Energy and Power Engineering, 2013, 5, 1347-1351
doi:10.4236/epe.2013.54B255 Published Online July 2013 (
Impact of Electric Vehicle Charging on Power Load
Based on TOU Price*
Yubo Fan, Chunlin Guo, Pengxin Hou, Zheci Tang
State Key Laboratory for Alternat e El ectrical Power Sy stem with Renewable E n ergy Sources,
North China Electric Power University, Beijing, China
Received February, 2013
Large-scale electric vehicle charging has a significant impact on power grid load, disorderly charging will increase
power grid peak load. This article proposes an orderly charging mechanism based on TOU price. To build an orderly
charging model by researching TOU price and user price reaction model. This article research the impact of electric
vehicle charging on grid load by orderly charging model. With this model the grid’s peak and valley characteristics, the
utilization of charg ing equipment, the economics of grid operation can all be improved.
Keywords: Electric Vehicles; DSM; TOU Price; Orderly Charge; Power Load
1. Introduction
TOU price refers to a kind of tariff system. In that system,
a day will be divided into multiple time periods, and will
change different prices for different time periods of elec-
tricity consumption. TOU price is an important means of
Demand Side Management (DSM), it can also direct us-
ers to utilize electricity in reasonable way[1-6].
With economic development and social progress,
power has increasingly became a necessity of people’s
production and life, electricity demand continues to in-
crease which will further intensification of the contradict-
tion between electricity supply and demand. Generally,
the electricity change is very obvious in both corporate
users and residential customers every day, and most users
have low power consumption at night than during the
We call a time period Peak Time when power load is
greater than a certain value, on the contrary, when power
load is below a certain value, we call that time period
Valley Time, except this two kinds time periods, all other
times is called flat section. Valley time and flat section
are also called non-peak time. If electric energy can be
stored massively in a long time like ord inary good s, there
will be tiny differences in power supply cost between
peak time and other time period. However, as a special
commodity, massive storage is difficulty and costly for
power, its production and consumption needs it
Every day, in order to meet power demand in peak
time, power plants according to peak time power demand
to organize electric power production. Meantime,
according to the requirements of power load, power grid
coordinates power supply and installs plenty power
transmission and distribution equipment. Among them,
there must be part of equipment in idle or low load
condition when in non-peak period. In load peak period,
since all equipment is in operation condition, the cost is
higher. In other time, only a small amount of generating
set, power transmission and distribution equipment can
make a balance between power supply and demand, so
the cost is lower. Therefore, according to economic
principles, it is reasonable and feasible to charge
different power price in different consumption period.
From the implementation effect, we can see that TOU
price plays an effective role in power price leverage. It
can inhibit irrational electricity growth in peak time and
improve power con sumption in valley period, which will
enhance economic benefits of the whole power system
and ensure power supply and demand balance.
TOU price system will stimulate electric vehicle us-
ers reduce peak time urgency electricity demand, and
transfer it to flat section or valley period. For electricity
companies, TOU price will adjust users’ consumption
ways, so as to reduce power production cost and balance
power demand and supply. For electric vehicle users,
TOU price will make charging process happen in elec-
tricity price lower time, which will greatly reduce the
*This work is supported by: National High Technology R&D Program
of China (863Program) (2012AA050804), Key Project of the National
Research Program of China (2011BAG02B14), National High Tech-
nology R&D Program of China (863 Program) (2011AA05A109).
Copyright © 2013 SciRes. EPE
cost of using electric vehicles.
2. Effects of Disordered Charge to Power
2.1. Forecasting of EV Charging Load
The most important factor of electric vehicles charging
behavior is the beginning moment, the more concentrated
the charging begin time is, the more bigger power margin
is needed from the grid, and the equipment investment
cost is also greater[10- 13].
To study the distributed electric vehicle charging start
time, we can assume the return time of traditional fuel
vehicle for the start time of distributed electric vehicle
charging in the future.
