1. See discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/273451377
Impact of Photovoltaic (PV) Systems on Distribution Networks
Article in International Review on Modelling and Simulations (IREMOS) · February 2014
CITATIONS
3
READS
2,269
3 authors:
Some of the authors of this publication are also working on these related projects:
PV systems Installation View project
Master of Science Thesis View project
Wadhah Esmaeel
Baaquba Technical Institute
2 PUBLICATIONS 3 CITATIONS
SEE PROFILE
Chin Kim Gan
Technical University of Malaysia Malacca
125 PUBLICATIONS 1,172 CITATIONS
SEE PROFILE
Mohd Ruddin Ab Ghani
Technical University of Malaysia Malacca
113 PUBLICATIONS 1,132 CITATIONS
SEE PROFILE
All content following this page was uploaded by Wadhah Esmaeel on 16 October 2018.
The user has requested enhancement of the downloaded file.
2. International Review on Modelling and Simulations (I.RE.MO.S.)
Impact of Photovoltaic (PV) Systems on Distribution Networks
Wadhah Esmaeel Ibraheem 1
, Chin Kim Gan 2
, Mohd Ruddin Ab. Ghani3
Faculty of Electrical Engineering
Universiti Teknikal Malaysia Melaka (UTeM)1, 2,3
Department of Electrical Engineering, University of Diyala, Iraq1
Abstract – Traditionally, power systems are designed to operate in a unidirectional power flow.
In the past few years, solar Photovoltaic (PV) systems have grown rapidly driven by its potential
technical and economic benefits. These include higher network utilization, enhanced reliability
and loss reduction. However, the PV generation depends directly to the sun's radiation. Thus, the
intermittent fluctuations may potentially cause problems to the network operation especially in
high penetration levels. In addition, the voltage fluctuation and power quality issues may limit the
PV penetration level and hence mitigation measures are needed to alleviate the potential
problems. In this paper, the impact of PV on the distribution network in term of voltage
performance and losses has been investigated by using the OpenDss simulator tool. Mitigation
strategy has also been proposed to control the voltage fluctuation that caused by the PV plants.
IEEE 13-bus test system was used to perform the case study.
Keywords: Photovoltaic (PV), distribution network, loss reduction, OpenDss
Nomenclatures
RES Renewable Energy System
PV Photovoltaic
DG Distributed Generation
OPENDSS Open Distribution System Simulator
COM Component Object Model
VBA Visual Basic Applications
DLL Dynamic Link Library
Yprim Primitive y Matrix
I0 Diode's reverse saturation current
q Electron charge
A Curve fitting factor
K The Boltzmann constant
T The absolute temperature in Ko
Np Number of the parallel connected panels
Ki The temperature coefficient of short circuit
Solar radiation measured by (W/m2
)
Tr Reference temperature of the cell
Ego The energy of band gab
IL The line current
IPV The PV current
Vd Voltage drop
Voc Open circuit voltage
Isc Short circuit current
I. Introduction
These days the demand for electrical energy is
increasing to meet the load expansion in the electrical
power system. The rise of the world weather temperature
and the depletion of fossil fuel and the price of the fuel
had motivated more researches and development in
renewable energy system (RES). This is to reduce the
CO2 emission. Many countries have engaged in
renewable electricity and they have set ambitious target
for producing power from green sources to meet the
expected demand in the coming years.
Photovoltaic (PV) solar energy is one of the important
sources of the renewable energy; it has grown in many
countries steadily in the last few years. In Europe, the
combined target yield a total expected PV power
generation capacity is 84.4 GW by 2020, where the
maximum projected PV target is in Germany which is
around 51.8 GW from solar energy [1]. In Malaysia
according to the 10th Malaysian plan (2011-2015) the PV
generation in the country will reach 65 MW in 2015 and
it is projected to be 190 MW by 2020 [2]. U.S. imports
of PV products from South Korea are small, but the
country has a stated goal to capture 10% of the global PV
market by 2020 [3].
PV systems generation is not different of other
renewable energy resources (naturally replenished). Solar
energy is comparatively clean by owning a lower effect
on the environment as well as it saves the depletion of
fossil fuel and coal. But the variability in the output of
3. Wadhah Esmaeel Ibraheem, Chin Kim Gan, Mohd Ruddin Ab. Ghani
the PV panels is a natural behavior of these resources and
it is a significant issue. The rising and setting of the sun
leads to a regular variation of the PV panel generation
over the daily time period. In addition, the output can be
decreased to 50%-80% when the clouds pass over PV
plants [4]. In this case a backup power will be required to
cover the output variation to maintain the operation
voltage under the limit. If the response of the backup
elements is too slow to cover the problem, power quality
can adversely affect. Despite these difficulties, solar PV
plants still have the fastest growing of renewable energy
technology in 2012. The power was almost 150% has
increased of the installed capacity of 2010 bringing the
total generation about 100 GW [5]. As the rapid
increasing of the PV integration, it could potentially bring
problems in terms system operation where a reverse
power can be introduced by higher penetration levels
resulted a rise in buses voltages and feeder loses [6].
