthermore, the shades on the PV field
can be analyzed properly taking into account all the object that can interfere with the PV field.
Many simulation tools can be found to address the more precisely possible the losses due to obstacles. The more
accurate are the 3D modelling tools coupled with solar simulation software. For a preliminary approach, to
simulate the system quickly, to have a rough idea about the possible size of the system that can be placed on
the roof and to simulate the shading, SketchUp software with Skelion plugin has been found very useful. The 3D
modelling software by Google allows to create a precise drawing of the buildings with an interface easy and
intuitive as well as fast. 3D modelling is very useful as well to interact with the customers who are requesting
the project showing them the possible outcome of the future installation.
Figure 12 - Shading situation of an array where two modules on one to four strings are shaded, where the black curves = no shading, blue curves = 2 modules shaded, green curves = 4 modules shaded, orange curves = 6 modules shaded
and the red curves = 8 modules shaded [27].
23
Skelion plugin is a powerful tool that enables to put solar panels on SketchUp surfaces drawings. Once the model
of the building and the surrounding area have been created, selecting one or more surfaces, solar panels can be
added in the drawing. This tool is particularly useful since it has lot of panel models and gives the possibility of
adding new ones with dimensions and specifications. It allows the user to decide the number of panels, the
orientation, the relative tilt, the distance of each panel from the successive one and the distance of each row.
Skelion is also integrated with the PVWatts calculator by NREL that estimates the energy production of grid-
connected PV systems. PVWatts report is not a precise report. The system is defined only for the modules type
and disposition on the roof. String connection, inverter and cabling are not taken into account and thus cannot
be considered reliable for a proper project design but is useful to have a first idea about the project and the
possible modules layout. It is also possible to export the 3D drawing into PVsyst to better analyze the shading
impact in the final project design phase.
For what concern the case study, the buildings of the garment factory have been found very large, with a roof
area that is bigger than the space required for the PV system. The compound is characterized by three main
buildings where all the practical work is done and that have the best type of roof among the others for the PV
installation. The complex has also two smaller buildings with the two different transformers and all the electrical
cabinets inside and other three small buildings for administrative purposes (not represented in the 3D model
since small and not relevant). The aerial photo of the garment factory has been taken from Google Maps and is
shown in Figure 13 together with its location.
To optimize the future cable length, two of the main buildings have been chosen for the installation of the solar
system. Each building is positioned in front of the electrical room where the inverters will have to be positioned
Figure 13 - Aerial image of the garment factory (left) and its location in Cambodia (right).
24
and have a pre-existing cable duct that will facilitate the wiring procedures. The first building where the system
connected to T1 need to be placed is shown in Figure 14 while the roofs main characteristics have been listed in
Table 7. The roofs of the factory present a slop of 8°, have an azimuth angle of -4°and thus is facing almost
perfectly South on one side and North on the other side. The roofs area has been found completely empty from
nearby obstacles as can be seen from the pictures of the site visit in Appendix 2. Only one small antenna is
present on one of the roofs (single pole 2 m high) but it can be moved in another location easily and thus has
not been considered in the 3D simulation. Natural lighting of the buildings would not be influenced by the PV
modules since the roof does not have openings and the only windows are located at the first-floor level, below
the roof. The few trees present in the compound are lower than the buildings and they are not expected to grow
much more than actual situation. There are no chimneys or cooling systems exhausts on the rooftops neither
other higher building in the surrounding area. The roof is the typical metal roof of Cambodian factories. It is
quite robust and permits to install the rail structures for the solar panels or to mount them directly on its surface.
It is not easily accessible since there are no stairs to the top, but in case of installation the owners will provide
the necessary equipment. Monsoons in this area are not particularly strong and since the panels will be semi-
integrated on the roof with a small air duct behind, there is no special concern on monsoons winds in the PV
structure design.
Table 7 - Main roof characteristics of the garment factory.
Roof Characteristics
Factory roof area available for PV 26280 m2 (146m X 60m X 3 buildings) South facing roof area available for PV 13140 m2 (146m X 30m X 3 buildings)
Roof type Metal roof Roof inclination 8°
Roof orientation (azimuth) 4°/184° (depending on the side) Distance of the roofs of the buildings from the
electrical rooms ≈30m (each one from the respective electrical room taking into account the vertical gap)
Drone Mapping
Thanks to the development in the field and the everyday grater affordability of these systems, drones in the last
few years have found many applications in various fields. A mention in this study has been done since it has
been found a useful method for a fast and precise 2D and 3D design of the selected area for the PV installation.
In Cambodia as in other countries, it can happen that the roof is not easily accessible, at all or in part, that the
roof structure is particularly complicated and thus not easily measurable in all its part or it is not safe to walk on
it. For all these reasons a site visit with measurements can be difficult. Moreover, the site for the installation can
be too recent to have updated pictures of the area from Google Maps and thus the dimensions of the buildings
and the surrounding area need to be measured. Drones represents a cheap and easy tool to overcome the
problem making the measurements on the area fast and precise. Drone mapping principle is to take many
pictures of a selected location during the flight to then reproduce the area in a 2D and 3D. 2D maps can be
helpful to compute distance more precisely than with google maps to calculate for instance the cable length for
25
the installation. On the other hand, 3D models can be used to evaluate possible obstacles in the area as well as
roof dimensions and exposition.
Different capture mode for the camera and the drone flight can be used depending on the application used, on
the purpose of the flight and on the ability of the pilot as can be seen in Figure 15. For 2D maps all the flight
modes give good results while for 3D mapping the double grid and the circular flight has been used and
compared and has been found that the double grid mode presents better results. This flight mode captures more
photos than the circular one for the same area and allows the program to better process the model. Important
attention for the mapping has to be made on the camera tilt, the drone speed and elevation with respect to the
building and, in case of double grid, the front and back overlaps during the shooting. Camera angle (tilt) has
been found of primary importance for the quality of the 3D model. While for 2D maps a 90° angle (the camera
is pointing perfectly down from the aircraft) is excellent, in case of 3D model the tilt of the camera has to be
adjusted depending on the altitude of the aircraft flight in order to point at the subject with an angle that permit
then to the program to evaluate the dimensions in respect on all the axis correctly. The front and back overlaps
during the double grid flight, influence the area of the site that is shared by more pictures, i.e. the drone can
shoot every 2 m or every 10 m, having or not images sharing the same part of the landscape. Other entities such
as the position of the sun and the materials of the building need to be evaluated as well to avoid reflection that
can lead to worse models processing.
It is important to notice that the experience in this field has been done with free versions of some different
software (Pix4D and DroneDeploy), and thus the results are not perfect. The models created can be utilized for
the preliminary design and to measure distances more accurately than with Google Maps, but they cannot be
used for the final installation. Some programs provide paid options with improved computation capability and
Figure 14 - View of the first building, where the solar system connected to T1 should be installed, and the corresponding electrical room.
26
thus more precise results with the possibility to export the 3D models into other programs such as Photoshop,
SketchUp or AutoCAD in order to modify and improve them. Since they have not been tested, the performances
and thus the reliability of these options cannot be evaluated in this study and need further investigation.
In Appendix 4 some 3D model elaborations that have been performed in Cambodia can be found. The results
vary considerably depending on the settings and on the site. Some of them have been used to make the
preliminary design of the successive installation while other ones were only tests. The garment factory analyzed
in this study is not present in the pictures since the 3D of the area has not been performed due to the drone
unavailability at the time of the site visit.
Figure 15 - Example of different flight modes for drone mapping in Pix4D app.
27
Chapter Four - System Design with PVsyst
PVsyst is considered one of the most comprehensive software for grid-connected systems. It is capable of
evaluate almost all the aspects of a PV project. It is divided into two main section: preliminary design and project
design. Preliminary design section has been intended for the pre-sizing step of a project that can be grid-
connected, stand alone or for water pumping. This part of the program it is only used for a quick evaluation of
the system’s and component’s sizes, it will not contribute with significant information compared to the project
design simulation option and has therefore not been considered in this study. The project design section is the
most advanced part of the program. It is divided into grid-connected, stand alone, water pumping and DC grid.
The grid-connected section of the project design of PVsyst has been used in this study to evaluate the production
of the PV system. The performances of the system have been investigated to see the impacts of a grid-connected
solar system in Cambodia. The outcomes of the simulation have been then used to produce an economic
assessment as well as an environmental analysis, both presented in Chapter Five.
The structure of this section of the work follows the process of a grid-connected simulation in PVsyst before to
summarize the simulations results. Each sub-chapter describes a specific section of the program with a deep
analysis on the settings chosen for the simulation and the definition of the models used by the program.
4.1. Geographical Location and Meteorology
The geographical location of the project and the local weather conditions influences the optimal tilt angle of the
modules as well as their yearly specific production. The PV outcome relies primarily on the amount of radiation
that it receives and on the temperature at a precise location. The geographical coordinates are therefore critical.
To compute the effective irradiance reaching the modules plane, PVsyst can use the Hay model or the Perez-
Ineichen model as explained further in this section. Both models take into account the orientation of the PV
system (tilt angle and azimuth), its location on the Earth’s surface (latitude and longitude), the available global
solar radiation and the time when this energy is available. The time of year influence the declination, which is
the angle between the earth's rotation axis and the earth-sun line, while the time of the day influence the hour
angle, that is the position of the sun in the sky. PVsyst compute the declination as:
𝛿 = 23.45° sin (360°
284 + 𝑛
365) (1)
Where 𝛿 is the declination and n is the number of day of the year (between 1 and 365).
The hour angle is computed by PVsyst as:
𝜔 = 15°(𝑡𝑠 − 12) (2)
Where 𝜔 is the hour angle and 𝑡𝑠 is the solar time.
28
Solar time differs from the standard time we use to measure the hour of the day. PVsyst compute the solar time
with equation (3):
𝑡𝑠 = 𝑡𝑠𝑡 +
𝜆
15− 𝑍 + 𝑇 (3)
Where 𝑡𝑠𝑡 is the standard time in hours corresponding to the midpoint of the time step, 𝜆 is the longitude of the
location, Z is the time zone in hours and T is the equation of time defined as:
𝑇 = 0.072 cos (2𝜋
𝑛) − 0.0528 cos (
4𝜋
𝑛) − 0.012 cos (
6𝜋
𝑛) − 0.1229 sin (
2𝜋
𝑛) − 0.1565 sin (
4𝜋
𝑛) − 0.0041 sin (
3𝜋
𝑛) (4)
The sun’s height hs, meaning the angle between the sun direction and the horizontal plane, is defined by PVsyst
with Equation 5:
sin ℎ𝑠 = sin 𝜙 sin 𝛿 + cos 𝜙 cos 𝛿 cos 𝜔 (5)
Where 𝜙 is the latitude of the location.
The sun’s height is related to an important angle in solar geometry called the zenith angle. The zenith angle is
the angle between the vertical from the ground and the sun rays and can be defined as:
𝜃𝑧 = 90° − ℎ𝑠 (6)
Where 𝜃𝑧 is the zenith angle.
The azimuth angle 𝛾, meaning the angular displacement from south of the projection of beam radiation on the
horizontal plane is defined with Equation 7:
sin 𝛾 = cos 𝛿 sin 𝜔/ cos ℎ𝑠 (7)
Finally, the incident angle 𝜃 defined as the angle between the sun’s ray and the normal to a plane is defined as:
cos 𝜃 = cos(𝛾 − 𝛾𝑝) cos ℎ𝑠 sin 𝛽 + sin ℎ𝑠 cos 𝛽 (8)
Where 𝛾𝑝 is the is the plane azimuth and 𝛽 is the plane tilt.
When the radiation from the sun enters the atmosphere, it can collide with clouds and air molecules and the
radiation can then scatter or be absorbed. The beam or direct radiation is the radiation which is not reflected or
absorbed and reaches the surface of a PV module in a direct line from the sun. The radiation after a scattering
can either be re-emitted into the atmosphere or reach the surface of the module, this is called diffuse radiation.
Albedo radiation is the radiation reaching the module surface after being reflected by the ground. Global
radiation consists of all three components: beam, diffuse and albedo radiation [28]. A scheme of beam, diffuse
and albedo radiation is presented in Figure 16. In this context is useful to define two quantity: the air mass AM
29
and the clearness index KT. The air mass represents the mass of atmosphere through which beam radiation
passes to the mass it would pass through if the sun were at the zenith and is defined by:
𝐴𝑀 = 1cos 𝜃𝑧
⁄ (9)
The clearness index is defined as the ratio of the horizontal global irradiance to the corresponding irradiance
available out of the atmosphere and represents the clearness of the atmosphere. The clearness index can be
expressed with:
𝐾𝑇 =
𝐼
𝐼𝑜
(10)
Where 𝐼 is the global horizontal radiation averaged on the time step and 𝐼𝑜 is the extraterrestrial horizontal
radiation expressed as:
𝐼𝑜 =
12 ∙ 3600
𝜋𝐺𝑠𝑐 (1 + 0.033
360𝑛
365) [cos 𝜙 cos 𝛿 (sin 𝜔2 − sin 𝜔1) +
𝜋(𝜔2 − 𝜔1)
180sin 𝜙 sin 𝛿] (11)
Where 𝐺𝑠𝑐 is the solar constant equal to 1367 W/m2 and 𝜔1 and 𝜔2 are the hour angles at the beginning and at
the end of the time step respectively [29].
The project location is the first to be defined when performing a project design in PVsyst. The location can be
chosen directly selecting it from the world map or entering the name of the country and the city. Many weather
databases are available with the compatibility of PVsyst, some are free and other must be purchased.
Meteonorm 7.2 have been found to be the most complete, updated (1991-2010) available for free and already
integrated in the program and thus has been used in this study.
Figure 16 - Beam, diffuse and ground reflected radiation [29].
30
Once meteo database has been chosen and the data imported, latitude, longitude, altitude and time zone are
displayed and a synthetic table with the monthly averages is shown. In the table, horizontal global irradiation,
horizontal diffuse irradiation, temperature, wind speed, Linke turbidity and relative humidity are displayed as
shown in Figure 17. If only monthly meteorological data are available for the chosen location, PVsyst performs
a generation of hourly synthetic meteo data using Meteonorm 7.2 algorithm. The generation of hourly value of
global horizontal irradiation from daily values is based on the Collares-Pereira algorithm that consists of two
parts: the first part calculates an average daily profile while the second part simulates the intermittent hourly
variations. DirInt model is used in PVsyst to separate global into diffuse and beam. The DirInt global to direct
model is based on a quasi-physical model, the DISC model, developed by Maxwell (1987), which has the form of
a clear sky irradiance based on a Linke turbidity factor equal to 2.2, attenuated by a function of the clearness
index KT. This beam component is then corrected by a function of the modified clearness index KT, the solar
zenith angle, the atmospheric water vapor column and a stability index that accounts for the dynamics of the
time series. The corresponding coefficients are obtained from a four-dimensional lookup table consisting of a
6 × 6 × 5 × 7 matrix [30]. This model has not been investigated in this study, but considering the clearness
index variability, it has shown improvements in the precision of about 10–20% compared to the simple diffuse
fraction Erbs model that was previously used from the software.
