This paper analyzes the effect of voltage sag on distribution systems due to the connection of Electric Vehicles (EVs). In order to study the impact of the voltage sag on the power system, two scenarios have been selected in this paper. The distribution system and EVs are modeled using the Electro Magnetic Transients Program (EMTP). The numbers of EVs are predicted based on the number of vehicles in distribution system of Seoul. In addition, the number of EVs is set up using realtime traffic in Seoul to simulate Scenario I and II. The simulation results show that voltage sag can occur if the distribution system has more than 30% of the total number of vehicles.
1. Introduction
To work on solving the global warming problem, countries have enacted policies for reducing carbon dioxide through the Kyoto Protocol in 1997 and the Copenhagen Climate Change Conference in 2009. Accordingly, the need for environmentallyfriendly paradigm changes is on the rise. With the concerns of lowcarbon green growth, global environmental problems, and fossil fuel shortages, public interests in electric vehicles such as Electric Vehicles (EVs), Plugin Hybrid Electric Vehicles (PHEVs), and Hybrid Electric Vehicles (HEVs) are increasing. If unprecedented numbers of EVs are connected to the local power systems, it could result in the deterioration of the power quality, overload, and other system issues. Therefore, the impact of EVs on the distribution system should be analyzed.
Most of previous studies analyzed voltage sag according to the number of EVs. Also, it focused primarily on determining whether the existing or planned generation capacity will be sufficient to supply enough power to meet demand result by EV charging
[1

4]
. Similar studies are found in
[5

9]
. However, these studies do not consider real conditions such as actual traffic volume, actual distribution system etc. for the charging of EVs.
In this paper, the effects of voltage sag on the distribution system caused by the connection of EVs are analyzed by the realtime traffic volume consideration in South Korea and the number of electric vehicles. Therefore, this paper is more realistic then previous paper. In section 2, the EV models such as chargers of EVs and batteries are discussed. Two scenarios are described in section 3; Scenario I assumes a uniform charging scenario in which a constant number of EVs are connected to the distribution system in the charging time, and scenario II assumes that the number of charging EVs changes according to real time traffic volume in the charging time. In section 4, we explain realistic simulation conditions. In section 5, the modeling of South Korea Electric Power Corporation`s (KEPCO) distribution system is discussed. Various simulations for each scenario are performed, and the simulation results are analyzed. Finally, the conclusions derived from our work are discussed in section 6.
2. Modeling of the Electric Vehicle
 2.1 PWM AC/DC converter
The electric charger for an EV is modeled by using the EMTP
[10]
. The PWM converter consists of an input resistor, inductor, switches, antiparallel diodes, and a DC link capacitor for the stabilization of the output voltage
[11

12]
.
 2.2 Lithiumion battery
In this paper, a lithiumion battery is used for the EV because this battery is very promising and it increases the driving range up to 322 or even 483km
[13

15]
. The lithiumion battery for an EV is modeled by using EMTP
[16]
.
Electric Vehicle Models
3. Scenarios
 3.1 Flow chart for scenario design
This section explains the flow chart of scenario design by using realdata.
Fig. 2
shows the flow chart for the designed scenario. As shown in
Fig. 2
, it is comprised of a total of 8 steps.
Flow chart for scenario design
Through the above process, it is possible the create scenario for charging EVs. Therefore, the applied method has been restructured to fit the reality of South Korea by using plans to increase South Korean EVs and traffic volume. In this paper, two scenarios were set up to analyze the impact of EVs on the grid. As shown in
Fig. 3
and
4
, some parameters, such as the number of EVs and electricity consumption are used in scenario I and II. However, there is only one different parameter, which is the simultaneous charging rate. Scenario I has a simultaneous charging rate of 20%, and scenario II depends on the actual traffic volume, where the simultaneous charging rate varies hourly
Flow chart of scenario I
Flow chart of scenario II
 3.2 Scenario I
Scenario I assumes the uniform charging after the close of office hours, where office hours are ranging from 7:00 a.m. to 6:00p.m. according to the transport and maritime statistics annual report from 2011 in MLTM
[18]
.
