Increasing Small Scale Solar Power in Sri Lanka A feasibility study of

Increasing Small Scale Solar Power in Sri Lanka A feasibility study of
June 14, 2014
A feasibility study of
Increasing Small Scale
Solar Power in Sri Lanka
Hannes Hagmar
Bachelor’s thesis
Electrical Engineering, Electric Power Technology
Department of Engineering Science
BACHELOR’S THESIS
Increasing Small Scale Solar Power in Sri Lanka
Summary
The following report is conducted as a feasibility study, aimed to objectively uncover the
advantages and challenges of increasing the amount of small scale solar power in Sri Lanka.
The demand for electricity in Sri Lanka has been steadily increasing the last few years and
there is an urgent need to find new ways of generating electricity. To not further increase
the already high dependency of foreign oil and to decrease the impact on the environment,
a transition from traditional combustion of fossil fuel to new renewable energy is required.
The report shows that there exists substantial potential for generating solar energy in Sri
Lanka. Calculations show that an investment in a photovoltaic system can be economically
favourable and that the investment often is paid back within a few years. Current
regulations and electricity pricing increases the economic incitement for high electricity
consumers to invest in small scale solar power. Furthermore, the report demonstrates that
there are likely no technical obstacles of increasing small scale solar power at this period. In
contrary, the report shows that small scale solar power in general decreases line losses,
voltage drops, and the peak demand of electricity.
At present, it is probably not the lack of economic incitement but rather socio-economic
factors that limit the development of small scale solar power. Sri Lanka is still a relatively
poor country and the long years of civil war have prevented development and wealth. Lack
of funds and a high ratio of low-income earners are probably the main reason for the slow
development.
Date:
Author:
Examiner:
Advisor:
Advisor:
Main area:
Credits:
Keywords
Publisher:
June 14, 2014
Hannes Hagmar
Mikael Ericsson
Lars Holmblad, University West
Manjula Fernando, University of Peradeniya, Sri Lanka
Electrical Engineering, Electric Power Technology
15 HE credits
Education level: first cycle
Sri Lanka, distributed generation, photovoltaic power, technical feasibility,
economic feasibility.
University West, Department of Engineering Science,
S-461 86 Trollhättan, SWEDEN
Phone: + 46 520 22 30 00 Fax: + 46 520 22 32 99 Web: www.hv.se
i
A Feasibility Study of Increasing Small Scale Solar Power in Sri Lanka
Preface
This report is a result of a Minor Field study carried out in Sri Lanka between April to May
2014, with the funding by the Swedish International Development Cooperation Agency
(SIDA). To enhance the understanding and ease the comprehension of the figures and
tables it is recommended that the report is reprinted in colour. The author is the creator of
all figures unless specifically stated otherwise.
I would like to thank my supervisors for all inputs and advices. All colleges at the Electrical
Engineering faculty in the University of Peradeniya have my greatest gratitude for being
most welcoming and taking great care of me during my stay.
I would then like to give a special thanks to my father for all the discussions and support
throughout this period. Then of course, my highest gratitude is towards Lina who has been
my best support and the best travel companion imaginable.
Hannes Hagmar, 14th of June 2014
ii
A Feasibility Study of Increasing Small Scale Solar Power in Sri Lanka
Contents
1
2
3
4
5
6
7
Introduction ........................................................................................................... 1
1.1 Background ....................................................................................................................... 1
1.2 Purpose and Scope........................................................................................................... 3
1.3 Methodology ..................................................................................................................... 4
1.4 Overview of project ......................................................................................................... 5
Current conditions in Sri Lanka ............................................................................6
2.1 Energy situation of Sri Lanka ......................................................................................... 7
2.1.1 Energy supply ..................................................................................................... 7
2.1.2 Energy distribution ............................................................................................ 8
2.1.3 Electricity pricing ............................................................................................... 9
2.2 Solar resources in Sri Lanka ......................................................................................... 10
Photovoltaic cells ................................................................................................. 12
3.1 Photovoltaic technology ............................................................................................... 12
3.2 Calculating possible generation of a PV system ........................................................ 15
3.2.1 Radiation on an inclined surface .................................................................... 15
3.2.2 Temperature effects ......................................................................................... 19
3.3 Variability of output....................................................................................................... 20
Distributed generation and photovoltaic power ................................................. 22
4.1 Distributed generation and grid protection................................................................ 23
4.1.1 Islanding ............................................................................................................ 24
4.1.2 Voltage regulation with DG ........................................................................... 25
4.2 Electric power quality .................................................................................................... 25
4.2.1 System frequency .............................................................................................. 26
4.2.2 Fast voltage fluctuations and flicker .............................................................. 26
4.2.3 Harmonic distortion ........................................................................................ 26
4.3 Voltage level and unbalance ......................................................................................... 27
4.3.1 Single-phase/three-phase connections.......................................................... 31
4.4 Line loss reductions ....................................................................................................... 33
4.5 Peak demand reduction and diversified demand....................................................... 34
Simulations and calculations ............................................................................... 36
5.1 Economic feasibility ...................................................................................................... 36
5.2 Technical feasibility........................................................................................................ 39
5.2.1 Increased voltage level ..................................................................................... 39
5.2.2 Decreased line losses ....................................................................................... 40
5.2.3 Typical case grid ............................................................................................... 41
Results.................................................................................................................. 42
6.1 Economic feasibility ...................................................................................................... 42
6.2 Technical feasibility........................................................................................................ 44
6.2.1 Increased voltage level ..................................................................................... 45
6.2.2 Decreased line losses ....................................................................................... 47
Analysis and discussion ....................................................................................... 50
iii
A Feasibility Study of Increasing Small Scale Solar Power in Sri Lanka
7.1 Analysis of economic feasibility ................................................................................... 50
7.2 Analysis of technical feasibility .................................................................................... 51
7.3 Discussion of method and results ............................................................................... 52
7.4 Future work .................................................................................................................... 53
8 Summary of conclusions...................................................................................... 54
References ................................................................................................................. 55
Appendices
A. Investment calculations and discount rate factors
B.
C.
D.
E.
Increased voltage levels
Decreased line losses
Cable data and impedance calculations
MATLAB-script
iv
A Feasibility Study of Increasing Small Scale Solar Power in Sri Lanka
Abbreviations
AC
Alternating Current
CEB
Ceylon Electric Board: National power company of Sri
Lanka
DC
Direct Current
DG
Distributed Generation
Feeder
Electric line
HH
Households
IPCC
Intergovernmental Panel on Climate Change
kWh
Kilowatt hour
LCC analysis
Life-Cycle Cost analysis
LV
Low Voltage
NRE
New Renewable Energy
NREL
National Renewable Energy Laboratory
Penetration level (of DG)
Amount of distributed generation in comparison to the
power load of that sub-grid.
PCC
Point of Common Coupling. Point in the network closest
to other customers
POG
Point Of Generation. Location in grid with distributed
generation
PV
Photovoltaic
PV-DG
Distributed Generation of Photovoltaic Systems
PVWatts
An application designed to calculate the output of a
standardized photovoltaic system
Rs.
Sri Lankan rupee
SIDA
Swedish International Development Cooperation Agency
v
A Feasibility Study of Increasing Small Scale Solar Power in Sri Lanka
Terminology
and
Temperature coefficients
Phase angle
Solar declination
Geographical latitude
Solar azimuth
Constant albedo
Sunrise hour angle
Annual savings
Tilt of an array with respect to the horizon
Radiation received over a day by a horizontal area outside Earth's
atmosphere
Global radiation on a horizontal surface
Current year
Beam radiation on a plane perpendicular to the radiation
Diffuse radiation on a horizontal surface
Diffuse radiation on an inclined surface
Ground-reflected radiation on an inclined surface
Beam radiation on an inclined plane as followed
Short-circuit current
Line current
Clearness index
Economic lifetime of the investment
Total active line losses
Active power at load point
Discount rate
Reactive power at load point
Resistance of the line
Solar constant
Rated power of the transformer
Open-circuit voltage
vi
A Feasibility Study of Increasing Small Scale Solar Power in Sri Lanka
Voltage level at the beginning of the line
Voltage level at the load point
Phase-to-phase voltage
Rated voltage of the transformer
Voltage drop along the line
Complex impedance of the line
Impedance of the feeding grid
Short-circuit impedance for the feeding transformer
and
Phase- and neutral impedance
Relative short-circuit impedance of the transformer
Complex reactance of the line
vii
A Feasibility Study of Increasing Small Scale Solar Power in Sri Lanka
1 Introduction
The following report is performed within the framework of the Minor Field Studies
Scholarship Programme, funded by the Swedish International Development Cooperation
Agency, SIDA. The programme is intended to enhance students’ knowledge and
understanding of development perspectives in a country of choice. The field study has
been performed during 8 weeks in Sri Lanka, in cooperation with the University of
Peradeniya and the solar power company JLanka Technologies.
The report is a part of a bachelor thesis within the electrical engineering program at
University West in Trollhättan, Sweden. The report is aiming to investigate the possibilities
of increasing the amount of small scale solar power in Sri Lanka.
1.1 Background
Creating electricity in an environmentally friendly and sustainable way is one of the greatest
challenges of today. The temperature of the Earth has seen a successively increase due the
emissions of greenhouse gases. According to the United Nations climate panel (IPCC) the
atmospheric concentration of carbon dioxide has increased by 40% since pre-industrial
times, primarily due the combustion of fossil fuels [1]. To reach the aim of a temperature
increase of maximum two degrees, actions has to be taken immediately. Scenarios designed
to stabilize the climate verifies that the global emissions of greenhouse gases need to
culminate at the year 2020 and then rapidly decrease [1]. Despite the vast knowledge of the
problem and a general consensus that the amount of emitted greenhouse gases needs to be
decreased - the emissions only seem to increase.
The developed countries have for a long time accounted for the largest part of the
emissions of greenhouse gases [1]. However, as developing countries increase their welfare
and living standards, their amount of used fossil fuel is increasing. At the same time, it is
generally accepted that developing countries have the same right to the high living
standards as in the already developed countries. Consequently, there exists a conflict
between the goal of reducing the emissions of greenhouse gases and increasing the living
standards in developing countries. An important challenge is therefore how the developing
countries are to increase their living standards without harming the environment.
The answer to the problematization might be sustainable development. Sustainable
development was initially defined by the Brundtland report as: "…development that meets the
needs of the present without compromising the ability of future generations to meet their own needs” [2]. If
welfare and living standards could be increased without harming the environment or
1
A Feasibility Study of Increasing Small Scale Solar Power in Sri Lanka
compromising the possibilities of future generations, the development would be made in a
sustainable way.
Sri Lanka, situated south east of India, has for a long time been troubled with civil war as
well as poverty, and the country suffered hard from the tsunami in 2004 [3]. However,
since the end of the civil war in 2009 the country has had an increase in both tourism as
well as domestic development. With increased growth, the demand of electricity has
increased as well and now more than 50% of all electricity is generated from the
combustion of fossil fuels. A transition from fossil fuels to more renewable energy sources
would not only reduce the impact on the environment but might also bring other economic
benefits. Sri Lanka is currently heavily dependent of the import of foreign oil and a longterm transition to more renewable energy would both decrease that dependency and result
in an increased amount of qualified jobs within the country.
Renewable energy can be generated in a numerous ways, but the perhaps most obvious and
especially for countries close to the equator is using solar power. Solar power is not only a
fully renewable energy source, but it also allows Sri Lanka to create energy without
disturbing the sensitive natural areas and biodiversity. As the market for solar systems has
expanded, the technology is both getting more efficient and less expensive. As solar cells
are getting more and more economically competitive to install they start to challenge the
more conventional ways of generating electricity.
Solar cells, or photovoltaic cells (PV), are often implemented in small scale on residential
rooftops or as utility owned units. However, the development and the increase of
distributed PV systems (PV-DG) are not all uncomplicated. First of all, if a substantial
increase of installed solar power is to actually take place, the investment has to be
economically favourable. The reduced impact of the environment and the reduced
dependency in changes in electricity prices could of course be a sufficient incitement for an
investment. But in developing countries with a less strong economic situation the
economic incitement of doing an investment has to be sufficiently high.
Furthermore, small scale solar power is generated within the so called distribution grid.
Distribution grids are generally not dimensioned for generation and an increase of PV-DG
will affect factors such as power quality, line losses, and grid protection in numerous ways.
Therefore, for an increase of PV-DG to even be technically possible the grid has to be
sufficiently developed.
2
A Feasibility Study of Increasing Small Scale Solar Power in Sri Lanka
1.2 Purpose and Scope
The following report is conducted as a feasibility study, aimed to objectively uncover the
advantages and challenges of an increased amount of small scale solar power in Sri Lanka.
The report aims furthermore to, from the current conditions in Sri Lanka, develop general
guidelines that could serve as a guide in the future development of solar power in Sri
Lanka. In order to achieve the objective, two main areas are thoroughly examined and
discussed:

Economic feasibility of an increased amount of distributed generation by
photovoltaic cells in Sri Lanka. Is an investment in a photovoltaic system
economically beneficial in Sri Lanka? What factors is the investment reliant on to
become economically beneficial? What comparative advantages does Sri Lanka
have in increasing small scale solar power? In what manner does the possible power
output differ from other countries and how is it calculated?