Irradiance and temperature, as previously mentioned, are the ones that influence the most the behavior of a
cell/module. Assuming the cell as a diode (same assumption that also PVsyst software does in the simulation),
the short circuit current of the cell is almost directly proportional to the irradiance while the voltage only slightly
Figure 17 - Synthetic table with monthly average values for the site location defined in PVsyst.
31
decreases with the irradiance. The temperature, on the other hand, mostly influence the voltage of the cell, that
decreases as the temperature increases, while the current remains approximately constant. As a result, the
power of the cell/module decreases while the temperature increases and/or the irradiance decreases. The main
equation used by PVsysy to describe the behavior of a PV panel is
𝐼 = 𝐼𝑝ℎ − 𝐼0 (𝑒
𝑞(𝑉+𝐼∙𝑅𝑠)𝑁𝑠∙𝑚∙𝑘∙𝑇𝑐
⁄− 1) −
𝑉 + 𝐼 ∙ 𝑅𝑠
𝑅𝑠ℎ
(13)
Where I is the current supplied by the module, q is the charge of the electron, V is the voltage at the terminals
of the module, 𝑅𝑠 is the series resistance, 𝑁𝑠 is the number of cells in series, m is the diode quality factor, k is
the Bolzmann’s constant, 𝑇𝑐 is the effective temperature of the cells in K, 𝑅𝑠ℎ is the shunt resistance, Iph is the
photocurrent and 𝐼0 is the reverse saturation current.
The photocurrent can be expressed in function of the reference conditions as:
𝐼𝑝ℎ =
𝐺
𝐺𝑟𝑒𝑓
[𝐼𝑝ℎ,𝑟𝑒𝑓 + 𝐼𝑠𝑐(𝑇𝑐 − 𝑇𝑐,𝑟𝑒𝑓)] (14)
Where G and Gref are the effective and reference irradiance, 𝐼𝑝ℎ,𝑟𝑒𝑓 is the reference photocurrent and 𝑇𝑐,𝑟𝑒𝑓 is
the reference cell temperature.
The reverse saturation current can be expressed as:
𝐼0 = 𝐼0,𝑟𝑒𝑓 (
𝑇𝑐
𝑇𝑐,𝑟𝑒𝑓
)
3
𝑒[𝑞𝐸𝑔
𝑚𝑘∙(
1𝑇𝑐,𝑟𝑒𝑓
−1𝑇𝑐
)] (15)
Where Eg is the energy gap of the material.
The performances of a cell/module under the influence of irradiation and temperature are illustrated in the I-V
curves presented in Figure 18.
For each project, in PVsyst is possible to define the albedo values, the transposition model to be used and the
reference temperatures for the array design by respect to the inverter input voltages.
Figure 18 - Cell/module I-V characteristic curves depending on the irradiance (left) and on the temperature (right).
32
Albedo
Albedo values are set to an average value of 0.2 for every month by default in PVsyst. The installation site is near
the city with few grass fields around. Since the albedo values for urban areas are between 0.14 and 0.22 and
between 0.15 and 0.25 for grass surroundings, the value of 0.2 has been considered adequate and thus has not
been varied from the default value.
Transposition model
The transposition model is the model to transform the horizontal data in the meteo file, to the tilted ones of the
modules surfaces. PVsyst can use the Hay model or the Perez model. The transposition model influences the
transposition factor of the system and thus the optimization of the tilt angle as well as the yearly yield.
As mentioned before in this chapter, solar radiation is divided into beam (or direct), diffuse and albedo. While
the beam component is the result of a geometrical calculation and thus does not present any intrinsic error
being not a model, the two models differs by the treatment of the diffuse component. The diffuse radiation is
divided into three part. The first is an isotropic part, received uniformly from the entire sky dome. The second is
circumsolar diffuse, resulting from forward scattering of solar radiation and concentrated in the part of the sky
around the sun. The third, referred to as horizon brightening, is concentrated near the horizon and is most
pronounced in clear skies [31]. A schematic is presented in Figure 19. The incident solar radiation is thus the sum
of the beam, the three components of the diffuse and the reflected radiation. It can thus be written as:
𝐼𝑇 = 𝐼𝑇,𝑏 + 𝐼𝑇,𝑑,𝑖𝑠𝑜 + 𝐼𝑇,𝑑,𝑐𝑠 + 𝐼𝑇,𝑑,ℎ𝑧 + 𝐼𝑇,𝑟𝑒𝑓𝑙 (16)
Where 𝐼𝑇 is the total radiation on a tilted surface, 𝐼𝑇,𝑏 is the beam component, 𝐼𝑇,𝑑,𝑖𝑠𝑜 + 𝐼𝑇,𝑑,𝑐𝑠 + 𝐼𝑇,𝑑,ℎ𝑧 are the
three component of the diffuse, respectively the isotropic, circumsolar and horizontal contribution and 𝐼𝑇,𝑟𝑒𝑓𝑙 is
the reflected radiation.
This equation can be rewritten for a collector surface area Ac, in terms of the beam and diffuse radiation on the
horizontal surface and the total radiation on the surface (assumed one, horizontal and large in extend ground)
that reflects to the tilted surface [31]:
𝐼𝑇 = 𝐼𝑏𝑅𝑏 + 𝐼𝑑,𝑖𝑠𝑜𝐹𝑐−𝑠 + 𝐼𝑑,𝑐𝑠𝑅𝑏 + 𝐼𝑑,ℎ𝑧𝐹𝑐−ℎ𝑧 + 𝐼𝜌𝑔𝐹𝑐−𝑔 (17)
Where 𝐹𝑐−𝑠 , 𝐹𝑐−ℎ𝑧 and 𝐹𝑐−𝑔 are respectively the view factor from the collector to the sky, from the collector to
the horizon and from the collector to the ground, 𝐼 is the total radiation, 𝜌𝑔 is the general reflectance of the
ground, 𝐼𝑏 is the beam radiation and 𝑅𝑏 is the mean tilted to horizontal beam irradiance ratio, calculated as cos 𝜃
cos 𝜃𝑧
where 𝜃 is the angle between the beam radiation on a surface and the normal to that surface and is called
incidence angle and 𝜃z ,the zenith angle, represents the angle of incidence of beam radiation on a horizontal
surface.
33
Hay’s model to compute 𝐼𝑇 is an improvement of the isotropic model that tended to underestimate 𝐼𝑇 . It is
based on the assumption that all diffuse radiation can be presented by two parts: the isotropic and the
circumsolar. Equation 17 can thus be written neglecting the fourth term. 𝐼𝑑,𝑇 can be defined as:
𝐼𝑑,𝑇 = 𝐼𝑑 [(1 − 𝐴𝑖) (
1 + cos 𝛽
2) + 𝐴𝑖𝑅𝑏] (18)
Where 𝐼𝑑,𝑇 is diffuse radiation on a tilted surface, 𝐼𝑑 is the total diffuse, 𝐴𝑖 is an anisotropy index which is a
function of the transmittance of the atmosphere for beam radiation and it is considered to be incident at the
same angle as the beam radiation and results equal to 𝐼𝑏 𝐼𝑜⁄ and (1 + cos 𝛽)/2 is the resulting view factor of
the sky. The total radiation on the tilted surface is then:
𝐼𝑇 = (𝐼𝑏 + 𝐼𝑑𝐴𝑖)𝑅𝑏 + 𝐼𝑑(1 − 𝐴𝑖) (
1 + cos 𝛽
2) + 𝐼𝜌𝑔 (
1 − cos 𝛽
2) (19)
Where (1 − cos 𝛽)/2 corresponds to the view factor of the ground.
Written in this form, the first term of the Equation 19 represents the direct radiation component while the
second term is the combination of both isotropic and circumsolar component of the diffuse radiation and the
third term is the albedo radiation due to ground reflection.
The Perez-Ineichen model is a more “sophisticated model”, as described in PVsyst, and is based on a more
detailed analysis of the three components of the diffuse radiation. In this model the diffuse radiation on a tilted
surface is expressed in the form:
Figure 19 - Schematic of the distribution of diffuse radiation over the sky dome showing the circumsolar and horizon brightening components added to the isotropic component [29].
34
𝐼𝑑,𝑇 = 𝐼𝑑 [(1 − 𝐹1) (
1 + cos 𝛽
2) + 𝐹1
𝑎
𝑏+ 𝐹2 sin 𝛽]
(20)
Where 𝐹1 and 𝐹2 are circumsolar and horizon brightness coefficients, they are function of three parameters that
describe the sky conditions: the zenith angle θz, a clearness ε, and a brightness ∆ that depends also from the air
mass; and 𝑎 and 𝑏 are terms that account for the angles of incidence of the cone of circumsolar radiation on the
tilted and horizontal surfaces calculated as:
𝑎 = 𝑚𝑎𝑥(0, cos 𝜗), 𝑏 = 𝑚𝑎𝑥(cos 85, cos 𝜗𝑧) (21)
From the analysis of the two models appears that the Hay model is the simpler one, slightly more complex than
the basic isotropic model but with results that are closer to measured values. The Perez model is more complex
and has usually a 2% (or more) higher yield as reported also by PVsyst developers in [31]. In [29] it is suggested
that Perez model is more suitable for modules oriented far from 0° azimuth angle in the northern hemisphere
and that Hay model presents better results when predicting utilizable radiation when the critical radiation levels
are significant. Moreover, Hay model gives good results even when the knowledge of the diffuse radiation is not
perfect while Perez model requires better well measured horizontal data. Since the meteo file available for the
chosen location in Cambodia is not well updated, reaching 2010 data only and the data are extrapolated from
interpolation of nearby stations, since Cambodia has high radiation levels and the roofs of the garment factory
have an azimuth of -4°, the Hay model has been chosen as a more appropriate and conservative option for this
study.
Reference temperatures
For what concerns the reference temperatures, Pvsyst considers four different temperatures: the lower
temperature for absolute voltage limit computation, the summer operating temperature for the
𝑉𝑀𝑃𝑃,𝑖𝑛𝑣𝑒𝑟𝑡𝑒𝑟 𝑚𝑖𝑛 (maximum power point voltage minimum limit of the inverter) design, the usual operating
temperature under 1000 W/m2 of irradiance and the winter operating temperature for 𝑉𝑀𝑃𝑃,𝑖𝑛𝑣𝑒𝑟𝑡𝑒𝑟 𝑚𝑎𝑥
(maximum power point voltage maximum limit of the inverter) design. The first temperature, by default set at
a value of -10°C, is the minimum cell temperature at which the system can work in order to respect the upper
voltage limit of the inverter range. In Cambodia, in Phnom Penh area, the lowest temperature recorded have
been found to be 12.8°C [32]. For this limit the cell temperature has been considered equal to the ambient
temperature since the colder temperature are registered during the night and at the first light in the morning
the sun has not had the time to heat up the modules. In a conservative perspective has thus been set as 10°C.
The maximum operating temperature for the calculation of 𝑉𝑀𝑃𝑃,𝑖𝑛𝑣𝑒𝑟𝑡𝑒𝑟 𝑚𝑖𝑛, by default set as 60°C, has been
raised due to the high temperatures of the Cambodian environment. The highest temperature recorded in
Cambodia has been found to be 42.6°C [33]. To compute the cell temperature for this condition, Equation 22
has been applied:
35
𝑇𝑐𝑒𝑙𝑙 = 𝑇𝑎𝑚𝑏 + (
𝑁𝑂𝐶𝑇 − 20
800) 𝐺 (22)
Where 𝑇𝑐𝑒𝑙𝑙 is the temperature of the cell in °C, 𝑇𝑎𝑚𝑏 is the ambient temperature in °C, NOCT is the module type
dependent test temperature at an irradiance of 800 W/m2 with an ambient temperature of 20°C and G is the
effective solar irradiance in W/m2.
The NOCT has been found to be in the range of 41-47°C for the modules considered for the installation and thus
has been considered 47°C as the worst-case scenario for high temperatures. For what concerns the effective
irradiance, examining the Meteonorm meteo data file in PVsyst, a maximum irradiance of 1066 W/m2 has been
found. A value of 1100 W/m2 has been considered in a conservative perspective, taking into account that higher
values can be achieved but only for short periods of time and thus their influence on the system can be
neglected. The maximum cell operating temperature, following the calculation, was thus found to be 78.35°C
and thus has been set at a value of 80°C in the program.
Other two temperature parameters are also defined in the program: the usual operating temperature under
1000 W/m which has 50°C as default value and the winter operating temperature for 𝑉𝑀𝑃𝑃,𝑖𝑛𝑣𝑒𝑟𝑡𝑒𝑟 𝑚𝑎𝑥 design
parameter, which describes the minimum cell temperature during operating conditions in winter and is set at a
value of 20°C. The usual operating condition have been raised up to a value of 62°C following Equation 6. For
the value of the winter operating temperature, the computation using the extreme value of 12.8°C and a low
effective irradiance of 100 W/m2 has given as result 16°C. Nevertheless, temperature in Cambodia during winter
has an average of 26°C and hardly never goes below 22°C. Furthermore, winter in Cambodia is the season with
less precipitation and less cloud cover and thus the irradiation level is usually high. Examining the meteo file,
has emerged that even during the first lights in the morning and at evening rarely goes below 150 W/m2. For
these reasons, the default winter operating temperature of 20°C has been considered a safe value for the project
settings and has thus not been changed. Table 8 summarizes the reference temperatures of the project.
Table 8 - Reference temperature for the garment factory site.
Site dependent design parameters
Low temperature absolute limit 12°C Winter operating temperature for VmppMax design 20°C
Usual operating Temperature under 1000 W/m2 62°C Summer operating temperature for VmppMin design 80°C
4.2. Orientation
The first practical settings of the project design in PVsyst are chosen in the Orientation section of the program.
In this project the fixed tilt inclination of the panels was chosen between the various options. One or two axis
tracking systems are mostly used for large scale ground mounted systems and are very uncommon in country
such as Cambodia. Their high costs compared to the better yields that can be achieved with this solution make
those system not economically convenient, especially for internal consumption purpose. In tropical countries
36
such as Cambodia, the sun path variability during the year is not as high as in European countries, the days have
almost a constant length of 12 hours, the sun rise faster in the sky than in Europe as well as it lowers down on
the horizon and thus tracking systems are not justified and have not been considered in this study.