It was presumed to last over one hour (the average commute time is 45 minutes), after which the vehicle will immediately begin charging at home. It is also assumed that each EV is charged for a total of 11 hours from 7:00p.m. to 6:00a.m. In this scenario, the total battery capacity of 24.4kWh is the same for both compact vehicles and midsize vehicles
[19]
, and it assumes that the EVs are used for commuting only.
Fig. 3
shows the flow chart of scenario I. It decides the battery capacity, the number of EVs, and the connection time for the EVs, and then runs the simulation using the EMTP program. In this paper, we used a 5.7p6 version of EMTP that allows largecapacity simulations. Previous versions of EMTP limit the number of switches. Therefore, many EVs which require switches for connection cannot be allowed in the program.
 3.3 The number of charging EV in scenario I
The simultaneous charging rate for scenario I is 20% according to reference
[17]
. The charging starting time of the EVs does not depend on the total number of vehicles in the hourly volume of traffic. Therefore it is set by uniform charging system.
Table 4
in Appendix shows the charging of 10% of the number of EVs according to scenario I and II.
In the cases of 15%, 20%, and 30% of the total number of vehicles, the number of vehicles charging can be obtained in the manner described above.
 3.4 Scenario II
Scenario II has the same settings as scenario I in terms of the battery capacity and the number of EVs. However, scenario II has a different set point for the simultaneous charging rate.
According to the volume of traffic, the simultaneous charging rate of scenario II varies hourly.
Fig. 4
shows the flow chart for scenario II, which is the same as the flow chart for scenario I, but with a different connection time. The simultaneous charging rate, which depends on the volume of traffic, is applied to scenario II. Hence, the number of charging EVs during per hour changes according to the volume of traffic. The opening rate of the number of recharging vehicles according to the volume of traffic is shown in
Table 1
[17]
. Opening rate means the ratio of EVs that start charging. The meaning of recharging vehicles is the number of changing EVs.
where, V
_{c}
: The number of EVs which are start charging
V
_{t}
: The total number of EVs
Hourly traffic volume according to vehicle opening rate of recharging
Hourly traffic volume according to vehicle opening rate of recharging
 3.5 The number of charging EV in scenario II
In this paper, scenario II assumes the charging start time according to hourly traffic. In scenario II, the number of EVs at each charging initiation time is calculated in
Table 1
.
Table 4
in Appendix shows charging rates when 10% of the numbers of vehicles are EVs according to Scenario II in parentheses. In the cases in which 15%, 20%, and 30% of the total number of vehicles are EVs, the number of vehicles charging can be obtained in the manner described above. The simultaneous charging rate and the number of EVs that are charged hourly in the cases in which 15%, 20%, and 30% of the total number of vehicles are EVs can be obtained in the manner described in Section 3.3.
 3.6 Results of the number of charging EV
The numbers of EVs, according to the scenario, are connected to the connecting point of
Fig. 10
in appendix, and then the voltages at points A and B are measured.
Fig. 5
shows the numbers of EVs that charge per hour according to the simultaneous charging rate and penetration level.
Charging in number of EVs in Scenario I, II
Fig. 5(a)
shows the numbers of EVs that charge per hour with Scenario I.
Fig. 5(b)
shows the numbers of EVs that charge per hour with scenario II. However simultaneous charging rate of
Fig 5(b)
changes according to real traffic volume of South Korea, so the number of EVs being charged keeps changing. According to scenario II, as shown in
Fig. 5(b)
, the number of EVs starting to charge is the highest between 8p.m ~ 9p.m.
4. Simulation Condition Considering Domestic Data
 4.1 Determination of the number of electric vehicles
According to the Vehicle Care Status of the MLTM (Ministry of Land, Transport and Maritime Affairs) in December 2011, the numbers of vehicles registered in South Korea.
Table 2
shows the numbers of vehicles in South Korea and Seoul
[20

21]
.