The injected power by the PV plant modules at the
load side buses will decrease the demand of the local load
which leads to a loss reduction and voltage profile
improvement [7],[8]. Obviously, this case is true as long
as the real power flows from the substation to the
customer side (when the load is less that PV power). If
the PV generation is more than the load downstream of
the PV location, the power flow may be reversed towards
the substation. Consequently, a voltage rise can be
expected along the distribution system feeder as a result
of the reverse power flow. The rise of the voltage at the
end-user limits the amount of the penetration level which
wanted to be installed in the distribution network. The
natural behavior of the solar source makes the generation
of the PV plant systems in fluctuated profile. The rapid
variation of the PV power introduces a voltage
fluctuation along the PV working time and hence its
effects on the voltage regulation in some cases. All these
issues will be evaluated and mitigation control will be
proposed in this paper to mitigate the voltage rise which
caused by PV interconnection and to investigate the
regulation benefits in term of daily time.
II. Open Distribution system Simulator
(OpenDss)
The Open Distribution System Simulator (OpenDss) is
an inclusive tool for the electrical distribution system
simulation. It supports the frequency domain analysis
usually performed for distribution analysis and planning.
Furthermore, many features in this simulator were
intended for backing distribution generation analysis
needs, it has been designed as flexible to expand which
can easily to amend for supporting a several analysis that
meet the future studies. OpenDss is executed by the
solution engine through a basic text-based interface to
assist the researchers in developing scripts and exploring
solutions. The program can be implemented either stand-
alone program or driven by other existing software
platforms as a component object model (COM) server
DLL. By the (COM) interface, researchers have the
ability to run the solution modes from external programs
that can handle COM. Commonly, the DSS can be driven
by the MS office tool through VBA, C#, MATLAB and
other languages. This is providing wide capability of
analysis and excellent graphics for exploring results and
solutions in power flow, energy efficiency, harmonics,
smart grids and other studies.
The power flow is a most common popular problem
which can be solved by OpenDss program. The largest
advantage of this simulator is the unique and powerful
capabilities of performing all aspects of (DG), distributed
generation, including harmonic analysis. The program is
designed to execute a basic distribution style power flow.
However, OpenDss varies from other radial system
solvers where it solves the meshed distribution networks
easily. Thus, it can be utilized to perform small to
medium sized system's power flow execution. Number of
solution models can be implemented by power flow in the
program including the standard snapshot mode, daily
mode, duty cycle mode and several modes in the case of
the load is changing with the time. Commonly, 24-hour a
day, a month, a year time period is used for planning
purposes. The power flow solution can be divided into
two types; Iterative Power flow solution and Direct
solution. In Iterative power flow the treatment of DGs
and Loads model is considering them as an injection
sources where Direct solution is a direct solution without
iterations through the system admittance matrix.
In this research OpenDss will be used in modeling of
each of distribution network circuits and PV time series
generation to execute the grid connected PV system and
evaluate the PV impact on the distribution grid [9]. The
architecture of the program to model the circuit and PV
plant will be as:
II.1 Distribution Network Model
The main object simulation engine of OpenDss has a
structure consist of executive program that control the
simulation. The different distribution components are
classified into five objects classes which are:
• Power delivery elements.
• Power conversation elements.
• Control.
• Meters.
• General.
All those components of the architecture of the
OpenDss engine as shown in Figure 1. Power delivery
elements are the components that transfer the electrical
power from node to another. They can be either single or
multi phased elements including transformers and
transmission lines. Electrical energy is converted to
another form once achieve the power conversation
elements. Load and generators are considered common
4. Wadhah Esmaeel Ibraheem, Chin Kim Gan, Mohd Ruddin Ab. Ghani
power delivery elements are modeled by introducing their
impedances or injected currents. The other circuit
elements are assistance elements that associated with the
main objects to achieve solutions; some of them will be
used in this research.
DSS Executive
Circuit
Solution
V | [Y] | I
Controls Meters General
PCElement
PDElement
Line
Transformer
Capacitor
Reactor
Load
Generator
Vsource
Isource
Storage
RegControler
CapControl
Relay
Reclose
Fuse
Monitor
Energy Meter
Sensor
LineCode
LineGeometry
Wiredata
LoadShape
GrowthShape
Spectrum
TCCcurve
XfmrCode
Figure 1 OpenDss Architecture of Models [9]
All elements in the architecture of the program and
system bus bars are managed as a creation and
modification of (Yprim) primitive Y matrices. The results
collection will be through meter components. The system
Y matrix is constructed according to feeding the spars
matrix solver by Yprim matrices.
II.2 PV Panel Model
OpenDss has implemented PV model as a combination
of PV array modeling and the PV inverter into one model
to be utilized for distribution grids impact studies. Figure
2 shows the structure of the PV system model which
presents the circuit of the PV panel as a generator. The
output generation power, P, will be as a function of:
Irradiance: the net outdoor irradiance applied as a load
with a certain time step resolution.
Temperature: the net weather temperature which applied
as a load with a certain time step
resolution.
P-T Factor: this factor is applied as a curve to base the
Pmpp at different degree of temperature.
Efficiency: this curve is a number of efficiencies of the
inverter based on the operating power.
Other functions are needed in the model such as rated
voltage and power, as well as an average Pmpp for the
panel at 1kW/m2
irradiance at constant panel temperature
as 25Co
.(see Figure 2)
AC
DC
V
I
Eff.
P
Pmpp
T
0 100C
kV Conn. Kvar PF
Pmpp
1kW/m2
T
Irradiance
Yearly
Daily
Irradiance,
Temperature
loadshapes
P*Eff.
Figure 2 Block diagram of the PVsystem model [9]
III. Data Gathering
In any kind of study, data preparation is very important
to assess and evaluate the case studies. In this paper, load
profile, irradiance, temperatures and distribution test
system have been selected to execute the impact study
and evaluate the results. Following are the types of
selected data and their description.