The amount of radiation collected on the solar modules should be as large as possible and the tilt angle of the
panels highly influence this value. The tilt of fixed modules can be maximized with regards to seasonal
performance or annual performance. The choice depends on the user needs and on the type of installation
(stand alone or grid connected). In this study the yearly optimization has been chosen. The optimum tilt angle
𝛽𝑜𝑝𝑡 depends mainly on latitude ϕ. The general rule with regards to the highest annual performances is to apply
a tilt angle equal to the latitude. In case of Phnom Penh in Cambodia this value is 11.56°. The common rule is
confirmed by computing the optimal tilt using Equation 23 [26]:
𝛽𝑜𝑝𝑡 = 3.7 + 0.69|𝜙| (23)
The result is 11.68°, with almost no difference from the latitude of the location. The computation result differs
more for high latitude values, i.e. for locations further from the equator. Furthermore, being on the northern
hemisphere, the panels should be oriented towards South (0° azimuth) in order to collect as much radiation as
possible.
In PVsyst, an optimization tool for azimuth and tilt angle is present and has been taken into account for this
study. The garment factory roofs are tilted of an angle of 8° and have an azimuth of -4° and thus result almost
in the optimal orientation. PVsyst computed an optimum tilt angle of 10° and an optimum azimuth of 0°.
Nevertheless, the small differences from the installation site characteristics do not lead to high losses as shown
in Figure 20 and thus a different orientation has not been considered.
Figure 20 - Possible losses in respect to the optimum situation computed with PVsyst for the garment factory case study.
37
4.3. Matching PV arrays and Inverters
The second section of the program consists in the system components and size definition. PV arrays and
inverters need to match in three areas: power, voltage and current. First, the input power of the inverter limits
the maximum number of panels that can be connected to it. Second, the voltage range of the array should be
within the MPPT voltage range of the inverter. Finally, the current of the strings connected in parallel cannot
exceed the maximum input current of the inverter. To describe the behavior of each module, PVsyst utilizes the
single-diode model which implies that all the cells are identical.
More complicated models such as the two-diodes model exist. The single-diode model has been considered
sufficiently accurate in PVsyst, since is considered well-suited for Si-crystalline modules and mismatch losses are
taken into account separately [34]. Mismatch losses will be explained in Section 4.4. The solar array and inverter
have to be optimally matched to each other's output values. The nominal power of inverters can be ±20% of the
PV peak power, depending upon the inverter and module technology and the local conditions. In general, the
following power range can be specified for the sizing range [25]:
0.8𝑃𝑃𝑉 < 𝑃𝐼𝑁𝑉,𝐷𝐶 < 1.2𝑃𝑃𝑉
Where 𝑃𝑃𝑉 is the nominal power of the PV system and 𝑃𝐼𝑁𝑉,𝐷𝐶 is the maximum DC input power of the inverter.
To determine the minimum number of PV modules that can be connected in series in a string, the maximum cell
temperature is computed and used to evaluate the minimum voltage of the string. The reason in that the
minimum voltage of the string should not fall below the minimum operating voltage of the inverter. If the
operating voltage of the system drops below the minimum MPP voltage of the inverter, this would no longer
feed the maximum possible power and, in the worst-case, would even switch itself off. The minimum number
of modules in series can be computed using Equation 24 [25]:
𝑛𝑚𝑖𝑛 =
𝑉𝑀𝑃𝑃,𝑖𝑛𝑣𝑒𝑟𝑡𝑒𝑟 𝑚𝑖𝑛
𝑉𝑀𝑃𝑃,𝑚𝑜𝑑𝑢𝑙𝑒 max 𝑡𝑒𝑚𝑝
(24)
Where 𝑉𝑀𝑃𝑃,𝑖𝑛𝑣𝑒𝑟𝑡𝑒𝑟 𝑚𝑖𝑛 is the minimum MPP voltage of the inverter and where:
𝑉𝑀𝑃𝑃,𝑚𝑜𝑑𝑢𝑙𝑒 max 𝑡𝑒𝑚𝑝 = (1 + ∆𝑇 ∙ 𝛾𝑉𝑀𝑃𝑃)𝑉𝑀𝑃𝑃,𝑆𝑇𝐶 (25)
Where ∆𝑇 is the temperature difference between the highest cell temperature and the cell temperature under
STC (for example in the case study is 80°-25°C), 𝛾𝑉𝑀𝑃𝑃 is the voltage maximum temperature coefficient of the
panel in %/°C and 𝑉𝑀𝑃𝑃,𝑆𝑇𝐶 is the MPP voltage of the module under STC.
To determine the maximum number of modules that can be connected in series the minimum cell temperature
must be used. The highest voltage that can occur in an operating condition is the open-circuit voltage at lowest
temperature. Thus, the maximum number of series-connected modules is derived from the quotient of the
maximum input voltage of the inverter and the open-circuit voltage of the module at -10°C [25]:
38
𝑛𝑚𝑎𝑥 =
𝑉𝑚𝑎𝑥 𝑖𝑛𝑣𝑒𝑟𝑡𝑒𝑟
𝑉𝑂𝐶,𝑚𝑜𝑑𝑢𝑙𝑒 min 𝑡𝑒𝑚𝑝
(26)
Where 𝑉𝑚𝑎𝑥 𝑖𝑛𝑣𝑒𝑟𝑡𝑒𝑟 is the maximum DC input voltage of the inverter and where:
𝑉𝑂𝐶,𝑚𝑜𝑑𝑢𝑙𝑒 min 𝑡𝑒𝑚𝑝 = 𝑉𝑂𝐶,𝑆𝑇𝐶 − (𝛾𝑉𝑂𝐶∙ 𝑉𝑂𝐶,𝑆𝑇𝐶 ∙ ∆𝑇) (27)
Where 𝑉𝑂𝐶,𝑆𝑇𝐶 is the open circuit voltage of the panel under STC, 𝛾𝑉𝑂𝐶 is the open circuit temperature coefficient
of the module in %/°C (that has to be negative) and ∆𝑇 is the temperature difference between the STC of 25°C
and the lowest temperature.
Even if the current is only slightly influenced by the temperature, it is needed to ensure that the maximum PV
array current does not exceed the maximum inverter input current. The maximum number of strings in parallel
can be computed in accordance with:
𝑛𝑠𝑡𝑟𝑖𝑛𝑔 𝑚𝑎𝑥 =
𝐼𝑚𝑎𝑥 𝑖𝑛𝑣𝑒𝑟𝑡𝑒𝑟
𝐼𝑆𝐶,𝑆𝑇𝐶 + (𝛾I𝑆𝐶∙ ∆𝑇)
(28)
Where 𝐼𝑚𝑎𝑥 𝑖𝑛𝑣𝑒𝑟𝑡𝑒𝑟 is the maximum input current of the inverter, 𝐼𝑆𝐶,𝑆𝑇𝐶 is the short circuit current of the
module under STC, 𝛾I𝑆𝐶 is the short circuit current temperature coefficient of the module in %/°C and ∆𝑇 is the
temperature difference between the STC and the highest cell temperature.
In PVsyst, the inverter sizing is based on an acceptable overload loss during operation, and therefore involves
estimations or simulations in the real conditions of the system (meteo, orientation, losses).
With all modern inverters, when the maximum power of the array overcomes its DC nominal power limit, the
inverter will stay at its safe nominal power by displacing the operating point in the I-V curve of the PV array.
Therefore, it will not undertake any overpower, simply the potential power of the array is not produced. There
is no power to dissipate, no overheating and therefore no supplementary ageing of the system [35]. In this way,
the only energy loss is the difference between the "potential" maximum power (the power that theoretically
can be achieved by the PV system) and the nominal DC power limit effectively drawn. Even when the inverter's
power is slightly under the maximum power attained by the array in real operation, this results in very little
power losses. For instance, in the first transformer part of the garment factory (T1) simulation, the annual power
losses due to overload have been estimated by PVsyst to be 4 kWh with a nominal power array/inverter ratio of
1.14 (1323 Jinko Solar modules of 340 Wp and 11 Huawei inverters of 36 kW). An extra inverter (or one inverter
of higher power) would result in higher investment cost for the installation that is not justified by this small
amount of power losses. The example of array voltage sizing for T1 is reported in Figure 21. The inverter
maximum current limit is represented by the horizontal violet line and the inverter maximum DC input power is
represented by the dotted curve while the voltage range are shown by the vertical lines.
39
Module choice
The modules choice is very wide today and will not be deeply investigated in this study. The selection of the
modules has been restricted to three manufacturers: Jinko Solar, Canadian Solar and Antaris. Jinko Solar is a
first-tier supplier from China with interest in solar installations in Cambodia. It has been chosen since it produces
high quality poly- and monocrystalline panels at a very cheap price compared to other comparable brands.
Canadian Solar, another first-tier supplier, has been chosen for its quality of panels at an affordable price and
because it was already collaborating with Kamworks for another big project in Cambodia. Antaris, a German
solar company, has been chosen because it has been the supplier of solar modules for many Kamworks projects.
It produces quality poly-crystalline panels at an affordable price and is already familiar with the company. From
each company, one panel model has been chosen: JKM 340PP-72-V polycrystalline module from Jinko Solar (340
Wp and 17.52% efficiency), CS6U-335P polycrystalline module from Canadian Solar (335 Wp and 17.23%
efficiency) and AS P72 320 polycrystalline module from Antaris (320 Wp and 16.49% efficiency). No
monocrystalline modules have been taken into account since the available space on the roof was enough to fit
all the power needed with the polycrystalline modules and the higher efficiency of the monocrystalline was not
justified by their higher cost for the case study installation. Further investigation should be made in order to
address the possible higher savings for what concern the CO2 emissions in a cradle-to-grave perspective of this
type of solution.
Inverter choice
The choice of the inverter is as wide as the module choice and has been done shrinking the long list of
manufacturers to two: SMA and Huawei. SMA is one of the world leader of inverters technologies and has been
already used for many installations by Kamworks. Their inverters are reliable and durable and power range is
very wide. Huawei is quite new in the solar market but have already worked with Kamworks for medium-size
installations with great results. Moreover, Huawei has recently established a Cambodian branch of the company
and thus the order of the materials, the shipping and spare parts pick up can be easier and faster than with other
Figure 21 - Inverter/array sizing in PVsyst with 63 strings of 21 modules model JKM 340PP-72-V by Jinko Solar and 11 inverters SUN2000 36-KTL by Huawei.
40
companies. Both of the companies do not have an inverter perfectly sized as the garment installation would
need and thus smaller inverters models have been chosen. Moreover, a central inverter for a large application
is not the preferred solution. The shades impact will be greater with all the strings connected to one inverter
and in case of failure, all the system would not work resulting in high energy production losses. On the other
hand, the cost related to one inverter only would be lower than the costs of several smaller inverters.
One model has been chosen from each company: SunnyBoy Tripower 25000TL JP-30 with a peak power of 25
kW from SMA and SUN2000 36KTL from Huawei with a peak power of 36 kW.
All the technical specifications of the three different modules and the two inverters can be found in Appendix 5.
4.4. Detailed Losses
Array losses can be defined as all events which penalize the available array output energy with respect to the
PV-module nominal power as quoted by the manufacturer for STC conditions [31]. In order to facilitate
comparisons between several PV installations and take into account all the losses, JRC (European Joint Research
Center) introduced the Normalized Performance Index and the Performance Ratio (PR), now fixed in the IEC EN
61724 norm. The ideal yield of 100% considers that a PV array would produce one kWh energy under one kWh
of incident energy on the collector plane for every kWp of solar PV installed (as stated in STC). This ideal Yield is
diminished by the following losses:
▪ Shading losses and electrical effect
▪ Incident Angle Modifier (IAM)
▪ Irradiance losses
▪ Thermal behavior
▪ Real performance losses
▪ Mismatch losses
▪ Soiling losses
▪ Ohmic losses
In normalized performance index, all these array losses are accounted for in the "Collection Losses" (Lc) that is
the difference between the ideal array yield at STC (Yr) without taking into account any losses and the effective
yield as measured at the output of the array (Ya), referred to the nominal power and expressed in kWh/kWp/day.
The system yield (Yf), is defined as the system daily useful energy, referred to the nominal power and is
expressed in kWh/kWp/day. With these quantities is possible to define the PR as expressed in Equation 29:
𝑃𝑅 =
𝑌𝑓
𝑌𝑟
(29)
41
For a grid-connected system, the available energy produced with the solar system is expressed as Egrid in PVsyst
and Equation 29 can be written as:
𝑃𝑅 =
𝐸𝑔𝑟𝑖𝑑
𝐻𝑖𝑛𝑐 ∙ 𝐴𝑃𝑉 ∙ 𝜂𝑚
(30)
Where Egrid is expressed in kWh, 𝐻𝑖𝑛𝑐 is the total irradiation on the panels plane expressed in kWh/m2, 𝐴𝑃𝑉 is
the total panels area in m2 and 𝜂𝑚 is the modules efficiency.
The PR includes all the optical losses, the array losses and the system losses, being Egrid measured at the PV
output, and is therefore a useful indicator to compare the system quality for different installations. Being the
losses dependent on the environment conditions, the PR is not constant along the year, but an average value is
produced in PVsyst to evaluate the overall performances.
PVsyst has a dialog box to define all the detailed losses of the system. Some default values are already set but
all of them can be modified by the user. The specific definition of each loss is presented in this section.
Shading losses and electrical effect
PVsyst distinguishes two different types of shades: far shadings described as a horizon line below the one the
sun is not visible anymore and near shadings that are produced by object close by the panels. The first type of
shades acts on the PV installation in a global way, obscuring the sun. Typically, the distance of these objects
should be more than ten times the PV field size to avoid serious losses. For the second type of shades, the ratio
of the shaded area, with respect to the total sensitive area of the PV field for a given position of the sun, is called
shading factor. The impact of these shades cannot be evaluated without a 3D description of the installation and
the surrounding area. The shading calculations have to be computed at each hour, and applied differently on
the beam, diffuse and albedo components. This analysis can be done accurately exclusively with the help of a
simulation software and the definition of a 3D scene. PVsyst have four option that can be chosen for the shading
computation: no shading, linear shading, according to the strings and according to the module layout. The first
option does not consider shading while the second option uses the 3D scene to compute the table of shading
factors as a function of the altitude of the sun and azimuth. In the third option, the PV field is split into
rectangular areas, each representing a whole string of modules. When calculating the shading factors the
corresponding string becomes unproductive if a rectangle is shaded. The fourth option is the more accurate and
it has been used in this study. It takes into account the real position of the modules on the 3D scene and their
connections. Because of that it is performed at the end of the design process when the module layout is
completely defined.
The shading factors for the beam will be dependent on the sun position while the shading factors for diffuse and
for albedo will be independent on the sun position. The shading factor for the beam is a purely geometric
calculation. It is necessary to project, at a certain time, the object’s shade on the PV field and then compute the
ratio of the shaded area in comparison to the whole PV area (factor from 0, with no shading, to 1 when total
shaded). For the beam component, two kind of losses must be considered: irradiance losses, which correspond
42
to the deficit of irradiance on the cells (also called "linear shading losses") that are computed from the shading
factor previously explained and electrical losses, resulting from the mismatch of electrical response of the
modules in series and strings in parallel: in a string of modules or cells, the total current is always determined
by the current in the weakest cell, if the current imposed in the string is higher than the max. current of a shaded
cell, the by-pass diode of the concerned sub-module will be activated. The modules layout (landscape or portrait
position) only affects the electrical losses due to the different relative position of the shade in respect to the
bypass diodes connection.