The number of fossil fuel vehicles in South Korea and Seoul
The number of fossil fuel vehicles in South Korea and Seoul
South Korea has a plan for supplying EVs and implementation strategies, which was established in 2010, and the plan is divided into three cases as follows:
1) 10% of total compact vehicles in South Korea: 152,000 units
2) 10% of total compact and intermediate vehicles in South Korea: 1,000,000 units
3) Electric vehicles: 1,080,000 units
The South Korean government set up the diffusion of EVs of one million, which is 10% of the production goal for compact and intermediate vehicles
[17]
. Accordingly, this study sets up the number of EVs of X S/S ~ Z D/Lin Seoul, and electric vehicle of the X S/S ~ Z D/L is 896 units. X S/S refers to “X Substation” and Z D/L refers to “Z Distribution Line”. In addition, X S/S and Z D/L are the names of actual distribution system of KEPCO (Korea Electric Power Corporation).
 4.2 Capacity of electric vehicle battery
Vehicles in South Korea that run on fossil fuels are segregated according to the size of the displacement (CC: Cubic Centimeter). There is no standard for the classification of EVs. Hence, the EVs are classified after the fossil fuel vehicles are examined
[19]
.
Table 3
shows a comparison of the capacities of EVs and fossil fuel vehicles. Only the compact vehicles and midsize vehicles are separated, because the South Korean government is set to produce 1,000,000 vehicles until 2020, which was 10% of the compact and intermediate vehicles in 2011.
Comparison of capacity of fossil fuel vehicles and EVs
Comparison of capacity of fossil fuel vehicles and EVs
 4.3 Consumption of EV battery capacity
This calculation shows the capacity consumption of EVs considering the average distances driven by automobiles that use fossil fuels. The daily average mileage of an EV is set to 40km
[22]
. The EV battery power consumption is 6.2kWh
[17]
. We can assume that this power energy can be charged completely within 2 hours when we set the home charger capacity to 3.3kW.
5. Simulation
 5.1 System model
In this paper, the modeled distribution system is shown in
Fig.7
in Appendix. The total active load is 28.6MW, and the reactive load is 13.9Mvar. Overhead ground wire and distribution wire are modeled by using LCC (Line Cable Constant) device of EMTP. The “X S/S Y D/L” section in the upper portion is designated as Part 1 and the “X S/S Z D/L” section in the lower portion is designated as Part 2. There are 24 threephase loads in the system
[23]
. Applied distribution system in this paper has a Part1 and Part 2. However, in this paper, the simulation was carried out by considering only Part 2 because simulated conditions are actual conditions corresponding to the area Part 2. In order to analyze the effect on distribution system due to EVs charging, we set the point A and B as charging point. Therefore, in order to consider the worst condition, we choose the B position which has the most amount of load. In otherwise, we also choose the A position which has the least amount of load to consider inverse situation. Also, in order to choose a place where it is possible to charge a large amount at one point, the EVs are connected at one point. In this paper, these places are assumed as supermarkets, parking lots, and airport parking space.
 5.2 Voltage drop analysis method
General analysis for voltage drop uses the increase of the number of EVs
[24

26]
. Therefore, we can find that the value of voltage sag changes according to the increased impedance in eq.2
[27]
.
where,
V
is preevent voltage,
Z_{S}
is existing load and
Z_{D}
is adding load. Subsequently, when the number of charging EVs increases, we can see that the value of voltage drop increases as well.
☞ Our new contributions are as follows;
Impedance
Z_{D}
of existing analysis increases in ascending order, because existing analysis do not consider real conditions of charging of EVs. However, impedance
Z_{D}
used in this paper is decided according to the number of charging EVs ba sed on scenario, and fluctuates according to real conditions such as RealTime Traffic, battery capacity and the number of EVs.
 5.3 Simulation results
Voltage Sag is a short duration reduction in RMS voltage. A voltage sag happen when the RMS voltage drops below 90% for longer than 0.5~30cycles. Therefore, voltage sag in
Fig. 6
~
8
is measured according to reference
[28]
.
RMS voltage
Simulation results in scenario I
Simulation results in scenario II
Fig. 6
shows RMS voltage when penetration level of electric vehicle is 10% according to scenario I and II. A sharp voltage drop occurs instantaneously when EVs are connected to distribution system. After measure the voltage in scenarios I and II, the lowest RMS voltages are selected and it is shown in
Fig. 7
,
8
in p.u.unit.