III.1 Distribution Network
To determine the impact of photovoltaic (PV) system
on of distribution grids. A 13-bus system test feeder has
been used which includes the most of the important
electrical components that are commonly used to analyse
voltage impact in case of grids-connected PV system.
The single line diagram in Figure 3 illustrates the 13-bus
distribution system which had been suggested by IEEE
Power and Energy Society (PES) [10].The test feeder
contains:
• Different levels of voltages that are transformed by 2
connected transformers (115 kV, 4.16 kV and 0.48
kV).
• 9 node-connected loads with 2 capacitors at 2 bus
bars (611,675), the loads are unbalanced.
• 3 voltage regulators are connected to the 3 phases of
the main sub-transformer.
632
650
633 634
692 675
671
646 645
611 684
680
652
Voltage
Regulator
Figure 3 IEEE 13-Node Distribution Network without PV System
5. Wadhah Esmaeel Ibraheem, Chin Kim Gan, Mohd Ruddin Ab. Ghani
This system contains a comination of three phases and
two phases feeders. Some of bus bars are connected with
one or two phases, thus, the PV installation in this case
study will be only on buses that are supplied by three
phases feeders.
III.2 Irradiance and Temperature Data
In this research PV power generation module requires
a variability of the irradiance time history and load data.
Commonly, utilities records a data for the historical load
time series in term of 1 hour resolution, but this type of
time resolution is too poor to analyze some important
aspects in impact studies on grid-connected. Therefore
high resolution in order to 15 minute term has been
collected to drive the time series of power flow and
voltage profile simulation. The current paper assumes
that the location of the grids connected PV system is in
Loughborough city in United Kingdom at latitude and
longitude, 52.8 N and 1.2 W respectively. The global
outdoor irradiance and weather temperature are shown in
the Figure 4. Integrated High-resolution Modeling of
Domestic Electricity Demand and Low Voltage
Electricity Distribution Networks tool has been used for
irradiance modeling in any orientation at any location in
the world [11]. 16th of June 2012 is the collected date for
the historical temperature data at PV installation site [12].
Those data will be used in time series power generation
model of PV system by OpenDss simulator.
Figure 4 Global Outdoor Irradiance and Temperature at Selected Site
III.3 Daily Demand (Load Profiles)
A typical load profile has been used in this research as
shown in Fig 5. The considered load in the case studies is
normalized to the maximum load during the daily
demand. It is worth to mention, generally, the load starts
to increase during the night. As can be noted from the
figure the The load drops to its minimum demand during
the time from 8:00 AM to 5:00 PM and starts to increase
to its peak value at 7:00 PM until midnight. This
behavior can be understood by mentioning that the daily
consumption of the customers is low at noon time
because most of the people are in their work. On the
other hand the load becomes high at the night because all
adult and children in the homes at that time [13].
Figure 5 Daily Load Profile
All the previously mentioned data are used to perform
the impact study analysis by using Open Distribution
Simulation Software (OpenDss) program in several
scenarios
IV. PV Power Generation Model
The electrical specification data sheet for the
photovoltaic panel module in table 1 provides the
electrical characteristics of SolarWorld SW 255 (mono)
type. Commonly, the important characteristics are
maximum generated power (Pmax), voltage and current
at maximum power point (Vmp) and (Imp) respectively,
open circuit voltage (Voc) and short circuit current (Isc).
All these parameters are under Standard Test Condition
(STC),1000 W/m2 and 25 Co, but the ecological weather
always changes with time in general real life [14]. Thus,
Data sheets provide other parameters in order to accurate
the calculation; these parameters are represented by
Nominal Operating Cell Temperature (NOCT) with the
temperature coefficients for each (Voc) and (Isc). The
following table represents to the simulation parameters of
the collected PV model under (STC) conditions,
irradiation 1KW/m2, spectrum of 1.5 air mass and cell
temperature equal to 25Co
.
TABLE I THE ELECTRICAL CHARACTERISTICS DATA OF SOLARWORLD
SW 255 PV PANEL
Parameter Value
Rated Power 255 W
Voltage at Maximum power (Vmp) 37.8 V
Current at Maximum power (Imp) 8.66 A
Open circuit voltage (Voc) 31.4 V
Short circuit current (Iscr) 8.15 A
Total number of cells in series, parallel
(NS, Np)
60
Mathematically, photovoltaic has been modeled
according to equations from (1) to (4). In view of the PV
cell equivalent circuit in Figure 6, the mathematical
model of output current has formed by equation (1)
[15],[16]:
6. Wadhah Esmaeel Ibraheem, Chin Kim Gan, Mohd Ruddin Ab. Ghani
1
)
(
exp
AKT
R
V
q
I
Np
I
Np
I s
o
ph
PV (1)
Where, I0: diode's reverse saturation current.
q: electron charge (1.602×10-19 C).
A: curve fitting factor normally greater than 1.
K is the Boltzmann constant (1.38×10-23 J/K) and T is
the absolute temperature of the solar cell.
Np: number of parallel connected panels (in this section
Np is 1).
Figure 6 Equivalent circuit of a single solar cell [17]
As mentioned before the output voltage is very
small to use, thus the PV array should be assembled as a
collection of series and parallel panels to generate a
usable voltage and current. Therefore, the photovoltaic
model in this research has been used is a string of 255W
in series or parallel connection to get the require power.
