Electric Power Systems Research 140 (2016) 107115

Contents lists available at ScienceDirect

Electric Power Systems Research

journal homepage: www.elsevier.com/locate/epsr

Impact of dynamic and static fast inductive charging of electric

vehicles on the distribution network
I. Karakitsios , E. Karfopoulos, N. Hatziargyriou
National Technical University of Athens, School of Electrical and Computer Engineering, Greece

a r t i c l e i n f o a b s t r a c t

Article history: Fast inductive charging technologies allow the exchange of high power quantities (>20 kW) between an
Received 12 May 2015 Electric Vehicle and the electrical grid in contactless way. This demand can signicantly modify the load
Received in revised form prole of a distribution network and affect its operation and planning. Thus, it is necessary to quantify the
15 November 2015
grid impact of a network of fast inductive chargers and dene the maximum allowable deployment level
Accepted 18 June 2016
which does not violate the technical constraints of the network. This paper introduces a methodology
Available online 30 June 2016
for grid impact analysis of fast inductive charging technologies into distribution networks. The proposed
methodology is implemented in a realistic model of a Greek MV distribution feeder providing indicative
Keywords:
Electric vehicles
qualitative and quantitative results.
Inductive charging 2016 Elsevier B.V. All rights reserved.
Distribution network

1. Introduction
The charging duration which is one of the major electric vehicle
(EV) drivers concerns, can be signicantly reduced by fast charging.
There are two fast charging alternatives, namely conductive and
inductive charging. Conductive charging of electric vehicles implies
the direct contact between the EV and the station established with
the interconnection of a charging cable [1]. An extended analysis of
conductive charging technologies (i.e. unidirectional and bidirec-
tional power electronic topologies) is presented in [1]. The main
drawback of such solutions lies in the need of active involvement of
the EV user in the initiation of the charging process, while the cable
used could present a risk of electrocution, when used in rainy or
snowy environments. On the contrary, inductive charging infras-
tructures enable the transfer of power between the station and
the vehicle without the need of a physical connection [13]. The
operational principle of a typical inductive charging infrastructure
lies in the wireless transfer of energy between two magnetically
coupled coils: the primary coil which supplies the power and it is
placed at the charging stations side and the secondary coil which
receives the power and it is placed onboard of the car. The opera-
tion principle of an inductive charging station is similar to the one
of a typical transformer. The main difference lies in the coupling
Corresponding author. Tel.: +30 2107723696.
E-mail addresses: jkarak@power.ece.ntua.gr
(I. Karakitsios), ekarfop@power.ece.ntua.gr (E. Karfopoulos), nh@power.ece.ntua.gr
(N. Hatziargyriou).
medium between the two coils (i.e. air) resulting in a lower cou-
pling co-efcient between them in the case of inductive charging
compared to the typical transformer. For this reason, the induc-
tive power transfer system is referred in the literature as loosely
coupled system [3].
Inductive charging technologies are classied into two sub-
categories: the static chargers [49], where EVs charge during
non-commuting hours and the dynamic (or on-route) ones, which
enable EV battery charging, while the EV is moving on the road
[4,1017]. The wireless power exchange between the EV and the
grid which enables the dynamic charging of the battery increases
the driving autonomy of electric vehicles satisfying the drivers
demand for maximum travel distance.
The positioning (i.e. the air gap and the misalignment) of the
inductive charging equipment of the electric vehicles over the
inductive part of the charging station denes the charging power
rate as a fraction of the nominal power of the charging station. Dif-
ferent EV placement approaches are introduced in the literature
[18,19] aiming to achieve the maximum charging power rate.
In the literature, there are several studies concerning the grid
impact of fast conductive chargers [2029], but there is limited
research concerning fast inductive charging, especially dynamic
one [3034]. The studies in [2024] examine the operation of
fast conductive chargers and their real-time impact on the grid,
in terms of voltage or current variations. Other studies [2529]
assess the impact of fast conductive charging on the daily load
prole. This is estimated based on refueling needs [25,26], home
arrival time of commuters [27] or real/statistical mobility patterns
for conventional vehicles [28]. In [28,29], a lowest bound for the

http://dx.doi.org/10.1016/j.epsr.2016.06.034
0378-7796/ 2016 Elsevier B.V. All rights reserved.
108 I. Karakitsios et al. / Electric Power Systems Research 140 (2016) 107115

Fig. 2. Amount of time the vehicles remain on the stations.

Fig. 1. Charging events occurring during the day, and the relevant approximated Table 1
distribution. Gaussian distribution parameters regarding charging session initiation.