The shading factors for the diffuse can be computed as an integral of the shading factor, calculated for each
direction of the space. The integral must be performed over the sky vault, as "seen" by the collector plane. This
integral, is not dependent on the sun position, but only on the geometry of the system. It is therefore constant
along the whole year and even independent from the latitude.
For the albedo component can be assumed that for a given azimuth, the albedo contribution is proportional to
the part which would be "seen" below the azimuth point, between the horizontal plane and the prolongation of
the collector plane. The albedo losses have a weak contribution on the yearly yield but a higher influence on the
performance ratio. The albedo is indeed part of the transposition value (𝐺𝑔𝑙𝑜𝑏,𝑖𝑛𝑐) and the PR is referenced to it
as expressed in Equation 30.
Incident Angle Modifier
The incidence effect, expressed as the Incidence Angle Modifier (IAM), is the decrease in the irradiance that
effectively reaches the cells’ surfaces. It can be expressed as reflection loss and it is due to the reflection on the
glass cover which increase with the incident angle. Theoretical models, based on the Fresnel’s law, have been
developed for clean surfaces. The most popular formulation, and the one used in PVsyst, is the ASHRAE
parametrization expressed with Equation 31:
𝐹𝑇𝐵(𝜗𝑆) = 1 − 𝑏0 (
1
cos 𝜗𝑆
− 1) (31)
Where 𝐹𝑇𝐵(𝜗𝑆) is the relative transmittance, normalized by the total transmittance for normal incidence, 𝑏0 is
an adjustable parameter that can be empirically determined for each type of photovoltaic module and 𝜗𝑆 is the
incident angle [26]. If the value of 𝑏0 is unknown, PVsyst default value for crystalline module is 0.05. The effect
of the angle of incidence on the successfully collected solar radiation can be calculated by applying Equation 31
to the direct and circumsolar irradiances, and by considering an approximated value, FT = 0.9, for the isotropic
diffuse and reflected radiation terms [26]. The ASHRAE model is simple to use but has some disadvantages. It
cannot be used for 𝜗𝑆 > 80° and it cannot take into consideration the effects of dust. Dust is considered
separately into PVsyst as well as the soiling losses and will be subsequently discussed in this chapter. To better
estimate the IAM, PVsyst gives the possibility to specify the type of glass cover for the new modules (normal,
plastic or with anti-reflective coating) in order to apply the Fresnel’s law. Is even possible to set a specific IAM
profile defined by the user. Figure 22 shows the different parametrization curves. It has been noticed that the
ASHRAE parametrization underestimates the IAM value at medium angles (30° to 60°) and overestimates them
43
over these values [31]. In this study, since all PV modules considered have normal tempered glass without anti-
reflective coating, since more specific data where absent and since high incident angles occur only for short
periods of time in Cambodia being the country near the equator, the Fresnel parametrization for normal glass
has been chosen for all the three panels models simulated.
It should be stressed that angular-dependent reflection become significant in many practical situations, for
example, where vertical (facade-integrated PV generators) or horizontal (N–S horizontal trackers) surfaces are
concerned. Furthermore, they help to explain the observed low irradiance effects in PV module performance.
This is because low irradiance just happens when the incidence angle is large or when solar radiation is mainly
diffuse. In both cases, angular losses are particularly important [26].
Thermal losses
Thermal losses describe the loss as a result of the temperature difference between the modules in the array and
at 25°C (STC). The operating temperature of the modules is usually much higher than the STC one and thus highly
influences the output of the array. In Cambodia those losses account for the majority in a PV system due to the
high temperatures that occur during the whole year therefore the estimation of the thermal behavior is
fundamental.
In PVsyst there are two different ways to consider thermal losses: defining the field thermal loss factor or the
standard NOCT coefficient. Despite the fact that NOCT coefficient is usually specified in the modules catalogues
by the manufacturers, it does not include any information about the mounting mode (free ventilated, integrated
or insulated). It is always given for a "nude" module. Therefore, the NOCT is useless for the module temperature
evaluation in any conditions during the simulation [31] and its use is highly discouraged in PVsyst.
The thermal behavior of the field can be expressed as the energy balance between the ambient temperature
and the cell’s heating up due to the incident irradiance:
Figure 22 - Incident Angle Modifier for different parametrizations.
44
𝑈(𝑇𝑐𝑒𝑙𝑙 − 𝑇𝑎𝑚𝑏) = 𝛼 ∙ 𝐺(1 − 𝜂𝑚) (32)
Where 𝑈 is the thermal loss factor, 𝑇𝑐𝑒𝑙𝑙 is the cell temperature computed with Equation 22, 𝑇𝑎𝑚𝑏 is the ambient
temperature, 𝛼 is the absorption coefficient of solar radiation, 𝐺 is the irradiance on the module and 𝜂𝑚 is the
module efficiency. The absorption coefficient has a default value of 0.9 in PVsyst but can be changed in the
module settings. The default value as been assumed for this study.
The thermal loss factor is defined in PVsyst as:
𝑈 = 𝑈𝑐 + 𝑈𝑣 ∙ 𝑣 (33)
Where 𝑈𝑐 is the constant component of the thermal loss factor expressed in 𝑊/𝑚2 ∙ 𝐾 and 𝑈𝑣 is the factor
proportional to the wind velocity 𝑣. These U-factors depend on the mounting of the modules that in PVsyst can
be set as: “free” mounted module with air circulation, semi-integrated with air duct behind or integrated with
fully insulated back. The determination of the parameters 𝑈𝑐 and 𝑈𝑣 is indeed not easy. In the absence of reliable
measured data, PVsyst proposes some default values depending on the mounting system and not depending on
the wind velocity (i.e. assuming average wind speed). For semi-integrated system with air duct on the back of
the panels, as in the case of the garment factory analyzed, the default value is 𝑈 = 20 𝑊/𝑚2 ∙ 𝐾. This value has
not been changed due to the lack of data from the manufacturer and the absence of measurement of wind
speed inside the air duct for similar installations in Cambodia.
Real performance losses
Real performance losses or quality losses express the confidence of the user to the real module’s performance,
with respect to the manufacturer’s specifications. PVsyst initializes the "Module Quality Loss" according to the
PV module manufacturer's tolerance specification choosing a quarter of the difference between these values.
For example, with -3...+3%, it will be 1.5%, and with positive sorting 0..+3%, it will be -0.75% (i.e. a negative loss
value, representing a gain). During the simulation, this factor will induce a loss on the array Pmpp production,
constant (in percentage) over all operating conditions [31]. While in the past most of PV modules didn't match
the manufacturer nominal specifications, now with the panels having guaranteed power and flash test
assertions, the specifications should be considered more reliable. This value has thus been set to 0 in this study.
Moreover, LID losses are accounted in another section of the detailed losses tool. LID (Light Induced
Degradation) losses are a loss of performances arising in the first hours of exposition to the sun that affects
crystalline modules. Only p-type doped silicon suffers of this loss that is due to traces of Oxygen included in the
molten Silicon during the Czochralski process. Under the light exposition effect, these positive-charged O2 dimers
may diffuse across the silicon lattice and create complexes with boron dopant acceptors. The boron-oxygen
complexes create their own energy levels in the silicon lattice and can capture electrons and holes which are
lost for the PV effect [31]. LID losses are normally in the range of 1% and 3% but their precise estimation is hard
due to the lack of information from the manufacturers. For the case study the PV modules manufacturers have
45
been contacted and all of them stated that for their modules LID losses are absent or already included in the
first-year degradation. Because of this they have not been taken into account for the simulations.
Mismatch losses and degradation
Mismatch losses are due to the fact that in a string of modules the lowest current drives the current of the whole
string. In a real installation, the characteristics of the modules, also if they are of the same model, will slightly
differ one from the other.
PVsyst has made a tool to statistically estimate the corresponding power loss. This tool first creates a statistical
sample of modules, setting a random dispersion of the characteristics of VOC and ISC for each module according
to a Gaussian or square distribution. Then it adds the I-V characteristics of each module in each string (add
voltages) and then gathers the strings in the array (add currents). Finally, it draws the resulting I/V curve of the
array and identifies the MPP value which may be compared to the MPP value of an array with identical modules
[31]. This loss result is different for every statistical sample and a histogram can be visualized in order to see the
different samples results. The default value of PVsyst is set as 1% and has been used in this study. The reason is
that the mismatch due to the degradation of the panels have been considered separately as well as the soiling
losses that could increase this value. A low value has thus been considerate appropriate to avoid overestimation
of the losses.
Degradation, or ageing, of the panels can be considered in the simulations and has been used in this study.
PVsyst gives the possibility of including degradation losses that give rise to a progressive loss of the efficiency
into the simulation. The degradation means a decrease in the system yield and may have sometimes positive
effects due to the decrease of overpower losses when the inverter is undersized. Moreover, not all the modules
will degrade at the same rate and thus the mismatch will increase with the time. This simulation of the increasing
mismatch as a function of the year is performed by the program with a Monte-Carlo method (choice of a great
number of random distributions) with the hypothesis that the degradation rate of each module is constant over
the years. This method does not have much references for the average module degradation and for the real
ageing discrepancies between modules. Nevertheless, considering the module warranty as a lower limit for the
modules degradation it can be considered sufficiently reliable. Figure 23 shows the degradation of the Jinko
Solar panels as an example.
Soiling losses
Soiling losses are defined as the losses due to the accumulation of dirt on the panels. The effects on the system
is an uncertainty that strongly depends on the environment of the system, on the raining conditions, on the tilt
angle of the panels etc.
In medium-rainy climates (like middle of Europe) and in residential zones, this is usually low and may be
neglected (less than 1%) [31]. In Cambodia, dust and dirt can be a serious problem during the dry season, when
rainy events will not occur for long time and dust from the dirt roads will be easily spread around. Another
recurrent problem in garment factories is the exhaust chimneys that are often placed on the roof. However, in
the garment factory presented in this study, no chimneys were present on the roof. Moreover, during the rainy
46
season daily precipitations can clean the panels from the dust making this loss neglectable. After the installation
of the PV system, a specific training for a designed person of the factory will be done as well. This person will
take care of the PV field, cleaning it and ensuring that everything is working properly. Because of all these
considerations and taking into account an analysis of reviews of reference articles presented in [36], the default
value proposed by PVsyst of 3% has been lowered down to 2%.
Unavailability losses
Unavailability losses is a tool used by PVsyst to foresee failures in the system or maintenance stops. It can be
defined as a fraction of time or a number of days of the year when the system will be unavailable. PVsyst set a
default value of 2% and gives the user the possibility to set a random value or specify a precise period. It is clear
how this value is completely random and not easily approximated. It is important to mention that in Cambodia
grid fails are extremely common and occur almost every day in rural areas, for variable period of time ranging
from few minutes to more serious blackouts of several hours. This faults in the grid highly influence the PV
energy that can be fed to the user (Egrid) and thus the PR of the system as explained in Equation 30. It is clear
that these events are not predictable and their influence on the system should be computed separately since it
is not a quality index of the system. For these reasons the unavailability losses have been neglected in this study.
Further investigation should be done in order to evaluate the impact that those events can have on the yearly
production of PV systems, especially for developing countries with unreliable grids.
Ohmic losses
The energy losses due to the ohmic resistance of the wires is proportional to the square of the current as
expressed in Equation 34:
Figure 23 - Graph of the increasing mismatch of the PV system simulated in PVsyst for the Jinko Solar panels model JKM 340PP-72-V. The green dot represents the losses simulation at the 10th year of the system
considering the increasing mismatch.
47
𝑃𝑙𝑜𝑠𝑠 = 𝑅𝑤 ∙ 𝐼𝑠𝑐2 (34)
Where Ploss is the power lost due to the ohmic resistance, Rw is the wiring ohmic resistance and Isc is the short
circuit current of the modules.
PVsyst takes those losses into account and has also the option to consider an external transformer if present.
When determining the global wiring resistance of the DC circuit it can be done by either setting the ohmic loss
ratio or by explicitly defining the loss in mOhm. A detailed computation of the global wiring resistance is also
available in PVsyst. The Rw value evaluation will highly depend on the sub-array structure, thus on the module
layout, and is therefore the last evaluation about the losses that have to be done. Only when all the system is
defined it can be evaluated properly.
PVsyst gives the possibility to choose between two different wiring layouts as shown in Figure 24 where blue
lines represent the string module connections, the pink lines the connections to the main box (or connection
box) and the green lines represent the wiring between the connection box and the inverter. The wiring
connection that better represents the garment factory case study is the parallel string schema. In the garment
factory analyzed, the strings will be connected to the MPPT inputs of the inverters directly and thus, without the
need of a junction box.
The ohmic loss ratio for an early evaluation of the wiring losses is defined in PVsyst as the ratio between Ploss
and the nominal power of the array Pnom,array = Rarray · Isc2. Where 𝑅𝑎𝑟𝑟𝑎𝑦 =
𝑉𝑚𝑝𝐼𝑚𝑝
⁄ at STC. Rw in Equation 34 is
the global wiring resistance and is obtained by placing the sub-array wiring resistances in parallel. The cross
section of the cables can be adjusted in order to match with an upper limit of wiring loss that is set to 1.5% by
default.
In the latest stage of the project, the length of the cables should be evaluated in order to compute the equivalent
wiring resistance that is dependent on the real wires length and section. The total wires resistance is computed
multiplying the wires length for the resistivity of the cables (different if copper or aluminum is chosen). The
Figure 24 - Parallel strings (left) and group of parallel strings (right) wiring connection schemes in PVsyst.
48
average length of each circuit, i.e. from minus to plus poles of each loop, has to be specified in the program as
well as the cables cross sectional area. The average length for each loop has been calculated taking into account
that the strings have to be connected to the inverters that will be located inside the electrical rooms of each
building (T1 and T2). To compute the minimum cross section of the cables Equation 35 has been used:
𝐴𝐷𝐶 =
𝐿𝐷𝐶 ∙ 𝐼𝐷𝐶 ∙ 𝜌
𝐿𝑜𝑠𝑠 ∙ 𝑉𝑚𝑝𝑝,𝑠𝑡𝑟𝑖𝑛𝑔
(35)
Where 𝐴𝐷𝐶 is the minimum cross-sectional area of the DC cables, 𝐿𝐷𝐶 is the total route length of the cables
(from minus to plus) from the first panel of the string to the inverter (or connection box), 𝐼𝐷𝐶 is the string current,
𝜌 is the resistivity of the wires, 𝐿𝑜𝑠𝑠 is the percentage of maximum voltage loss in the conductor and 𝑉𝑚𝑝𝑝,𝑠𝑡𝑟𝑖𝑛𝑔
is the maximum power point voltage of the string. In this case a cross section of less than 4mm2 has been
computed in any of the layouts examined and thus the standard cables section of 4mm2 have been chosen for
the DC part of the system.