At present, there is no coll ection statistics about return
time of traditional fuel automobile, we can use statistics
of America transportation department as our reference.
According to National Household Travel Survey (NHTS)
in 2001, the probability statistical results of household
vehicles’ return moment shows in column Figure 1.
According to electric vehicle users’ tradition[14-15],
we assume that the owner start charging his car
immediately after he back home, the above probability
distribution namely for electric vehicles normal charging
start time. After analysis the column figure, we can see
that without any limit or guide, there must be kind of
charging concentration of electric vehicles (as shown in
figure in rush hour of 16:00-18:00).
At present, the rated battery capacity of mainly used
electric vehicle is 20 kW·h--30 kW·h, we assume it as 25
kw-h, meanwhile, the car charger power is about 2-3 kw,
we assume it as 2.5 kw. In this way if the efficiency of
charging machine is 1, the electricity charge of battery
from 0% to 100% needs ten hours. Therefore, we can
assume the charging time TL as standard normal distribu-
tion, whose probability func tion expression is
() 2
ft e
We assume that NHTS finding is applicable to Chi-
nese household automobile users, so based on the prob-
ability distribution of individual household vehicle and
electric vehicle car charger power, we can get ordinary
charging power expectation of single electric vehicle
within a day, which is sho wn in Figure 2.
Single electric vehicle charging expectation only ex-
press the charging possibilities in certain time, it has no
actual meaning with charging power. However, when
large-scale(set to N) electric vehicles connected to the
grid simultaneously, the product of single charge expec-
tation and number N can be considered as electric vehi-
cle’s charging load at this moment. When N equals to
500,000, the electric vehicle charging load carve is
shown in Figure 3.
2.2. The Impact of EV Disordered Charging
As a modern city, Beijing has a large number of cars,
highly developed traffic and well-equipped infrastructure;
all these show the potential of electric vehicle promotion.
We consider Beijing grid load as the original value to
study the impacts of large-scale electric vehicle access on
the load curve. When the access scale of electric vehicle
N equals to 500,000, the grid load is shown as in Figure
4. When the access scale N equals to 1000,000, the grid
load shows in Figure 5.
Figure 1. Probability distribution of household automobile
return time.
Figure 2. Charging expectations of single electric vehicle.
Figure 3. Charging expectations of 500,000 electric vehicles.
Copyright © 2013 SciRes. EPE
Y. B. FAN ET AL. 1349
Figure 4. Grid load after 500,000 electric vehicles connected
to grid.
Figure 5. Grid load after 1,000,000 electric vehicles assess.
From Figure 4, Figure 5, we can see that without
TOU price, there will be an obvious elevation on grid
load when large-scale electric vehicle charging load
connected to the grid. When electric vehic le charging load
not connected to the grid, the highest peak of original grid
load appears in around 18:00, which is a concentrated
period of the residential electricity consumption. Accord-
ing to our statistics of car owners return time, electric
vehicle charging start time also centered in this period. In
this case, there must be a grid load problem. However,
the original grid load valley is also the electric vehicle
charging load valley, which increases the difference
between grid load peaks and valleys. With the increase
grid load from electric vehicle, this phenomenon will be
more and more obvious.
From the Table 1, we can see that there are peak load
increases of 3.95% and 7.94%, and valley load increase
of 0.81% and 1.51% when 500,000, 1,000,000, electric
vehicle connected to grid.
When electric vehicle connected to grid, the peak load
increase is much higher than valley load increase, the gap
between p eak load and v alley load become deeper.
We defined the ratio of the grid peak and valley as
peak-to-valley rate. Peak-to-valley rate is an important
parameter of power equipment, it is a reflection of power
equipment utilization status.
The installed capacity of the generator is designed
according to grid load expectation, therefore, when the
peak and valley difference is big, there will be a lot of
generator sets and other equipment stay in low-loaded or
stop condition, which will greatly reduce the utilization
of electrical equipment and cause unnecessary waste.
