Content uploaded by Morsy Nour
Author content
All content in this area was uploaded by Morsy Nour on Feb 25, 2019
Content may be subject to copyright.
2019 INTERNATIONAL CONFERENCE ON INNOVATIVE TRENDS IN COMPUTER ENGINEERING (ITCE’2019)
978-1-5386-5261-9/19/$31.00 ©2019 IEEE
Smart Charging of Electric Vehicles According to
Electricity Price
Morsy Nour1,4, Sayed M. Said2,4, Abdelfatah Ali3,4, and Csaba Farkas4
1Department of Electrical Engineering, Faculty of Energy Engineering, Aswan University, Aswan 81528, Egypt
2Department of Electrical Engineering, Faculty of Engineering, Aswan University, Aswan 81542, Egypt
3Department of Electrical Engineering, Faculty of Engineering, South Valley University, Qena 83523, Egypt
4Department of Electric Power Engineering, Faculty of Electrical Engineering and Informatics, Budapest University of Technology and
Economics, Budapest 1111, Hungary
morsy.abdo@aswu.edu.eg, sayed.said@aswu.edu.eg, abdelfatah.mohamed@vet.bme.hu, farkas.csaba@vet.bme.hu
Abstract— The growing popularity of private vehicles’
electrification will have a negative impact on the electric power
system, especially on the distribution networks, if electric vehicles
(EVs) charging is not managed properly. In this paper, a new
technique for smart charging of EVs is proposed and tested with
simulation. A fuzzy logic controller is used to control and manage
the EV charging process to maximize electric utility and EV
owner benefits. The electric utility’s benefit is to mitigate the EV
charging impacts on the distribution network by shifting EV
charging to the off-peak period, while EV owners’ benefit is to
charge the EV at low cost. The controller regulates and controls
the EV charging power depending on electricity price signal
provided by the electric utility and EV battery state of charge
(SoC). This controller needs basic communication with the
electric utility to receive the electricity price signal every 1 hour.
The objective of the controller is to charge EVs at low cost while
keeping the normal operating conditions of the distribution
network. MATLAB/SIMULINK is used to perform simulations
and test the effectiveness of the proposed smart charging method.
The results demonstrated that the proposed smart charging
method reduced the impacts of EVs charging on the distribution
network compared with uncontrolled charging.
Keywords—Uncontrolled charging; smart charging; electric
vehicles; fuzzy logic control.
I. INTRODUCTION
According to the International Energy Agency (IEA) global
EV outlook of 2017 [1], there is a continuous decrease in the
price of EVs which will result in the acceleration of EV
deployment. The decline of battery cost is the reason for EVs
prices reduction. Mass production in addition to research and
development (R&D) results in rapid decay of battery costs. The
continuous improvement of EVs and battery technology will
narrow the price gap between EVs and conventional internal
combustion engine (ICE) vehicles which will increase
competitiveness. Evaluations of the countries’ targets and
equipment manufacturers’ announcements seem to confirm
these positive signals of the massive increase of the EVs stock
market. EVs stock crossed 1 million in 2015 and surpassed 2
million in 2016. It is expected that the EVs stock market will
be between 9 million and 20 million in 2020 and at 2025 will
be between 40 million and 70 million.
Because this large number of EVs will be charged from the
electric power grid, it is important to develop techniques for the
optimal integration of EVs. There are three main types of EV
charging; uncontrolled charging, delayed charging and smart
charging. Uncontrolled charging is also called unregulated
charging, uncoordinated charging, or dumb charging.
Uncontrolled charging means that the cars start charging at the
instant of arrival to home or workplace. This is the type of EV
charging that is used nowadays. In this case, EV charging
usually occurs at peak load hours which leads to severe grid
impacts. Various studies found that uncontrolled EV charging
may limit the acceptable penetration level of EVs in the
distribution network [2]. The highest impacts on the
distribution networks are expected if this charging technique is
used. Usually, this happens when the utility has fixed tariff
structure, so there are no incentives for EV owners to delay the
charging.
In delayed charging, electricity utility companies use two
tariff structures with high prices at peak period and lower
prices at the off-peak period to motivate EV owners to charge
at these times. With proper choice of the tariff structure, this
can result in a flattened load profile and a decrease of the
voltage drop caused by EV charging. If the off-peak and peak
hours in the tariff structure are not set in an optimal way, the
impact of EV charging may get worse [3]–[5]. This may occur
since the low prices at off-peak period may motivate a huge
number of EV owners to charge simultaneously which may
result in higher voltage drop and load demand with the
possibility of second peak formation at the first hours of off-
peak hours.