Radiation and temperature from the most important and
effective factors in calculation of PV modeling, the
changes consideration of these factors is given by
following equation:
1000
298
T
i
K
scr
I
ph
I (2)
Where Ki is the temperature coefficient of short circuit
and is solar radiation measured by (W/m2
). Reverse
saturation current and its variation with temperature can
be calculated and modeled by equation (4) and equation
(3) respectively.
1
/
exp
kAT
s
N
oc
qV
scr
I
rs
I (3)
T
r
T
Bk
go
E
q
r
T
T
rs
I
o
I
1
1
exp
3
(4)
Tr is a reference temperature of the cell. Some of
modeling parameters not obtainable in the photovoltaic
manufacture’s datasheets like curve fitting and the energy
of band gab (Ego). These parameters are considered
different from type to another of semi-conductors
materials as mentioned in [18]. The obtained results in
this section are I-V and P-V characteristics of
photovoltaic (PV) panel modeling with different
irradiation at constant temperature as shown in Figure 7
and Figure 8. Power is multiplying VPV with the Iph and
it has been found current and voltage are scaled based on
solar irradiation and temperature, where the generated
current has recorded 8.6 A at the times with 1kW/m2 of
irradiance value and 1.7 A with 0.2kW/m2 of irradiance.
The parameters magnitude will be increased according to
connecting parallel and series PV cells to meet the set up
requirements.
Because of the fixed temperature in this case the power
increases with irradiation rising (the voltage has not
significant change with decreasing on irradiance).
Conversely it the power has inversely proportional with
temperature due to the lessening of the voltage with
temperature rising and fixed irradiation as shown Figure
9. The validation of these results has been investigated
through comparing them with the manufactured I-V
curves of SolarWorld SW 255 (mono) which provided in
[19]. Those results have been modeled by using
MATLAB/SIMULINK toolbox based on the equations
that has been provided in this section.
Figure 7 I-V characteristics of SolarWorld SW 255 (mono)
photovoltaic cell (25 Co
)
Figure 8 P-V characteristics of SolarWorld SW 255 (mono)
photovoltaic cell (25 Co
)
7. Wadhah Esmaeel Ibraheem, Chin Kim Gan, Mohd Ruddin Ab. Ghani
Figure 9 P-V Characteristics of SolarWorld SW 255 (mono) hotovoltaic
cell (1000 kW/m2
)
As shown in the previous results, The PV power output
generation is varied with temperature of the panel surface
and the change of solar irradiation. For the impact study
purpose, the hourly generation of PV system as a function
of solar irradiation at different values can be estimated.
From the collected data regarding to the irradiance and
temperature, the OpenDss software has been used to
model the daily time series PV generation. The explored
results showed that the clearing index which varies from
time to time introduced a fluctuation in global irradiance.
Thus the PV current will be affected by that; thus again
PV power generation will be in oscillation profile.
A three phase of photovoltaic system with 6.12 kW
power and has been tested by DSS with time series
execution. The obtained results in figures below are
based on applying the temperature and irradiance profiles
to the program; they are the daily generating AC
voltages, currents and three phase power at each 5
minutes in the selected day.
Figure 10 PV Power Generation Profile
The capacity of generating power from the solar PV
system at sunny time is greater than a cloudy time.
Obviously, photovoltaic panels have not electricity
generation at night and less in the evening with early
morning compared with the middle of the day. As shown
in the power Fig 10, the PV generation is reduced to the
zero from 6:20 AM until 7:40 AM due to the high density
of clouds that occurred at that time, but PV power has
been generated at maximum at midday from 11 AM until
12.30 AM. After midday until 7 PM thin and thick clouds
had obscured PV panel introduced a generation power
from 25% to 90% at time of heavy and light clouds
respectively. This fluctuation will impact the customer's
voltages in some cases with higher penetration levels as
would be shown later.
The voltage in Figure 11 has the same impact by
irradiation with current and power but with very small
values of volts as proven in section V-I characteristics.
The generating voltage is fluctuating in a range of 1.5
volts.
239
239
240
240
241
241
242
0
1
2
3
4
5
6
7
8
9
1
13
25
37
49
61
73
85
97
109
121
133
145
157
169
181
193
205
217
229
241
253
265
277
Current
(A)
Daily Time (5-minutes)
Current Voltage
Figure 11 PV Panel Daily Voltage and Current Profiles
V. Impact of the PV Plants on Network’s
Voltage and losses
PV plant systems have some unique characteristics
which have to be considered when connecting a high
penetration of such distributed generators into the
distribution system. For preferable investigation of the
impacts that are accompanied with connecting the PV
system into the distribution grids; a simulation of IEEE
test feeders has been done by OpenDss simulator. The
13-bus IEEE test systems has been considered with
specific load profile data to evaluate the effect of
photovoltaic system on distribution grids in terms of
losses and voltage issues with penetration level as well.
V.1 Voltage Assessment of Grid-connected PV Systems
This section identifies the improvement of voltage
which has been occurred due to connect solar PV system
in distribution grids. For this purpose a program code has
been developed to execute the power flow study and
record the voltage improvement through different levels
(10%-50%) of penetration. Penetration level has been
calculated as the percentage of installed power by PV to
the total load which is supplied by the system. The solar
PV system has been connected to the 13-bus IEEE
distribution network at buses (671, 675 and 680). The
expected results have been obtained and the voltage
8. Wadhah Esmaeel Ibraheem, Chin Kim Gan, Mohd Ruddin Ab. Ghani
improvement has been investigated as shown in Figure 12
which shows the variation of the 3 phases of buses
voltages along the feeder as PV penetration increasing.