 5.3.1 Simulation results in scenario I
Table 5
in Appendix shows the simulation results of scenario I.
Fig. 7
shows the plots from
Table 5
in Appendix. In this result, because the simulated conditions of scenario I use a uniform charging from 7:00p.m. to 6:00a.m., all values have the same p.u. for voltage. However in this case, it is within the allowed range of IEEE and the South Korean Standard
[28

29]
. When 30% EVs of the total number of vehicles are connected to the power system, the results are barely within the allowed acceptable range.
 5.3.2 Simulation Results in Scenario II
Table 6
in Appendix shows the simulation results of scenario II.
Fig. 8
shows the plots from
Table 6
in Appendix; it shows the lowest voltage when the simultaneous charging rate is the highest value (9p.m. ~ 11p.m.). In cases in which the EVs are 10% or 15% of the total number of vehicles, voltage sag does not occur at any point. Therefore the results are within the acceptable range of IEEE and the South Korean standard.
However when it is 20% or 30%, voltage sag occurs at certain points. In the cases of 20% and 30% penetration level, the results do not satisfy the acceptable range of IEEE and the South Korean standard from 9p.m. ~ 11p.m. First, in the cases of 20% penetration level, the voltage sag occur during 9p.m. ~ 11p.m. in
Table 6
, when the number of opening rate of recharging vehicle is the highest.
The voltage sag that occurs at this point is lower than the IEEE and the South Korean standard
[28

29]
, which is 0.9p.u., by 0.0047p.u. at point A, and by 0.0048p.u. at point B. In addition, when EVs are 30% of the total number of vehicles, voltage sag occurs from 8p.m. to 11p.m. for three hours, when the number of EVs being charged is the highest. The largest voltage drop that occurs at this point is lower than the IEEE and the South Korea standard, which is 0.9p.u., by 0.0473p.u. at point A, and by 0.0483p.u. at point B. In particular, the voltage sag is very serious in case of the EVs are 30% of the total number of vehicles and are connected to the power system, as shown in
Table 6
.
Fig. 9
is a comparison of scenario I and II. It shows voltage and the number of EVs when electric vehicle penetration level is 30%. This figure clearly illustrates the change of voltage according to the number of charging electric vehicle. Bar graph means the number of EVs and line graph means voltage value (p.u.).
Compare charging number of EVs and voltage
6. Conclusion
This paper confirms the effects of voltage sag on the distribution system when EVs are connected. The numbers of EVs are determined by considering the actual traffic volumes and scenarios that are appropriate to the setting of South Korea. By using traffic data of South Korea, the simulation conditions are set to meet real conditions of South Korea in order to set up the simultaneous charging rate and to initiate the charging of EVs. From the simulation results of scenario I, the voltage drop appears uniformly from 8:00p.m. to 4:00a.m. when vehicles are uniformly charged. The voltage drops are all included within the acceptable range of IEEE and the South Korean standard. In scenario II, the voltage is lowest when the simultaneous charging rate is the highest from 8:00p.m. to 11:00p.m. In the cases in which the EVs are considered to be 10% and 15% of the total number of vehicles, the voltage is within the allowed range of IEEE and the South Korean standard. However, in the cases in which the EVs are considered to be 20% and 30% of the total number of vehicles, the voltage drop does not satisfy the acceptable range of IEEE and the South Korean standard. With these tendencies, it could be concluded that as the popularity of EVs in South Korea continues to rise, appropriate measures need to be taken to prepare for these changes in charging distribution during high initiation time. Moreover, as shown above, we can find that voltage drop that deviates from the IEEE and the South Korean standard occurs at 20% and 30%. Thus, in the future, solution has to be provided to improve the power quality when the number of EVs connected to distribution system. Further research on the solution to power quality improvement will be continued the basis of this research.
Acknowledgements
This work was supported by the Human Resources Development program (No.20124010203300) of the Korea Institute of Energy Technology Evaluation and Planning (KETEP) grant funded by the Korea government Ministry of Trade, Industry and Energy.
BIO
HyoSang Go was born in Korea, 1985. He received his B.S in School of Department of Physics from Chungbuk National University, Korea, 2010. At present, he is working on his Master and Ph.D thesis at Sungkyunkwan University. His research interests include power system transients, protection and power quality with electric vehicles.