Component Parameter
state-of-charge (SOC) of an EV battery is considered as a criterion p  
for charging request. The demand prole of static inductive charg-
1 0.267656 9.197 0.987121
ers and its impact on the daily load curve are presented in [30],
2 0.032443 10.3 0.443215
while the energy needs of dynamic ones are analysed in [31,32]. In 3 0.000872 14.05 0.179605
[33,34], the provision of real-time V2G services, regarding dynamic 4 0.333774 16.73 1.636245
inductive chargers, is examined. It should be mentioned, that in 5 0.137854 11.93 0.697561
6 0.015256 15.95 0.429638
all the aforementioned studies [3034], the impact of the charg-
7 0.036219 15.05 0.640073
ing load prole of inductive chargers on network operation is not 8 0.177336 13.66 0.909339
assessed.
This paper introduces a methodology for estimating the charg-
ing demand of fast static and dynamic inductive chargers and The charging sessions, as a percentage of all charging events
assesses its impact on the operation of a distribution network, reported in the examined period, are presented in Fig. 1. It is evident
in terms of voltage prole, line loading and network losses. The that all charging events occur between 7 am and 9 pm. Furthermore,
demand due to conductive and inductive charging are considered two peak periods can be observed during the day: a peak in the
in parallel and/or separately, simulating more realistic conditions morning demand (8.0010.00 a.m.) and a peak in the middle-day
for the 24-h grid impact analysis. demand (12.0018.00 am).
In Section 2 the methodology for estimating the charging In order to dene when a particular charging event is expected
demand prole of fast inductive chargers is introduced. The EV to occur during the day, a Gaussian probability distribution, with n
charging demand is superimposed on the given network load pro- components, is assumed. The probability density function of such
le of a realistic MV distribution feeder operating in Greece. The a distribution is dened as:
technical characteristics of the network and the simulation sce-
narios are presented in Section 3. The results of the analysis are 
n
1 2 2
presented and analyzed in Section 4. Conclusions are drawn in f (x) = pi e((x) /2 ) (1)
 2
Section 5. i=1

2. Estimation tool
2.1. Static inductive charging
Inductive charging is a recently developed technology with few
actual applications, thus, there is no real world operational expe-
rience. For the purpose of this paper, the operational behaviour of
fast stationary inductive charging can be assumed similar to the
one of fast conductive charging. Consequently, the power prole
analysis can be performed based on an equivalent demand prole
derived from real fast conductive charging stations. In this respect,
data on more than 20 actual fast conductive charging stations for a
period of a month were processed, in order to observe the hour of
the day and the duration of the charging sessions. This data was pro-
vided by DBT, a French company offering charging station solutions,
within the FastInCharge1 project. Due to the commercial nature of
this information, only limited aggregated and processed data can
be published (Figs. 1 and 2).
1
http://www.fastincharge.eu/.
where  and  are the mean and standard deviation of each com-
ponent (the variance is, therefore,  2 ) and p species the mixing
proportions of each component. A Gaussian distribution with 8
components, with values presented in Table 1, is a good approx-
imation to the distribution depicted in Fig. 1.
The second parameter, which is necessary to estimate the
energy demands for static inductive charging, is the expected dura-
tion of each charging event. The available data about fast conductive
charging stations (Fig. 2) indicates that half of the EVs remain in the
stations around 30 min, while none of them remains more than 1 h.
Around 30% of the vehicles charge for 3045 min, while 12.5% and
8.5% of the charging events have durations of 015 and 4560 min,
respectively. The Gaussian distribution of the duration of the charg-
ing events is approximated by the four parameters presented in
Table 2.
The tool developed in order to dene the total EV demand is
presented in Fig. 3.
Two main EV users can be distinguished:
1. The users adopting home charging (Level 13.6 kW or Level
211 kW)
2. The users exploiting fast static inductive charging options.
 I. Karakitsios et al. / Electric Power Systems Research 140 (2016) 107115
Table 2
Gaussian distribution parameters regarding the EV parking duration.
Component Parameter
p 
1 0.300336 1.8590
2 0.017057 2.0100
3 0.667935 2.6190
4 0.020148 0.5588
Fig. 3. Estimation tool of the total EV demand.
When considering home charging, the energy needs are esti-
mated by the left-hand side of the owchart in Fig. 3 [35]. Two
charging levels are considered: Level 13.6 kW and Level211 kW,
while the vehicles examined are of classes L7e and M1, according
to Directives 2002/24/EC2 and 2007/46/EC.3 Home charging is con-
sidered to take place after the vehicle has completed the last trip of
the day. The duration of charging depends on the daily travel dis-
tance and the battery consumption. In order to dene the energy
needs of home charging, the arrival time of the vehicle and the dura-
tion of the charging process, are randomly dened according to the
class of the vehicle. Due to the stochastic nature of the EV mobil-
ity parameters, Monte Carlo is applied for extracting the estimated
daily charging prole.
The energy needs of the fast static inductive chargers are esti-
mated as seen in the right-hand side of the owchart of Fig. 3.
2
http://eur-lex.europa.eu/legal-content/EN/TXT/PDF/
?uri=CELEX:32002L0024&from=EN.
3
http://eur-lex.europa.eu/legal-content/EN/TXT/PDF/
?uri=CELEX:32007L0046&from=EN.
109