Technical feasibility of an increased amount of distributed generation by
photovoltaic cells. In what ways does an increased amount of distributed
generation affect line losses, electric power quality, and the grid as a whole? Is
distributed generation an obstacle to achieve a more environmentally friendly
energy sector or in fact an asset?
Due to time and resource limitations, the report will only focus on the purely technical and
economic aspects of increasing small scale solar power. However, the increase of solar
power is of course not only restrained by merely technical aspects, but perhaps even more
by the lack of sufficient funds, political will, and other socio-economic aspects. These
aspects will not be taken into account in the report, but instead be briefly discussed in
context to the results of the report. Furthermore, only calculations concerning low voltage
grids will be taken into account. Due to the limitations, no regard is paid to the effects of
PV-DG in medium to high voltage grids.
3
A Feasibility Study of Increasing Small Scale Solar Power in Sri Lanka
1.3 Methodology
The following report is partly consisting of gathered information from the solar power
company JLanka Technologies and the University of Peradeniya. Due to the difficulty of
predicting exactly what kind of information that would be available in Sri Lanka, a
thoroughly literature study was performed prior to the field study. There is a vast amount
of literature covering the aspects of photovoltaic power. However, the concept of
distributed generation is still rather unexplored. Most of the information for distributed
generation is thus gathered from various reports and journals.
Being a feasibility study, there is no attempt to exactly calculate the effect of an increased
amount of PV-DG. A feasibility study is a way to uncover the strengths and weaknesses of
a project and thus provide knowledge of what factors that limits the prospects of the
project. The choice to design the report as a feasibility study came mainly from the fact that
the project has fairly strict time and resource limitations. It would not be possible to cover
all the aspects in detail and a larger scope was thus necessary. Instead of examining the
effect of an increase of PV-DG at a large scale, it would be possible to investigate a single
object in more detail. However, such a report would not be able to draw the same general
conclusions as a feasibility study. A more general study was found to be of more interest
from a development perspective which is suitable as the report is written under the
framework of the Minor Field Studies Scholarship Programme.
The main two areas of the feasibility study; the technical and economic feasibility are
examined separately. From the specific conditions of Sri Lanka, a typical case grid is
constructed. This case grid is then used to evaluate and estimate the effects of an increased
amount of PV-DG. Calculations are performed only on the aspects of slow voltage
variations and changes in line losses. The restriction is made due to time and resource
limitations and other aspects of an increased amount of PV-DG are instead discussed
thoroughly. Some of the calculations are performed with the aid of MATLAB. MATLAB is
used both because some of the simulations require a large number of calculations and to
ease the plotting and presentation of the results. The economic feasibility is studied with
the aid of a life-cycle cost analysis. The analysis is implemented with the aid of the PVWatts
application, an application programmed to analyse the possible output for a standardized
photovoltaic system [4].
The combination of in-depth literature studies and experience from the field provides a
good mixture of theory and practice. Since the report is supposed to investigate the
feasibility of actually increasing the amount of small scale solar power in Sri Lanka, a purely
4
A Feasibility Study of Increasing Small Scale Solar Power in Sri Lanka
theoretical work would not be optimal. At the same time, a practical study with no
theoretic background would not be scientifically viable.
1.4 Overview of the report
The report is mainly divided into four separate, although coherent parts: background and
theory, simulations and calculations, results, and finally discussion and conclusions. The
report commence by establishing a background of the function of the photovoltaic cell and
how the possible output of solar power is calculated. Furthermore, the concept of
distributed generation is explored and aspects such as line losses and electric power quality
are examined. The background is later used to aid the comprehension and put the results in
a context.
The calculations and simulations part introduces the reader to the examined issues. The
used models and the typical case grid is constructed and described. The main part of the
calculations is then presented within the appendixes. The results for each part are then
presented and briefly explained within the result section. Finally, the results are being
discussed and conclusions are made.
5
A Feasibility Study of Increasing Small Scale Solar Power in Sri Lanka
2 Current conditions in Sri Lanka
Sri Lanka is situated south east of India and has for a long time been troubled with civil
war as well as poverty. In the mid-1970, tension between the two largest populations in Sri
Lanka, Sinhala and Tamil, began to increase [3]. The conflict intensified and violence from
both sides escalated into a civil war that would eventually claim upwards 100.000 lives.
After several attempts at peace, in 2002 finally a ceasefire agreement was signed. However,
an event beyond predictions struck the island that would eventually shatter the long
awaited peace attempts. The 26th of December 2004, the waves of a tsunami had a
devastating effect of the whole coastline with more than 30 000 casualties and many more
homeless and injured. The conflict increased yet again and not until 2009, after 26 years,
the civil war finally ended.
However, during the last few years the economy has boomed and Sri Lanka now shows a
higher gross domestic product per capita, higher educational level, and expectancy of life
than many neighbouring countries [3]. It has seen a significant increase in both tourism as
well as national development. Furthermore, Sri Lanka has a rich but sensitive climate with a
high biodiversity and sensitive natural areas. It has been identified as one of the planet’s 25
biodiversity hotspots with very high level of endemism (species unique to the area). One of
the greatest environmental challenges and the greatest threat to the biodiversity has been a
rapid deforestation of the island. Land needed for agriculture and biofuel for heating and
cooking are the main drivers for the deforestation [3].
Due to advantageous geo-climatic conditions, several forms of different energy sources are
abundant in Sri Lanka [5]. Being located in the equatorial belt, Sri Lanka receives a high
supply of solar radiation year around. The radiation over the island show a small seasonal
variation, however, significant spatial differentiation is possible to observe between the
mountain and lowland regions. The temperature in Sri Lanka is high year around and the
mean annual temperature is ranging from 26.5 to 28.5 degrees Celsius in the lowlands [6].
In the highlands the temperature decreases rapidly as the altitude increases.
The weather is characterized into four climate seasons with two separate monsoon seasons
[6]. The southwest monsoon hit the southwest coast during May to September but mainly
leaves the eastern coast dry. During December to February the northeast monsoon creates
heavy rainfall in the along the eastern coast but leaves the west and south coast relatively
dry. Although clouds are consistent during monsoon periods the sun is still strong and
sunny days are not uncommon [6].
6
A Feasibility Study of Increasing Small Scale Solar Power in Sri Lanka
2.1 Energy situation of Sri Lanka
The Ministry of Power and Energy is the main authority and responsible for the energy
sector in Sri Lanka [5]. The Sri Lanka Sustainable Energy Authority is a governmental
agency operating to increase the amount of renewable energy sources within the country
and reducing wasted energy. Subordinate to the Ministry of Power and Energy is the
national power company, Ceylon Electricity Board (CEB). With the mission to develop and
maintain an efficient and economical system of electricity supply in Sri Lanka, the CEB is
empowered to generate, distribute and transmit electricity within the country. In order to
reduce the impact on the environment, the CEB has also adopted a long term plan to
achieve a 20% of the country’s total electricity through renewable energy sources.
2.1.1 Energy supply
With the booming and growing economy the demand for electricity has increased
significantly the last few years [5]. Between the year 2000 and 2011, the generated electricity
increased with more than 82% (see table 2.1). The power sector in Sri Lanka has for a long
time been heavily dependent on hydro power. However, the capacity of the hydro power is
almost at a maximum and other means of power generation are necessary. As a result of
the high demand for electricity the last few years, the demand for petroleum have also
increased drastically. The amount of electricity produced by fossil fuelled thermal power is
more and more significant, from producing 37.3% of all electricity the year 2000 to 53.8%
the year 2011. Due to an increase of the price of petroleum, the expenditure on oil imports
consumes a growing share of the foreign earnings of Sri Lanka. For example, the
petroleum import bill in 2011 estimated to about 44.2% of the Sri Lanka’s non petroleum
export earnings [5]. Seeking to decrease the dependency of petroleum, the first coal power
plant was commissioned in 2011 with a capacity of 300 MW.
The development of new renewable energy (NRE) has increased rapidly the last few years
(see table 2.1). At present, NRE is found in many forms such as small hydro, solar, wind
and biomass power plants [5]. Solar energy is mostly used in non-commercial forms such
as for drying and heating water and the total usage of solar energy may not be quantified
properly. The first grid connect solar power plant was established in 2011 with a capacity
of 1.237 MW and is the largest contribution to NRE industry so far. The development of
different power sources and the generated electricity to the grid is presented in table 2.1.
7
A Feasibility Study of Increasing Small Scale Solar Power in Sri Lanka
Table 2.1 Total grid connected capacity 2011 in Sri Lanka [5].
Generated electricity to CEB
Grid [GWh]
2000
Major Hydro
2,812.8 3,222.5 3,700.5 3,355.6 4,988.5
4,017.7
Thermal (Oil)
3,512.4 5,339.3 5,848.8 6,062.5 5,063.3
5,857.5
Thermal (Coal)
-
-
-
-
-
1,027.6
CEB Wind
3.4
2.4
3.2
3.5
3.0
2.7
New Renewable Energy
43.3
279.7
434.6
548.5
728.5
722.3
Total generation to grid
6,371.8 8,844.0 9,987.1 9,970.1 10,783.2 11,627.8
Year-on-year growth rate
11.5%
2005
9.6%
2008
1.4%
2009
-0.2%
2010
8.2%
2011
7.8%
The average per capita electricity consumption in Sri Lanka was estimated to 480 kWh per
person at the year of 2011. The highest demand for electricity occurs at the dry periods,
when most tourists visit Sri Lanka. These periods are often the warmest and air
conditioning and other appliances need generally more electricity [5].
2.1.2 Energy distribution
Sri Lanka has a regulated grid and the CEB has a monopoly over electricity transmission
[5]. The national grid is consisting of 220, 132 and 33 kV lines providing electricity to
almost all households in Sri Lanka. It consists of overhead transmission lines
interconnecting the large scale power plants (mainly situated in central and western regions)
and grid substations where the distribution network is connected. At the end of year 2011,
91% of the total households were electrified and the major part of the not yet electrified
parts were to be found in the poorer north and northeast regions [7].
The grid of Sri Lanka has during the last decade suffered from high energy losses, in the
range of 15-20% (see figure 2.1). However, the efficiency has gradually improved during
the last few years and was estimated to 11.7& in year 2011[7].
8
A Feasibility Study of Increasing Small Scale Solar Power in Sri Lanka
25
System Energy Losses [%]
19,69
20
19,2
18,44
17,11
17,27
16,58
15,67
14,99
15
13,9
12,97
11,72
10
5
0
2001
2003
2005
2007
2009
2011
Year
Figure 2.1 System energy losses in Sri Lanka between 2001 and 2011 [7].
2.1.3 Electricity pricing
In April 2013 a new electricity tariff for domestic consumers was implemented, which
caused generally higher electricity prices for all demand profiles [5]. The electricity tariff
method is a highly progressive method that not only depends on the number of units
(kWh) consumed, but also the rate of consumption. For example, the charge for a
consumed amount of electricity over a period of 10 days will cost more than the same
amount of electricity consumed over a period of 20 days [8]. The type of consumer is also
affecting the price model and domestic, religious, industrial, and commercial consumers all
have different tariff models.
In table 2.2 below, the tariffs for domestic customers with different monthly consumptions
are listed with both unit charge as well as other incremental costs [8]. As can be noted from
the table, the difference in the charge per unit is ranging from 3 Sri Lankan rupee (Rs.) per
kWh for those with the lowest monthly consumption to 42 Rs./kWh for those with the
highest (>180 kWh/month).
The CEB has also recently implemented the possibility of net energy metering, allowing
micro producers of electricity to be compensated if their generation exceeds their
consumption [9]. If the export is larger than the import, the consumer will receive an
9
A Feasibility Study of Increasing Small Scale Solar Power in Sri Lanka
export credit that can be used for the next billing period. Thus, there will be no financial
compensation for the excess energy exported by the consumer. Instead, all exports will be
set of the consumers own future consumption.
Table 2.2 Electric tariffs for domestic users currently used (2014 April) by the CEB [8].
Monthly
Consumption
[kWh]
Unit Charge
Fuel Adjustment Fixed
Charge
Charge
[Rs./kWh]
[%]
[Rs./Month]
0-60
61-90
10.00
12.00
10
90.00
91-120
121-180
>180
26.50
30.50
42.00
40
40
40
315.00
315.00
420.00
2.2 Solar resources in Sri Lanka
A quantitative knowledge of the distribution and the extent of the solar resources are
necessary in order to estimate and make appropriative decisions regarding the applications
of solar power. Either to properly size a new connection to meet the current load, or to
investigate and analyse the economic benefits of an investment, the need of a solar data is
of high importance. A study performed by the U.S. government-owned National
Renewable Energy Laboratory (NREL) show that ample solar resources exists throughout
the year for almost all locations in Sri Lanka [10]. In figure 2.2 an estimated solar resource
map shows the monthly and annual average radiation on a flat plate tilted at latitude.
10
A Feasibility Study of Increasing Small Scale Solar Power in Sri Lanka
Figure 2.2 Annual average radiation on a flat plate tilted at latitude in Sri Lanka. Public domain,
reprinted with permission [10].
The resource map shows that the solar radiation varies from 4.5-6.0 kWh/m2/day. The
highest amount of solar radiation can be found in the south eastern as well as the northern
parts of the country. The study shows furthermore that the variability in global horizontal
solar resources is relatively small across the country and varies spatially about 20% to 30%
during different seasons [10]. Thus, it exist a substantial potential for generating solar
energy and especially in the dry zones of Sri Lanka. During the southwest monsoon the
highest solar resources occur in the north-eastern parts of Sri Lanka, while during the
northeast monsoon the southern and western parts receive the highest resources [10].
11
A Feasibility Study of Increasing Small Scale Solar Power in Sri Lanka
3 Photovoltaic cells
In the following section the technology behind PV cells is briefly investigated. The major
part is examining the amount of possible electric generation and how it is limited by
various factors. The calculation of incident radiation on an inclined surface is thoroughly
examined. The variation of the output of a PV system is furthermore examined and the
effect of a varying energy source is briefly discussed.
3.1 Photovoltaic technology
PV cells are made of semiconductors and resemble a lot with other electronic devices such
as diodes and transistors [11]. Crystalline silicon cell hold the majority part of the market
(<80%) and the technology has become well established. The efficiency of a PV cell is
depending on the type of material and the design but is generally increasing for all types of
cells. The efficiency of crystalline silicon is currently approaching 18% whereas the cheaper
thin-film solar cell has an efficiency of about 8-10%. There are efforts to improve the
efficiencies and in laboratory the efficiency of the most advanced cells are approaching
30% [12]. Due to no moving parts, the mechanical wear is very small and consequently
does PV systems has low required maintenance and a long estimated lifetime (25 years or
more) [11]. Most of the grid connected PV generation is installed as either utility owned
units or as residential rooftops PV systems.
During operation, no emissions and low upkeep makes PV almost non-malignant [11].
Silicon is a stable material and imposes no threats on the environment at the end of the
technical life-cycle. However, the production of the cells is fairly energy intensive and the
environmental hazards are similar to those encountered in the microelectronics industry.
The energy required for the production is often assumed to be regenerated within 3-6 years
depending on the material and properties of the location. As the technology has developed
and both the prices of modules has decreased and the efficiency increased, PV systems are
starting to compete with more conventional methods of creating electricity. PV has
furthermore the advantage that the fuel source is free and it is available almost worldwide
without any need for fuel or power distribution infrastructure [12].
A PV cell consists most commonly of two assembled layers of silicon, much like a PN
(positive-negative) junction diode [11]. By doping two intrinsic layers of silicon, commonly
by boron and phosphorus, two layers with different electric properties are created. The Ndoped layer is doped negatively and has a surplus of free electrons, while the P-doped layer
is doped positively and has free openings for electrons. The two layers are then assembled
and a diffusion of electrons move from the N-doped layer to the P-doped layer creating a
12
A Feasibility Study of Increasing Small Scale Solar Power in Sri Lanka
junction, or an insulating barrier between the layers. As the electrons moved to the Pdoped side, a strong electric field has been created with a surplus of negative loaded
particles at the P-doped side and respectively a surplus of positively loaded particles at the
N-doped side [11].
When light, in the form of photons, reaches the photovoltaic cell, the energy breaks the
structure of the electron-hole pairs [11]. Each photon with sufficient energy will free one
electron and thus creating an electron hole. The electric field will then send the electrons to
the N-side and the electron holes to the P-side. Because of the junction, it is impossible for
the electrons and electron holes to reverse the displacement. However, if an external circuit
is connected, the electrons will start flowing through the path creating electricity as they go.
Consequently, the electron flow generates a current and the electric field creates a voltage,
which together is the definition of a source of electric power. The dynamics of a PV cell is
illustrated in figure 3.1.
Figure 3.1 Model of how electric power is generated from a solar cell. Incident radiation breaks the
structure of the electron-hole pairs in the solar cell thus creating a flow of electrons through the
circuit.
A photovoltaic system is consisting of several solar cell modules, together creating the
photovoltaic array [12]. The solar cells are interconnected in the modules and these are in
turn connected in series in order to increase the voltage output, see figure 3.2. A PV system
generates direct current (DC) and in order to generate alternating current (AC) it is thus
13
A Feasibility Study of Increasing Small Scale Solar Power in Sri Lanka
necessary to connect a DC/AC converter (inverter). The converter can be designed as
either single-phase or three-phase depending on the conditions.
Figure 3.2 A photovoltaic system installed by JLanka Technologies. Several solar cell modules
assembled together making up the photovoltaic array. Reprinted with permission [14].
The possibilities in output of solar energy vary considerably from location to location. To
maximize the collection of solar radiation the orientation of the array has to be optimized
[13]. The PV system can be installed either as a fixed system or with a tracking array
system. The tracking array system is consisting of moving support frames that follow the
suns movement throughout the day. The tracking device could either be designed as a oneor two-axis system. The two-axis system is maximizing the gained energy as the array is
constantly kept perpendicular to the radiation of the sun. A one-axis system will generate
about 15 – 20% more energy than the same size PV system with a fixed axis [12]. A twoaxis system will generate even more, about 25 – 33% compared to a fixed axis PV system.
The tracking array systems increases thus the energy output of a PV system, but has instead
higher installation and maintenance costs. The theory of calculating the radiation on an
inclined surface is described in section 3.2.1.
14
A Feasibility Study of Increasing Small Scale Solar Power in Sri Lanka
3.2 Calculating possible generation of a PV system
Calculating the output of a PV system in fully detail takes a meticulous effort due to a
complicated model in need of highly specified geometrical and metrological data to be
precise [11]. However, the results are often fairly easy to estimate and provide mostly a
good approximation. In the following section a model of calculating the radiation from the
sun to an inclined surface (e.g. a PV system on a rooftop) is presented.
There is a huge amount of factors affecting the possible power output, such as reflection
losses, equipment losses, and conversion losses [11]. However, these energy losses are
dependent on the type of material and technology and behave more or less the same on all
PV-systems. For a more detailed account regarding the technical factors surrounding the
efficiency losses of a PV cell, the reader is referred to [11]. In section 5.1, a calculation of
the economics on a standardized PV-system in Sri Lanka is presented. The calculations are
founded on the weather and radiation data for a location in Sri Lanka and are calculated in
a comparable way as presented in section 3.2.1.
3.2.1 Radiation on an inclined surface
The PV cell is creating electricity by using the energy from the sun. The amount of
radiation reaching the solar cell is in direct relation to the amount of possible generation.
To a good approximation, the energy flux on a unit area perpendicular to the sun beam
outside Earth's atmosphere is known as the solar constant, [11]:
(1a)
The total energy flux reaching the Earth is dependent on the disc area, i.e.
where is
the radius of the Earth. The average flux is then obtained by dividing the total energy flux
with the total surface area of the Earth,
, resulting in a net average flux incident of:
(1b)
The radiation is then scattered and absorbed by the atmosphere resulting in a total incident
radiation ( ) depending on both direct beam radiation, diffuse radiation, and groundreflected radiation. The amount of radiation that reaches the ground is not constant, but
depends on a number of different factors such as: regular daily and yearly variations,
irregular variations caused by climatic conditions, and the composition of the atmosphere
[11]. The incident radiation on a PV cell is illustrated in figure 3.3.
15
A Feasibility Study of Increasing Small Scale Solar Power in Sri Lanka
Figure 3.3 The different types of incident radiation on a photovoltaic cell.
The total incident radiation on an inclined surface, e.g. a rooftop, is as previously stated
depending on both the direct beam radiation ( ), the diffuse radiation ( ), and the
ground-reflected radiation ( ). In order to calculate the total radiation on an inclined
surface, a number or angles and correlations are first needed to be defined. To ensure
understanding of the following calculations, refer to the list below. The following
calculations are derived from the studies of [11] and [13].

