You are on page 1of 10

Energy Conversion and Management 83 (2014) 337346

Contents lists available at ScienceDirect

Energy Conversion and Management


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

Electric vehicle charge patterns and the electricity generation mix and
competitiveness of next generation vehicles
Taisuke Masuta , Akinobu Murata 1, Eiichi Endo 2
National Institute of Advanced Industrial Science and Technology (AIST), 1-2-1, Namiki, Tsukuba, Ibaraki 305-8564, Japan

a r t i c l e

i n f o

Article history:
Received 10 October 2013
Accepted 1 April 2014

Keywords:
Demand-side management
Electricity generation mix
Next-generation vehicle
Electric vehicle (EV)
Energy system analysis
Renewable energy sources

a b s t r a c t
The nuclear accident of 2011 brought about a reconsideration of the future electricity generation mix of
power systems in Japan. A debate on whether to phase out nuclear power plants and replace them with
renewable energy sources is taking place. Demand-side management becomes increasingly important in
future Japanese power systems with a large-scale integration of renewable energy sources. This paper
considers the charge control of electric vehicles (EVs) through demand-side management. There have
been many studies of the control or operation methods of EVs known as vehicle-to-grid (V2G), and it
is important to evaluate both their short-term and long-term operation. In this study, we employ energy
system to evaluate the impact of the charge patterns of EVs on both the electricity generation mix and the
technology competitiveness of the next generation vehicles. An advanced energy system model based on
Market Allocation (MARKAL) is used to consider power system control in detail.
2014 Elsevier Ltd. All rights reserved.

1. Introduction
The nuclear accident caused by the 2011 Tohoku earthquake
had a great impact on Japanese energy policy. The issue of whether
to phase out nuclear power plants and install renewable energy
sources, such as wind power or photovoltaic (PV) generation, on
a large-scale is being debated. In such a situation, attention is
focused on the installation of new energy technologies, such as
energy saving appliances or energy storage. Energy technology
based on electricity and those based on other energy mediums
often compete, as with electric heat pump water heaters and
gas-latent heat-recovery water heaters; electric vehicles (EVs)
and hydrogen fuel cell vehicles (FCVs); and battery energy storage
and hydrogen energy storage. An energy technology based on electricity tends to be superior to one based on other energy mediums.
In conventional power systems where the ratio of generation from
nuclear power plants is high, the generation cost is supposed to be
Abbreviations: EDC, economic-load dispatching control; EV, electric vehicle;
FCV, fuel cell vehicles; ICEV, internal combustion engine vehicle; IEA, International
Energy Agency; LFC, load frequency control; LNG, liqueed natural gas; LPG,
liqueed petroleum gas; MARKAL, market allocation; PV, photovoltaic; V2G,
vehicle-to-grid.
Corresponding author. Tel.: +81 29 862 6680; fax: +81 29 851 7523.
E-mail addresses: taisuke.masuta@aist.go.jp (T. Masuta), aki.murata@aist.go.jp
(A. Murata), endo.e@aist.go.jp (E. Endo).
1
Tel.: +81 29 861 5828; fax: +81 29 851 7523.
2
Tel.: +81 29 861 5826; fax: +81 29 851 7523.
http://dx.doi.org/10.1016/j.enconman.2014.04.001
0196-8904/ 2014 Elsevier Ltd. All rights reserved.

cheap, and controllable power plants, such as hydro and thermal


power plants, stably maintain the supply and demand of electricity. It has been seen as natural that the economy and environment
can be enhanced by increasing the ratio of electrication [1]. However, the trend will be impacted by the change of the electricity
generation mix and power system control and operation because
of the decrease of nuclear power plants and increase of renewable
energy sources [2]. However, demand-side management, which
applies electric appliances to power system control and operation
as controllable loads, has gained much attention. An energy
technology based on electricity, which has controllability, may be
superior to that based on other mediums precisely because of the
large-scale integration of uncertain renewable energy sources.
Control methods for demand-side management using controllable
appliances have been investigated [36]; however, both the shortterm evaluation of the effectiveness of such methods in daily or
weekly operation [35] and the long-term evaluation of the impact
of them on power system generation planning [6] and the diffusion
of such appliances themselves are also important.
Vehicle-to-grid (V2G) [7], a supply and demand control using
the charge/discharge of EVs, typies such demand-side management. Various objectives have been targeted in the study of V2G
[614], including the minimization of operational cost, the leveling
of electricity load demand, the reduction of emissions, the regulation of power system frequency, and so on. Researchers [610]
have focused on the contribution of the charge/discharge of EVs