0.553311
0.133643
0.927724
0.238436
Fig. 4. Process to dene the number of fast inductive chargers in the grid required
to serve the charging needs of EV users.
The arrival of each vehicle to the charging stations and the dura-
tion of the charging process are randomly generated according to
a probability distribution function.
Another parameter is the number of fast static inductive charg-
ers required to full the charging needs of the EVs. The method
dening the number of fast inductive chargers is described in Fig. 4.
The hour of the day that each vehicle arrives at a station and the
duration of charging are dened according to the relevant distri-
bution functions described before, and the process is followed for
all the vehicles charging on fast inductive charging stations. Apply-
ing the Mode Carlo method, an average number of vehicles A(t),
occupying the stations can be dened for each one of the timeslots
t examined. The number of fast static inductive chargers required
to full the charging needs of EV users is dened by the following
formula
C = maximum{A(t)} (2)
2.2. Dynamic inductive charging
For dynamic inductive charging, the denition of the exact
energy needs of electric vehicles requires several stochastic param-
eters: trips performed by each vehicle during a day, emergency
charging needs, trafc on the roads, speed of the vehicle etc. More
than one type of dynamic stations may exist, each one with individ-
ual set of parameters: stations placed at cross sections or before a
110 I. Karakitsios et al. / Electric Power Systems Research 140 (2016) 107115
Fig. 5. Process to dene the maximum number of on-route chargers.
trafc light, stations placed at a normal speed road within an urban
area, or high-speed stations in a highway. Such detailed analysis
from the grid perspective would be impossible under the scenario
of a mass deployment. In this respect, a more simplistic approach is
adopted for dening the maximum number of dynamic inductive
stations that can be installed in a grid, without causing any grid
related issue, as illustrated in Fig. 5.
At each iteration of the process, the maximum number of charg-
ers is increased by one and the number of chargers operating at each
hour is dened, taking into account the charging needs and the
trafc on the roads. The users are not expected to use the chargers
for emergency charging, neither immediately before or after night
home charging, or in other terms too early in the morning or too late
in the evening. Therefore in order to evaluate the charging needs
of the drivers, while they are away from home and in case of an
emergency situation, we rely on the distribution dened in Fig. 1.
Regarding trafc, the number of vehicles moving on the roads at
particular hours of the day was considered based on real data pro-
vided by the Department for Transport (DFT) in Great Britain.4 The
average percentage of vehicles moving on the roads for a weekday
is illustrated (Fig. 6).
Let Pemergency be the probability of a charger being occupied due
to the charging needs as dened in Fig. 1, and Ptrafc the proba-
bility of a charger being occupied due to the trafc on the roads,
4
https://www.gov.uk/government/collections/road-trafc-statistics.
Fig. 6. Percentage of vehicles moving on the roads.
Fig. 7. Charging probability regarding the on-route charging stations.
as dened in Fig. 6. Then the probability of an on-route inductive
charger serving a vehicle is:
Pon-route (h) = Pemergency (h) Ptraffic (h) (3)
where h is the hour of the day examined.
The daily probability Pon-route is depicted in Fig. 7. It is clear
that only a small amount of chargers are expected to operate very
early in the morning and very late in the evening, while an increased
amount of vehicles are expected to charge between 9 and 10 am,
and between 2 pm and 6 pm. At these time intervals the trafc on
the roads is increased, while an increased need for fast charging is
also observed. Assuming that M is the maximum number of charg-
ers installed in the grid, the number of chargers Nh , operating at
hour h will be:
Pon-route (h)
Nh = M (4)
Pon-route (hmax )
where hmax is the hour where the maximum charging probability
is observed.
The dynamic charging demand is added to the grid demand
together with the total EV demand, comprising both home charging
and static inductive charging. Power ow analysis is performed, in
order to check the following constraints: thermal limits regarding
loading of lines and transformers and voltage limits regarding
the maximum (Vmax ) and minimum (Vmin ) voltages at all nodes
(5% of the nominal value, Vnom ), and voltage deviation: Vdev =
V Vmin
100 max 2Vnom
%, (below 3%). In case any of the above constraints
is violated, the process is terminated.
The simulation tools do not aim to optimize the energy pro-
le of the inductive charging infrastructure but rather to estimate
 I. Karakitsios et al. / Electric Power Systems Research 140 (2016) 107115
Fig. 8. MV distribution line in the area of Katerini, Greece.
its additional demand introduced in the network load considering
both stochastic (i.e. the probability of the occurrence of a charg-
ing event during a specic hour in the day, the probability for
the duration of each charging event and the mobility prole) and
deterministic parameters (i.e. total number of EVs, type of elec-
tric vehicle, battery consumption and the nominal power of each
charging station type).
Moreover, the proposed demand estimation tools for inductive
and conductive charging can be used as input to any power ow
analysis tool, commercial or custom-made. In this manuscript, the
load ow analysis tool for radial distribution networks developed
in [36] is employed. The proposed simulation tools update the
demand prole of each bus of the studied distribution network.
The additional demand of inductive chargers is over-imposed to
the non-EV load prole of the network which is afterwards used
for load ow analysis.
3. Study case
For simulation purposes, a single radial feeder of a primary
transformer station in the area of Katerini is considered (Fig. 8). The
network comprises 35 MV/LV substations, 3 of 160 kVA installed
capacity, 1 of 250 kVA, 3 of 400 kVA, 1 of 500 kVA, 22 of 630 kVA, 1
of 1000 kVA and 4 of 1260 kVA. The installed capacity of the primary
HV/MV substation is 25 MVA. The single phase simulation model of
the distribution line is presented in Fig. 9.
The reference case scenario (Scenario A, as depicted in Table 3)
assumes no EVs in the grid.
Table 3
Description of the scenarios under study.
Scenario Sub-case Users charging at home
Scenario A 0 0
Scenario B 1000
Scenario C I 1000
II 1000
Scenario D I 1000
II 1000
111
Fig. 9. The single phase simulation model of the distribution line of Katerini.
In Scenario B, an EV eet of 1000 electric vehicles is considered
that uses home charging. Two classes are considered: L7e and M1.
The percentage of each class in the eet is presented in Table 4.
The majority of the eet comprises M1 commercial vehicles, while
a small number of smaller L7e vehicles are considered. The drivers
use two charging levels: Level 1 (3.6 kW) and Level 2 (11 kW). The
probability of a class driver charging at a particular level, as well
as the consumption per vehicle class is illustrated in Table 4. The
distance traveled is considered as a stochastic parameter dened by
the normal distribution with a mean value of 35 km and a standard
deviation of 5 km. The arrival time for each vehicle is also randomly
dened according to the normal distribution, with a mean value
observed at 19:00 pm and a standard deviation of 1 h.
Fast inductive chargers are expected to deliver an average power
of 30 kW to the electric vehicle, as explained in [4]. An average
efciency of 90% and 85% is considered, for static and dynamic
inductive charging, respectively. For static inductive charging, two
sub-cases are considered. In the rst case, 100 users are considered
to daily use fast static inductive chargers, while a greater penetra-
tion of static inductive chargers is considered in the second case,
where 300 EV users daily rely on such chargers. The hour of the day
each vehicle enters a static inductive charger, as well as the dura-
tion of each charging event is randomly dened according to the
distributions described in Section 2.
Users relying on fast inductive charging solutions Dynamic inductive chargers
No
0 No
100 No
300 No
100 Yes
300 Yes
112 I. Karakitsios et al. / Electric Power Systems Research 140 (2016) 107115