The solar declination ( ): a measure of the angle between the equator and radiation
from the Sun, where is the number of the day in the year. Approximately given
by (in degrees):
.
/
(2)

The geographical latitude of the location, :

The tilt of the surface with respect to the horizontal, :

The solar azimuth,
: a measure of the tilted surface orientation relative south and
is given by:
16
A Feasibility Study of Increasing Small Scale Solar Power in Sri Lanka

The sunrise hour angle,
: the angular displacement of the sun relative to the local
meridian, afternoon positive and zero at noon:
(

.
)
(3)
The radiation received over one day by a horizontally area outside Earth's
atmosphere
*
:
(
)+ (
)
(4)
3.2.1.1 Direct beam radiation
The direct beam radiation is the radiation from the sun that directly, without being
scattered or reflected, reaches the PV cell. With knowledge of the above stated parameters,
it is possible to calculate the angle of incidence ( ). The angle of incidence is the angle
between the radiation on the surface to the normal of the inclined surface and is calculated
according to equation (5):
…
...
(5)
…
…
The final step is to calculate the zenith angle ( ) which represents the radiation on a
horizontal plane, i.e. when = 0. By assuming that = 0 it is possible to simplify equation
(5) to the followed:
(6)
Defining the zenith angle, it is now possible to express the beam radiation on an inclined
plane as followed:
(7)
Where
= Beam radiation on a plane perpendicular to the radiation
17
A Feasibility Study of Increasing Small Scale Solar Power in Sri Lanka
The correlation is valid under the condition that both
the sun is above the horizon.
and
is
0, that is when
3.2.1.2 Diffuse radiation
Diffuse radiation is the radiation that reaches the PV cell after first being scattered by the
molecules in the atmosphere. Various formulas are available for the calculation of diffuse
radiation. In the following report a simpler, although sufficiently accurate, method will be
adopted according to [11]. The diffuse radiation on a horizontally surface ( ) can be
expressed as:
(8)
Where
= Global radiation on a horizontal surface
= Clearness index, calculated according to equation (9):
(9)
The diffuse radiation on an inclined surface (
as:
(
) is then calculated from the above results
)
(10)
3.2.1.3 Ground-reflected radiation
The amount of ground-reflected radiation is dependent on the albedo (measure of the
reflexivity of the landscape). The ground-reflected radiation on an inclined surface (
) is
calculated in a manner closely related to the diffuse radiation. The irradiance reflected is
generally small and thus the following simple model is proved sufficient:
(
)
(11)
Where
= Constant albedo. Typically ranging from 0.2 for dry bare ground and up to 0.5-0.8
for snow [11].
18
A Feasibility Study of Increasing Small Scale Solar Power in Sri Lanka
3.2.2 Temperature effects
The output of a PV system is not only dependent on the amount of radiation but also
several other factors [15]. The surrounding temperature is affecting the efficiency of the PV
cell and an increasing temperature generally reduces the output. An increased temperature
has two effects on the PV cell: the open-circuit voltage decreases and the short-circuit
current increases. The effect on the possible power output is calculated by evaluating the
effects separately [15]. If the reference temperature is changed by , the new short-circuit
current and open-circuit voltage are given by equation (12) and equation (13):
(
)
(
(12)
)
(13)
Where
= Short-circuit current
= Open-circuit voltage
and
= Corresponding temperature coefficients
The operating current and voltage change approximately as the short-circuit current and
open-circuit voltage does. Hence, it is possible to define the new power output as:
(
)
(
)
(14)
By simplifying and ignoring a small term in equation (14), the expression can be simplified
into the followed:
,
Where
(
and
)
-
(15)
is generally about 20µ units/°C respectively 5m units/°C, for a typical
single-crystal silicon cell [15]. Thus, the power in terms of change in temperature is given
by the following equation:
,
(
)
-
,
-
(16)
As equation (16) indicates, an increase of the temperature by one degree Celsius decreases
the power output by 0.5%. The decrease in power output is thus resulting from the fact
that the increase of short-circuit current is much less than the decrease of the open-circuit
voltage.
19
A Feasibility Study of Increasing Small Scale Solar Power in Sri Lanka
3.3 Variability of output
The variability of output is an important feature for most renewable energy resources and
is often considered as a major obstacle for increased penetration of renewable energy. The
variability differs typically due to two major factors [13]:

Seasonal and diurnal fluctuations: Variability due to the earth’s movement
around the sun and its own axis. As seen in section 3.2, the motion can be
calculated accurately and is also highly dependent on location. At regions around
the equator, the sun sets and rises at approximately the same time all year, whilst at
higher latitudes the days are considerably longer in the summer than during the
winter period.