PVsyst gives the possibility to specify also the AC wiring circuit, meaning from the inverter to the injection point.
If it is of significant length it should be taken into account. In this case, neither the external transformer losses
neither the AC circuit losses as been taken into account. The injection point is few meters far from the inverters
installation site and thus the losses have been considered neglectable while no data about the transformers
were available.
4.5. Module Layout
The module layout presented in this section is the layout of the optimal solution found after the simulations.
The layout of the system influences the modules disposition on the roof and thus has many consequences.
Repercussions can be found on the ohmic losses due to the different cables length because of the different
dispositions and dimension of the strings. Shading can have different impacts depending on the position of the
modules on the roof (portrait position or landscape) as well as their disposition in the available space with
respect to the shaded area. Because of that, PVsyst have a dedicated section where, after the definition of a 3D
scene is possible to simulate the modules layout. In the “near shading” section is possible to select the
calculation “Detailed, according to module layout” that takes into account the real position of the modules and
their connections. The shadings are computed for each hour and are projected on the field defined for the 3D
scene. In this study, all the shadings have been computed after the definition of the 3D scene and the complete
module layout definition. Modules layout and 3D scene are plotted in the final report by the program as can be
seen for the optimal solution reports in Appendix 6 and Appendix 7. Moreover, the economics of the project is
highly influenced by the module layout since the position of the installation with respect to the inverters
location, the number of mounting structure used as well as the wires length count on the budget. The choice of
the module layout has been guided both by the limits of the modules and inverters matching and by economic
considerations. The selected modules for the final layout, JKM 340PP-72-V from Jinko Solar, have one-meter
49
cable connectors on the back of the panel that allow the leapfrog connection only in portrait mode. This
connection method, presented in Figure 25, consent to drastically reduce the costs of DC cables, especially when
considering a big installation such as the one presented in this study. With leapfrog connection, the length of
the cables, and thus the related costs, has been estimated to be reduced as much as 26% for the 450 kWp section
of the garment factory connected to T1. For the garment factory in Phnom Penh, taking into account that near
shading problem is absent on the roofs selected for the installation, the modules position has been chosen to
be the portrait one. The final layout of the two systems for the garment factory have been summarized in Table
9. The PV system connected to T1 has 63 strings of 21 modules connected to 63 inputs of the total 88 available
on the inverters. The system connected to T2 has 36 strings of 21 panels connected to 36 inputs of the 48
available on the 6 inverters. In both the roofs layout, each group of two strings has been positioned at one-
meter distance to the one next to it on the horizontal and 0.5 m far from the next one on the vertical as
presented in Figure 26 and Figure 27. A safety distance of 1.6 m from the roofs margins have been considered
in the layout.
4.6. Simulation Results and Methodology
All the simulations done to find the best solution have been performed varying the model of panels used and
the model of inverters used. The related detailed losses dependent on the model’s characteristics as well as the
implications of different module layouts (such as average cable length) have been taken into account and varied
for each simulation. Since the system is divided into two parts, T1 and T2 connected to the respective
transformers, the optimal solution has been found for the best modules/inverters combination for T1 and then
applied to T2. The two systems present identic roofs orientation and tilt angles and thus an optimal components
combination for T1 has been assumed optimal also for T2. Both of the systems have been designed to stay below
the half transformer size limit imposed by the regulation. While T1 transformer was slightly oversized in
comparison to the load, T2 has been found largely oversized. Because of this, T1 has been sized at the maximum
Figure 25 - Different modules connection methods: conventional method (1a) and leapfrog wiring (1b) [44].
50
allowed by the regulation and thus to 450 kWp. The reason is that the transformer is 1000 kVA and, to comply
with the new regulation, a PV system cannot exceed half of the transformer size, with a correction factor of 0.9
from kVA to kW. T2 PV system instead, has been dimensioned and optimized in order to minimize the energy
exported to the grid. Its size is 1000kVA as T1, but the peak consumption has been found to be 168 kW only.
Therefore, multiple simulations have been performed to find a system with a good solar fraction but with energy
export below a reasonable percentage.
The most significant outputs of the different simulations for T1 have been summarized in Table 10 while the
simulations for T2 have been listed in Table 11. All the simulations for T1 present a nominal power of 450 kWp
in order to complain with the regulation. Only the system with Jinko Solar panels and SMA inverters presents a
peak power of 449 kW because of the string design of the system. The modules layout as well as the 3D scene
have been carefully set for every simulation in order to compute shading losses in the most precise way possible.
It has been found that no losses due to near shading were present in none of the simulations since no near
obstacles were on the field. Tilt angle of 8° and azimuth of -4° is common to all simulations both for T1 and T2
systems.
Figure 26 - Roof layout for T1 solution drawn in Sketchup.
Figure 27 - Roof layout of T2 drawn in SketchUp.
51
Table 9 - Summary of the module layout characteristic of the garment factory case study in Phnom Penh area for both installation regarding T1 and T2.
Module Layout T1 T2
Total numer of modules 1323 756 Number of modules per string 21 21
Number of strings 63 36 Horizontal space each two strings [m] 0.5 0.5
Vertical space each string [m] 1 1 Safety margin from the roof edge [m] 1.6 1.6
Optimal solution for T1
All the simulations reveal similar behavior and performance of the system as expected. Modules and inverters
do not have remarkably different specifications. The higher difference is in the inverter sizes: higher peak power
of 36 kW for the Huawei model while 25kW for the SMA inverters.
The optimal solution has been found to be with 1323 JKM 340PP-72-V modules from Jinko Solar combined to 11
SUN2000 36KTL inverters from Huawei. The modules are grouped in 63 strings of 21 modules each as described
in section 4.5. The PR has been used to compare the different solutions and to determine the best one. The
preferred solution presents a PR of 80.13% while the second-best solution presents a slightly lower PR of 79.92%
and is obtained with the same panels model from Jinko Solar coupled with SMA inverters. From the simulations
emerge that JKM 340PP-72-V modules from Jinko Solar and AS P72 320 polycrystalline modules from Antaris
perform better in high temperature situations with less losses due to temperature while CS6U-335P panels from
Canadian Solar maintain higher performances under lower irradiance level. Therefore, it appears clear that
Antaris and Jinko Solar panels achieve better PR results since Cambodia’s climate present high irradiation levels
and high temperatures during all year. Inverter losses has been found lower and constant at a value of 1.5% for
the Huawei inverters with no dependence on the panels used. SMA inverters have shown worse performances
with all the modules. This can be explained with the lower number of inputs and the limited voltage range. While
SUN2000 36KTL inverters from Huawei presents 4 MPPT with 2 inputs each and a VMPP range of 200-1000 V,
SunnyBoy Tripower 25000TL JP-30 inverters from SMA have 2 MPPT with 3 inputs each and a smaller VMPP range
of 390-800 V. With this lower voltage range, the layout flexibility diminishes, the strings must have lower
maximum number of panels and this will result in higher number of strings. Furthermore, higher number of
strings will also result in higher average cables length with higher related losses and costs.
52
Table 10 - Summary of the outputs of the PVsyst simulations. P_nom is the nominal power of the PV system under STC, E_load is the energy needs of the user, E_user is the solar energy supplied to the user, E_grid is the excess energy exported to the grid, E_total is the total energy produced with the PV system, Grid Export Percentage represents the percentage of solar energy, in respect to the total solar energy produced, that is lost because exported. All the losses are computed as explained in section 3.4.4.
T1 Jinko & Huawei
Jinko & SMA
Canadian & Huawei
Canadian & SMA
Antaris & Huawei
Antaris & SMA
P_nom (STC) [kWp] 450 449 450 450 450 450
E_load [MWh/y] 1419.00 1419.00 1419.00 1419.00 1419.00 1419.00
E_user [MWh/y] 596.23 593.86 590.90 589.76 594.56 594.11
E_grid [MWh/y] 108.15 107.09 106.10 105.58 107.61 107.35
E_total [MWh/y] 704.4 700.9 697.0 695.3 702.2 701.5
Grid Export Percentage 15.35% 15.28% 15.22% 15.18% 15.32% 15.30%
Specific Production [kWh/kWp/y]
1566 1562 1549 1546 1561 1559
Solar Fraction 42.02% 41.85% 41.64% 41.56% 41.90% 41.87%
Performance Ratio 80.13% 79.92% 79.28% 79.09% 79.86% 79.78%
IAM on global -3.6% -3.7% -3.6% -3.6% -3.7% -3.7%
PV losses due to irradiance level
-0.6% -0.6% -0.5% -0.5% -0.5% -0.5%
Temperature losses -12.0% -12.0% -12.2% -12.2% -12.0% -12.0%
Shading (detailed module calc.)
0.0% 0.0% 0.0% 0.0% 0.0% 0.0%
Mismatch losses -1.1% -1.1% -1.1% -1.1% -1.1% -1.1%
Ohmic losses -0.6% -0.6% -0.8% -0.8% -0.7% -0.7%
Inverter losses -1.5% -1.7% -1.5% -1.7% -1.5% -1.6%
Optimal solution for T2
Five simulations for T2 were done initially: 200 kWp, 228 kWp, 250 kWp, 278 kWp and 300 kWp systems. All the
simulations were done following the optimal results of T1 simulations and thus with JKM 340PP-72-V modules
from Jinko Solar and SUN2000 36KTL inverters from Huawei. After these simulations, the grid exports have been
compared to the total production from solar. Because of the net zero export imposed by the regulation on solar
installation, all the energy generated in excess should be prevented and would represents a loss for the system
both in energetic and economic sense. Being the garment factory working only on 6 days per week, all the energy
produced by the solar system during the Sundays and the holiday days will be wasted. T2 section of the garment
factory has an average weekly load of 9521.58 kWh, with an average load of 1563.74 kWh for the working days
and 208.87 kWh of average load during Sundays. An ideal system covering 100% of the load during the working
days would lead to an over production during the Sundays of around 1355 kWh, representing a percentage of
overproduction of more than 14%. If lunch brakes are considered, that make the consumption of the factory to
decrease at least of 45 kWp during the highest production hour of the solar system, this value can grow up to
more than 17%. Considering the holidays during the year it can be even higher. Because of this, the limit of 20%
of grid export has been considered acceptable and has been imposed has limit for the optimization of the
53
system. Other simulations have been thus performed to maximize the solar fraction and to stay below 20% of
grid exports. It has been found that maintaining the design of the 250 kWp option and adding a string of 21
modules would have led to an overproduction of 19.6%. This system has thus been considered the optimal. The
summary of the simulations results has been plotted in Table 11.
The optimal solution presents 756 Jinko Solar modules organized in 36 strings of 21 modules each and 6 Huawei
inverters. The number of inverters has not been increased from the 250 kWp system. This would correspond in
a slightly higher nominal power ratio of the inverter (1.16 for 250 kWp system and 1.19 for 257 kWp system)
but not in an increase of the losses. Moreover, the limit of 𝑃𝐼𝑁𝑉,𝐷𝐶 < 1.2 ∙ 𝑃𝑃𝑉 already expressed in section 4.3
has been respected. The selected option presents also the highest PR of 80.31% together with the 250 kWp
solution. A high solar fraction of 60.01% is reached with this installation and is the highest among the other
option below the export limit.
The reports of the final solutions can be found in Appendix 6 for T1 and in Appendix 7 for T2.
Table 11 - Summary of T2 simulations in PVsyst.
T2 simulations 1 2 3 4 (opt.) 5 6
P_nom (STC) [kWp] 200 228 250 257 278 300
E_load [MWh] 540.55 540.55 540.55 540.55 540.55 540.55
E_user [MWh] 271.89 300.42 318.89 324.39 339.24 352.18
E_grid [MWh] 41.798 58.029 73.316 79.010 97.71 118.17
E_total [MWh] 313.7 358.4 392.2 403.4 437.0 470.3
Grid Export Percentage 13.32% 16.19% 18.69% 19.59% 22.36% 25.13%
Specific Production [kWh/kWp/y] 1569 1569 1569 1569 1569 1568
Solar Fraction 50.3% 55.58% 58.99% 60.01% 62.76% 65.15%
Performance Ratio 80.29% 80.28% 80.31% 80.31% 80.30% 80.26%
54
Chapter Five – Economics and Environment
In order to determine whether the project is feasible both from the economic and environmental point of view,
an economic assessment on the project and a carbon balance analysis have been performed. In the economic
assessment of the project, the prices of the balance of system (BoS), meaning all components of a PV system,
have been computed together with the installation and engineer costs. Moreover, an analysis of the impacts of
the new solar regulation has been done to show the consequences of such measure on the economics of a
project in Cambodia. Furthermore, an environmental analysis has been performed with the help of PVsyst
software to address the CO2 emission reduction potential of a big PV installation in Cambodia.
5.1. Economic Assessment of the Project
Table 12 shows the approximate investment cost for the PV installation in the garment factory case study. The
prices of modules, inverters, inverter manager and mounting structures have been asked directly to the
manufacturers or to the local dealers. The shipping cost is referred to those items that are shipped from different
companies and locations in China. It has been computed as a percentage of the items costs (10%) and thus is
not precise. It is indeed important since it gives an idea of the high impact that has on the economics of the
project. The DC cables price is from a local producer and thus has no shipping costs. The price of installation and
engineering is an average price from preview installations of Kamworks and is referred to systems installations
of more than 100 kWp, it is not a standard and can vary according to the projects. It is important to notice that
the price of the solar panels vary widely and often, due to the fluctuation of the silicon price in China. Therefore,
the price shown was valid at the time of the quotation but can vary abundantly becoming even 5% to 10% higher.
Inverter and mounting structure have a more stable price but also those ones can vary from time to time, also
according to the numbers of items ordered. The maintenance cost is referred to the standard price of Kamworks
for a big installation. The salinity of the Gulf of Thailand, more than 150 Km far from the installation site, would
not influence the maintenance of the system. It can be noticed that the modules cost accounts for almost 50%
on the total cost including VAT and 56% of the total cost of the BoS, representing the most expensive part of the
project. Inverters represent the 15% on the total price while the DC cables the 9% of the total and 10% of the
entire BoS. It is important to notice that for big system this can represent an important part of the expenses,
especially when is not possible to mount the inverters directly on the roof or closely to the strings. A correct
disposition of the modules can make the difference for this part of the expenses. Without the leapfrog
connection method used in this project for instance, the DC cables length (and thus cost) could have been up to
26% higher. Decrease the DC cables length decreases the ohmic losses and consequently the energy waste and
the cost of the system. The installation of the inverters on the roof should be evaluated when the site is suitable
for the installation. In the garment factory the inverters installation on the roof was not possible and has thus
not been taken into account in this study.