Table 4
Parameters for the evaluation of the home charging needs.

Parameter Class

L7e M1

Percentage of EVs 20% 80%

Probability of charging level L1 L2 L1 L2
90% 10% 85% 15%
Consumption (kW h/km) 0.13 0.16
Distance traveled (km) Mean value Standard deviation Mean value Standard deviation
35 5 35 5
Arrival time Mean value Standard deviation Mean value Standard deviation
19:00 1h 19:00 1h

Fig. 11. Loading of the lines at the peak hour load, for Scenarios A and B.
Fig. 10. Load curves for Scenarios A and B.

Dynamic inductive charging is considered in Scenario D. In order

to evaluate the maximum number of dynamic inductive charging
stations that can be installed in the grid, the tool described in Fig. 5
is applied, considering two sub-cases for static inductive charging
preferences, as depicted in Table 3. In Scenarios D-I and D-II 100 and
300 EV users, respectively daily exploit static inductive charging
solutions.
The present grid impact analysis aims to identify the capac-
ity limits of the existing network infrastructures to host inductive
charging stations. In this respect, the worst case scenario means
that the charging infrastructures are evenly placed in the most
loaded buses of the grid.

4. Results
The load curves for the case of Scenarios A and B are presented
in Fig. 10. For Scenario A, the peak of the load curve (7.28 MW) can
be observed between 18:00 and 19:00 and the demand remains
high for the next few hours. In Scenario B, an increase of 27.88% is
observed in the evening demand due to the synchronization of the
additional charging demand of 1000 EVs with the high household
consumption. More than 60% of the total EV demand is covered
during the hours, when high network consumption occurs (i.e.
between 18:30 and 20:30).
Fig. 11 indicates the maximum loading of the lines expressed as
percentage of their thermal limit in a 24-h based simulation period.
The line sections closer to the substation are loaded higher than the
furthest ones, since they are own by the aggregated demand of the
downstream nodes. It should be noted that lines serving approx-
imately the same demand appear with different loadings, due to
the differences in their thermal limits. The increase in the net-
work demand introduced by home charging in Scenario B results
in increased line loading, as illustrated in Fig. 11. The highest line
loading in Scenario B is 87.7% and it is observed in line 12 (con-
necting buses 12 and 13). Moreover, due to the additional home
Fig. 12. Voltage deviation for Scenarios A and B.
charging demand, an increase of 7.75% is observed in the total daily
active losses of the grid compared to Scenario A.
Due to the EV deployment a decrement is noticed in the voltage
values without however violation of the allowable regulation lim-
its. More specically, the minimum voltage value (0.959 p.u.) and
the maximum voltage deviation (1.19%) are noticed in buses 72 and
73, which are the furthest buses from the distribution transformer.
An increase in the voltage deviation is observed in all buses of the
network compared to the non-EV case (Scenario A), as illustrated in
Fig. 12, due to the additional demand introduced by home charging
in Scenario B.
The methodology described in Section II for the two sub-cases
of Scenario C (Scenarios C-I and C-II), results in 6 and 19 static
inductive chargers daily required to serve 100 and 300 EV users,
respectively. The EV demand for Scenarios C-I and C-II is illustrated
in Fig. 13. The implementation of fast static inductive charging
introduces an additional demand during the morning and middle-
day hours, which is added to the evening demand introduced by
 I. Karakitsios et al. / Electric Power Systems Research 140 (2016) 107115
Fig. 13. EV Demand for Scenarios B, C-I and C-II.
Fig. 14. Percentage of static inductive charging stations operating each hour of the
day.
Fig. 15. Load curves for Scenarios C-I and C-II.
the home charging (Scenario B). More specically, when all the sta-
tions are operating at the same time (i.e. between 9:30 and 10:00,