From the graphs below, it can be seen the impact of PV
generation on regulation zone, where PV installation with
high penetration levels has increased voltages the end
point of the feeder. The regulator has been excited by that
to return back to lower settings which are required to
maintain system voltages within limits. It can be noticed
the voltages at regulator bus in three phases are reduced
at lower than base case (without PV) due to increasing
the PV penetration level which mean regulator changed
the tap position to lower settings as shown in Figure
5.13.
(a) Phase A
(b) Phase B
(c) Phase C
Figure 12 Three Phase Voltages at 13-bus Feeder with PV
Figure 13 Tap-changer Position According to Penetration level
The previous results were modeled by considering that
the system loads are a full load and the meet the full
generation of the PV power output. But in the real life,
the improvement of the voltage of the PV system is
during the daytime where starts with working time of the
PV power generation. In this task the same test system
has been executed with daily time series power flow with
5 minute resolution control mode. A random load profile
is shown in Fig 5 has been selected and applied to the
program after normalization. The injected PV generation
has been modeled by the OpenDss program as shown in
Fig 10 . In this system a monitor has been installed beside
bus bar 671 and the regulator bus to record the voltage at
each 5 minutes at customer end. The execution includes
two scenarios, the first of them is voltage performance of
the customers voltage at bus 671 and on the regulator bus
without PV installation. The second one was with 1800
kW of PV power divided into 3 plants (each one has 600
kWp) at buses (671, 675 and 680). Fig shows the voltage
behavior before and after PV penetration.
Figure 14 Voltage Profile of Bus-Bar 671
As shown in the previous figure, generally, the voltage
of the system has a directly proportional to the amount of
PV power and the improvement starts with working of
PV system from 7AM to 6:30 PM. The peak voltage on
the profile without PV system during that time was 1.02
pu at the bus 671. The 50% of penetration level has
9. Wadhah Esmaeel Ibraheem, Chin Kim Gan, Mohd Ruddin Ab. Ghani
increased the voltages to 1.033 pu at bus 671. But the
voltage of the substation is decreased by PV penetration
due to tap setting which had set the baseline of the
substation voltage in the range of 1.0 pu to 1.0625 pu as
shown in Figure 15. The regulator by this range trying to
maintain the voltage work within the acceptable limit of
standards (+5, -5 pu voltage), but in the night in this case
the voltage drops to minimum due to the full load
consumption. It is worth to mention, the negative impact
may occur due to the voltage fluctuation which is caused
by PV system. Voltage behavior acts the response of the
demand profile with PV integration including a local
oscillation at each minute of irradiation. This fact may
effects the system in case of higher penetration on PV
installations through violating the voltage constrains or
may impacts the tap changer operation by increasing the
number of position changes. In this thesis a mitigation of
voltage fluctuation as a control strategy has been
investigated and will be explored later.
Figure 15 Voltage Profile of Main Substation
V.2 Losses Impact on Grid-connected PV Systems
The feeder losses in distribution systems with PV
integration have been reduced. In this section, the three
phase power flow has been executed in 13-bus IEEE test
feeder system to show the loss reduction by connecting
1800 kW PV system. The results have shown that the
percentage of loss reduction differs with the changing of
PV system location. The maximum reduction of system
losses has been obtained with this capacity of PV
installation is when the PV connection at bus 671 and
675, where it is decreased until 1.37% from total losses
before PV injection which was equal to 3.21%. Figure
5.18 illustrates the percentage of system losses at
different connection buses in 13-bus IEEE system.
Figure 16 Percentage of Losses at Various PV Locations
From that, it can be understood that the place of PV
connection which supplies more loads than others will
introduce more reduction of losses in the system. But it
should be far from the main source because it will
compensate the demand current that would come from
the main source and would be loosed by line parameters.
Thus, the amount of the current which is compensated by
PV system in case of bus 632 connections, is only the
currents that flow from buses 650 to 632, but in case of
bus 671 installation a large loosed current has been
compensated, thus again a lowest losses had been
recorded by that.
From hour to another throughout the daily time, the
losses change with a variation of the load demand. The
time step load flow by OpenDss simulator has applied on
the 13-bus IEEE test feeder to calculate the hourly losses
that occur based on the local load profile. It has
calculated each individual interval losses in the feeder,
the time step interval in this case is one hour. In this task,
3 plants of PV systems, each one is rated with 600 kWp.
Figure 17 shows the hourly demand profile of the loads
that are supplied by 1800 kWp PV system. Solar
generation based on the irradiation at the collected site.
Figure 17 The Load Demand and the PV Generation Profile
As it can be seen, a part of demand load will be served
by the PV power from 7AM until 6PM by a range of
(500kW - 1000kW) power consumption. Thus a loss
reduction will be recorded at that time due to decreasing
10. Wadhah Esmaeel Ibraheem, Chin Kim Gan, Mohd Ruddin Ab. Ghani
of the demand currents that are needed from the main
source. The previous figure illustrates the reverse power
has been occurred by installing this capacity of the PV
power. This power will flow through the system feeders.
In the highest penetration levels, if it will not be
consumed from other loads, it will introduce power losses
in the system.