The final investment cost for this installation can be considered as a good price for Cambodia. Installations prices
for big system normally range between 800 USD/kWp and 900 USD/kWp. Being this system cost equal to 825
USD/kWp means the design and the components have been carefully selected.
55
Table 12 - Investment cost and maintenance cost for the installation at the garment factory (707 kWp).
Jinko & Huawei Unit Tot Unit
Module $ 390.00 USD/kWp $ 275,730.00 USD
Inverter $ 135.69 USD/kWp $ 83,045.00 USD
Inverter manager $ 1,200.00 USD/piece $ 2,400.00 USD
Mounting structure (rails) $ 9.20 USD/m $ 38,253.60 USD
DC cables $ 3.65 USD/m $ 49,951.12 USD
Shipping $ 56.50 USD/kWp $ 39,942.86 USD
Total BoS $ 692.11 USD/kWp $ 489,322.58 USD
Installation and Engineering $ 58.00 USD/kWp $ 41,006.00 USD
Total incl. VAT (10%) $ 825.12 USD/kWp $ 583,361.44 USD
Maintenance $ 1,000.00 USD/y
Profitability of the project
The profitability of the PV installation of the garment factory case study have been performed and is shown in
this section. As already presented in section 2.4.1, Cambodia has approved a new regulation about solar that
highly influence the profitability of PV grid-connected projects. It is not completely clear and defined and
because of that, three different scenarios have been compared.
The first scenario analyzes the profitability of the project if the regulation would not have been approved. It thus
represents the situation of PV projects before January 2018. The second scenario examines the project’s
profitability with the new regulation. It is of common habit in Cambodia to oversize the transformer for MV and
HV consumer and the consequent contract type. Before the new regulation was approved this measure had no
major impacts except for a higher initial installation cost. After the approval of the new regulation, and only in
case of solar installations, with the contracted load charge split from the consumption charge, the impact of the
transformers oversizing is significant. The third scenario has thus been computed taking into account a possible
modification in the electricity contract, considering possible a re-sizing of the transformers. In the case study of
the garment factory where two different transformers were present, only the second transformer has been
considered to be changed. Indeed, T2 is much more oversized with respect to the peak load in comparison to
T1. Moreover, solar installation cannot exceed half of the contracted power. This restriction has been taken into
account and thus T2 have been considered double the optimal size of the system simulated in Chapter Four. In
Table 13 the electrical characteristics of the garment factory site and the contracted load charge in case of solar
installation have been summarized.
In order to evaluate the profitability of the project for the three scenarios, the yearly savings, the Pay Back Time
(PBT), the Net Present Value (NPV) and the Return on Investment (ROI) have been computed for a lifetime of
the project of 25 years. The LCOE of the electricity produced has also been computed to compare it with the
grid electricity cost and with the PPAs rate. Table 14 summarizes the results obtained and gives an overview of
the garment factory characteristics.
56
Savings
The yearly savings have been assumed equal to the differences between the grid consumption charge before
the solar installation and the grid consumption charge after the solar installation, taking into account the change
in the tariff. As can be seen from Table 14, the savings thanks to the solar PV installation would have been more
than 150,000 USD/y with the preview regulation while with the new regulation there will be no savings but more
expenses. For the third scenario, in case of T2 resizing, the annual savings would be diminished to less than
10,000 dollars. Considering an average monthly price of the electricity bills of the garment factory of more than
20,000 USD, it appears clear the inconsistency of the savings compared to the investment.
Table 13 - Characteristics of the garment factory. The EDC consumption charge and contracted load charge are referred to the connection from a MV (22kV) sub-transmission or distribution network of the national grid as referred in the solar regulation.
Characteristics of the garment factory
Transformer 1 (T1) size [kVA] 1,000.00
Transformer 2 (T2) size [kVA] 1,000.00
Transformer 2 new possible size [kVA] 572.00
EDC current tariff [USD] 0.165
EDC consumption charge after solar installation [USD] 0.12
EDC contracted load charge [USD/kW/month] 10.00
Simple Pay Back Time
The simple PBT has been computed for the three scenarios. The investment, assumed to occur entirely at the
beginning of the project, has been divided by the amount of net cash inflow generate by the project per year,
that in this case are the yearly savings minus the maintenance O&M cost for each year. It can be noticed from
Table 14 that the PBT increased from less than 4 years for the old regulation to more than 65 years for the new
best scenario. In case of no resizing of the system, the investment would never be recovered. The simple payback
method is inaccurate since it does not take into account replacement, inflation and discount rate. However, it
gives an idea of the payback time, which would never occur for the new regulation scenario and which is above
the assumed lifetime of the modules for the new scenario with re-sizing.
Net Present Value
The NPV of the project has been computed as well to take into account also the maintenance cost and the
discount rate. To compute the NPV, the simplified model has been used:
𝑁𝑃𝑉 = (𝑅 − 𝑐𝑜𝑚)𝑘𝑎 − 𝐼𝑇 (36)
Where R are the revenues that in this case are the yearly savings, 𝑐𝑜𝑚 are the operation and maintenance costs
in respect to the total investment, 𝐼𝑇 is the total investment cost of the project and
57
𝑘𝑎 =
(1 + 𝑎)𝑛 − 1
𝑎(1 + 𝑎)𝑛 (37)
Where a is the discount rate and n is the lifetime of the project in years.
The simplified model for the computation of the NPV assumes a constant annual utilization factor of the system,
constant operation and maintenance costs and that the total investment cost is concentrated in year 0. These
assumptions can be considered adequate for the case study except for the constant operation and maintenance
costs. While the maintenance contract has a constant price during the years, the eventual replacement of the
items is not considered. For PV installations, while panels are not considered as a replacement issue because of
the long performance warranty of 25 years that most of producers offer, inverters are the weak point of the
system. Huawei inverters have 5 years warranty, but they can last for 10 years or more if operating properly.
Since the exact lifetime of an inverter is not predictable, an average lifetime of 7 years has been considered in a
conservative perspective. The NPV formula have been thus modified to take into account the replacements.
Equation 36 can be rewritten as:
𝑁𝑃𝑉 = (𝑅 − 𝑐𝑜𝑚)𝑘𝑎 − 𝐼𝑇 −
𝐶𝑖𝑛𝑣
(1 + 𝑎)𝑛1−
𝐶𝑖𝑛𝑣
(1 + 𝑎)𝑛2−
𝐶𝑖𝑛𝑣
(1 + 𝑎)𝑛3 (38)
Where 𝐶𝑖𝑛𝑣 is the replacement cost of the inverters and 𝑛1, 𝑛2 and 𝑛3 are the years at which the replacement
will take place, in this case 7, 14 and 21 year respectively.
The NPV index represent the feasibility or not of a project. When the NPV is positive, the investment is
recovered, the minimum rate of return of capital is achieved, a surplus is obtained and thus the project is
feasible. When the NPV is negative, the project is not profitable since the minimum rate of return is not achieved.
The higher the discount rate, the lower the NPV. In this case the discount rate has been fixed at 7% in a
conservative perspective, while in Cambodia is usually higher in case of a PPA since the investors want to invest
into more profitable projects. The project lifetime has been fixed at 25 years considering the panels warranty.
As can be noticed from the NPV results presented in Table 14, the NPV is positive only for the scenario of the
old regulation while is deeply negative for both new regulation scenarios.
Return on Investment
Return on Investment (ROI) is a performance measure to evaluate the efficiency of an investment. ROI measures
the amount of return on an investment, relative to the investment’s cost. It is a useful index to compare
investment. To calculate ROI, the benefit of an investment, in this case the cumulative savings during the lifespan
of the project, is divided by the cost of the investment, in this case the initial investment. The result can be
expressed as a percentage or a ratio.
The formula that express the ROI can be written as:
𝑅𝑂𝐼 =
𝐵𝑒𝑛𝑒𝑓𝑖𝑡
𝐼𝑛𝑣𝑒𝑠𝑡𝑚𝑒𝑛𝑡 𝑐𝑜𝑠𝑡
(39)
58
The benefit is equal to the cumulative net discounted cash-flow of the project, computed as the sum of the net
cash-flows of each year (yearly discounted savings minus the O&M, comprehensive of replacement cost when
they occur) and the initial investment (with negative value).
As can be seen in Table 14, the old regulation scenario presents a ROI of 1.85. The new regulation scenarios with
and without re-sizing of the transformers present a ROI per year of -0.98 and -2.01 respectively. The results show
how the old regulation was profitable for similar investment while with new regulation solar projects present
losses of money instead of gains. The cumulative net cash-flows of each scenario has been plotted in Figure 28
in order to visualize the different situations.
Levelized Cost of Energy
The Levelized Cost of Energy (LCOE), or discounted average cost, has been computed for the garment factory
installation in order to compare the cost of the electricity produced with solar and the price of the electricity
from the national grid.
The LCOE can be expressed as:
𝐿𝐶𝑂𝐸 =
𝐿𝑖𝑓𝑒 𝐶𝑦𝑐𝑙𝑒 𝐶𝑜𝑠𝑡
𝐿𝑖𝑓𝑒𝑡𝑖𝑚𝑒 𝐸𝑛𝑒𝑟𝑔𝑦 𝑃𝑟𝑜𝑑𝑢𝑐𝑡𝑖𝑜𝑛 (40)
The life cycle cost represents the discounted investment cost plus the discounted O&M cost while the lifetime
energy production is the energy produced by the solar system during the lifespan of the project.
-1500000
-1000000
-500000
0
500000
1000000
1500000
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25
Cumulative Net Cash-Flow
Figure 28 - Cumulative net cash flow of the old regulation scenario (green), of the new regulation scenario (red) and of the new regulation with T2 re-sizing (yellow)
59
Assuming the investment cost fully concentrated at the initial time and taking into account the inverters
replacement costs, Equation 40 can be written as:
𝐿𝐶𝑂𝐸 =
𝐼𝑇 + 𝐼𝑇 ∑𝑐𝑜𝑚𝑗
(1 + 𝑎)𝑗 + ∑𝐶𝑖𝑛𝑣𝑗
(1 + 𝑎)𝑗𝑛𝑗=1
𝑛𝑗=1
𝐸𝑑
(41)
Where 𝐼𝑇 is the total investment cost, n is the number of years of lifetime, j is the year in which the expenses
occurred, 𝑐𝑜𝑚𝑗 is the O&M cost incurred in the year j as percentage of 𝐼𝑇 , 𝐶𝑖𝑛𝑣𝑗 is the inverters replacement
costs incurred in the year j in USD and 𝐸𝑑 is the present value of the energy that can be expressed as:
𝐸𝑑 = ∑
𝐸𝑎𝑗
(1 + 𝑎)𝑗
𝑛
𝑗=1
(42)
Where 𝐸𝑎𝑗 is the annual energy produced by the system in year j.
For the computation of the LCOE the same discount rate of the NPV of 7% have been assumed. An average
system degradation of 0.4% per year has been considered and thus the energy produced has not been
considered constant during the lifetime of the project. Inverters replacement have been considered to occur in
year 7, 14 and 21. The results are reported in Table 14 and are off course equal for all the scenarios. The LCOE
present a value of 67.82 USD/MWh corresponding to 6.78 cent/kWh. This price is 60% lower with respect to the
contracted price of the electricity for the garment factory presented in this study, and even one third of most of
other contracted price of electricity for MV consumers. However, with the new regulation this price has to
compete with 0.12 USD/kWh, or even less than 0.10 USD/kWh for certain contracts as shown in Table 3 in
Section 2.4.1. The LCOE would thus become only 30-40% lower than the grid price becoming less competitive
and making solar PV installation no more advantageous both for customers and investors.
60
Table 14 - Scenarios comparison summary. All the data about PV production have been taken from the simulations of the optimal systems in PVsyst.
Old regulation New regulation New regulation
(T2 smaller)
Total PV System Size T1+T2 [kWp] 707.00 707.00 707.00
Solar Production
T1 Solar Production [MWh/y] 704.00 704.00 704.00
T2 Solar Production [MWh/y] 403.00 403.00 403.00
Total Solar Production [MWh/y] 1,107.00 1,107.00 1,107.00
Grid Consumption
T1 Energy Consumption per year [MWh], simul.
1419 1419 1419
T2 Energy Consumption per year [MWh], simul.
541 541 541
Tot Energy Consumption per year [MWh], simul.
1960 1960 1960
Grid Sales
T1 Grid Export [MWh/y] 108.15 108.15 108.15
T2 Grid Export [MWh/y] 79.01 79.01 79.01
Tot Grid Export [MWh/y] 187.16 187.16 187.16
% Grid Export compared to total PV generation
16.91% 16.91% 16.91%
Percentage of PV Usage [%], T1+T2 47% 47% 47%
Economics
System Cost [USD] 583,361.44 583,361.44 583,361.44
Electricity Cost before PV installation [USD/y] 323,400.00 323,400.00 323,400.00
Electricity Cost after PV installation [USD/y] 171,626.40 364,819.20 313,459.20
Yearly Savings [USD/y] 151,773.60 -41,419.20 9,940.80
Simple PBT [years] 3.87 - 65.25
NPV [USD] 36,429,072.14 -19,855,636.97 -4,892,436.17
ROI 1.85 -2.01 -0.98
LCOE [USD/MWh] 67.82 67.82 67.82
5.2. Environmental Analysis
In order to address the possible CO2 emissions reduction of the project an analysis has been done with the help
of PVsyst software. The carbon balance tool integrated in PVsyst permits an analysis of the expected savings in
CO2 for the PV installation taking into account the degradation of the system.
The basis of this calculation are the so called Life Cycle Emissions (LCE). These values include the total life cycle
of a component or energy amount, including production, operation, maintenance, disposal etc. This calculation
61
helps to understand the impacts of a solar PV installation. The reason is that the energy produced with the solar
system, will replace the same amount of electricity that would have been drawn from the grid. While the PV
installation will not lead to carbon emissions during operation, it is important to notice that is highly impactive
during the production and disposal phases.
The total carbon balance depends on many factors:
• The system production, or energy yield
• The system lifetime
• The LCE of the electricity produced by the grid
• The LCE of the electricity produced by the PV system
The system production comes from the PVsyst simulation while the system lifetime has been set to 25 years as
for the economic assessment.
The most difficult calculations are the ones regarding the LCE both for the grid and for the PV system. LCE
method is highly sensitive to the region and specific item analyzed and thus cannot be generalized without
implicit errors. The technology of the power plants used to produce the energy of the grid, or the PV panels
materials and technology production are determinant in order to establish correctly the CO2 emissions. LCE takes
long time and a deep study about all the life of a product or unit of energy from the materials extraction to the
disposal, or the use of the energy. PVsyst has some default values for what concern both the grid LCE and the
PV system LCE depending on the country in which the installation take place. The default values for the grid LCE
are from a 2012 IEA study that presents the CO2 emission in 2010 per kWh of electricity production for different
countries. For Cambodia this value amounts to 813 gCO2/kWh. The default values for the PV system LCE is based
on the 50th percentile LCE value for photovoltaic electricity production published by the International Panel on
Climate Change (IPCC). This value amounts to 46gCO2/kWh and it represents a very coarse global average.