as indicated in Fig. 14), the EV demand is increased by 0.2 MW and
0.63 MW in Scenarios C-I and C-II, respectively. Furthermore, the
operation of static inductive charging stations provokes a slight
increase in the peak of the total EV demand between 19:15 and
19:30. At this time interval and considering Scenarios C-I and C-
II, the percentage of operating charging stations is 33% and 26%,
respectively (Fig. 14), resulting in an EV demand increase of 3.19%
and 7.97% compared to Scenario B.
Due to the operation of a small number of static inductive charg-
ers at the peak hour (19:1519:30), a slight increase of 0.72% and
1.79% (compared to Scenario B) is noticed in the peak load of the
network in Scenarios C-I and C-II, respectively (Fig. 15). Addition-
ally, it is evident in Fig. 15 that the operation of static inductive
113
Fig. 16. Load Increment for Scenarios C-I and C-II compared to Scenario B.
chargers during the day introduces an increase in the networks
load during the morning and middle-day hours. When considering
a small number of charging events at the static inductive charg-
ing stations (i.e. Scenario C-I and 100 charging events) a smooth
increase of approximately 3% is observed in the networks load
between 8:30 and 16:00 (Fig. 16). However, a more volatile varia-
tion is noticed in the load increment when considering an increased
number of charging events on static stations (Scenario C-II). More
specically, in the case of Scenario C-II, an increase of approxi-
mately 9.9% is noted in the systems load between 9:00 and 10:00,
while the relevant increase between 14:00 and 16:00 is approxi-
mately equal to 9%.
The additional demand of static inductive chargers increases the
active network losses by 2.42% and 7.26% in Scenarios C-I and C-
II respectively, compared to Scenario B. A slight increase is also
noticed in the loading of line 12 (connecting buses 12 and 13), which
is loaded at 88.3% and 89.4% of its thermal limit for the case of
Scenarios C-I and C-II.
The voltage deviation and the maximum/minimum voltage val-
ues remain within acceptable limits for both sub-cases of Scenario
C. The minimum voltage value at buses 72 and 73 is 0.959 p.u., while
the maximum voltage deviation (at the same buses) is 1.20% and
1.21% for Scenarios C-I and C-II, respectively.
The impact of dynamic inductive charging in the EV demand is
examined in Scenario D. For the two cases of Scenario D (where 100
and 300 EV users daily use static inductive charging) the resulting
maximum number of simultaneously operating chargers is 6 static
and 68 dynamic inductive chargers for Scenario D-I, and 19 static
and 61 dynamic inductive chargers for Scenario D-II.
Dynamic inductive charging introduces a high demand in the
morning and middle day hours, as illustrated in Fig. 17. Between
18:00 and 19:00 in the evening an increased dynamic inductive
charging demand (i.e. 1.3 MW and 1.2 MW for Scenarios D-I and
D-II, respectively) is noticed due to the operation of a high number
of dynamic chargers serving the users returning home from work.
This dynamic charging demand coincides with the increased home
charging demand resulting in a considerable peak in the total EV
demand (3.22 MW and 3.21 MW in the case of Scenarios D-I and
D-II respectively).
An increase of 44.1% (compared to Scenario A) is also noticed in
the total peak demand, which is noted between 18:45 and 19:00,
for the two cases of Scenario D (Fig. 18). This is justied by the fact
that the EV peak demand observed between 18:45 and 19:00, is
dened mainly by the home charging, as well as by the operation
of inductive chargers (static and dynamic). Moreover, the operation
of static and dynamic inductive chargers results in a considerable
increase in the total demand during the morning and middle-day
hours. Comparing the load increment introduced by Scenarios D-
I and D-II with the respective one in Scenario B (Fig. 19), it can
be concluded that the operation of inductive charging stations is
114 I. Karakitsios et al. / Electric Power Systems Research 140 (2016) 107115