Take Beijing power grid as an example, when 1,000,000
electric vehicles connected to grid, the peak load is
985MW higher than the original one. If the capacity of
distribution is the only factor we consider, we assume
capacity-load ratio as 2.0, power factor as 0.9, then the
increase distribution transformer capacity is 2188.89
MVA, and however, the valley load is only increase 111
MW at the same time. In this case, the utilization rate of
distribution transformer is only 5.07% at the lowest mo-
3. The Impact of EV Ordered Charging
Based on TOU Price
3.1. Ordered Charging Model Based on
TOU Price
From the above analysis, we can see that there will be a
significant impact of grid peak load when large-scale
electric vehicle connected to grid. Therefore, there must
be an effective and direct method or economic lever
guide to change people’s charging habit. In this paper,
our main object is household electric vehicle, with its
disperse and slow charging characters, it will be more
difficult to charge them in a central way. Therefore, TOU
price will guide users charging their vehicle in valley
period, and this is a convenient charging way.
Suppose the charge capacity of electric vehicles in the
peak period before the implementation of TOU pricing
for W1, the charge capacity of the flat period for W2, the
charge capacity of Valley period for W3.
After implementation of TOU price, due to the guiding
role of the value, the charge level of the peak period
transferred 12 1
to non-peak period, 13 1
to Valley
period. Non-peak charging amount transferred 23 2
Table 1. Grid load changes after electric vehicle connected
to the grid.
peak loadValley load Peak and valley
difference Peak-to-vall
ey rate
Original load 12400MW7370
MW 5030MW 40.56%
After 500,000
electric vehicles
connected to grid12890MW 7430
MW 5460MW 42.36%
After 1,000,000
electric vehicles
connected to grid13385MW 7481
MW 5904MW 44.11%
Copyright © 2013 SciRes. EPE
the Valley period, here is the transfer matrix.
 
W, , are respectively for the implementation of
TOU price after the charge level of the peak, non-peak,
valley period.
Electric vehicle users’ reaction to charging tariff is
shown in Figure 6. Point is corresponding to the user's
reaction blind the spot that the difference of electricity
price is less than another, user does not adjust the origi-
nal charging habits. Po int b is corresponding to the maxi-
mum of user’s reaction, which is the maximum percent-
age of changing the charging habits.
Assuming users can achieve the maximum degree of
reaction after the implementation the peak and valley
price as the price difference between peak and valley not
smell .Set the user reaction b of the peak to level segment
value of 0.6, the peak to valley segment value of 0.8, the
level to valley segment value of 0.6. We can see that the
price does not impact the 20% users’ charging habits of
charging o n peak period.
12 13
From the formula we can know: 12
= 0.342, 13
0.457, 23
= 0.6charging expectation of Single Elec-
tric Vehicle show as Fi gure 7.
3.2. The Impact of EV Ordered Charging
When electric vehicle access scale N equals to 500,000,
the grid load status as follows in Figure 8. When electric
vehicle access scale N equals to 1,000,000, the grid load
status as follows in Figure 9.
Figures 8, 9 tell us that after the effect of TOU price,
sensitive tariff users transfer their charg ing time to valley
period, thus they can enjoy cheaper tariff. But there are a
few people who are not sensitive to tariff or care more
about charging time, their charging time are not changed.
The above grid peak load is still higher th an original load,
because most users are more sensitive for electricity
tariff, so they will be guided by TOU price and then the
valley load will be improved greatly. We can see this
clearly in Table 2.
From the Table 2, we can see that with TOU price
system there are peak load increases of 1.97% and 4.02%,
and valley load increase of 6.23% and 11.68% when
500,000, and 1,000,000, electric vehicle connected to
grid, this shows that TOU price system plays a guiding
way in electric vehicles charging time. Meanwhile, with
TOU price, the peak-to-valley rate is smaller than the
original one that is to way, after electric vehicle con-
nected to grid, the equipment utilization is higher, which
makes an economic grid operation.
Figure 6. Users reflect model.