Although the use of delayed charging can reduce the EV
charging impacts on the distribution networks, it is observed
that the network capacity is not used in an optimal manner.
Various studies concluded that the controlling of charging start
time and charging rates of EVs using a smart charging
(controlled charging or coordinated charging) algorithm may
use the power system in a more optimal and efficient way.
Many studies presented coordinated charging algorithms [6]–
[8] to control EV charging to maximize EV owners’ benefits
by reducing the charging cost or to maximize the electric utility
benefits by shifting the EV charging to off-peak hours which
reduces the impact on the electric power system. This is a part
of the smart grid concept. In this type, the communication
between EV and electric utility or Distribution System
Operator (DSO) is continuous.
In this paper, three operation scenarios are studied and
compared. In the first scenario, the LV distribution network is
supplying residential consumers only, and no EVs are plugged
in for charging. In the second scenario, EVs with 50%
penetration level are connected to the grid for charging. In this
scenario, the EVs are charged in uncontrolled charging manner.
In the third scenario, the EVs are charging based on the
proposed smart charging method.
II. DISTRIBUTION NETWORK MODELING AND SIMULATION
A. Distribution Network
A 400 V LV distribution network is selected as a case study
in this paper to execute simulations. It is a real distribution
network located in New Toshka city, Aswan, Egypt. Fig. 1
shows the one-line diagram of the distribution network. It has a
500 kVA distribution transformer with 22 kV primary and 400
V secondary voltage. It supplies 96 residential consumers
distributed equally at eight buildings. The system is assumed to
be balanced, and the residential consumer's load is the same in
all the 3 phases. All the system data is provided in [9]. Fig. 2
shows the load profile (power consumption variation) during
the day for each consumer [10]. 1 p.u. is equivalent to 4 kVA
which is the highest power consumption in the 24 hours. The
load power factor is 0.9 lagging.
Nissan Leaf [11] is used to evaluate the impacts of its
charging on the distribution network. It is a battery electric
vehicle (BEV). This type of EVs is powered entirely by
electricity and batteries are used as the electric power source,
and it supplies an electric motor which drives the vehicle. No
internal combustion engine (ICE) is used in this type of EVs.
Its battery capacity is 24 kWh. The EVs are assumed to have
20% SoC when connected to the charger (initial SoC). A three-
Fig. 1 LV distribution network one-line diagram.
0
0.2
0.4
0.6
0.8
1
1.2
0 2 4 6 8 10 12 14 16 18 20 22 24
Demand [pu of maximum]
Time of day [h]
Fig. 2 Daily load curve of residential consumer.
phase charger with 90% efficiency and 6.6 kW power rating is
used to charge EVs.
B. Uncontrolled Charging
Three operation scenarios are studied; base case,
uncontrolled charging, and smart charging. For uncontrolled
charging, EVs start charging at the instant of home arrival (1)
at the maximum charging power of the charger. This case has
no coordination or management of EVs charging and the
highest impacts is expected in this case. Fig. 3 shows that the
start time of EVs charging follows a Gaussian distribution (2)
with a mean (µ) equal to 18:00 and standard deviation(σ) equal
to 5 hours [12]. The actual number of EVs that begin charging
at each hour is shown in Fig. 4. A 50% penetration level of
EVs is investigated which is equal to 48 EV. The penetration
level of EVs can be calculated by (3).
arrivalingch tt =
arg
(1)
et
tf )2(
)(
2
2
2
2
1
),,(
−−
=
(2)
100[%] = LoadsofNumber EVsofNumber
LevelnPenetratioEVs
(3)
C. Smart Charging
In this charging method, the EV charging is controlled by a
fuzzy controller with electricity price signal and SoC as inputs
and the charging power as an output. Most of the proposed
smart charging techniques depend on the availability of
sophisticated real-time communication between EVs and
electric utility, but the proposed controller needs only basic
communication to receive the energy price which is provided
by the electric utility every 1 hour. The electricity price is
0
0.01
0.02
0.03
0.04
0.05
0.06
0.07
0.08
0.09
6 8 10 12 14 16 18 20 22 24 2 4 6
Probability
Time of day [h]
Fig. 3 Probability distribution of EVs charging start time.