Table 5
Summary results for the scenarios examined.

Parameter Scenarios

A B C-I C-II D-I D-II

Peak demand (MW) Total 7.28 9.31 9.37 9.47 10.49 10.49
EV 0.00 2.09 2.16 2.26 3.22 3.21
Hour 18:0019:00 19:1519:30 19:1519:30 19:1519:30 18:4519:00 18:4519:00
Daily active losses (MW) 11.5 12.4 12.7 13.3 16.0 16.3
Max. voltage deviation (%) 0.89 1.19 1.20 1.21 1.41 1.42
Min. voltage value (p.u.) 0.965 0.959 0.959 0.959 0.955 0.955
Max. line loading (%) 68.5 87.7 88.3 89.4 99.9 99.9

Fig. 19. Networks load increment, compared to Scenario B, in the case of Scenarios
D-I and D-II.

the main responsible for the considerable increase in the morning

Fig. 17. EV Demand for Scenarios (a) D-I and (b) D-II.
demand (9:009:30). At this time interval, the load increment is
approximately 33% and 37% for Scenarios D-I and D-II, respectively.
Additionally, the inductive charging demand signicantly increases
the networks load during the middle-day and early evening hours.
More specically, between 16:00 and 17:00 an average load incre-
ment of 40% and 41%, respectively is noted for Scenarios D-I and
D-II, as shown in Fig. 19.
Furthermore, the daily active losses are increased by 39.13% and
41.74% for Scenarios D-I and D-II, compared to Scenario A, due to
the operation of dynamic inductive chargers.
As far as the voltage quality is concerned, no violations in the
voltage operation constraints are observed in Scenario D. The max-
imum voltage deviation observed in buses 72 and 73 is equal to
1.41% and 1.42% for Scenarios D-I and D-II, while a slight decrease
of 1.04%, compared to Scenario A is noticed in the minimum voltage
values of the same buses.
The simulation results for the different scenarios are summa-
rized in Table 5.