Weather conditions: As seen in section 3.2, part of the sunlight is absorbed and
reflected before reaching the surface of the Earth. On a clear day, about 20% of all
sunlight reaching the surface may be diffuse, while on a cloudy day almost all
radiation may be diffuse. Fast moving clouds may affect radiation, and thus output,
to vary from full generation to more or less zero output. The output of a PV
module during a sunny day is shown in figure 3.4.
Figure 3.4 Output of a PV module on a sunny day with low variability of output, situated in
Colombo, Sri Lanka. Reprinted with permission [14].
An important factor concerning variability, and perhaps especially for PV-systems, is the
smoothing effect of several geographically outspread points of generation (POG). Many
20
A Feasibility Study of Increasing Small Scale Solar Power in Sri Lanka
studies have come to the same conclusion; that geographic dispersion decreases
fluctuations of output and increases availability and power quality [13]. The impact of
variability affects both the owner of the system and the total production and load balance
of the grid as a whole. For the owner, unfavourable variability may cause mismatch of the
local demand and production. In areas with high penetration of PV-systems, the variability
may affect the total production-load balance and thus bring implications for the voltage
management of the grid.
In order to deal with variability of production, the power system has to be dimensioned to
handle such changes in the demand and production balance [13]. Some kind of generation
reserves has to be available and automatically and instantaneously activated when the
system frequency of the grid drops due to mismatch in generation and production. An
important consideration of newly installed capacity is whether the generation is available
when it is needed the most. Further discussion concerning this topic is to be found in
section 4.5.
21
A Feasibility Study of Increasing Small Scale Solar Power in Sri Lanka
4 Distributed generation and photovoltaic power
Distributed small-scale PV systems, installed at end users such as residential buildings or
companies, are examples of the general concept called distributed generation (DG) [13].
DG is characterised by a fairly small amount of power generation, located within the
distribution grid. In the last few years, due to government policies, deregulation, and
environmental aspects, there has been a rapid growth of DG [16]. The distributed
generation could be in the form of for example wind power plants, small-scale hydro plants
or as in the scope of this report; photovoltaic systems. DG is characterised by a relatively
small power output; usually far below 40 kW and PV systems are often dimensioned for 1
kW up to 10 kW.
PV systems are highly suitable for small scale DG, and have major advantages compared to
other technologies [13]. Most conventional technologies of generating power have
significant economy of scale, making small scale DG not very cost-effective. PV has
instead almost no economy of scale, and thus making a centralised power plant just barely
more cost effective. PV are also perfectly suited for mounting at walls or rooftops, and is
thus not in need for land preparation nor affecting the landscape.
The implementation of PV-DG is unfortunately not all uncomplicated [12]. The traditional
way of distributing electricity is to generate power at large centralized power plants and
then transport the power via the transmission grid to the distribution network and the
consumer. DG is instead located at the consumer, closer to the load and thus reducing
both line losses as well as in many cases increasing the electric power quality. However, the
distribution grid is traditionally dimensioned and constructed for being a so called passive
grid, where the power flows in only one direction [12]. If the amount of generated power
increases above the consumption, the electricity starts to flow at the reverse direction. An
active grid where the power is not only flowing in one direction has increased requirements
on the structure of the grid. With an active grid not only the power consumed at the load
need to be taken into account, but also the amount of generation.
One of the few studies conducted of areas with high penetration of PV is presented in the
PV-UPSCALE project [17]. The project studied locations with a PV penetration level
ranging from 110% to 33% of the rated transformer power. The results showed that even
systems with a high ratio of generation capacity, i.e. 80% and higher, did not in general
deteriorate the quality of the grid. Limits to the amount of connected power were not set
by the voltage rise due to excess generation, but it was rather the power rating of the
distribution transformer that limited the amount of reversed power. Furthermore, the
power quality met all requirements of the European standard and the power quality was
22
A Feasibility Study of Increasing Small Scale Solar Power in Sri Lanka
only affected by an increasing of the voltage levels at the end of the feeders. In one case,
the voltage harmonics exceeded permitted values. However, the increased harmonics could
easily be avoided by selecting inverters with a low input capacitance. A rule of thumb
developed by the report, is that no trouble should be caused if the connected power is
limited to 70% of the rated power of the feeding transformer [17].
The following sections will try to investigate how and in which manner DG affects the
above mentioned problems and possibilities. The major focus is to establish a method of
calculating the change of voltage levels and line losses due to the implementation of DG.
Furthermore, the concepts of electric power quality and peak demand reduction are
examined and discussed.
4.1 Distributed generation and grid protection
DG can have a significant impact on power flows that occur on transmission and
distribution grids [12]. The degree of the impact will depend not only on the size of the
connected DG, but also the location, the existing load at the POG, and at what time the
DG system is operating. An increased share of DG within a distribution grid might cause
the power to flow from low into medium voltage grids. To ensure a safe operation and
high protection, different protection methods may be required at the event of a reversed
power flow. Furthermore, to ensure a high reliability and availability the protection system
need to be selective. However, since the fault current is not always originating from one
location, the detection and the selectivity becomes a far more complicated matter than at
the conventional power flow.
The protection problems are illustrated by figure 4.1 below. If a short-circuit occurs at S1
or S2 the short circuit is supplied by not only the main grid, but also the connected
generators on this feeder as well as other DG systems in nearby feeders [18]. If the shortcircuit current is mainly generated from the generators G1 and G2 the current through the
circuit breaker/fuse of B1 might be too low to be able to detect the short-circuit in the
feeder.
23
A Feasibility Study of Increasing Small Scale Solar Power in Sri Lanka
Furthermore, if the contribution to the short-circuit current from other adjacent generators
is significant, for example from G3, these healthy feeders might be disconnected before the
faulty feeder and thus remove the selectivity.
Figure 4.1 Grid with high DG penetration and possible safety problems.
4.1.1 Islanding
Another possible problem connected to grid protection is islanding. Islanding refers to event
when a DG continues to generate power to an area even though the power from the
electrical grid is no longer present [19]. In the event of islanding, a portion of the grid is
thus electrically separated from the main system. The phenomenon is possibly dangerous
for both utility workers and equipment, who may not realize that the circuit is still
powered. The most important aspects concerning islanding are [12]:

Voltage problems.

Public and utility worker safety.

Damage to equipment due to out-of-phase reclosures.