It is also possible in PVsyst to compute manually both the LCE taking into account the energy mix of the local
grid, for the grid LCE, and all the items of the installation with their origin.
In order to have a more averaged vision of the carbon balance, two different simulation have been performed
in PV syst. The first amount of carbon emissions has been computed with all the default value of PVsyst and a
degradation of the PV system of 0.4 %/y (as in the simulations). The second computation takes into account the
energy mix of the Cambodian electricity sector. Since the study from where the data of IEA are taken from is
from 2010, and the grid production mix of Cambodia has changed largely in the last years, the standard value
has been substituted taking into account the energy mix of 2017 from the study of EAC [11]. The import has
been assumed all from Vietnam because of lack of information from the grid of Thailand. Since Vietnam
represents the majority of the imports, this assumption can be considerate reasonable. Table 15 shows the
energy generation mix and energy imports for Cambodia for the year 2017. The LCE conversion factors have
been taken from the 50th percentile of the IPCC. Those value are 1001 gCO2/kWh for coal power plants, 4
gCO2/kWh for hydropower, 469 gCO2/kWh for natural gas plants, 840 gCO2/kWh for oil and 46 gCO2/kWh for
renewables (it has been used the value of solar PV generation since is the predominant renewable technology
in Cambodia).
62
Table 15 - Energy generation mix and energy import mix from 2017 [11].
Electricity Generation
[GWh] Vietnam Imports
[GWh] Tot [GWh] %
Coal 2829.12 547.57 3376.69 42.44%
Oil 381.14 - 381.14 4.79%
Hydro (also pumped) 3217.79 547.57 3765.36 47.32%
Renewable Energies 48.61 - 48.61 0.61%
Gas 0 384.78 384.78 4.84%
Total 6486.25 1479.92 7956.58
For what concerns the PV system installation LCE, the standard value from IPCC of 46 gCO2/kWh has been
changed according to the project specification. The items shipped has been considered all shipped from Hong
Kong (China) to Sihanoukville (Cambodia) by boat, and then to Phnom Penh, where the the garment factory is
located, by truck. This assumption is considered a good assumption since Hong Kong is the main departure
harbor from china and Sihanoukville is the main harbor in Phnom Penh. Replacement of the inverters has been
considered while for the end of life, recycling has been assumed to compute the emission per kWh. The value
of 29 kgCO2/kWh of panel recycled has been taken considering the study [37]. This value is uncertain and not
referred to Cambodia. Nevertheless, literature on this topic is quite scares and values vary widely for locations,
types of panels and methodology of recycling. This value has been used in a perspective of development in this
area. Since the panels would not have to be recycled before 20 years, it has been assumed that technology
improvement would have reached Cambodia as well. All the values regarding the detailed system LCE that have
been inserted into PVsyst can be found in Appendix 8.
To compute the carbon balance of the installation, Equation 43 has been used for each year, taking into account
also the degradation of the system:
𝑇𝐶𝑂2 = (
𝐸𝑔𝑟𝑖𝑑 ∙ 𝑛 ∙ 𝐿𝐶𝐸𝑔𝑟𝑖𝑑
1000) − 𝐿𝐶𝐸𝑃𝑉 (43)
Where 𝑇𝐶𝑂2 is the total amount of CO2 avoided thanks to solar PV installation in tons, 𝐸𝑔𝑟𝑖𝑑 is the annual amount
of energy produced with solar in MWh, 𝑛 is the lifetime of the project in years, 𝐿𝐶𝐸𝑔𝑟𝑖𝑑 is the LCE of the national
grid of Cambodia in gCO2/kWh and 𝐿𝐶𝐸𝑃𝑉 is the total amount of CO2 emissions caused by the PV installation in
the construction and transportation phases in tons. An average degradation of 0.4 %/y have been taken into
account as well by the program. Table 16 shows the different results obtained.
It is important to notice that grid LCE from EIA 2012 overestimate the grid emissions for each kWh, while it
underestimates the PV installation impact. It looks reasonable since national grid of Cambodia has changed
largely since 2010 when the data are referred to. The underestimation of the PV installation can be addressed
to the high impact of the shipping of the items as well as the length of the cables that is higher for the garment
factory installation than the ones considered in the study by IPCC where the value is taken by PVsyst.
Nevertheless, the total amount of CO2 saved results much lower for the detailed calculations.
63
Dividing the total amount of CO2 for the lifetime of the project, a result of 452 tCO2/y is achieved for the detailed
calculation case. Considering an average modern car emission of 120 gCO2/km [38] and 12000 km/y of travelled
distance [39] the impact would be the same of 313 cars taken out of the road each year for a total of 7825 cars
taken out of the road for the 25 years of the project lifetime.
The emissions reduction does not look much if compared with the total CO2 emissions of the country of more
than 6.5 million tons [40], representing less than the 0.1%, but with a larger deployment on the industry sector
in Cambodia can represents a good way to reduce toxic emissions.
Table 16 - Carbon balance results for the whole system of the garment factory both with PVsyst default values and with values insert manually taking into account the energy mix of Cambodia for 2017 and the specific PV installation at the garment factory.
E_grid [MWh]
Lifetime [years] LCE grid [gCO2/kWh] LCE PV [tCO2]
Carbon Balance [tCO2]
T1 default 704.4 25 813 972 12677.958
T2 default 403.4 25 813 464 7171.077
T1+T2 default 1107.8 25 813 1436 19849.035
T1 manual 704.4 25 488 1031.9 7164.945
T2manual 403.4 25 488 562.8 4131.604
T1+T2 detailed 1107.8 25 488 1583.2 11296.549
64
Further Work
This thesis should be considered as a feasibility assessment for a project as well as a complete design of the
optimal system. This work presented a focus on both technical aspects of the design and the analysis of the
regulation in place at the present time in Cambodia. The project analyzed and described in the thesis has not
been implemented. Further work such as single line diagram and a deeper evaluation on the electrical aspects
of the garment factory should be evaluated before the installation. Furthermore, it would be interesting to
investigate the system performances after the installation in order to match the simulation and the real data to
see the deviation.
Important aspects such as unavailability of the system due to grid fault should be better examined since it has
high impacts on the yield of grid-connected system in Cambodia. The connection with a generator to maintain
the production of solar energy should be evaluated addressing benefits and drawbacks.
With regards to the economics, the new solar regulation approved in Cambodia should be better investigate.
The regulation is not clear in various aspects and could change due to economic and private sector pressure. In
that case new economic possibilities would be opened that should be taken into account for further installations.
Moreover, a specific grid LCE specific for Cambodia can help to better determine the possible emissions
reduction.
65
Conclusion
The analysis carried out for the garment factory with PVsyst has demonstrated the high potential that solar
energy has in Cambodia. With the simulations have been found that high yield of more than 1550 kWh/kWp/y
can be achieved. The system with the best PR has been found to be the one with JKM 340PP-72-V polycrystalline
module from Jinko Solar and SUN2000 36KTL inverters from Huawei. The two different systems for the factory
have been sized at 450 kWp for T1 and 257 kWp for T2. The first system has a solar fraction of 42.02% while T2
reaches a solar fraction of 60.01%. The total system size would be 707 kWp with a yearly total energy production
of 1107.8 MWh and a solar fraction of 47%. The size of the systems has been limited by the regulation constrains
and thus its solar fraction. Thanks to the PVsyst analysis, all the technical aspects of the system have been
analyzed. The main source of losses of the system has been found to be the heat. Thermal losses amount to 12%
of the total ideal production of the system and thus the modules selection should point at the ones with better
high temperature performances. The correct module layout has been found of primary importance due to its
influence on both losses and economics of the project.
The uncertainties about the new solar regulation have slowed down the solar market since the beginning of
2018. It appears clear how the regulation highly affects solar installations in the country, both from the technical
and economical point of view. With a low price for the electricity consumption and an elevated price for the
connection charge, the grid-connected PV projects would not be profitable anymore. Even in the scenario with
a re-sizing of the transformers to diminish the fixed tariff, the PV installation would not be capable of recovering
the capital cost necessary for the installation. The size limit of half the contracted capacity, together with the no
export to the grid limit, restrict the possible designs of the system, the amount of energy produced with solar
and thus the yearly savings. In case of PPA, the investors that own the system, have to compete with a
consumption charge that is less than half the price of the preview regulation one. The new consumption tariff is
lower than the usual charge of PPA contracts in Cambodia. With an LCOE of almost 7 cent/kWh the margin is
minimum if a PPA tariff of 8-9 cent/kWh is applied and would not be capable of recovering the costs in the
lifetime of the project. Moreover, the customer will not save money, on the contrary he would lose money due
to the high fixed tariff.
It is not clear why solar power has been targeted in the country, but without a revision of the reform about solar,
the private investments would not be directed anymore towards renewable sector, solar in particular, and the
CO2 emissions will be hardly lowered in the country. Moreover, the economic sector that is growing rapidly in
the country, will slow down due to the high costs and poor quality of the electricity from the grid.
66
References
[1] Yale University, “Environmental Performance Index,” 2018.
[2] IPCC, Intergovernmental Panel on Climate Change, Climate Change 2014 - Mitigation on Climate Change,
2014.
[3] The World Bank Group, ESMAP, “CAMBODIA Beyond Connections: Energy Access Diagnostic Report
Based on Multi-Tier Framework,” 2018.
[4] M. Sarraf, B. Rismanchi, R. Saidur, H. W. Ping, and N. A. Rahim, “Renewable energy policies for
sustainable development in Cambodia,” in Renewable and Sustainable Energy Reviews, vol. 22, 2013,
pp. 223-229.
[5] The World Bank, “Doing Business Report,” 2018.
[6] SolarGIS, The World Bank, ESMAP, “Solar Resource Map,” [Online]. Available:
https://solargis.com/maps-and-gis-data/download/cambodia.
[7] Electricity Authority of Cambodia, Annual Report on Power Sector of the Kingdom of Cambodia, 2017.
[8] E. C. A. William Derbyshire, Cambodia – In Depth Study on Electricity Cost and Supplies, 2015.
[9] The World Bank, “Access to Electricity - Cambodia,” [Online]. Available:
https://data.worldbank.org/indicator/EG.ELC.ACCS.ZS?end=2016&locations=KH&start=1990&view=ch
art.
[10] The world Bank, “World Development Indicators: Sustainable energy for all,” [Online]. Available:
http://wdi.worldbank.org/table/3.13.
[11] Electricity Authority of Cambodia, “Salient Features of Power Development in Kingdom of Cambodia for
year 2017,” 2018.
[12] International Energy Agency, Southeast Asia Energy Outlook 2017 - World Energy Outlook Special
Report, 2017.
[13] General Department of Energy and General Department of Petroleum, Ministry of Mines and Energy of
Cambodia, Cambodia Energy Statistics 2016, 2016.
[14] Ministry of Economy and Finance, National Council for Sustainable Development, Addressing Climate
Change Impacts on Economic Growth in Cambodia, 2018.
[15] EDC, Report on Transferring the Benefits Resulting from the Development of Electricity to the Population
in Rural Area for the Year 2016, 2017.
[16] Electricity Authority of Cambodia, “On General Conditions For Connecting Solar Generation Sources To
The Electricity Supply System Of National Grid Or To Electrical System Of A Consumer Connected To The
Electricity Supply System Of National Grid,” 2018.
67
[17] Electricity Authority of Cambodia, “Announcement by the Electricity Authority of Cambodia (EAC) On
Publicdiscussion on Tariff Plans for Bulk and Large Purchase Consumers Installing Solar Power to the
Distribution Network of the National Grid,” 2018.
[18] Global Withness, “HOSTILE TAKEOVER - The corporate empire of Cambodia's ruling family,” 2016.
[19] Central Intelligence Agency, “The World Factbook - Cambodia - Economy,” 2018. [Online]. Available:
https://www.cia.gov/library/publications/the-world-factbook/geos/cb.html. [Accessed 5 September
2018].
[20] Royal Government of Cambodia, “National Strategic Development Plan 2014-2018,” 2014.
[21] Asian Development Bank, Asian Development Outlook 2018: How Technology Affects Jobs, 2018.
[22] Bike Europe, “Bike Europe - Cambodia Now EU’s Leading Bicycle Supplier,” 2017. [Online]. Available:
https://www.bike-eu.com/sales-trends/nieuws/2018/06/cambodia-now-eus-leading-bicycle-supplier-
10134013?vakmedianet-approve-cookies=1&_ga=2.186818925.432973189.1530239937-
885009858.1530239937?vakmedianet-approve-cookies=1&_ga=2.186818925.432973189.15302399.
[Accessed 5 September 2018].
[23] The World Bank, “Cambodia Economic Update - Cambodia Climbing Up the Manufacturing Value Chain,”
2017.
[24] European Commission, “Cambodia: EU mission assesses human rights and labour situation,” [Online].
Available: http://europa.eu/rapid/press-release_IP-18-4467_en.htm. [Accessed 5 September 2018].
[25] The German Energy Society, Planning and Installing Photovoltaic system - A guide for installers, architets
and engineers - Second Edition, 2008.
[26] S. H. Antonio Luque, Handbook of Photovoltaic Science and Engineering, 2003.
[27] S. H. Nordahl, “Design of Roof PV Installation in Oslo,” 2012.
[28] T. Markvart, Solar Electricity, John Wiley and sons, ltd., 2000.
[29] W. A. B. Jhon A. Duffie, Solar Engineering of Thermal Processes, Fourth Edition, 2013.
[30] P. Ineichen, Comparison and validation of three global-to-beam irradiance models against ground
measurements, 2008.
[31] PVsyst, “PVsyst contextual help,” 2018. [Online]. Available: www.pvsyst.com/help.
[32] “Wheatherbase - Phnom Penh, Cambodia,” [Online]. Available:
https://www.weatherbase.com/weather/weather-
summary.php3?s=19984&cityname=Phnom+Penh,+Cambodia. [Accessed 20 September 2018].
[33] “The Phnom Penh Post - Scorching hot season sets new all-time highs,” [Online]. Available:
https://www.phnompenhpost.com/national/scorching-hot-season-sets-new-all-time-highs. [Accessed
2018 September 20].
[34] PVsyst, “PVsyst help - Phisical model used,” [Online]. Available: www.pvsyst.com/help.
[35] PVsyst, “PVsysyt help - Inverter sizing,” [Online]. Available: www.pvsysyt/help.
68
[36] D. Thevenard, UNCERTAINTY IN LONG-TERM PHOTOVOLTAIC YIELD PREDICTIONS, 2010.