The provided results in the Figure 18 have shown that
the losses have been reduced during the working time of
PV system. The results have shown that the system was
consuming 1518.7kW/day before PV penetration where a
50% of penetration level has decreased the energy
consumption to 1121.4kW/day. The energy saving in this
case was 397.3kW/day. When the penetration level
increased to 70% the losses was still less than the base
case (before PV connection) but the daily energy saving
decreased to 383.5kW/day. That means the reverse power
as shown in Fig 17 has introduced line losses entire the
system. For this purpose, losses corresponding to the
penetration level has been prepared and it showed that
the U-shape trajectory had been obtained, where the
behavior of losses falls until minimum at 90% and 100%
penetration, that because, at that level there is no power
flowing inside the system, power of PV is equal to load
power. After 100% the losses starts again to rise due to
the revers power from the bus tied PV. Figure 19
illustrates that the benefit of the PV system connection is
until 170% of penetration level, where the losses after
that, will be more than that before PV installation.
Figure 18 System Losses with and wthout PV Instalation
Figure 19 U-shape Trajectory of The Loses Behavior in IEEE 13-bus
System
V.3 Penetration Level Evaluation
Although solar photovoltaic systems are considered
green energy have a wide ranges of positive benefits
applications, there is a limit in penetration level in case of
grid-connected. Maximum penetration level is a
maximum solar power allowable to inject into the
distribution grids without any acceding in voltage or add
extra system equipment. IEEE 1547-2003 recommends
that the penetration level want to not violate electrical
system voltage (Normal system should be inside range of
(0.95 pu – 1.05 pu) voltage), and the electrical system
equipment must not be overloaded [20]. In this paper, the
sensitivity of the voltage magnitude is the way to
calculate the maximum allowable penetration level can be
applied on 13-bus IEEE test system. As has been shown,
the generation power of the connecting PV has reversed
to the grid during the daylight. The reverse power is
introduced when the load downstream is less than PV
plant power generation. This power associates in a
voltage rise in the system [21]. This rise of the voltage
will limit the penetration level in the system. The
penetration level is different from system to another, in
some cases the penetration level is more than 100%
without causing any voltage violation [13]. In this case
also the penetration level can be achieved more than
100% without voltage violation. But it can be calculated
by monitoring the system losses. The injected power at
which the losses of the system would be more than the
base case in Figure 18 is considered as a maximum
penetration level. The maximum penetration level in this
system is 90% (3150 kWp as total capacity of three
plants) where at above this value, losses have been
occurred at 12:00 PM as shown in Figure 20 below. This
case has not violated the voltage where it was still under
limit but it will introduce an extra operation cost.
11. Wadhah Esmaeel Ibraheem, Chin Kim Gan, Mohd Ruddin Ab. Ghani
Figure 20 System Losses without and with Maximum Penetration Level
of PV
Voltage rise depend on the voltage regulation in the
system, where the voltage drop is a function of feeder
length [22]. For this purpose, the length of the feeder
633-671 has been increased from 2000 ft to 6000ft. It has
been found the difference between the voltage profile
(see Figure 21 ) with PV connection and voltage profile
without PV in case of 6000ft length is higher than that in
case of 2000ft length (see Figure 14). This case has been
evaluated by (5) and (6) [22].
))
sin(
)
cos(
(
.
X
R
I L
VdwithoutPV
(5)
)
sin(
.
)
)
cos(
.
(
X
I L
PV
I
R
I L
VdwithPV
where is Vd is voltage drop
IL, IPV are the line and PV currents
and:
Difference = VdwithoutPV VdwithPV
(6)
According to the equations and Figure 21the base case
(without PV) voltage is dropped by increasing the length.
But when connecting the 1800 kWp of PV penetration,
the voltage has been increased with percentage more than
before if compared with Fig 14.
Figure 21 Voltage Profile with and without PV in Case of Longer
Feeder
That means the voltages with PV interconnection
behaves according systems design and topologies. In this
case, the voltage violation will be obtained at higher
penetration level. Thus, the maximum penetration level
can be connected in this system is 3000 kWp which equal
to 86% of total load connected to the grid. The voltage
profile behavior with this capacity of PV power is shown
in Fig 22.
Figure 22 Voltage Profile with Maximum Penetration Level
VI. Mitigation of Voltage Fluctuation by
STATCOM Compensation
The current research considered the reactive power
compensation by STATCOM as control approach to
mitigate the voltage fluctuations. Generally, this device
provides a capacitive or reactive power in cases of under
or upper voltage limit respectively. But in this research it
has been used to ramp the voltage dips that produced by
PV power fluctuation and mitigate the over voltage
causes. MATLAB environment has been interfaced with
OpenDss simulator to develop a strategy of voltage
control. The methodology of this part is taking a
reference (mean) voltage between the upper and lower
value in the voltage profile. IEEE 13-bus system has been
used as a case study for the voltage mitigation with 5
minutes daily PV generation data, the load demand data
were assumed for the test requirements. This method has
been applied to test the system by connecting a 20
MVAR of compensation to ramp the voltage profile that
is caused by installing three PV plants with 3000 kW
solar system at bus 671,675 and 680, the mitigation on
voltage violation and fluctuation has been obtained
The installing of reactive power compensators in
distribution systems is according to its exigency in system
and operation economics. This study has used
STATCOM after considering that the main requirement
is to regulate feeder voltage with power fluctuation which
is produced by connected solar systems. The expected
result is the compensator will mitigate the voltage
fluctuation and the number of Tap changing positions
will be reduced as well. The penetration level of this case
will be increased by using STATCOM mitigation. The
steps of the block diagram in Fig 23 represent the
12. Wadhah Esmaeel Ibraheem, Chin Kim Gan, Mohd Ruddin Ab. Ghani
mitigation approach of voltage fluctuation in this
research.