High overvoltage to equipment caused by neutral shifts or ferroresonance.
24
A Feasibility Study of Increasing Small Scale Solar Power in Sri Lanka
For this reason, the DG systems need to automatically disconnect generation to ensure
person and equipment safety even when faults do occurs. The above stated problems
require a more "active" and advanced protection system with some sort of communication
in order to keep the safety at a high level in the future. The most common way today of
preventing islanding is to use frequency and voltage relays on the DG-system, which are
programmed to trip if the frequency or the voltage varies over a predetermined limit [12].
This method of protection, known as passive protection, prevents islanding in most cases.
4.1.2 Voltage regulation with DG
An increased amount of DG could affect the voltage regulation of the transmission and
distribution grid [12]. As an increased amount DG affects voltage drops, the voltage levels
along the feeder also change. In the case of a DG connected in close proximity to a
distribution transformer, the increased voltage due to the generation may actually cause low
voltage at the end of the feeder. As the voltage regulation is often based on the amount of
the line current, a generator just downstream of the transformer will decrease the observed
current on the feeder. The perceived reduced load will cause the regulator decrease the
voltage boost of the grid, hence leading to lower voltage further down the feeder [12].
Reverse power flows may also cause high voltages [12]. During the event of light load for a
location with high primary voltage, the increased voltage due to DG may be too high
according to regulations. The aspects of increased voltage levels are discussed in section 4.3
and analysed further in appendix A. Changes in the output of PV-DG is common and if
the output varies enough, it may change the voltage levels enough to cause a regulator tap
change or similar. Similarly, a DG with voltage feedback regulation may interact
undesirably with utility regulation equipment. In these events, adverse cycling of regulation
devices may occur with negatively impacts on power quality as an effect.
4.2 Electric power quality
Electric power quality is generally defined as the quality of the voltage wave shape and its
frequency, the current wave shape, the voltage regulations and the levels of impulses, noise,
and the absence of momentary outages [19]. Photovoltaic cells and distributed generation
both affects the quality of the electric power in varying way - and perhaps surprisingly not
always in a negative way. The following part will try to briefly discuss how electric power
quality is affected by DG and especially PV-DG systems, and how the problem can be
dealt with.
25
A Feasibility Study of Increasing Small Scale Solar Power in Sri Lanka
4.2.1 System frequency
A stable system frequency is of high importance to the operation of many industrial and
household applications and need to be kept within very narrow margins [19]. The number
and size of connected DG are likely to affect the system frequency and it needs to be taken
into consideration by the grid operator when connecting and installing new DG systems
[19]. To protect the grid, inverters connected to solar cells are configured to shut down if
the measured frequency increases above a certain value. If the same problem occurs in a
grid with high penetration of DG, problems might occur if these shut down
simultaneously.
4.2.2 Fast voltage fluctuations and flicker
Irregular solar radiation, for example due to moving clouds, produces fast power output
fluctuations in PV-DG. The irregularities may affect the voltage levels and cause
fluctuations in the grid, and is especially significant in weak residential and rural grids [19].
These fast voltage fluctuations is the cause of light flicker. Flicker causes disturbing
variations of the brightness or colour of the electric powered lightning.
The flicker is dependent on the frequency of the voltage variation, the amplitude, and the
shape of the waveform. Even a small change in voltage can cause disturbing and noticeable
light flicker. Flicker is generally worse closer to the fluctuating DG and will increase if the
DG is large compared to the power load of the feeder [12]. The disturbances could either
be diminished by decreasing the power output of the source or by improving the grid.
4.2.3 Harmonic distortion
Harmonic distortion is not to be confused with transients as these dissipates within a few
cycles while harmonics take place in a steady state and are integer multiplies of the
fundamental frequency [19]. Harmonic distortion is created by non-linear devices in the
distribution system, such as the DC-to-AC converter connected to a PV-system. All of
these devices produce currents that are not perfectly sinusoidal and may therefore
contribute to higher harmonics in the grid.
Figure 4.2 shows how a perfect sinusoidal voltage curve affected by harmonics is resulting
in a new far from sinusoidal curve. Single-phase line commutated inverters are used when
integrating a PV-DG system with the grid. These inverters create low-order odd-numbered
harmonics, beginning with the third harmonic (see figure 4.2) [12].
26
A Feasibility Study of Increasing Small Scale Solar Power in Sri Lanka
Figure 4.2 Harmonics of different frequencies resulting in a distorted sinusoidal curve.
Harmonics causes higher power losses in transformers and feeders and other loads as e.g.
electric motors [12]. They also cause resonance in power systems, interference in
communication circuits, and may also cause abnormal operation of control equipment.
Fortunately, it is possible to reduce these undesirable harmonics if they do cause too much
harm. There are some traditional solutions to reduce the harmonic distortion, where the
most common is to use some sort of filters [12]. The most traditional solution is using a so
called shunt filter bank. The filter would consist of a series inductor and capacitor tuned in
to the harmonic that is wished to reduced, such as the 3rd harmonic. Using active filters are
a newer method that consists of fast-switching power electronics that creates and inject
harmonics to cancel out the source harmonic.
4.3 Voltage level and unbalance
For radial power systems without the impact of DG, the voltage regulation practice is
based on a single source of power and that power taking one path from the substation to
the loads [12]. That one-way power flow leads to the assumption that voltage will drop on
the feeder as the distance from the substation increases (with the exception of installed
capacitor banks in the grid). The introduction of DG may change the classical voltage
decrease along the feeder and the following section will try to investigate in what manner
the voltage variations are calculated.
27
A Feasibility Study of Increasing Small Scale Solar Power in Sri Lanka
Figure 4.3 shows the event of a feeder with loads connected at the different nodes. The
voltage drop is illustrated in the event of no DG, DG at node 3, and DG at node 4 and 6.
The figure shows that the implementation of DG at a feeder may result in less significant
voltage drop along the feeder. The case of DG installed at node 4 and 6 results even in an
increased voltage rate along these nodes.
1,02
Voltage level [p.u]
1,01
1
0,99
0,98
0,97
1
2
3
4
5
6
Node number
No DG
DG at node 3
DG at node 4 and 6
Figure 4.3 Voltage drop along a feeder in the cases with or without distributed generation.
The effect that a high amount of DG has on the power quality is hard to evaluate. Slow
voltage variations occur with varying load or with varying generation, for example different
load situations during day and night or changes in the amount of generation of a PV system
during a cloudy day. The voltage level will vary thus not only due to the size of the DG but
also due to the present load situation. Two typical cases are of most interest, maximal load
and no generation and no load and full generation.
It is often the slow voltage variations that restrain the size of the production and that is
examined first if a new installation of DG is possible [20]. The strength of a distribution
network is often measured by the size of transformers, the cross-sectional area of cables
and corresponding factors in overlying grids [20]. A related way of defining the strength of
a distribution grid is to measure the change in voltage levels that occurs by connecting a
load to the grid. In table 4.1, the recommended maximal voltage variations in Sweden are
presented. No corresponding regulations for Sri Lanka have been found.
28
A Feasibility Study of Increasing Small Scale Solar Power in Sri Lanka
Table 4.1 Allowed voltage variation for DG at different points of connection [20].
Point of connection
Allowed voltage variation
[%]
Connection point to customer
5%
Point of common connection to other customers
3%
When connecting a DG system it is important to verify that the grid will not suffer from a
deteriorated electric quality and assure that the limits in table 4.1 will not be exceeded.
Connections to DG systems are generally quite short (less than 80 kilometres) and a simple
model is often sufficient. Figure 4.4 is demonstrating a short transmission line with only
resistance and reactance included [21]. In the case of longer transmission lines the so called
π-circuit is used instead, and consideration to the line capacitance is required.
Figure 4.4 Electric model of a short transmission line.
The terminology used in figure 4.4 is described below:
= Voltage level at the beginning of the line
= Voltage level at the load point
= Complex impedance of the line
= Resistance of the line
= Complex reactance of the line
To calculate the highest voltage variations due to DG, the case of no load and full
generation is examined. To examine the highest voltage variations it is necessary to first
establish a model for calculating voltage drop along a short transmission line. The voltage
29
A Feasibility Study of Increasing Small Scale Solar Power in Sri Lanka
drop along the line is equal to the line current multiplied with the equivalent impedance of
the system [22]:
(
)
(17a)
Where
= Voltage drop along the line
= Line current
It is possible to calculate the exact voltage drop along the line. However, by neglecting a
small term, the equation can be simplified significantly. Figure 4.5 demonstrates the voltage
drop along a short transmission line. The exact voltage drop is as previously stated
calculated as (
) and is found as the difference of
and
in the figure.
Figure 4.5 Phasor diagram demonstrating the voltage drop over a short transmission line.
It is possible to divide the voltage drop phasor into different components, each
contributing to the voltage drop [22]. By using common trigonometric rules, the different
components of the voltage drop can be stated according to equation (17b):
(
)
(
)
(17b)
Where
= Phase angle
The second term in equation (17b) is generally very small, and it is possible to neglect its
impact on the amount of voltage drop. The difference, or the error, by neglecting this term
is visualised in figure 4.5. According to the figure, the actual voltage drop will thus only be
30
A Feasibility Study of Increasing Small Scale Solar Power in Sri Lanka
marginally larger than the calculated voltage drop. By neglecting this second term, the
voltage drop can be simplified into equation (17c):
(
)
(17c)
Furthermore, by rewriting the equation in terms of power instead of current, it is possible
to simplify equation (17c) into the following:
(
) (
)
(17d)
The voltage drop expressed in percentage is achieved by dividing equation (17d) with
(
) (
)
:
(17e)
However, as the voltage drop is generally small on distribution grids it is possible to assume
that
. When calculating voltage variations, this is a safe method as the voltage
variations then will always be higher than in reality [22]. The grid will thus be slightly overdimensioned, but it is always possible to calculate the possible voltage increase with the
more detailed method. This assumption allows equation (17e) to be further simplified and
expressed in percentage as followed:
(
) (
)
(18)
If the calculated voltage rise could be a problem there are several options to reduce the
impact of the DG. The perhaps simplest option, but sometimes the most undesirable, is to
limit the size of the DG. It may also be necessary to relocate the DG or it may require
improvement of the feeder. Another more unconventional method is to limit the operation
of the generator to peak periods of the day. This is however not desirable for the investor
of the DG as the benefits of the production will decrease. If the probability of overvoltage
is very low it could also be possible to rely on the overvoltage relay to disconnect the DG
during high voltages [12].
4.3.1 Single-phase/three-phase connections
The maximal possible output of the connected PV system is depending on whether the PV
system is connected to the grid as single-phase or as a three-phase system. In the case of a
single-phase connection, the total loop impedance of the connecting cable is calculated
twice for both R and X, since the line current in the neutral-line is the same as in the phaseline. In the case of a symmetric three-phase connection it is possible to assume that no
31
A Feasibility Study of Increasing Small Scale Solar Power in Sri Lanka
current will flow in the neutral-line. The total impedance is thus calculated only once for
the connected cable [20].
The followed is a simplified method of calculating the total impedance of the line. The
absolute values are summarized for each part and the calculated impedance will always
exceed the actual value, thus guaranteeing that the PV system will not be over
dimensioned. In the case of uncertain border line problems, the total impedance can be
calculated with the more detailed method and the complex impedance can be summarized
for each part. According to equation (19) and equation (20) the major difference between a
single-phase and a three-phase connection is what kind of impedance that is taken into
account [23].
(19)
(20)
Where
= Impedance of the feeding grid
= Short-circuit impedance for the feeding transformer
and
= Phase- and neutral impedance.
is often very small in comparison of the other impedances and is therefore often
neglected. The short-circuit impedance of the feeding transformer can be calculated
according to the followed equation [23]:
(21)
Where
= Rated voltage of the transformer
= Rated power of the transformer
= Relative short-circuit impedance of the transformer
The loop impedances of distribution transformers used in the case of a single-phase
connection are presented in table D1 in appendix D. Generally, unless the feeder from the
transformer is very short, the impedance will be approximately twice as high for a singlephase connection compared to a three phase connection. The impedance of the
transformer is usually far less than the impedance of the feeder and the feeding grid is as
previously stated often negligible [23].
32
A Feasibility Study of Increasing Small Scale Solar Power in Sri Lanka
4.4 Line loss reductions
With suitable location and output levels, the implementation of DG may reduce energy
losses on the overall transmission and distribution grid [12]. As the DG is located close to
the load the requirement of power from overlying grids decreases and thus also the line
losses. As mentioned in section 2.1, the line losses in Sri Lanka were in 2011 estimated to
11.7%. It is though important to note that 11.7% is the national average and losses on
heavily loaded distribution systems can be significantly higher, especially during periods of
peak demand.
The following part will try to briefly investigate how line loss reduction is calculated and
what factors that might affect the losses. The model will be restrained to only investigate
loss reductions at the low voltage grid. The same model as in section 4.3 (see figure 4.4) is
used for calculating line loss reductions.
The total complex power is defined as
the load ( ) can be stated as followed:
√
and hence the current absorbed by
(22)
√
Where
= Active power at load point
= Reactive power at load point
= Phase-to-phase voltage
The total active line losses ( ) are calculated according to equation (23) [22]:
(23)
Where
= Total resistance on the line
The amount of line losses depend on size of the current flow and the line resistance. By
installing a DG system, the current, and thus the line losses, could decrease. Combining
equation (22) with equation (23) gives us the following expression of total active line losses
[22]:
.
√
/
. /
. /
33
. /
(24)
A Feasibility Study of Increasing Small Scale Solar Power in Sri Lanka
Now, consider the case where a DG is introduced at the point of load. If the reactive
power is considered constant, the change in line losses can be expressed as equation 24:
.
/
. /
(25)
The reduction of line losses due to DG is dependent on a number of factors. Heavily
loaded lines are ideal targets for an increased amount of PV-DG since the line losses can be
reduced significantly with a relatively small amount of DG [12]. Due to the non-linear
relationship between the phase current and the line losses, the relative gain of reduced line
losses is larger if DG is installed at locations with a high load. However, DG does not
always reduce the line losses. If dimensioned too large, the DG systems might even increase
the line losses. This indicates that the rating and location of DG need to be taken into
account prior to the installation.
It is possible to applying DG to optimally reduce line losses. However, in reality it is hard
to implement due to complex feeder branching and uncontrollable demand cycles [12].
Furthermore, the main purpose of installing DG is commonly not to reduce the amount of
line losses, but rather as a way to reduce the dependency of electricity or by environmental
concerns. The effects that DG has on line losses are further examined and illustrated in the
section concerning simulation and calculations.
4.5 Peak demand reduction and diversified demand
Diversified demand is the name of the phenomena when the maximal demanded load from
a diversified number of customers is less than the sum of their total maximal demand [22].
Diversified demand depends on the fact that the customers demand curve differs and that
the maximal demand does not occur at the exact same time. The result is that feeders and
transformers can be dimensioned for a lesser maximal power than the sum of the
maximum power of the connect loads.
Concerning a load that has installed its own DG, it is instead most favourable if the
generation and the consumption curves overlap, i.e. that the generation coincides with the
consumption. When discussing DG, this feature is generally denoted as peak demand
reduction [12]. In many countries, the demand for electricity peaks at more or less the same
time every day. If the DG could be installed in such manner to produce maximum output
at the times of maximal demand, not only would it reduce line losses significantly but it
would also reduce the need of grid improvement due to lower maximal demand.
34
A Feasibility Study of Increasing Small Scale Solar Power in Sri Lanka
Figure 4.6 illustrates the event of a peak demand reduction, where a well dimensioned DG
reduces the peak demand during day time. As is discussed in section 5.4, the line losses of a
feeder are proportional to the quadrant of the line current. Thus, reducing the line current
by 50% would decrease the losses by 75%.
6
5
Load
Power [kW]
4
Output of DG
Demand with DG
3
2
1
0
0
4
8
12
16
20
24
Time at day
Load
Output of DG
Demand with DG
Figure 4.6 Distributed generation contributing to reduce peak demand at consumer.
In areas with warm weather, energy consumption increases with high temperature due to
appliances such as air conditioning and other cooling systems [24]. The peak demand tends
to be at its highest during day time and the afternoon and evening, while the demand is low
during the night and morning. In Sri Lanka, the highest demand for electricity is during the
dry periods when the tourism is at its highest. Conveniently, it is during the dry seasons
that the output of PV-DG is at its highest due to sunny days and lesser cloud cover. An
increase of PV-DG might therefore not only increase the energy capacity but also reduce
the peak demand at end-users.
Reducing peak energy demand has several benefits. Not only do line losses decrease, but
also the need for improving the grid decreases. Furthermore, the energy used at peak
demand is commonly more expensive than out-of-peak energy. Peak energy is often
needed to be generated by inefficient and expensive energy sources, such as gas or diesel
turbines [24].
35
A Feasibility Study of Increasing Small Scale Solar Power in Sri Lanka
5 Simulations and calculations
In order to analyse the effect of an increased amount of PV-DG, a series of calculations
and simulations are performed. The calculations and simulations are divided mainly into
two separate parts: one examining the economic feasibility of a new PV-DG system in Sri
Lanka and the other investigating the effects and the technical feasibility of increasing the
amount of PV-DG. The used methods and applications are briefly presented in context to
each part. The presentation of data and calculations are performed in the appendixes and
the results are presented in section 6.
5.1 Economic feasibility
The economic feasibility is examined by using a life-cycle cost analysis (LCC analysis) for a
standardized PV system in Sri Lanka. Calculating the output of PV system is made with the
aid of the NREL developed program PVWatts [4]. NREL is the U.S Department of
Energy’s primary national laboratory for renewable energy and energy efficiency research
and development, and has performed several studies concerning solar resources in Sri
Lanka.
PVWatts is an application that is based on a few inputs calculates the electricity production
of a PV system. Using an hour-by-hour simulation with solar data in Sri Lanka, the
application calculates the output for a standardized PV system with crystalline modules
over a period of one year. By a similar model as described in section 3.2, the PVWatts uses
the hourly incident radiation on an inclined surface to estimate the output. Furthermore,
the model is taking into account cell temperature losses, reflection losses, and inverter
losses. For a more detailed description of the PVWatts application, the reader is referred to
the PVWatts technical manual [25]. Six inputs are needed by the model and the following
are used for the calculations:
Table 5.1 Input data for the PVWatt application.
Solar resource data:
Anuradhapura
DC System Size:
2.5 kW
Array Type:
Fixed
DC-to-Ac Derate Factor:
0.8
Tilt:
8
Azimuth :
180
36
A Feasibility Study of Increasing Small Scale Solar Power in Sri Lanka
The size of the PV system is chosen to cover the demand of electricity for a high electricity
consumer in Sri Lanka. For simplicity, the array type is assumed to be fixed and the DC-toAC derate factor is chosen to the standard value of 0.8 [25]. The tilt angle is the angle from
the array of the PV system to the horizon (90° = vertical, 0° = horizontal). The tilt is
chosen to 8 degrees as this is the latitude of the PV system location. Note that setting the
tilt equal to the latitude does not always maximise the output as lower/higher tilt favour
generation during summer/winter [4]. The standard value of the azimuth angle is chosen to
180° (south-facing), as this normally maximizes energy production for locations in the
northern hemisphere.
The financial profit of an investment in a PV system is then compared to the event of no
investment. Every year the investment yields savings in electricity costs. In order to analyse
the actual value of these future savings, all savings are converted to present value with the
method of present discount value. Discounting is used in economic analyses in order to
analyse the present value of future benefits and costs [26]. All future savings are therefore
discounted to reflect its current value as if existed at the time of the investment. The sum
of these discounted savings, plus eventual rest value of the investment at the end of the
time period, represents the present value of the investment.
The choice of discount rate is essential for the outcome of the investment and particularly
in long time-scales [26]. Choosing a proper discount rate is not an easy task, but it is often
chosen in context of possible profits from other investments or in context to the risk of
the investment. The present value of future savings is calculated according to equation
(26):
(
)
∑
(
)
(26)
Where
= Rest value of the investment at the end of the time period
= Discount rate
= Economic lifetime of the investment
= Annual savings
In the following calculations, the rest value is assumed to be zero after 25 years. Hence, the
first part of equation (26) may be neglected. The annual savings is furthermore assumed to
be constant. Therefore, it is possible to calculate the discount rate factor for each year and
then multiply this with each of the annual savings. The discount rate factor is calculated
37
A Feasibility Study of Increasing Small Scale Solar Power in Sri Lanka
according to equation (27). All discount factors used to discount the annual savings in the
calculations in section 6.1 is found in table A1 in appendix A.
(
)
(27)
Where
= Current year
There are many different methods of measuring the value of an investment. However,
there is no “correct” way and it is common to perform several methods to ensure that the
result is correct [26]. Another more simplified method is the payback or break-even point
method. The payback period refers to the period of time required to regain the cost of an
investment, or to reach a break-even point. The payback period is usually expressed in
years and uses the net savings for each year. The method is often used for its simplicity;
every year the net saving is accumulated until the cumulative cash flow is higher than the
investment.
In the calculations for the LCC analysis, the method of present discount value is used. In
connection to the results, the payback period is also presented. In all calculations, the
effects of inflation are neglected as energy prices can be expected to increase more than
other commodities in Sri Lanka. The following data are used for the LCC analysis:
Table 5.2 Input data for the LCC analysis.
Type of customer:
Residential
Monthly consumption of electricity:
300 kWh
Initial investment1:
1 100 000 Rs.
Discount rate:
5% and 10%
Calculation period:
25 years
Electricity costs:
Varying
Value of investment at end period:
0 Rs.
1
The initial investment of a standardized 2.5 kW PV-system from JLanka Technologies. Including material,
installation and a warranty covering all repair and maintenance cost during 25 years [27].
38
A Feasibility Study of Increasing Small Scale Solar Power in Sri Lanka
The technical lifetime of a typical PV system ranges from 20 to 30 years. It is assumed in
this analysis that the rest value of the investment after 25 years is negligible. The initial
investment is based on data from JLanka Technologies for 2.5 kW standardized PV-system
[29]. The price includes material, installation and a warranty for 25 years covering all repair
and maintenance costs. The difference in electricity costs with or without the investment of
a PV system is analysed with the aid of the CEB Bill Calculator [27]. The LCC analysis is
calculated in Sri Lankan rupee (Rs.), where 1000 Rs. correspond to about 50 Swedish kr.2
[28].
5.2 Technical feasibility
Analysing the technical feasibility of an increased amount of PV-DG is made with the
developed typical case grid in section 5.2.2. The typical case grid is then used to illustrate
the impact of an increased amount of PV-DG in the grid. The typical case grid has a fairly
simple structure with only one point of generation and one point of common coupling
(PCC). Three households (HH1, HH2 & HH3) are assumed to be connected to the feeder
and the DG is assumed to be connected at HH1. The simplicity of the case grid is
intentional and is constructed to ease the comprehension of the effects of distributed
generation. The case grid is not dimensioned to be specific for Sri Lanka and conclusions
are not always possible to make at once. However, the result from the case grid computed
with the information of the grid in Sri Lanka allows a number of conclusions to be made.
Calculations concerning cable data can be found in appendix D together with data for
transmission transformers. The summarized data concerning the typical case grid is found
in table 5.3.
The calculations are based on the theory developed in section 4. Simulations and
calculations are computed only for the effects that an increased amount of PV-DG has on
increased voltage levels and decreased line losses. Other aspects of an increased amount of
PV-DG are discussed in section 4.
5.2.1 Increased voltage level
An increased amount of PV-DG could cause an undesired voltage rise along the feeder.
The highest voltage variations occur as the generation is at a maximum at the same time as
2
Currency information of LKR/SEK for the 24-hour period ending Monday, May 16, 2014 [28]
39
A Feasibility Study of Increasing Small Scale Solar Power in Sri Lanka
the load is at its minimum. The calculations in appendix B consider a worst case scenario
of full generation at HH1 and no load at all at load HH1, HH2 and HH3. The voltage rise
is examined both at the point of connection as well as at the PCC (the wiring closet). To
examine the effects of an increased voltage rise, the following two different scenarios are
studied:


Scenario I: Generation connected as three-phase
Scenario II: Generation connected as single phase
In the calculations, the amount of generation for the two scenarios is increased from zero
output to 5 kW output. In order to find the maximal amount of the DG connected to the
case grid, the results is plotted and presented in section 6.2.1.
5.2.2 Decreased line losses
With a suitable location and output levels, the implementation of DG may reduce the
amount of line losses of the grid. As the DG is located close to the load the requirement of
power from overlying grids decreases and thus also the line losses. However, if the DG
increases more than the load, the line losses may actually increase. The concept is discussed
in section 4.4 and the calculations are presented in appendix C. To examine the effects that
DG has on line losses, the following two scenarios are studied:


Scenario I: Fixed load and varying amount of DG
Scenario II: Fixed amount of DG and a varying amount of load at HH1
40
A Feasibility Study of Increasing Small Scale Solar Power in Sri Lanka
5.2.3 Typical case grid
The typical case grid is used in the calculations
regarding technical feasibility. Figure 5.1 is
illustrating the structure of the grid. Table 5.3
summarizes technical data of the case grid.
Table 5.3 Summarized data for typical case grid.
Transformer:
Rated power:
Rated voltage (
200 kVA
):
Short-circuit impedance (
400 V
): 4%
Short-circuit impedance ( )
32 mΩ
Loop-impedance:
32 mΩ
Feeder data:
4x50 Alus3:
0.641 mΩ/m
4x25 FeAL3:
0.727 mΩ/m
192.3 mΩ
145.4 mΩ
218.1 mΩ
145.4 mΩ
Generation:
Power factor:
=1
Generation:
Varying
Consumption:
Varying
3.
The resistive part in the cable impedance is
dominating, it is possible to neglect X and assume
that Z = R. The chosen cable has the same area in
the return conductor and the loop impedance can
thus be assumed to be twice the value of the cable
impedance.
Figure 5.1 Typical case grid with
three customers and one PV-DG.
41
A Feasibility Study of Increasing Small Scale Solar Power in Sri Lanka
6 Results
In the following section the results from the described calculations in section 5.1-5.2 are
presented. The results are divided into two separate parts; one concerning the economic
feasibility and one concerning the technical feasibility. The results are then summarized and
further discussed in the next section.
6.1 Economic feasibility
The results from the PVWatts application are presented in table 6.1. The AC energy
generated per month is ranging from about 250 to 300 kWh. As the resident has an
assumed consumption of 300 kWh per month, the difference in generation and
consumption has to be bought from the CEB. The electricity tariff paid to the CEB with
and without PV-DG is based on the Bill Calculator constructed by the CEB [27]. In March
the generation exceeded the consumption by 21 kWh. The excess production is used as an
export credit for the coming billing periods, as described in section 2.1.3. In March, April
and May there is only a fixed cost paid and there is a reduced tariff for June.
Table 6.1 Result of the PVWatts application from a 2.5 kW PV system in Sri Lanka.
January
[kWh/m2/day]
5.22
AC
Energy
[kWh]
278
[kWh]
22
Tariff with Tariff without
PV-DG
PV-DG
[Rs.]
[Rs.]
113
12 310
February
5.89
280
20
105
12 866
March
6.26
321
-21
30
12 310
April
5.85
294
6
30
12 495
May
5.44
291
9
30
12 310
June
5.23
273
27
109
12 495
July
5.17
278
22
113
12 310
August
5.50
298
2
38
12 310
September
5.60
292
8
60
12 495
October
5.28
280
20
105
12 310
November
4.93
251
49
302
12 495
December
4.65
249
51
315
12 310
Month
Solar Radiation
Difference
Total [Rs]: 1350
Annual saving [Rs]:
42
149 016
147 666
A Feasibility Study of Increasing Small Scale Solar Power in Sri Lanka
The total electrical bill for one year with the investment in a PV system is estimated to
1,350 Rs (see table 6.1). The electrical bill for one year without the investment is estimated
to 149,016 Rs. The difference, i.e. the annual saving of installing a PV-DG system, is thus
estimated to 147,666 Rs. In order to analyse the actual value of these future savings, all
savings are converted to present value with the method of present discount value. In table
6.2 the present value for the annual savings are presented together with the cumulative
present value for the investment. The present value is calculated for two different values of
discount rate: 5% and 10%.
Table 6.2 Discounted annual savings and cumulative present value for a standardized PV system in
Sri Lanka. Calculated for two different discount rates: 5 and 10 %.
Year
1
Annual PV
[5%]
140 634
Cumulative PV
[5%]
140 634
Annual PV
[10%]
134 242
Cumulative PV
[10%]
134 242
2
3
4
133 937
127 559
121 485
274 572
402 131
523 616
122 038
110 944
100 858
256 280
367 223
468 081
5
6
7
115 700
110 191
104 943
639 317
749 507
854 451
91 689
83 354
75 776
559 770
643 124
718 900
8
9
10
99 946
95 187
90 654
954 397
1 049 584
1 140 238
68 887
62 625
56 932
787 787
850 412
907 344
11
12
13
86 337
82 226
78 310
1 226 575
1 308 801
1 387 111
51 756
47 051
42 774
959 100
1 006 151
1 048 924
14
15
16
74 581
71 030
67 647
1 461 693
1 532 723
1 600 370
38 885
35 350
32 136
1 087 809
1 123 159
1 155 296
17
18
19
64 426
61 358
58 436
1 664 796
1 726 155
1 784 591
29 215
26 559
24 145
1 184 511
1 211 070
1 235 214
20
21
22
55 654
53 004
50 480
1 840 245
1 893 248
1 943 728
21 950
19 954
18 140
1 257 164
1 277 118
1 295 258
23
24
25
48 076
45 786
43 606
1 991 804
2 037 590
2 081 196
16 491
14 992
13 629
1 311 749
1 326 741
1 340 370
Total:
2 081 196 Rs.
1 340 370 Rs.
43
A Feasibility Study of Increasing Small Scale Solar Power in Sri Lanka
According to table 6.2, the investment turns out to be financially favourable at both
discount rates. At the discount rate of 5% the investment has an estimated value of about
2.1 million Rs and at the discount rate of 10% the investment has an estimated value of
1.34 million Rs. The summarized discounted annual savings are therefore at both discount
rates significantly higher than the initial investment of 1.1 million Rs.
The payback period or break-even point is the point in time where the investment has been
paid back and is another measure of the financial benefit of the investment. According to
table 6.2, the break-even point is met after only 8 years. Note that the annual savings are
not discounted in the method of calculating the break-even point. The LCC analysis is
calculated in Sri Lankan rupee (Rs.), where 1000 Rs. correspond to about 50 Swedish
kronor (SEK) [28]. The initial investment of 1 100 000 Rs. correspond thus to about 55
000 SEK. The annual savings is corresponding to about 7 400 SEK and the estimated value
of the investment is 104 000 SEK at 5% discount rate and 67 000 SEK at 10% discount
rate.
It is important to note that by changing some parameter in the simulations, the results may
be altered in a very high degree. Since the technical life-time of a PV system is relatively
long, the value of the discount rate is essential. The results show that the estimated value of
the investment differs with more than 700 000 Rs. when calculated with the two discount
rates. A 5 % discount rate is often considered normal in discounting calculations
concerning safe investments, whereas higher risk investments require a higher return of
capital [26].
Furthermore, as described in section 2.1.3, the tariff system currently used by the CEB is
highly progressive making high electricity consumption drastically more expensive. With a
higher monthly consumption the electricity costs will increase drastically, resulting in
making the investment in a PV system even more financially beneficial. The opposite is
also true, with a low monthly electricity consumption, the electricity costs decrease
considerably and the benefit of investing in a PV system decreases. It is furthermore
important to remember that the savings in electricity costs are not taxable, which may be
the case for revenue from other investments. This has to be taken into account in
comparison and may act as a further benefit of investing in PV-DG.
6.2 Technical feasibility
In the following section, the results concerning the technical feasibility are presented. The
simulations examine both the effect that DG has on increased voltage levels as well as on
44
A Feasibility Study of Increasing Small Scale Solar Power in Sri Lanka
the amount of line losses. Other aspects concerning the technical feasibility are discussed in
the next section.
6.2.1 Increased voltage level
In appendix B, calculations are computed for two events: a DG system connected with a
single-phase connection and DG system connected with a three-phase connection. The
DG is connected at HH1 (see figure 5.1) and all loads are assumed to be zero. The output
of the DG is gradually increased to illustrate the voltage increase and to find the maximal
output for each connection. The maximum allowed voltage variation at each connection
point is found in table 4.1, and is 3% for PCC and 5% for the households.
Figure 6.1 demonstrates the voltage increase at a single-phase connection with an
increasing output of DG. As can be seen in the figure (and is complementary exactly
calculated in the appendix), the output of the connected DG needs to be restricted to less
than 3.74 kW. That maximum allowed connected output is restricted by the calculated
voltage variations at HH1. According to the calculations in appendix B, the maximum
connected output at the PCC due to voltage variations would be 3.81 kW.
Figure 6.1 Voltage increase at PCC and HH1 with an increasing amount of DG connected in one
single-phase.
45
A Feasibility Study of Increasing Small Scale Solar Power in Sri Lanka
Figure 6.2 displays the voltage increase at a three-phase connection with an increasing
output of DG. By comparing figure 6.1 and figure 6.2, it is easy to conclude that installing a
DG system with a three-phase connection allows a considerable larger amount of DG than
in the case of a single-phase connected DG. From the calculations in appendix B, the
maximum allowed output to not exceed voltage variations at the PCC is 21.4 kW. Similarly,
the maximum allowed output to not exceed voltage variations at HH1 is 21.6 kW.
Figure 6.2 Voltage increase at PCC and HH1 with an increasing amount of DG connected in
three-phases.
It is possible to verify with calculations that almost six times higher output may be allowed
in the case of a three-phase connection compared to a single-phase connection. As
discussed in section 4.3.1, the reason is that the total loop impedance of a single-phase
cable is generally twice the amount of the impedance of a three-phase cable. The rule of
thumb of allowing six times higher output is correct in most cases; unless in the case of
unusual conditions such as very short connecting feeders where the grid impedance and
transformer impedance are comparatively high.
The quality of the grid plays an important role in the amount of the voltage variations due
to DG. A strong grid with low impedance is not as vulnerable to high penetration of DG
as a weak grid. In order to analyse whether a new DG connection is advisable, it is thus
important to investigate both the size of the connected DG as well as the quality of the
46
A Feasibility Study of Increasing Small Scale Solar Power in Sri Lanka
grid. In the calculations of voltage variations, it is also important to note which part of the
feeder that is mostly contributing to the increased voltage levels. A badly dimensioned
feeder may not only rule out an installation of DG but may also considerably increase line
losses.
6.2.2 Decreased line losses
In appendix C two different scenarios are examined. In scenario I, the amount of line
losses are examined in the case of a fixed load and a varying amount of DG. In figure 6.3,
the graph to the left shows that as the amount of DG increases, the line losses of feeder
ZPCC decreases. The right graph shows that although the line losses initially decrease for
feeder ZHH1, they actually start to increase as the generation exceeds the load at HH1. If the
output of DG would increase further to the point where it exceeds the total load for all the
households connected to system, the line losses for ZPPC would start to increase yet again.
Figure 6.3 Line losses for feeder ZPCC and ZHH1 as the amount of DG varies.
The total line losses for the whole typical case grid are displayed in figure 6.4. As can be seen
in the figure, the lowest line losses for the whole system occur as the DG reaches an output
of about 12 kW. Hence, increasing the amount of DG above the load at the point of
generation would still decrease the total line losses of the grid, despite that the line losses
47
A Feasibility Study of Increasing Small Scale Solar Power in Sri Lanka
for ZHH1 increase. The reason is that since power now flows from the DG to HH2 and
HH3, the ZPCC feeder does not have to support the same amount of power. However, the
feeder from the PCC to the consumer (ZHH1) is generally smaller dimensioned than the
feeder from the substation to the PCC (ZPCC). With higher impedance on ZHH1 than on
ZPCC the total line losses will start to increase again already as the DG reaches an output of
12 kW. The optimal output of DG in the sense of minimizing line losses is thus dependent
on a number of factors. Not only are the cable impedances in the connecting as well as
overlying feeder affecting line losses, but also other factors such as the reactive power load
and active power load at PCC.
Figure 6.4 Total line losses for typical case grid with varying amount of DG
In scenario II, the amount of line losses are examined in the case of a fixed amount of DG
and a varying amount of load at HH1. The results are then compared with the line losses at
the varying load if no DG is connected at HH1. The results are presented in figure 6.5. In
the graph to the left, the line losses for both scenarios are plotted. As can be seen in the
figure, the line losses are increasing in a higher extent for the case of no installed DG. The
difference in line losses is further demonstrated in the right hand graph in figure 6.5. In the
case of a load of 15 kW at HH1, the line losses are reduced with almost 380 W. This can be
compared with the case of a small load (5 kW) whereas the line losses only are reduced
with about 170 W. The results prove that the line loss reduction due to the DG becomes
48
A Feasibility Study of Increasing Small Scale Solar Power in Sri Lanka
more and more significant as the amount of load increases. DG has therefore a greater
impact on reducing line losses if they are installed in a location with high electricity
consumption.
Figure 6.5 Line losses with and without DG computed with a varying amount of load.
49
A Feasibility Study of Increasing Small Scale Solar Power in Sri Lanka
7 Analysis and discussion
The economy in Sri Lanka has been steadily growing the last few years at a fast rate. In
order to meet the increased demand of energy in the future there is an urgent need to find
new ways of increasing the amount of generated electricity. To not further increase the
already high dependency of foreign oil and to decrease the impact on the environment, a
transition from traditional combustion of fossil fuel to new renewable energy methods is
required. No emissions and low upkeep makes PV-DG almost non-malignant to the
environment and the sensitive natural areas of Sri Lanka. In the following section the
results of the report are analysed and the used methods briefly discussed. As a final point, a
suggestion of related topics and future work is presented.
7.1 Analysis of economic feasibility
The report shows that there exists substantial potential for generation of solar energy in Sri
Lanka and especially in the dry zones. The solar radiation on a flat plate tilted at latitude
varies from 4.5 - 6.0 kWh/m2/day and is at its highest during the dry periods. Calculations
show that an investment in a PV system can be highly economically beneficial and the
investment is often paid back within a few years. However, for the investment to be
economically favourable the installer has to have high electricity consumption. The
electrical tariff system in Sri Lanka is highly progressive and heavy electricity consumption
is priced multiple times higher than low electricity consumption. Whether the progressive
tariff system is increasing or reducing the amount of PV-DG is debatable. Low electricity
consumers have almost no incitement at all of increasing PV-DG, but at the same time
there exists a high incitement for high electricity consumers of investing. As Sri Lanka is
still a quite poor country, the possibilities for low electricity consumers to invest in
expensive PV-DG systems are probably almost non-existent. Consequently, the
progressive tariff system is probably not preventing any development of PV-DG, but
rather just increasing the incitement for high electricity consumers.
To further increase the incitement of investing in PV-DG there are many available options.
Primarily, increased efficiencies and decreased prices may itself increase the incitement of
increasing PV-DG. Furthermore, as the price of oil is expected to increase in the future,
PV-DG may become a more attractive option. Another possibility is using economic
management control issues, such as carbon taxes or subventions for PV systems. The
concept of external effects states that all costs and benefits from a product should be
measured and quantified, and then the costs or benefit of this product should be placed on
the consumer of the good. For example, the combustion of fossil fuels has negative
50
A Feasibility Study of Increasing Small Scale Solar Power in Sri Lanka
external effects in the sense of increasing greenhouse gases and other harmful particles
such as NOx and SO4. In comparison, a PV-DG system has almost no negative external
effects on the environment. Thus, from an economic view, the combustion of fossil fuels
should be under the influence of some sort of emission tax or alike. As Sri Lanka is also
struggling to increase its welfare and development, they need to keep their competiveness
high compared to other countries. Therefore, it would be preferable if these economic
control issues would be implemented worldwide (e.g. such as improved certificate of
emissions or the implementation of global carbon tax).
The technical feasibility study also showed that an increase of PV-DG often reduces line
losses in the distribution grid. However, the decreased line losses are beneficial only to the
net owner (CEB). If at least a part of the value of these reduced line losses were given to
the installer of the PV-DG system it would further increase the economic incitement of
investing. Moreover, the value of goodwill and green advertising for hotels and other
businesses with installed PV-DG may further increase the will to invest.
7.2 Analysis of technical feasibility
The report has investigated the different aspects that an increased amount of PV-DG has
on grid connected factors such as electric power quality and line losses. When determining
the amount of output for an installation of a new PV-DG system it is often the slow
voltage variations that are restraining the size of the system. Simulations show that the
quality of the grid has an important part in the possible output of a system. A badly
dimensioned grid with high impedance is more sensitive to voltage variations. At the
moment, Sri Lanka is still relatively undeveloped and may not have the same standard in
the distribution grid as developed countries. Therefore, it is possible that an increase of
PV-DG could have a negative effect on the voltage variations of the distribution grid.
However, as PV-DG generally is installed as small generation systems (2-6 kW) they will
not affect the grid as much a larger generation systems. The report also shows that if a
higher amount of generation is desired, connecting this PV-DG with a three-phase
connection would allow almost six times higher output (or correspondingly reduce the
voltage variations with almost six times).
The PV-UPSCALE project (see section 4) showed that systems even with a high ratio of
generation capacity, i.e. 80% and higher, did not deteriorate the quality of the grid. The rule
of thumb developed by the project was that no trouble should be caused if the connected
power was limited to 70% of the rated power of the feeding transformer. Sri Lanka has at
present a negligible amount of PV-DG in the distribution grid and it is therefore possible
to conclude that there are no technical obstacles of increasing the amount of PV-DG.
51
A Feasibility Study of Increasing Small Scale Solar Power in Sri Lanka
However, if the amount of PV-DG would increase it is important to recognise that there
may require measures to ensure a safe operation and a good grid protection.
The report also demonstrates that an increased amount of PV-DG actually has many other
benefits apart from creating electricity in an environmentally friendly way. The simulations
have shown that PV-DG generally reduces the line losses in the grid if not dimensioned to
large. Sri Lanka has for a long time been suffering from high energy losses and the line
losses are especially high on badly dimensioned grids or on heavily loaded grids. Installing
PV-DG at certain locations with high electricity demand (utilities or larger hotels with
many appliances) could drastically decrease these energy losses.
The highest output of PV systems occurs during dry periods and during day time. This is
also the period in Sri Lanka with the highest electricity consumption as the tourism is at its
peak and appliances such as air conditioners require lots of electricity. As the demand and
generation coincides it results in a peak demand reduction, effectively reducing the highest
demand of electricity. Not only is peak demand electricity more expensive to generate, but
it also increases the amount of line losses. With an increased amount of PV-DG in Sri
Lanka not only would the line losses decrease but it would also result in reducing the
highest demand of electricity. The relatively undeveloped grid in Sri Lanka is also sensitive
for voltage drops and especially during periods of heavy load. With the implementation of
PV-DG the peak loads would be reduced and thus also the highest voltage drops (see
figure 4.3).
7.3 Discussion of method and results
At present, it is probably neither the economic incitement nor the technical feasibility that
decreases the development of PV-DG. The technology is available and for high electricity
consumers the pay-back period for the investment is just a few years. Then; why is the
development of PV-DG not taking place? The plausible answer is that at present it is not
the lack of financial incitement and benefits but rather socio-economic factors that limit
the development of PV-DG. Sri Lanka is still a relatively poor country and the long years
of civil war have prevented development and wealth. Lack of funds and a high ratio of lowincome earners are therefore probably the main reason for the slow development. The
average annual consumption of electricity was merely 480 kWh per person per year in Sri
Lanka in 2011; an electricity consumption that does not result in any financial incitement in
investing in PV-DG. It is probable to assume that with increased development and
prosperity in Sri Lanka, the development of PV-DG will undoubtedly occur.
52
A Feasibility Study of Increasing Small Scale Solar Power in Sri Lanka
To investigating the feasibility of increasing small scale solar power in Sri Lanka may seem
an impossibly task requiring a huge amount of work. However, the intention has not been
to exactly determine the possibilities of increasing PV-DG, but rather to uncover the
strengths and weaknesses and provide knowledge of what factors that limits the project.
The choice of using the rather simplified simulations and applications is also originating
from the fact that the report is intended as a feasibility study. The report is not aiming to
develop new guidelines in calculating output of a PV system or establishing methods of
calculating voltage variations in complicated grid structures. Such methods would require
too much work leaving the aim of the report not fully examined. Despite that all factors
concerning the feasibility of increasing small scale solar power in Sri Lanka are not fully
investigated, the main aspects remain carefully examined and hence achieving to fulfil the
purpose of the report.
7.4 Future work
There are many related aspects of an increased amount of PV-DG that could be further
examined. PV-UPSCALE used data from grids designed for a high penetration of PV-DG
and it is perhaps not surprising that the results are favourable. It would be more interesting
to investigate the actual effects in grids not designed for a high penetration of PV-DG.
These results would then be even more suitable for determining the generally highest
allowable penetration of PV-DG.
Furthermore, it would be interesting to examine and measure the economic value of the
changed power quality with the implementation of PV-DG. How high is the economic
value of the decreased line losses and voltage drop in general in Sri Lanka? This result
could then be used to develop guidelines for the amount of compensation that the installer
of the PV-DG should receive for the enhanced grid qualities.
53
A Feasibility Study of Increasing Small Scale Solar Power in Sri Lanka
8 Summary of conclusions
The following section summarises the main conclusions of the report in relation to the
aims of the thesis.