0
1
2
3
4
12345678910 11 12 13 14 15 16 17 18 19 20 21 22 23 24
Number of EVs
Time of day [h]
50% penetration
Fig. 4 EVs charging start time.
Fig. 5 Electricity tariff structure.
Fig. 6 Inputs and output membership functions.
assumed to vary during the day as in Fig. 5. The electricity
price is 5 p.u. from 24:00 to 17:00, 7.5 p.u. from 17:00 to 19:00
and from 22:00 to 24:00, and 10 p.u. from 19:00 to 22:00.
According to the controller design, the EV can charge at the
maximum charging power when electricity price is low (5
p.u.). In addition, it charges at medium charging power
depending on battery SoC when electricity price is medium
(7.5 p.u.) and stop charging when electricity price is high (10
p.u.). The controller is designed to maximize EV owner benefit
by charging the EV at low cost. Also, the electric utility
benefits (operating the system within the acceptable limits) can
be maximized by proper choice of electricity tariff structure.
EV owners can be motivated to charge at off-peak hours which
have low electricity prices and stop charging at peak hours
which have high electricity price.
TABLE I. FUZZY CONTROLLER KNOWLEDGE BASE RULES
Price
SoC
L
M
H
VL
VH
H
VL
L
VH
H
VL
M
VH
M
VL
H
VH
M
VL
VH
VH
M
VL
Fig. 6 shows that the universe of discourse of the output
(charging power) and first input (SoC) is divided into five
membership functions sets (three triangular and two
trapezoidal); Very High (VH), High (H), Medium (M), Low
(L), and Very Low (VL). Furthermore, the second input
(energy price) is divided into three trapezoidal membership
functions sets; Low (L), Medium (M) and High (H). Because
the controller output signal is in p.u. so, it should be multiplied
by EV charger rated charging power. A saturation limit block
is added to the fuzzy controller output to guarantee the
unidirectional flow of power from the power grid to the
vehicle. Fig. 7 shows the MATLAB/SIMULINK fuzzy
controller model with electricity price and SoC inputs and per
unit charging power output which is multiplied by 6.6 kW gain.
Table I shows the controller knowledge base rules. Fig. 8
illustrate a three-dimensional (3-D) visualization of the control
space. The surface viewer shows the inputs and output relation.
The proposed fuzzy controller can be easily extended to
have more than two inputs or to include other distribution
network parameters such as total power, transformer loading or
feeders loading as well as EV owner preferences.
III. RESULTS AND DISCUSSION
A. Total Power Demand
For uncontrolled charging, the peak demand increased to
500 kVA with 100 kVA increase in peak demand at 20:00
compared with the base case. In opposition, for smart charging,
the peak demand was almost the same as the base case because
EVs stopped charging during the peak hours from 19:00 to
Fig. 7 MATLAB/SIMULINK model of fuzzy controller.
Fig.8 relation between inputs and output in 3-D.
Fig. 9 Total power demand for the base case, uncontrolled charging,
and smart charging.
Fig. 10 Transformer loading for the base case, uncontrolled charging,
and smart charging.
22:00 due to the high electricity price in these hours. EV
charging was shifted to off-peak period as shown in Fig. 9
which is called valley filling.
B. Transformer Loading
The transformer loading increased with EV integration and
reached 100% of its power rating at 20:00 as shown in Fig. 10
when uncontrolled charging was used to charge EVs. On the
other hand, for smart charging, the transformer loading was
limited to 80% of its power rating as in the base case. This
means a 20% reduction in transformer loading compared with
uncontrolled charging.
C. Cable Loading
The cable loading was very low during the whole day even
at peak demand hours as shown in Fig. 11. The cable loading
increased with uncontrolled EV charging and reached more
than 50% of its capacity at 20:00. For smart charging, the
highest loading recorded during the day was 40% of its power
rating similar to the base case. This means more than 10%
reduction in cable loading compared to uncontrolled charging.
D. Voltage at Cable Endpoint
Fig. 12 shows that uncontrolled charging of EVs results in
a worse voltage profile compared with the base case. The
voltage at the furthest point from the transformer reached the
lower limit (0.95 p.u. according to ANSI standard [13]) of the
acceptable operating conditions at peak time (20:00). On the
contrary, smart charging improved the voltage profile because
no EVs were charging during peak hours from 19:00 to 22:00
and the lowest voltage recorded was more than 0.96 p.u.
a.