DooUng Kim was born in Korea, 1986. At present, he is working on his Master and Ph. D thesis at Sungkyunkwan University. His research interests include power system transients, protection and stability with electric vehicles. He received his B.S in School of Electrical Engineering from Sungkyunkwan University, Korea, 2012.
JunHyeok Kim He received B.S and M.S degree in School of Electrical and Computer Engineering from Sungkyunkwan University, 2012, 2014, respectively. At present, he is working on his Ph. D course. His research interests include power system transients, protection and stability with electric vehicle.
SoonJeong Lee He received his B.S. degree in Department of Electrical and Electronics Engineering from Kangwon National University, 2010 and M.S. degree in College of Information and Communication Engineering from Sungkyunkwan University, South Korea, 2012 respectively. At present, he is working for his Ph.D. thesis in Sungkyunkwan University. His research interests are power quality, electric vehicle and electrical energy storage system
SeulKi Kim received B.S and M.S degree in electrical engineering from Korea University, Korea in 1998 and in 2000 respectively. Since 2000, he has been working as a researcher in power system research group of Korea Electro technology Research Institute (KERI). His research interests are grid connection of wind turbines, voltage stability analysis and power flow analysis.
EungSang Kim (M’00) received the B.S. degree from Seoul National University of Science and Technology, Korea, in 1988 and the M.S. and Ph.D. degrees in Electrical Engineering from Soongsil University, Korea, in 1991 and 1997, respectively. Since 1991, he has been with the Department of Smart Distribution Research Center at the Korea Electrotechnology Research Institute (KERI), Korea, where he is currently a Principal Researcher and serves as a member of Smart Grid Project Planning committee in Korea. His interests are new and renewable energy system designs and developments of wind power, photovoltaic and energy storage systems.
ChulHwan Kim was born in Korea, 1961. In 1990 he joined Cheju National University, Cheju, Korea, as a fulltime Lecturer. He has been a visiting academic at the University of BATH, UK, in 1996, 1998, and 1999. Since March 1992, he has been a professor in the School of Electrical and Computer Engineering, Sungkyunkwan University, Korea. His research interests include power system protection, artificial intelligence application for protection and control, the modelling/protection of underground cable and EMTP software. He received his B.S and M.S degrees in Electrical Engineering from Sungkyunkwan University, Korea, 1982 and 1984, respectively. He received a Ph.D in Electrical Engineering from Sungkyunkwan University in 1990. Currently, he is a director of Center for Power IT (CPIT) in Sungkyunkwan University
Bae Sungwoo
,
Kwasinski Alexis
2012
“Spatial and Temporal Model of Electric Vehicle Charging Demand,”
IEEE Transactions on Smart Grid
3
(1)
394 
403
DOI : 10.1109/TSG.2011.2159278
EtezadiAmoli Mehdi
,
Choma Kent
,
Stefani Jason
2010
“RapidCharge ElectricVehicle Stations,”
IEEE Transactions on Power Delivery
25
(3)
1883 
1887
Hadley S. W.
,
Tsvetkova A.
2008
“Potential Impacts of Plugin Hybrid Electric Vehicles on Regional Power Generation,”
Taylor J.
,
Maitra A.
,
Alexander M.
,
Brooks D.
,
Duvall M.
2009
“Evaluation of the Impact of Plugin Electric Vehicle Loading on Distribution System Operations”
IEEE
Power & Energy Society General Meeting, 2009. PES ’09
1 
6
Han Sekyung
,
Han Soohee
,
Sezaki Kaoru
2010
“Development of an optimal vehicletogrid aggregator for frequency regulation”
IEEE Trans on Smart Grid
1
(1)
65 
72
DOI : 10.1109/TSG.2010.2045163
Shimizu K.
,
Masuta T.
,
Ota Y.
,
Yokoyama A.
2010
“Load frequency control in power system using vehicletogrid system considering the customer convenience of electric vehicles,”
Proc. Int. Conf. PowerSyst. Technol. (POWERCON)
1 
8
Pillai J.
,
BakJensen B.