Time Series
Power Flow
If Voltage < Vreference
Yes
No
Start
OpenDss
Circuit Model
PV System
Installation
PV model
Generation
Profile
Calculate the Vreference
Value of Voltage Profile
Network parameter
· Transformers
· Regulators
· Lines
· Capacitors
· Meters
· Load profile
STATCOM
Installation
Switch ON
Capacitor
Voltage Profile
at Customer Bus
Switch ON
Inductor
Voltage Profiles Before/After
Mitigation
With Tap positions
End
If Voltage < V limit
Yes
Switch OFF
STATCOM
Figure 23 Flow Chart of STATCOM Compensation Methodology
According to the methodology the results of installing
3000 kW solar PV system at three locations in the IEEE
13-bus test system with a certain load profile indicated to
there are a fluctuation of the voltage at the customers’
buses and out of the limit as shown below.
Figure 24 Voltage Profile Before and After Mitigation
By this way, the over voltage has been reduced and the
under voltage has been increased as shown in Fig 24. The
voltages limits in this research are according to Wills,
they are -5% and +5% [23]. Introducing by that a profile
of tap-changer with less tap positions compared with
before mitigation. The curried out results of tap-changer
before and after mitigation showed that the numbers of
tap changing before mitigation was 12 times but after
mitigation it is reduced to 11 times during the day. This
reduction of the mechanical movements will increase the
life time of transformer regulator resulting economic
benefits in the power system.
Then, according to the analysis above, the STATCOM
can increase the penetration level of the grid-connected
PV systems. Therefore within STATCOM control can
install extra capacity of the PV panels without violating
the system constraints. In this case 200 kW at each plant
has been added to the system. The PV system capacities
have become 1200 kW at each plant where the obtained
results showed that the voltage with that amount of PV
generation is still under the limit and also the tap
changing reduction has been obtained as shown the Fig
25. The tap changer movement also has been reduced
from 14 to 11 tap position per day.
Figure 25 Voltage Profiles with Mitigation Control and 1200 kW
installation
13. Wadhah Esmaeel Ibraheem, Chin Kim Gan, Mohd Ruddin Ab. Ghani
VII. Conclusion
This paper studied the impact of the PV plants
installation on the distribution networks in term of
voltages and feeder losses. The case studies claimed that
there is a positive and negative impact of the PV
interconnection, that’s where the interest lies in the
voltage improvement at loads buses and the reduction in
the system losses which can save the energy of the main
source, but nevertheless, there are some concerns that
accompany this in terms of voltages fluctuation and
overvoltage conditions. The impact of the PV on the grid
connected depends on the PV size, install location, and
system topology. This report proved that the penetration
level calculations are very important in installation design
to determine the maximum allowable PV power can be
installed in certain system. This study suggested a control
method to mitigate the voltage fluctuation which may
effect on the regulation and hence on the regulator.
Voltage mitigation can give the opportunity to increase
the PV penetration level in some cases by damping the
overvoltage conditions during the PV generation
working.
Acknowledgements
The authors would like to gratefully acknowledge the
funding support provided by the Ministry of Education
Malaysia under the research grant NO.MTUN
/2012/UTEM-FKE/7 M00015.
References
[1] N. M. Pearsall, “PV research and development in Europe — A
view from the Technology Platform,” 2011 37th IEEE
Photovolt. Spec. Conf., pp. 200–205, Jun. 2011.
[2] A. H. Shamsuddin, “Development of Renewable Energy in
Malaysia-Strategic Initiatives for Carbon Reduction in the Power
Generation Sector,” Procedia Eng., vol. 49, pp. 384–391, Jan.
2012.
[3] M. D. Platzer, “U . S . Solar Photovoltaic Manufacturing :
Industry Trends , Global Competition , Federal Support,” United
State, 2012.
[4] J. Kleissl, M. Lave, M. Jamaly, and J. Bosch, “Aggregate solar
variability,” 2012 IEEE Power Energy Soc. Gen. Meet., pp. 1–3,
Jul. 2012.
[5] D. Mcginn, D. Green, R. Hinrichs-rahlwes, S. Sawyer, M.
Sander, R. Taylor, I. Giner-reichl, S. Teske, H. Lehmann, and D.
Hales, “Renewables 2013 Global Status Reprot 2013,” 2013.
[6] Y. P. Agalgaonkar, S. Member, B. C. Pal, R. A. Jabr, and S.
Member, “Distribution Voltage Control Considering the Impact
of PV Generation on Tap Changers and Autonomous
Regulators,” IEEE Trans. Power Syst., vol. PP, no. 99, pp. 1–11,
2013.
[7] N. Diabi, R.; Belizidia, “Distributed generation influence on the
electric network voltage level,” Int. Rev. Electr. Eng., vol. 3, no.
2, pp. 248–252, 2008.
[8] S. Porkar, S.; Poure, P.; Abbaspour-Tehrani-fard, A.; Saadate,
“Distributed Generation Planning for Losses, Voltage Profile,
Line Congestion and Total System Cost Improvement,” Int. Rev.
Electr. Eng., vol. 4, no. 3, pp. 434–440, 2009.
[9] R. C. Dugan, “The Open Distribution System Simulator (
OpenDSS ),” California, 2012.
[10] PES-Power and Energy Society, “IEEE Distribution Test
Feeders,” Winter Power Meeting, 2009. [Online]. Available:
http://ewh.ieee.org/soc/pes/dsacom/testfeeders/.