5. Conclusions

The simulation results presented in this paper provide insights

concerning the potential network problems which may be caused
by fast inductive charging deployment. Such technologies are
mainly expected to be used in urban environments, possibly cre-
ating network overloading problems. Thus, the additional demand
introduced by inductive charging into distribution networks should
be taken into account by distribution system operators when plan-
ning their network expansion strategy.
The grid impact analysis presented examines the ability of
an urban distribution network to host fast inductive (static and
dynamic) charging stations, according to the network capacity (i.e.
line thermal limits, transformers nominal power). The outcomes of
this analysis can be exploited to study the grid adequacy under any
charging stations placement scenario. The grid hosting capacity of
Fig. 18. Load curves for Scenarios D-I and D-II.
fast inductive charging stations can be exploited for the denition of
 I. Karakitsios et al. / Electric Power Systems Research 140 (2016) 107115
EV smart management aiming to facilitate their wide deployment
without premature grid reinforcements.
The authors next research step is the use of the proposed simu-
lation tools in order to assess the benets of exploiting renewable
energy resources and storage units (i.e. second-hand EV batter-
ies) for supporting grid operation during high EV charging and
for enabling higher deployment of inductive chargers considering
existing network infrastructures.
Acknowledgement
This work was supported by the FP7 funding through FastIn-
Charge project funding under grant agreement 314284.
References
[1] M. Yilmaz, P.T. Krein, Review of battery charger topologies, charging power
levels, and infrastructure for plug-in electric and hybrid vehicles, IEEE Trans.
Power Electron. 28 (5) (2013) 21512169.
[2] S. Li, C.C. Mi, Wireless power transfer for electric vehicle applications, IEEE J.
Emerg. Sel. Top. Power Electron. 3 (1) (2015) 417.
[3] S. Lukic, Z. Pantic, Cutting the cord: static and dynamic inductive wireless
charging of electric vehicles, IEEE Electrif. Mag. 1 (1) (2013) 5764.
[4] N.D. Madzharov, R.T. Ilarionov, A.T. Tonchev, System for dynamic inductive
power transfer, Ind. J. Appl. Res. 4 (7) (2014).
[5] M. Budhia, J.T. Boys, G.A. Covic, H. Chang-Yu, Development of a single-sided
ux magnetic coupler for electric vehicle IPT charging systems, IEEE Trans. Ind.
Electron. 60 (2013) 318328.
[6] M. Chigira, Y. Nagatsuka, Y. Kaneko, S. Abe, T. Yasuda, A. Suzuki, Small-size
light-weight transformer with new core structure for contactless electric vehi-
cle power transfer system, in: 2011 IEEE Energy Conversion Congress and
Exposition (ECCE), 2011, pp. 260266.
[7] M. Budhia, G.A. Covic, J.T. Boys, Design and optimization of magnetic structures
for lumped inductive power transfer systems, IEEE Trans. Power Electron. Soc.
26 (11 (November)) (2011) 30963108.
[8] Y. Nagatsuka, N. Ehara, Y. Kaneko, S. Abe, T. Yasuda, Compact contactless power
transfer system for electric vehicles, in: International Power Electronics Con-
ference (IPEC), 2010, pp. 807813.
[9] C.-S. Wang, O.H. Stielau, G.A. Covic, Design considerations for a contact-
less electric vehicle battery charger, IEEE Trans. Ind. Electron. 52 (5) (2005)
13081314.
[10] J. Shin, S. Shin, Y. Kim, S. Ahn, S. Lee, G. Jung, S.-J. Jeon, D.-H. Cho, Design and
implementation of shaped magnetic-resonance-based wireless power trans-
fer system for roadway-powered moving electric vehicles, IEEE Trans. Ind.
Electron. 61 (3 (March)) (2014) 11791192.
[11] K. Lee, Z. Pantic, S. Lukic, Reexive eld containment in dynamic inductive
power transfer systems, IEEE Trans. Power Electron. 29 (9 (September)) (2014)
45924602.
[12] W. Zhang, S.-C. Wong, C.K. Tse, Q. Chen, An optimized track length in roadway
inductive power transfer systems, IEEE J. Emerg. Sel. Top. Power Electron. 2 (3
(September)) (2014) 598608.
[13] G.A. Covic, J.T. Boys, Modern trends in inductive power transfer for transporta-
tion applications, IEEE J. Emerg. Sel. Top. Power Electron. 1 (1 (March)) (2013)
2841.
[14] G.R. Nagendra, J.T. Boys, G.A. Covic, B.S. Riar, A. Sondhi, Design of a double
coupled IPT EV highway, in: IECON 2013-39th Annual Conference of the IEEE
in Industrial Electronics Society, 2013, November, pp. 46064611.
[15] S. Choi, J. Huh, W.Y. Lee, S.W. Lee, C.T. Rim, New cross-segmented power supply