2011
“Integration of vehicletogrid in the western Danish power system”
IEEE Trans on Sustainable Energy
2
(1)
12 
19
Singh Mukesh
,
Kumar Praveen
,
Kar Indrani
2012
“Implementation of Vehicle to Grid Infrastructure Using Fuzzy Logic Controller”
IEEE Trans on Smart Grid
3
565 
577
DOI : 10.1109/TSG.2011.2172697
Maitra A.
,
Alexander M.
,
Brooks D.
,
Duvall M.
2010
“Evaluations of plugin electric vehicle distribution system impacts”
IEEE Power and Energy Society General Meeting
1 
6
Kim DooUng
,
Kim JunHyeok
,
Go HyoSang
,
Seo HunChul
,
Kim ChulHwan
,
Kim Eungsang
2012
“Modeling of Single Phase PWM AC/DC Converter for EV using EMTP/MODELS”
KIEE Summer Conference & General Meeting
Lim JeeWoo
,
Kwon BongHwan
1999
“A PowerFactor Controller for SinglePhase PWM Rectifiers”
IEEE Trans. on Industrial Electronics
46
(5)
1035 
1037
DOI : 10.1109/41.793353
Tipsuwanporn V.
,
Intajag S.
,
Tarasantisuk C.
2004
“Enhanced Control Design of Single Phase ACDC Converter Using Power Balance Calculator”
Power Electronics and Motion Control Conference
Gómez J. Vehiclelos
,
Morcos Medhat M.
2003
“Impact of EV Battery Chargers on the Power Quality of Distribution Systems”
IEEE Trans on Power Delivery
18
(3)
975 
981
Chen Min
,
Rinc´onMora Gabriel A.
2006
“Accurate Electrical Battery Model Capable of Predicting Runtime and IV Performance”
IEEE Trans on Energy Conversion
21
504 
511
DOI : 10.1109/TEC.2006.874229
Kim JunHyeok
,
Lee SoonJeong
,
Kim EungSang
,
Kim SeulKi
,
Kim ChulHwan
,
Prikler Laszlo
2014
“Modeling of Battery for EV using EMTP/ATPDraw”
Journal of Electrical Engineering &Technology
9
(1)
98 
105
DOI : 10.5370/JEET.2014.9.1.098
2010
“Establishment Guideline for EV Charging Infrastructure”
Seoul Development Institute
57 
68
2012
Road Traffic Authority
http://www.koroad.or.kr/
2012
“vehicle management legislation in Korea”
MLTM(Ministry of Land Transport and Maritime Affairs)
2012
“Statistical yearbook of MLTM”
2012
“VehicleVehiclee Status”
Statistics Korea
http://kostat.go.kr
Park KeonWoo
,
Seo HunChul
,
Kim ChulHwan
,
Jung Changsoo
,
Yoo YeonPyo
,
Lim YongHoon
2009
“Analysis of the Neutral Current for TwoStepType Poles in Distribution Lines”
IEEE Transactions on Power Delivery
24
1483 
1489
DOI : 10.1109/TPWRD.2009.2021031
Sortomme Eric
,
Hindi Mohammad M.
,
MacPherson S. D. James
,
Venkata S. S.
2011
“Coordinated Charging of PlugIn Hybrid Electric Vehicles to Minimize Distribution System Losses”
IEEE Transactions on Smart Grid
2
198 
205
DOI : 10.1109/TSG.2010.2090913
Singh Mukesh
,
Kumar Praveen
,
Kar Indrani
2012
“Implementation of Vehicle to Grid Infrastructure Using Fuzzy Logic Controller”
IEEE Transactions on Smart Grid
3
(1)
565 
577
DOI : 10.1109/TSG.2011.2172697
2011
“Smart load management of plugin electric vehicles in distribution and residential networks with charging stations for peak shaving and loss minimization considering voltage regulation”
877 
888
DOI : 10.1049/ietgtd.2010.0574
Bollen Math H. J.
2000
“Understanding Power Quality Problems: Voltage Sag and Interruption”
1995
IEEE Recommended Practice for Monitoring Electric Power Quality
2009
“Power electric system reliability and maintain quality standards”