[37] F. A. G. A. B. L. M. Cynthia E.L. Latunussa, “Life Cycle Assessment of an innovative recycling process for
crystalline,” Solar Energy Materials & Solar Cells, 2016.
[38] ACEA - European Automobile Manufacturers Assosiation, “European Automobile Manufacturers
Assosiation - CO2 Emissions of New Passenger Cars,” 2018. [Online]. Available:
https://www.acea.be/statistics/article/new-car-co2-emissions. [Accessed 2018 September 10].
[39] Enerdata, “ODYSSEE MURE - Sectoral Profile - Transport,” 2015. [Online]. Available:
http://www.odyssee-mure.eu/publications/efficiency-by-sector/transport/distance-travelled-by-
car.html. [Accessed 10 September 2018].
[40] T. W. Bank, “CO2 emissions Cambodia,” [Online]. Available:
https://data.worldbank.org/indicator/EN.ATM.CO2E.KT?locations=KH. [Accessed 18 September 2018].
[41] IST, “Solar Photovoltaic Course Slides,” 2017.
[42] PV syst, “PVsyst help - Shading Losses,” [Online]. Available: http://files.pvsyst.com/help/.
[43] PVsyst, “PVsyst program help - Transposition model”.
[44] C. Colp, “Solarprofessional - Cost-Saving PV Source-Circuit Wiring Method,” 2014. [Online]. Available:
https://solarprofessional.com/articles/design-installation/cost-saving-pv-source-circuit-wiring-
method#.W7YRbWgzZnI. [Accessed 2018 May 23].
69
Appendices
Appendix 1
Appendix 1 presents an overview of the company where my six months internship took place. The EPC company
is described as well as their business model.
Kamworks Ltd.
Kamworks Ltd. was established in Cambodia in 2006 by Dutch engineers. It was the first solar company in the
country and it started with the Moonlight lantern project, a small solar lantern designed and manufactured by
Kamworks for the Cambodian market. In 2014, Kamworks entered the market of Solar Home System (SHS) selling
more than 14000 units and one year later developed a pay-as-you-go (PAYGO) SHS with GSM-based machine-
to-machine (M2M) connectivity integrated with mobile money for payment collection. The PAYGO systems were
sold in different sizes and with different types of purchase/rent: perpetual rent, rent to own and direct purchase.
They achieved their goal of 500 systems and they create the PAYGO Solutions section of the company. In the
meanwhile, they continue to install solar systems and they start focusing on off-grid solar water pumping making
it the most successful activity of the company. In the last 2 years the company start focusing more on bigger
customers such as hotels, garment factories, food and beverage industries and schools. In 2017 the company
installed its biggest on-grid project until now commissioned, a 100 kW PV system for a garment factory near
Phnom Penh. The company is in contact with some big investors that can offer PPA for a good price for big
project (>100kWp) but is missing investors for smaller projects.
The company consists in a small team of engineers and some internal technicians that install both the off-grid
and on-grid systems. During the first months of 2018 the company experienced few changes in the team. The
on-grid project manager left on February and has been replaced two weeks later, but after few months also the
successive project manager left the company and the position has not been filled by someone else (until I left).
My role was to work on all the project phases of the on-grid projects of the company together with the electrical
engineer, a 23 years old Cambodian engineer under the guide of the project manager of the section. Due to the
establishment of the new solar regulation that have led to the uncertainty of both customers and investors,
many big projects that were in the pipeline have been temporarily suspended, waiting for news from the
government side. Because of that, I’ve been able to work also on off-grid projects such as the design of an off-
grid solar dryer and to manage some small installations such as the installation of 20 KW on-grid system for a
hospital in Battambang. Kamworks is trying to emerge as the leading company for solar in Cambodia and is also
betting for big projects (MW scale). The advantage of the company is to have a good projects portfolio and have
been the first solar company in the country. The company has also good connections with investors that can
finance big projects for PPA and with an agricultural association with many farmers and food company
associated.
70
Kamworks is currently in a transition period with the changing regulation on solar market. Once the uncertainties
about the solar regulation will be solved and the market will be less worried about the investment on solar, the
company can scale up with some big projects ensuring a good investment base for the future.
Kamworks’s Business Strategy
In Figure 29 has been shown the scheme of the business strategy adopted by Kamworks Ltd.
The solar market in Cambodia is various and started attracting more interest both from the investors and
consumers side. The trust in the technology is not yet completely built but the process is ongoing. Kamworks
has different approaches according to the types of projects as summarized in the scheme presented in Figure
29. Currently the small projects act as the engine of the company helping to sustain the business and finance
the bigger projects. Water pumping projects are the most common in the pipeline of the company followed by
the small and medium on-grid installations. The good portfolio of the company allows it to provide good
examples to potential clients and to reduce their apprehension regarding the solar market. The goal of the
company is now to secure a large project (MW scale) to extend at maximum the portfolio and built trust in the
customers for all sizes of projects.
71
Figure 29 - Scheme of the business strategy of Kamworks according to the market sector.
On-grid
Small (<50 kWp)
•Too small for PPA
•High cost per kWp
•Higher margin
•Fast and easy procedure
Medium (50-500 kWp)
•Kamworks has financers for PPA contracts
•Good partnership for panels at cheap price
•Good partnership for inverters
•Standard margin
•Possible installation with the current team
Large (> 500 kWp)
•Kamworks has financers for PPA contracts
•Long procedures for the approval both from clients and EDC side
•Lower margin
•Challenging for the finance of the company, advance payments and
•Technically challenging for Kamworks technical team, need to hire more technicians for the installation
Off-grid
Small (<10 kWp)
•Too small for PPA
•High cost per kWp
•Higher margin
•Fast and easyprocedure
Medium (10-50 kWp)
•Kamworks has financers for PPA contracts
•Good partnership for panels at cheap price
•Good partnership for inverters
•Lower margin
•Possible installation with the current team
Large (>50 kWp)
•Large and challenging
•Kamworks will try new technologies for the battaries (NaS or Redox Flow)
•Large off-grid installations are rare due to really elevated costs
Water pumping
• large experience in the field
•fast and easy procedure
•high margin
"Ad-hoc"
•off grid projects specifically designed for the customer's needs (i.e. solar dryer, street ligths)
•challenging engineer
•high margin
PAYGO
Various customisable options and sizes of the systems to meet all the
possible needs
Transition from business-to-consumer (B2C) to
business-to-business (B2B)
Looking for clients around the world (Africa,
Bangladesh, India etc.)
72
Appendix 2
In this section the pictures of the first site visit at Grace Glory Garment Factory are resented. The pictures show
the load profile measurement on the two different electrical cabinet connected to T1 and T2, one of the
transformers of 1000 KVA.
Figure 30 - T1 (left) and T2 (right) electrical cabinet measurements.
Figure 31 - Transformer T1 (left), power analyzer (center) and electrical cabinet T2 (right).
73
The roof, the rooftop cover, the small bridges cover and the panoramic view of the roofs of the building are
shown and has been done in order to better reproduce the site in the 3D design that follow the measurement
process. The electrical bill is finally presented to show the monthly consumption of the garment factory and has
also been used to compare the simulation results.
Figure 32 - Panoramic view of the first building roof of the garment factory.
Figure 33 - T1 building roof side top view (left), metal roof cover (center) and connection buildings cover (right).
74
Figure 34 - Garment factory electricity bill.
75
Appendix 3
Hourly data (kW) from a typical week, extrapolated from the measurements at the Garment Factory.
Table 17 - T1 and T2 load measurements at the garment factory.
T10
12
34
56
78
910
1112
1314
1516
1718
1920
2122
23
Mon
day
30,7
529
,60
30,0
329
,69
29,5
428
,37
62,5
641
6,87
465,
0447
0,11
450,
7629
1,16
465,
3448
7,13
470,
6542
6,42
79,8
760
,28
43,2
246
,98
43,7
043
,62
36,1
132
,66
Tues
day
30,0
428
,91
29,6
730
,40
31,1
327
,04
87,4
947
3,19
491,
6949
5,37
465,
1330
9,38
473,
9147
2,45
479,
7736
5,03
73,2
562
,38
37,4
736
,21
36,4
541
,33
31,5
529
,21
Wed
nesd
ay28
,77
30,1
032
,02
29,7
227
,42
26,9
889
,64
469,
6548
2,65
481,
5745
6,09
296,
8748
2,21
473,
9649
0,41
459,
9987
,62
66,4
443
,28
44,3
340
,05
40,2
735
,91
30,6
5
Thur
sday
29,0
630
,96
30,7
429
,40
27,4
428
,73
100,
5745
6,48
474,
0047
0,44
468,
5330
4,15
476,
1847
8,78
481,
0037
1,01
69,2
269
,11
35,7
537
,71
36,3
640
,85
31,5
629
,67
Frid
ay31
,97
30,2
231
,49
29,4
429
,24
30,8
996
,20
429,
7646
3,40
469,
5744
3,66
285,
4547
2,81
483,
1846
4,71
438,
0893
,22
60,0
145
,31
47,5
344
,07
42,9
338
,02
35,6
3
Satu
rday
30,1
430
,56
33,2
831
,23
31,9
928
,59
105,
0745
6,32
474,
9847
8,39
460,
8929
3,89
480,
4547
7,38
481,
0837
1,01
54,0
845
,61
39,3
437
,78
38,5
136
,21
35,5
731
,83
Sund
ay28
,24
27,2
428
,34
29,0
828
,86
24,5
722
,80
27,8
353
,91
53,3
552
,67
55,3
153
,01
53,3
556
,87
57,3
538
,18
30,3
933
,87
36,3
732
,00
31,2
029
,09
32,0
7
Avg.
Mon
-Fri
30,1
229
,96
30,7
929
,73
28,9
528
,40
87,2
944
9,19
475,
3547
7,41
456,
8429
7,40
474,
0947
9,10
477,
3141
2,11
80,6
463
,64
41,0
142
,55
40,1
341
,80
34,6
331
,56
T2 Mon
day
12,5
312
,51
12,7
712
,12
12,1
111
,85
30,3
996
,23
144,
6214
6,35
135,
1659
,71
121,
5716
1,30
160,
9917
1,44
168,
8315
3,42
12,4
610
,74
10,7
910
,87
10,9
010
,88
Tues
day
10,9
110
,92
10,9
910
,89
10,8
910
,51
27,7
510
3,97
165,
0417
5,01
105,
8659
,40
128,
1114
2,25
109,
7315
8,81
150,
2312
3,96
14,2
812
,75
12,5
712
,59
12,1
812
,67
Wed
nesd
ay12
,24
12,2
312
,32
12,1
912
,15
11,6
330
,87
95,0
514
5,79
147,
1514
1,55
42,9
313
8,43
169,
7112
9,20
191,
4515
8,09
130,
4411
,51
9,46
9,43
9,45
9,36
9,26
Thur
sday
9,27
9,25
9,28
9,17
9,10
8,69
29,5
411
9,76
118,
2517
3,18
136,
7457
,47
117,
1714
5,14
156,
4911
8,74
118,
0680
,68
9,19
9,47
9,48
9,44
9,48
10,6
8
Frid
ay11
,06
10,9
510
,81
10,8
710
,88
10,3
027
,73
105,
7715
6,98
181,
1311
6,57
57,3
612
2,13
140,
1415
2,93
127,
1211
2,19
85,7
712
,03
11,8
09,
769,
519,
479,
49
Satu
rday
9,54
9,24
9,38
9,15
9,18
8,76
30,1
112
1,31
123,
1416
8,27
135,
1357
,41
114,
9315
1,16
152,
2710
9,71
99,1
776
,68
9,34
9,45
9,59
9,45
9,38
9,48
Sund
ay9,
439,
469,
449,
409,
249,
938,
408,
228,
308,
218,
278,
918,
258,
398,
468,
499,
718,
188,
228,
308,
218,
278,
918,
25
Avg.
Mon
-Fri
11,2
011
,17
11,2
311
,05
11,0
310
,59
29,2
610
4,16
146,
1416
4,56
127,
1855
,37
125,
4815
1,71
141,
8715
3,51
141,
4811
4,85
11,8
910
,85
10,4
110
,37
10,2
810
,60
76
Appendix 4
The 3D maps of some area are reported in this section. As can be seen from the images the results can vary
widely.
Figure 35 - 3D design (up) and 2D map (down) of a residence of orphans in Phnom Penh, Cambodia.
77
Figure 36 - Rice milling factory 3D design (up) and 2D design (down) situated in Phnom Penh, Cambodia.
78
Figure 37 - 3D elaboration (up) and 2D map (down) of a factory manufacturing and distributing home decor, gifts, and holiday products in Sihanoukville, Cambodia.
It is important to notice how the 2D maps represent a much more defined and precise option in comparison to
the commonly used online maps. 3D designs have still lot of imperfections. The main problem is the reflection
of the sun on the metallic surfaces that worse the drawings. For instance, Figure 37 presents a good 3D design
with few imperfections while Figure 36 has different results depending on the building exposition due to the sun
reflections on the metal roof.
79
Appendix 5
The main technical specification of the modules under STC are summarized in Table 18 while inverters
datasheets are presented in Figure 38 and Figure 39.
Table 18 - Main technical specifications of all modules under STC and I-V curves.
Antaris P72 320
Canadian Solar CS6U-335P
Jinko Solar JKM340PP-72-V
Rated Power (Pmax) [Wp] 320 330 340
Optimal Operating Voltage (Vmpp) [V] 37.1 37.4 38.2
Optimal Operating Current (Impp) [A] 8.63 8.96 8.91
Open Circuit Voltage (Voc) [V] 45.7 45.8 47.5
Short Circuit Current (Isc) [A] 9 9.54 9.22
Efficiency [%] 16.49 17.23 17.52
Operating Temperature [°C] -40°C ~ +85°C -40°C ~ +85°C -40°C ~ +85°C
Antaris P72 320 I-V curve
Canadian Solar CS6U-335P I-V curve
80
Jinko Solar JKM340PP-72-V I-V curve
81
Figure 38 - SUN2000 36KTL inverter from Huawei specifications.
82
Figure 39 - SunnyBoy Tripower 25000TL JP-30 inverter from SMA specifications.
83
Appendix 6
The simulation report of the optimal solution, as presented by PVsyst, is reported in this section. Pvsyst version
6.75 has been used with the student license. All functionalities of the program were available. The printed
reports present the student version mark printed on to avoid business purposes.
84
85
86
Appendix 7
In this section the report of the optimal solution of PVsyst for the system connected to T2 is presented.
87
88
89
90
Appendix 8
Appendix 8 presents the values and the results for the carbon balance simulation with the accurate values insert
in the program for the calculation both for T1 and T2.
Figure 40 - Carbon balance settings for T1 system (first image) and the overview for T1 (second image).
91
Figure 41 - Carbon balance settings for T2 system (first image) and the overview for T2 (last image).
92
Figure 42 - The grid energy mix of Cambodia (including imports) common at T1 and T2.