rails for roadway-powered electric vehicles, IEEE Trans. Power Electron. 28 (12
(December)) (2013) 58325841.
115
[16] S.W. Lee, W.Y. Lee, G.H. Cho, C.T. Rim, Narrow-width inductive power trans-
fer system for online electrical vehicles, IEEE Trans. Power Electron. 26 (2011)
36663679.
[17] L. Sungwoo, H. Jin, P. Changbyung, C. Nam-Sup, C. Gyu-Hyeoung, R. Chun-Taek,
On-line electric vehicle using inductive power transfer system, in: 2010 IEEE
in Energy Conversion Congress and Exposition (ECCE), 2010, pp. 15981601.
[18] J. Hudece, J. Kufen, O. Langen, J. Dankerty, L. Ecksteinz, A system for precise
positioning of vehicles aiming at increased inductive charging efciency, in:
MedPower Conference, Athens, 2014, November.
[19] I. Karakitsios, E. Karfopoulos, N. Hatziargyriou, Static and dynamic fast induc-
tive charging: The FastInCharge project concept, in: MedPower Conference,
Athens, 2014, November.
[20] J.Y. Yong, V.K. Ramachandaramurthy, K.M. Tan, A. Arulampalam, J. Selvaraj,
Modeling of electric vehicle fast charging station and impact on network volt-
age, in: IEEE Conference on Clean Energy and Technology (CEAT), 2013, pp.
399404.
[21] R. Shi, X.P. Zhang, D.C. Kong, N. Deng, P.Y. Wang, Dynamic impacts of fast-
charging stations for electric vehicles on active distribution networks, in:
Innovative Smart Grid Technologies Asia (ISGT Asia), 2012, pp. 16.
[22] M. Etezadi-Amoli, K. Choma, J. Stefani, Rapid-charge electric-vehicle stations,
IEEE Trans. Power Deliv. 25 (3) (2010) 18831887.
[23] H. Seljeseth, H. Taxt, T. Solvang, Measurements of network impact from electric
vehicles during slow and fast charging, in: 22nd International Conference and
Exhibition on Electricity Distribution (CIRED), 2013, pp. 14.
[24] D. Aggeler, F. Canales, H. Zelaya-De La Parra, A. Coccia, N. Butcher, O. Apeldoorn,
Ultra-fast DC-charge infrastructures for EV-mobility and future smart grids, in:
Innovative 2010 IEEE PES Smart Grid Technologies Conference Europe (ISGT
Europe), 2010, pp. 18.
[25] B. Pea-da, S. Dechanupaprittha, Impact of fast charging station to voltage prole
in distribution system, in: 2014 International Electrical Engineering Congress
(iEECON), 2014, pp. 14.
[26] K. Yunus, H.Z. De La Parra, M. Reza, Distribution grid impact of plug-in electric
vehicles charging at fast charging stations using stochastic charging model, in:
Proceedings of the 201114th European Conference on Power Electronics and
Applications (EPE 2011), 2011, pp. 111.
[27] C. Corchero, M. Cruz-Zambrano, F.-J. Heredia, J.-I. Cairo, L. Igualada-Gonzalez,
A. Romero-Ortega, Optimal sizing of microgrids: a fast charging station case,
in: 2012 9th International Conference on the European Energy Market (EEM),
2012, pp. 16.
[28] G. Mauri, A. Valsecchi, Fast charging stations for electric vehicle: the impact on
the MV distribution grids of the milan metropolitan area, in: Energy Conference
and Exhibition (ENERGYCON), 2012, pp. 10551059.
[29] F. Lacu, S. Mocci, N. Natale, F. Pilo, G.G. Soma, Aggregated electric vehicles
load proles with fast charging stations, in: PSCC Conference, Wroclaw, Poland,
2014.
[30] X. Huang, H. Qiang, Z. Huang, Y. Sun, J. Li, The interaction research of smart grid
and EV based wireless charging, in: 2013 IEEE Vehicle Power and Propulsion
Conference (VPPC), 2013, pp. 15.
[31] T.-E. Stamati, P. Bauer, On-road charging of electric vehicles, in: Transportation
Electrication Conference and Expo (ITEC), IEEE, 2013, pp. 18.
[32] T.-E. Stamati, P. Bauer, Green energy for on-road charging of electric vehicles,
in: MECHATRONIKA, 15th International Symposium, 2012, pp. 19.
[33] P. Javanbakht, S. Mohagheghi, B. Parkhideh, S. Dutta, R. Chattopadhyay, S.
Bhattacharya, Vehicle-to-grid scheme based on inductive power transfer for
advanced distribution automation, in: 2013 IEEE Energy Conversion Congress
and Exposition (ECCE), 2013, pp. 32503257.
[34] S. Mohagheghi, B. Parkhideh, S. Bhattacharya, Inductive power transfer for elec-
tric vehicles: potential benets for the distribution grid, in: Electric Vehicle
Conference (IEVC), 2012 IEEE International, 2012, pp. 18.
[35] N. Hatziargyriou, E.L. Karfopoulos, K. Tsatsakis, The impact of EV charging on
the System Demand, Chapter 3 of book entitled Electric Vehicle Integration
into Modern Power Networks, Spinger, 2012, ISBN 978-1-4614-0134-6.
[36] M. Papadopoulos, N.D. Hatziargyriou, M. Papadakis, Graphics aided interactive
analysis of radial distribution networks, IEEE Trans. Power Deliv. PWRD-2 (4
(October)) (1987).