Figure 7. Charging expectation of single electric vehicle
Figure 8. Grid load after 500,000 electric vehicles connected
to Grid.
Figure 9. Grid load after 1,000,000 electric vehicles con-
nected to grid.
Copyright © 2013 SciRes. EPE
Copyright © 2013 SciRes. EPE
changes after electric vehicle conn
Peak Peak Peak and
Table 2. Grid loadected
to grid.
Load Valley Valley
ifference rate
Original Load 1240 40.
MW 7370
MW 5030
MW 56%
Traditional 42.36%
After 500,000
o TOU Price 126447829 4815 38.
Traditional 133857481 5904 44.
after 1,000,000
TOU PRICE 127898131 4658 36.42%
Power Triff 12890
MW 7430
MW 5460
connected t
grid MW MW MW 08%
Power Triff MWMW MW 11%
electric vehicle
connected to
grid MW MW MW
4. Conclusions
alysis, we can see that TOU price
[1] W. W. F. Sclectric Vehicles in
ctrum, April,
omez and M. M. Morcos, “Impact of EV Ba
From the above an
plays a guiding role in adjusting large-scale electric ve-
hicle charging time. However, due to users’ autonomy
transfer charging time from peak time to valley time is
unrealistic. TOU price system plays a positive way in
guiding disordered charging way, and there is no need to
build complex orderly charging control system, which
will improve the utilization of power equipment, reduce
investment and enhance charging reliability.
otland, “The Role of E
Scotland's Low Carbon Future,” May 2010.
[2] J. Voelcker, “Top 10 Tech Cars,” IEEE Spe
[3] J. C. Gttery
Chargers on the Power Quality of Distribution Systems,”
IEEE Transactions on Power Delivery, Vol. 18, No. 3,
2003, pp. 975-981. doi:10.1109/TPWRD.2003.813873
[4] P. T. Staats, W. M. Grady, A. Arapostathis and R. S.
Thallam, “A Statistical Analysis of the Effect of Electric
Vehicle Battery Charging on Distribution System Har-
monic Voltages,” IEEE Transactions on Power Delivery,
6. Vol. 13, No. 2, 1998, pp. 640-64
[5] J. T. Salihi, “Energy Requirements for Electric Cars and
Their Impact on Electric Generation and Distribution
Systems,” IEEE Transactions on Indu
1973, Vol. IA-9, No. 5, pp. 516-stry Applications,
[6] IET Transport desk, “UK Grid Capacity 'Sufficient' for
onference, AUPEC 2007, Australasian,
lletin-vehicle Licensing Statistics: 2008,'
of Batteries,”
Highlights and R&D
cid Battery,” Southern
battery, South-
, “Travel to
National Gird Operational Data, Accessed
in May, 2010.
Electric Vehicles,” Published on I May 2009.
[7] G. B. Shrestha and B. C Chew, “Study on the Optimiza-
tion of Charge Discharge Cycle of Electric Vehicle Bat-
teries in the Context of Singapore,” Universities Power
Engineering C
2007, pp. I-7.
[8] National Statistics, “Department for Transport,” Trans-
port Statistics Bu
9th April 2009.
[9] D. Linden and T. B. Reddy, “Hand Book
Third Edition, McGraw-Hili, Aug., 2001.
[10] G. Rogers, “Spectrum-Technology
Activities at FEV,” 2008, No. 39.
[11] A. Mendoza and J. Argueta, “Performance Characteriza-
tion - GM EVI Panasonic Lead A
California EDISON, April, 2000.
[12] C. Madrid, J. Argueta and J. Smith, “Performance Char-
acterization - 1999 Nissan Altra - EV with Lithium-ion
Battery,” Southern California EDISON, September, 1999.
[13] R. Solares and J. Argueta, “Performance Characterization
– 1999 Toyota RA V 4 - EV with NiMH
ern California EDISON,” October, 1999.
[14] National Statistics, Department for Transport
work-personal Travel Factsheet,” July 2007.
[15] National Grid,