IV. CONCLUSIONS
In this study, maximizing EV owners’ benefit by
controlling EV charging based on electricity price was
executed. Fuzzy controller with electricity price and SoC inputs
and charging power output was used to control EV charging. It
was concluded that:
• Smart charging reduced the maximum power demand by
100 kVA which is a 20% reduction compared with
uncontrolled charging. Apparently, the highest power
demand recorded during the day was the same as in the
base case. Smart charging resulted in valley filling.
• Smart charging reduced the maximum transformer loading
during the day by 20% compared with uncontrolled
charging. The highest transformer loading recorded for
smart charging was 80% which was the same as the base
case.
• Smart charging reduced the cable maximum loading by
more than 10% compared with uncontrolled charging.
• Smart charging improved the voltage profile compared with
uncontrolled charging and the lowest voltage recorded was
higher than 0.96 p.u. similar to the base case.
REFERENCES
[1] International Energy Agency, “Global EV Outlook 2017: Two
million and counting,” IEA Publ., p. 66, 2017.
[2] Y. Chen, A. Oudalov, and J. S. Wang, “Integration of electric
vehicle charging system into distribution network,” in 8th
International Conference on Power Electronics - ECCE Asia, 2011,
pp. 593–598.
[3] S. Shao, T. Zhang, M. Pipattanasomporn, and S. Rahman, “Impact
of TOU rates on distribution load shapes in a smart grid with PHEV
penetration,” 2010 IEEE PES Transm. Distrib. Conf. Expo. Smart
Solut. a Chang. World, pp. 1–6, 2010.
[4] Yajing Gao, Chen Wang, Zhi Wang, and Haifeng Liang, “Research
on time-of-use price applying to electric vehicles charging,” in IEEE
PES Innovative Smart Grid Technologies, 2012, pp. 1–6.
[5] S. Schey, D. Scoffield, and J. Smart, “A First Look at the Impact of
Electric Vehicle Charging on the Electric Grid in The EV Project,”
EVS26 Int. Batter. Hybrid Fuel Cell Electr. Veh. Symp., vol. 1.
[6] C. Jin, J. Tang, and P. Ghosh, “Optimizing Electric Vehicle
Charging: A Customer’s Perspective,” IEEE Trans. Veh. Technol.,
vol. 62, no. 7, pp. 2919–2927, Sep. 2013.
[7] J. de Hoog, T. Alpcan, M. Brazil, D. A. Thomas, and I. Mareels,
“Optimal Charging of Electric Vehicles Taking Distribution
Network Constraints Into Account,” IEEE Trans. Power Syst., vol.
Fig. 11 Cable loading for the base case, uncontrolled charging,
and smart charging.
Fig. 12 Voltage profile for the base case, uncontrolled charging,
and smart charging.
30, no. 1, pp. 365–375, Jan. 2015.
[8] E. Sortomme, M. M. Hindi, S. D. J. MacPherson, and S. S. Venkata,
“Coordinated Charging of Plug-In Hybrid Electric Vehicles to
Minimize Distribution System Losses,” IEEE Trans. Smart Grid,
vol. 2, no. 1, pp. 198–205, Mar. 2011.
[9] M. Nour, A. Ali, and C. Farkas, “Evaluation of Electric Vehicles
Charging Impacts on A Real Low Voltage Grid,” Int. J. Power Eng.
Energy, no. April, pp. 837–842, 2018.
[10] S. Papathanassiou, H. Nikos, and S. Kai, “A benchmark low voltage
microgrid network,” Proc. CIGRE Symp. power Syst. with dispersed
Gener., no. April, pp. 1–8, 2005.
[11] “NISSAN LEAF,” 2013. [Online]. Available: http://www.auto-
brochures.com/makes/Nissan/Leaf/Nissan_US Leaf_2013.pdf.
[12] Y. Cao et al., “An optimized EV charging model considering TOU
price and SOC curve,” IEEE Trans. Smart Grid, vol. 3, no. 1, pp.
388–393, 2012.
[13] ANSI Standard Publication, “Electric power systems and
equipment—Voltage ratings (60 Hertz),” Natl. Electr. Manuf.
Assoc., p. ANSI C84.1-2011, 2011.