Due to a fast growing economy the demand for electricity is rapidly increasing in
Sri Lanka. To avoid further dependency of foreign oil and decrease impact on the
environment, a transition from traditional energy sources to new renewable energy
sources is required.

PV-DG is increasingly more cost-effective as the technology increases and the
prices of PV systems decreases. There exists substantial potential for generating
solar energy in Sri Lanka. Current regulations and electricity pricing increases the
incitement of investing primarily for high electricity consumers.

There are probably no technical obstacles of increasing PV-DG yet. With a
significantly increased amount of PV-DG it is required to start to pay attention to
aspects such as electric quality and grid protection.

PV-DG generally decreases line losses and voltage drops on heavily loaded feeders.
As the highest output of PV-DG is during dry periods this result in a peak demand
reduction. Peak demand electricity is expensive to generate, has the highest line
losses and may in the long run require expansion of the grid.

To further increase the incitement of investing in PV-DG there are many available
options. Increased oil price, enhanced efficiencies and decreased module prices may
itself increase the incitement of increasing PV-DG. Other options are to use using
economic management control issues such as carbon taxes or subventions for PV
systems.

At present it is probably not the lack of economic incitement but rather socioeconomic factors that limit the development of PV-DG. Sri Lanka is still a
relatively poor country and the long years of civil war have prevented development
and wealth. Lack of funds and a high ratio of low-income earners are the probable
main reason for the slow development.
54
A Feasibility Study of Increasing Small Scale Solar Power in Sri Lanka
References
1. IPCC Work group II (2014). Climate Change 2014: Impacts, Adaptation, and vulnerability:
Summary for policymakers. IPCC. Available: <http://www.ipccwg2.gov/AR5/images/uploads/IPCC_WG2AR5_SPM_Approved.pdf> [2014-04-24]
2. Drexhage, John & Murphy, Deborah (2010). Sustainable Development: From Brundtland to
Rio 2012. New York: United Nations. Available:
<http://www.un.org/wcm/webdav/site/climatechange/shared/gsp/docs/GSP16_Background%20on%20Sustainable%20Devt.pdf> [2014-04-25]
3. Svenska FN-förbundet (2013). Sri Lanka. [Online]. Available:
<http://www.globalis.se/Laender/Sri-Lanka> [2014-04-20]
4. NREL (2014). PVWatts Calculator. [Online]. Denver. Available:
<http://pvwatts.nrel.gov/> [2014-05-11].
5. Sri Lanka Sustainable Energy Authority (2011). Energy Balance 2011: An analysis of energy
sector performance. Colombo: Sustainable Energy Authority. Available:
<http://www.energy.gov.lk/pdf/Balance_2011.pdf> [2014-04-29]
6. Department of Meteorology (2012-05-02). Climate in Sri Lanka. [Online]. Available:
<http://www.meteo.gov.lk/index.php?option=com_content&view=article&id=106&Item
id=81&> [2014-03-15]
7. Ceylon Electric Board (2011). Annual Report 2011. Colombo: Ceylon Electric Board.
Available: <http://www.ceb.lk/sub/publications/annual.aspx> [2014-03-25]
8. Ceylon Electric Board (2014). Tariff Plan. [Online]. Available:
<http://www.ceb.lk/sub/residence/tariffplan.html> [2014-04-12]
9. Ceylon Electric Board (2014). Net Energy Metering Manual. Colombo: Ceylon Electric
Board. Available:
<http://www.ceb.lk/download/business/Netmetering%20Manual%20Annex%201234.pd
f> [2014-04-12]
10. Renné, Dave et al (2003). Solar resource assessment for Sri Lanka and Maldives. Colorado:
National Renewable Energy Laboratory. Available:
<http://www.nrel.gov/docs/fy03osti/34645.pdf> [2014-04-22]
11. Markvart, Tomas (red.) (2000). Solar electricity. 2. ed. Chichester: Wiley
12. EPRI (2000). Engineering Guide for Integration of Distributed Generation and Storage into Power
Distribution Systems. Palo Alto, CA: EPRI. 1000419.
13. Widén, Joakim (2010). System Studies and Simulations of Distributed Photovoltaics in Sweden.
Uppsala: Uppsala University, Department of Engineering Sciences. Available:
<http://uu.diva-portal.org/smash/get/diva2:359601/FULLTEXT01> [2014-02-23]
14. JLanka Technologies (2014). Photo gallery. [Online]. Colombo. Available:
<http://www.jlankatech.com/photo-galler/> [2014-05-10]
55
A Feasibility Study of Increasing Small Scale Solar Power in Sri Lanka
15. Patel, Mukund R. (2006). Wind and solar power systems: design, analysis, and operation. 2. ed.
Boca Raton: Taylor &Francis
16. Mitra, Parag & Heydt, T. Gerald (2012). The Impact of Distributed Photovoltaic Generation on
Residential Distribution Systems. [Online]. IEEE. North American Power Symposium 2012.
Available: IEEE Xplore. DOI: 10.1109/NAPS.2012.6336330
17. PV-UPSCALE. Impact of photovoltaic generation on power quality in urban areas with high PV
population—results from monitoring campaigns. WP4 - Deliverable 4.3, 2008. Tillgänglig:
<http://www.pvupscale.org/IMG/pdf/WP4_D4-3_public_v1c.pdf> [2014-04-23]
18. Driesen, Johan & Belmans Ronnie (2006). Distributed generation: challenges and possibilites.
[Online] IEEE. Power Engineering Society General Meeting, 2006. Available: IEEE
Xplore. DOI: 10.1109/PES.2006.1709099
19. Gönen, Turan. (2008). Electric power distribution system engineering. 2 ed. Boca Raton: CRC
Press
20. Svensk Energi (2011). Anslutning av mikroproduktion till konsumentanläggningar. Utgåva 1.
Stockholm
21. Glover, J. Duncan, Sarma, Mulukutla S. & Overbye, Thomas J. (2008). Power system
analysis and design. 4. ed. Toronto, Ontario: Thomson Learning
22. Almgren, Åke & Blomqvist, Hans (2003). Elkrafthandboken. Elkraftsystem, 2. 2.uppl.
Stockholm: Liber
23.Svensk Elstandard SEK (1993). Ledningsnät för max 1000 V: dimensionering med hänsyn till
utlösningsvillkoret – direkt jordade nät skyddade av säkringar SS 424 14 05. SIS Förlag. Stockholm
24. Martinez Ibáñez, Alán & Calleja Hugo (2000). A simple, High-Quality Output PV System
Aimed at Peak Demand Reduction. [Online]. IEEE. Power Electronics Congress, 2000. CIEP
2000. VII. DOI: 10.1109/CIEP.2000.891430
25. Dobos, P, Aron (2013). PVWatts Version 1 Technical Reference. [Online]. Denver: NREL.
Available: <http://www.nrel.gov/docs/fy14osti/60272.pdf> [2014-05-11]
26. Brealey, Richard A., Myers, Stewart C. & Marcus, Alan J. (2012). Fundamentals of corporate
finance. 7. ed. New York: McGraw-Hill/Irwin
27. Ceylon Electric Board (2014). Bill Calculator. [Online]. Available:
<http://www.ceb.lk/sub/residence/billcalculator.aspx> [2014-05-12]
28. Oanda (2014). Currency Converter. [Online]. Available:
<http://www.oanda.com/currency/converter/> [2014-05-13]
Personal references
29. C. B. Mudhalige ([email protected]). Technical Manager, JLanka Technologies.
E-mail communication (2014-05-05).
56
A Feasibility Study of Increasing Small Scale Solar Power in Sri Lanka
A. Investment calculations and discount rate factors
The discount factors used to discount the annual savings in the calculations in section 6.1 is
found in table A1. The discount factors are calculated according to equation (27).
Table A1. Discount rate factors for the discount rates of 5% and 10%.
Year
Discount rate
factor at 5%
Discount rate
factor at 10%
1
0,952381
0,909091
2
0,907029
0,826446
3
0,863838
0,751315
4
0,822702
0,683013
5
0,783526
0,620921
6
0,746215
0,564474
7
0,710681
0,513158
8
0,676839
0,466507
9
0,644609
0,424098
10
0,613913
0,385543
11
0,584679
0,350494
12
0,556837
0,318631
13
0,530321
0,289664
14
0,505068
0,263331
15
0,481017
0,239392
16
0,458112
0,217629
17
0,436297
0,197845
18
0,415521
0,179859
19
0,395734
0,163508
20
0,376889
0,148644
21
0,358942
0,135131
22
0,34185
0,122846
23
0,325571
0,111678
24
0,310068
0,101526
25
0,295303
0,092296
Appendix
A:1
A Feasibility Study of Increasing Small Scale Solar Power in Sri Lanka
B. Increased voltage levels
The output of the installed DG at HH1 is varied from zero output to 5 kW output. No
reactive power is assumed to be generated. As the resistive part of the short-circuit
impedance is dominating, it is possible to assume that Z = R, and that X is negligible. The
voltage increase is calculated both at the point of common connection (PCC) as well as at
the load (HH1). The two following scenarios are examined:

Scenario I: Generation connected as three-phase

Scenario II: Generation connected as single phase
Scenario I:
From table 5.3, the short-circuit impedance of the transformer and the connecting cables
are found. The total short-circuit impedance at the PCC, (
) is calculated according to
equation (20) as:
The total short-circuit impedance at HH1, (
) is calculated accordingly in the following
equation:
By neglecting Q and X, equation (18) can be simplified into equation (28).
(
)
(28)
By then rewriting equation (28) to equation (29) it is possible to calculate the maximal
output for the connected DG according to voltage regulations.
(29)
By then creating a vector of different values of the amount of DG and using equation (28),
it is possible to calculate the voltage increase for each output value. These calculations are
performed separately for the voltage increase at HH1 and at PCC. The calculations are
performed with the MATLAB-script found in appendix E. The results and corresponding
plots are then presented in section 6.2.
To calculate the maximal possible output of a connected DG to the case grid, equation (29)
is used. The maximal output is calculated for both the PCC as well as for the HH. The
regulated voltage variations are presented in table 4.1. With the calculated values of the
Appendix
B:1
A Feasibility Study of Increasing Small Scale Solar Power in Sri Lanka
short-circuit impedance, the following maximal output can be calculated for each
connection point as followed:
The lowest maximal output is determining the maximum allowed connected output.
According to the above calculations, the maximum allowed output of a connected DG is
3.74 kW and is restricted by the voltage increase at PCC.
Scenario II:
From table 5.3, the loop impedance of the transformer is found. The loop impedance of
the cables are calculated as twice the short-circuit impedance in the case of a three-phase
connection. The total loop impedance at the PCC (
) is calculated according to
equation (19):
(
)
The total loop impedance at HH1, (
) is then calculated according to the following
equation:
(
)
(
)
In the same way as in scenario I, a vector of values of the amount of DG is created. Then,
by using equation (28), it is possible to calculate the voltage increase for each output value.
These calculations are performed separately for the voltage increase at HH1 and at PCC.
The calculations are performed with the MATLAB-script found in appendix E. The results
and corresponding plots are then presented together with the results of scenario I in
section 6.2.
The maximal possible output of a connected DG in the case of a single-phase connection
is calculated below. The same regulated voltage variations as before is used and are found
in table 4.1. With the calculated values of the loop impedance, the following maximal
output can be calculated for each connection point as followed:
Appendix
B:2
A Feasibility Study of Increasing Small Scale Solar Power in Sri Lanka
The lowest maximal output is determining the maximum allowed connected output.
According to the above calculations, the maximum allowed output of a connected DG is
3.74 kW and is restricted by the voltage increase at HH1.
Appendix
B:3
A Feasibility Study of Increasing Small Scale Solar Power in Sri Lanka
C. Decreased line losses
The increased voltage at the POG does not only reduce the voltage drop along the feeder,
but it may also reduce the line losses. However, if the DG increases more than the load, the
line losses may actually increase. Two different scenarios are examined in the following
section:


Scenario I: Fixed load and varying amount of DG
Scenario II: Fixed amount of DG and a varying amount of load at HH1
As previous concluded, the resistive part of the short-circuit impedance is dominating and
it is possible to assume that Z = R. MATLAB is used to simplify the plotting and the
presentation of the results.
Scenario I:
In the following scenario, the line losses are analysed for the feeder ZPCC and for ZHH1.
With a varying amount of DG (from 0 – 15 kW), the line losses for the two feeders are
calculated and then plotted. From table 5.3, the short-circuit impedance of ZPCC and ZHH1
are found. Line losses for ZHH2 and ZHH3 are neglected as these do not change due to a
change in the amount of DG at HH1. The losses in the transformer are neglected. The
following loads are used for this scenario:
Table C1. List of current loads for case grid
Total active effect
[kW]
HH1 6
HH2 5
HH3 5
Total 16
Total reactive effect
[kVar]
2
1.5
1.5
5
Equation (24) can be transformed into equation (30), now being dependent on the amount
of DG:
.
/
. /
(30)
By then creating a vector of values of the amount of DG and using equation (30), it is
possible to calculate the corresponding values for the line losses. These calculations are
performed separately for feeder ZHH1 and ZPCC and then finally summarized to the total
line losses. The calculations are performed with the MATLAB-script found in appendix E.
The results and corresponding plots are then presented in section 6.3.
Appendix
C:1
A Feasibility Study of Increasing Small Scale Solar Power in Sri Lanka
Scenario II:
In this scenario, the line losses are yet again analysed for the feeder ZPCC and for ZHH1.
However, with a varying amount of load at HH1 (from 5 – 15 kW), the total line losses for
the two feeders are calculated and then plotted. The same short-circuit impedance as
previous is used.
Two situations are examined:

Line losses with varying amount of load and DG

Line losses with varying amount of load and no DG
The difference in line losses between the two cases are the compared and plotted. Equation
(30) is used for the calculations (now with the load as the dependent variable) and the same
load data as found in scenario I is used. The calculations are performed with the
MATLAB-script found in appendix E. The results and corresponding plots are then
presented in section 6.3.
Appendix
C:2
A Feasibility Study of Increasing Small Scale Solar Power in Sri Lanka
D. Cable data and impedance calculations
Table D1. Loop-impedance for different sizes of distribution transformers [23]
Rated power of
Loop-impedance
distribution transformer
(mΩ)
(kVA)
20
320
30
213
50
130
63
102
100
65
125
51
160
40
200
32
250
26
315
20
400
16
500
13
630
11
800
10
1000
8
1250
6.5
1600
6.25
Appendix
D:1
A Feasibility Study of Increasing Small Scale Solar Power in Sri Lanka
E. MATLAB-script
The following .m-file is used to simulate and calculate the effects of an increased amount
of PV-DG. Two aspects are examined: Voltage variations due to DG and line loss
reduction due to DG.
Calculations concerning voltage variations
clear all; close all; clc
DG=linspace(0,5000); % Varying amount of DG
Z_T_PCC=0.2243; %Line impedances at PCC
Z_T_HH1=0.37; %Line impedances at Z_H1
Z_T_L_PCC=0.417; %Total loop impedances at PCC
Z_T_L_HH1=0.707; %Total loop impedances at HH1:
Calculating voltage variations at PCC and HH1, for both a single-phase
connection and a three-phase connection
dU_Z_T_PCC=(DG*Z_T_PCC)/(400^2)*100;
dU_Z_T_HH1=(DG*Z_T_HH1)/(400^2)*100;
dU_Z_T_L_PCC=(DG*Z_T_L_PCC)/(230^2)*100;
dU_Z_T_L_HH1=(DG*Z_T_L_HH1)/(230^2)*100;
Plotting the result
hold on
plot(DG,dU_Z_T_L_PCC,'r',DG,dU_Z_T_L_HH1,'b','LineWidth',2);
xlabel('Amount
of
DG
[kW]');
ylabel('Voltage
Increase
[%]');title('Single-phase
connection');
legend('PCC','HH1')
plot(DG,3,DG,5); axis([0 5000 0 7]) ; hold off
figure
hold on
plot(DG,dU_Z_T_PCC,'r',DG,dU_Z_T_HH1,'b','LineWidth',2);xlabel('Amount of DG [kW]');
ylabel('Voltage Increase [%]'); title('Three-phase connection'); legend('PCC','HH1')
plot(DG,3,DG,5); axis([0 5000 0 7]); hold off; shg
Line losses calculations with varying generation - Scenario I
Appendix
E:1
A Feasibility Study of Increasing Small Scale Solar Power in Sri Lanka
clear all; clc
U=0.400; % Voltage (kV)
DG=linspace(0,15); % Varying amount of DG
Load (P and Q) and line impedances at end-users HH1, HH2, and HH3
PHH1=6; PHH2=5; PHH3=5;
QHH1=2; QHH2=1.5; QHH3=1.5;
Ptot=PHH1+PHH2+PHH3; Qtot=QHH1+QHH2+QHH3;
%Line impedances for HH1, HH2, HH3 and ZPPC
ZPCC=0.1923; ZHH1=0.1454; ZHH2=0.2181; ZHH3=0.1454;
Line losses with increasing amount of DG
Pf_PCC=ZPCC*(((Ptot-DG)/U).^2)+ZPCC*((Qtot/U).^2);
Pf_HH1=ZHH1*(((PHH1-DG)/U).^2)+ZHH1*((QHH1/U).^2);
Pf_tot=Pf_PCC+Pf_HH1;
Plotting the results
figure
subplot(1,2,1); plot(DG,Pf_PCC,'LineWidth',2); grid on; xlabel('Amount of DG [kW]');
ylabel('Line losses [W]'); title('ZPCC');
subplot(1,2,2); plot(DG,Pf_HH1,'LineWidth',2); grid on; xlabel('Amount of DG [kW]');
ylabel('Line losses [W]'); title('ZHH1');
shg
figure
plot(DG,Pf_tot,'LineWidth',2); xlabel('Amount of DG [kW]'); ylabel('Line losses [W]');
shg
grid on
Line losses with varying generation – Scenario II
Pload=linspace(5,15); % Loads: 5-15 kW
Ptotal=Pload+10; % total P at point of common coupling
PV_G=5; %DG: 5 kW
Calculating the line losses with varying amount of load and DG
Pf_PCC_DG=ZPCC*(((Ptotal-PV_G)/U).^2)+ZPCC*((Qtot/U).^2);
Pf_HH1_DG=ZHH1*(((Pload-PV_G)/U).^2)+ZHH1*((QHH1/U).^2);
Pf_total_DG=Pf_PCC_DG+Pf_HH1_DG;
Calculating the line losses with varying amount of load and NO DG
Pf_PCC_0=ZPCC*(((Ptotal)/U).^2)+ZPCC*((Qtot/U).^2);
Pf_HH1_0=ZHH1*(((Pload)/U).^2)+ZHH1*((QHH1/U).^2);
Appendix
E:2
A Feasibility Study of Increasing Small Scale Solar Power in Sri Lanka
Pf_total_0=Pf_PCC_0+Pf_HH1_0;
Calculating the difference in line losses in the cases with or without DG
Difference_pf=Pf_total_0-Pf_total_DG;
Plotting the result
figure; subplot(1,2,1)
plot(Pload,Pf_total_DG,Pload,Pf_total_0,'LineWidth',2); grid on;
[kW]'); ylabel('Line losses [W]');
hleg = legend('Line losses with DG','Line losses without DG')
grid on; shg
subplot(1,2,2)
plot(Pload,Difference_pf,'LineWidth',2);
ylabel('Change in Line losses [W]');
shg
grid on
Appendix
grid
E:3
on;
xlabel('Load
xlabel('Load
at
HH1
at
HH1
[kW]');
Was this manual useful for you? yes no
Thank you for your participation!

* Your assessment is very important for improving the work of artificial intelligence, which forms the content of this project

Related manuals

Download PDF

advertisement