[11] I. Richardson and M. Thomson, “Integrated domestic electricity
demand and PV micro-generation model.” Loughborough, UK,
2011.
[12] Wunderground, “Weather History for East Midlands, United
Kingdom,” Weather Underground, 2013. [Online]. Available:
http://www.wunderground.com/history/airport/EGNX/2012/6/16
/DailyHistory.html?req_city=NA&req_state=NA&req_statenam
e=NA&theprefset=SHOWMETAR&theprefvalue=1. [Accessed:
15-Dec-2013].
[13] C. H. Tie and C. K. Gan, “Impact of Grid-Connected Residential
PV Systems on the Malaysia Low Voltage Distribution
Network,” 2013 IEEE 7th Int. Power Eng. Optim. Conf., no.
June, pp. 670–675, 2013.
[14] R. Bharti, J. Kuitche, and M. G. TamizhMani, “Nominal
Operating Cell Temperature (NOCT): Effects of module size,
loading and solar spectrum,” 2009 34th IEEE Photovolt. Spec.
Conf., pp. 001657–001662, Jun. 2009.
[15] R. M. Pandiarajan, “Development of Power Electronic Circuit-
Oriented Model of Photovoltaic Module,” Int. J. Adv. Eng.
Technol., vol. II, no. Iv, pp. 118–127, 2011.
[16] H. Andrei, T. Ivanovici, E. Diaconu, M. R. Ghita, O. Marin, and
P. C. Andrei, “Analysis and experimental verification of the
sensitivity of PV cell model parameters,” 2012 Int. Conf. Synth.
Model. Anal. Simul. Methods Appl. to Circuit Des., pp. 129–
132, Sep. 2012.
[17] K. Ding, X. Bian, and H. Liu, “Matlab-Simulink Based
Modeling to Study the Influence of Nonuniform Insolation
Photovoltaic Array,” 2011 Asia-Pacific Power Energy Eng.
Conf., no. 1, pp. 1–4, Mar. 2011.
[18] P. Hsiao and C. Chang, “Accuracy Improvement of Practical PV
Model,” SICE Annu. Conf. 2010, pp. 2725–2730, 2010.
[19] SolarWorld, “Sunmodule SW 255 mono,” USA, 2012.
[20] R. A. Shayani, S. Member, M. Aurélio, G. De Oliveira, and S.
Member, “Photovoltaic Generation Penetration Limits in Radial
Distribution Systems,” IEEE Trans. Power Syst., vol. 26, no. 3,
pp. 1625–1631, 2011.
[21] Y. Liu, J. Bebic, B. Kroposki, J. De Bedout, and W. Ren,
“Distribution System Voltage Performance Analysis for High-
Penetration PV,” IEEE Energy2030, no. November, pp. 1–8,
2008.
[22] G. K. A. and Y. Baghzouz, “Impact of High PV Penetration on
Voltage Regulation in Electrical Distribution Systems,” Clean
Electr. Power (ICCEP), 2011 Int. Conf., pp. 744–748, 2011.
[23] E. Liu and J. Bebic, “Distribution System Voltage Performance
Analysis for High-Penetration Photovoltaics,” Oak Ridge, TN,
United States, 2008.
AUTHORS’ INFORMATION
1,2,3
Faculty of Electrical Engineering, Universiti Teknikal
Malaysia Melaka, 76100 Durian Tunggal, Melaka,
Malaysia
1
Department of Electrical Engineering, Engineering
Collage, University of Diyala, Baaquba, Iraq
Wadhah Esmaeel Ibraheem received
his master of electrical engineering
from Universiti Teknikal Malaysia
Melaka (UTeM) in 2014. Prior to this
he had graduated his first degree from
Department of Electrical Engineering,
University of Diyala in Iraq (2005-
2009). His research interests lie predominantly in the area
of renewable sources of energy and the impact of PV
system in distribution grids.
14. Wadhah Esmaeel Ibraheem, Chin Kim Gan, Mohd Ruddin Ab. Ghani
Email: wii_eng2012@yahoo.com
Chin Kim Gan received his B.Eng
and M.Sc degrees both in electrical
engineering from the Universiti
Teknologi Malaysia (UTM) and PhD
degree from the Imperial College
London, UK. He is currently a Senior
Lecturer at the Universiti Teknikal
Malaysia Melaka (UTeM). His research interests are
distribution network design, integration of renewable
energy and smart grid.
Email: ckgan@utem.edu.my
Mohd Ruddin Ab. Ghani is a
professor and the Rector of the
Universti Teknikal Malaysia Melaka
(UTeM). Before coming to UTeM,
He was professor and the dean of the
Faculty of Electrical Engineering at
Univesiti Teknologi Malaysia
(UTM). Prof. Mohd. Ruddin Ab. Ghani obtained his
Ph.D. in Systems Engineering and Control from the
University of Manchester Institute of Science and
Technology in 1989. His current research interests
include: dynamic economic load dispatch, unit
commitment, distribution automation, optimization of
large scale power systems, system identification, expert
system applications and advanced control techniques to
power systems. He has published over 100 papers and
articles in the related fields. Besides actively involved in
research and publications, he is also a committee member
of various distinguished boards such as: committee
member of Malaysian International Electro-technical
Commission (IEC), Intensification of Research in Priority
Areas (IRPA) and IEEE Malaysian chapter. He is also
member of Advisory Council Member of Malaysian
Armed Forces Academy and a member of Energy
Technology under Economic planning unit.
Email: dpdruddin@utem.edu.my
View publication stats
View publication stats