338

T. Masuta et al. / Energy Conversion and Management 83 (2014) 337346

to the power system frequency control. Fazelpour and his colleagues [11] propose an optimization method of charge allocation
for plug-in hybrid electric vehicles to improve voltage proles and
to minimize power losses. Other scholars [12] suggest a minimization method for both total operational costs and emissions by the
decentralized control of EVs. The leveling of the electricity load
demand is proposed for the charge control of EVs [13,14]. This
paper focuses on demand-side management through the charge
control of EVs in a long-term evaluation case study. The charge
control of EVs, which is a kind of V2G, enables efcient power system operation by shifting the charge period of the peak cut or peak
shift of electric power demand. Although the number of EVs is currently still small (about 25,000 passenger EVs in 2012 in Japan
[15]), their dispersal can be enhanced by charge control if they
are cheaply charged and have cost advantages. In this study, we
evaluate the impact of the charge patterns of EVs on power system
operation, the future electricity generation mix, and the technology competitiveness of next generation vehicles by energy system
analysis. We simultaneously consider the effectiveness of EV
charge control and the competitiveness of EVs, which is an alternative approach to verify the feasibility of V2G. An advanced energy
system model [16], which was developed based on Market Allocation (MARKAL) [17] and considers the characteristics of power system control, including demand-side management, is used for the
analysis. By employing this model, we plan a future generation
mix of all energy sectors, in which we consider the balance
between electricity and the other sources, although studies on
power system generation planning usually only focus on the electric energy sector [1820]. Moreover, we can compare the competitiveness of EVs to vehicles based on other energy mediums. Two
charge patterns are assumed; one is a nighttime load pattern,
where EVs are charged during the nighttime; the other is a controllable load pattern, where EVs are charged during the optimal time
for total system cost minimization.
2. Energy system model
An advanced energy system model, which was developed by
expanding the functions of MARKAL in a previous study [16], is
used for the analysis. The characteristics of the advanced model
are presented in this section.
2.1. MARKAL
MARKAL was developed in the international cooperation project lead by International Energy Agency (IEA) [17,21]. It is an optimization model that considers all the energy sectors and models an
energy system as the network ow from supply to demand. MARKAL can describe energy technologies in detail through their technical characteristics. It determines the multi-period composition of
the energy technologies, including nal energy consumption and
the electricity generation mix. It has been widely applied to energy
system analysis studies [2224]. However, MARKAL does not consider power system control of both the power plants and the
demand side in detail, which must, however, be taken into account
in an energy system analysis that weighs the time change of the
output of renewable energy sources, electric power demand, and
demand-side management. Thus, an advanced energy system
model [16] is used in this study.
2.2. Indices and parameters
The index of the analysis period is represented by T (13 periods
from 1990 to 2050 by every 5 years), that of the season by Z, and
that of the time period by Y. Z is set as the following 9 patterns

(3 daily patterns in 3 seasons); weekday, weekend, and singular


day (a week of light-load) in winter; weekday, weekend, and singular day (three days of heavy-load) in summer; and weekday,
weekend, and singular day (a week of light-load) in spring/fall. Y
is set as the hourly time series (24 time periods in a day). The
parameters qhr(Z, Y) and fr(Z, Y, DM), which contain the factors Z
and Y, are also set as the time-series data. The terms qhr(Z, Y) represents the proportion of time Y in season Z to a year. The terms
fr(Z, Y, DM) represents the proportion of the demand DM used at
Y in Z to a year. qhr(Z, Y) corresponds to the proportion of weekday,
weekend, and singular day in each season. The setting of fr(Z, Y,
DM) is presented in Section 3.
2.3. Modeling of supply and demand control in power systems
The constraint equations related to the power system considered in MARKAL are the maximum generation output, the balance
between supply and demand, the capacity usage, the reserve
capacity on peak load, and the base load operation [17,21]. MARKAL is intrinsically designed as an optimization model for multitime periods in multi-seasons. Therefore, the constraint equations
of maximum generation output, the balance between supply and
demand, and the capacity usage are used without change in the
advanced model. In the advanced model, the reserve capacity constraint equation assumes that the reserve capacity must be more
than 5% of the electric power demand at any time.
The load following capability of the power plants is simply
approximated in the base-load operation constraint equation in
MARKAL by suppressing the output of the base power plants to a
certain level. The advanced model used in this study approximates
the load following capability in more detail by considering the economic-load dispatching control (EDC) and the load frequency control (LFC). The EDC and LFC are the demand and supply control
schemes of controllable power plants (thermal or hydro power
plants) in power systems. The EDC is the control for following
long-period load variation (several ten minutes-several hours) by
the slow change of the output of power plants, whereas the LFC
is that for suppressing short-period load variation (several ten seconds-several minutes) by the rapid change of the output of power
plants. In general, power plants change the output for the LFC in a
certain range (a few percent of the rated capacity, which is called
LFC capacity) and that for the EDC for the rest. The constraint equation of the rate of change of the output of the power plants is given
by (1), which stands for EDC. E(T, Z, Y, ELA) is the electricity generation of power plant ELA at time Y in season Z in period T. RCUP(ELA) is
the upper limit of the increasing rate of ELA, whereas RCDOWN(ELA) is
that of the decreasing rate. C(T, ELA) is the installed capacity of ELA
in T.

8
ET; Z; Y; ELA  ET; Z; Y  1; ELA
>
>
>
<  RC ELA  qhrZ; Y  CT; ELA
UP
>
ET;
Z;
Y  1; ELA  ET; Z; Y; ELA
>
>
:
 RC DOWN ELA  qhrZ; Y  CT; ELA
Eqs. (2)(4) stand for the LFC constraints [6]. The constraint equations of the upper limit of the LFC capacity are given by (2) and
(3). CLFC(T, Z, Y, ELA) is the LFC capacity of ELA at Y in Z in T. RLFC(ELA)
is the proportion of the LFC capacity of ELA to the rated capacity.
RPmin(ELA) is the proportion of the minimum output of ELA to the
rated capacity; af(ELA) is the availability factor of the ELA. The Eq.
(4) indicates that the total LFC capacity must be more than a certain
proportion to the total electric power demand. ED(T, Z, Y) is the total
electric power demand at Y in Z in T. RD is the proportion of the LFC
capacity to the total electric power demand, which is set to 2% in
this study. The LFC capacity for the output uctuation of the renewable energy sources is not considered.

339

T. Masuta et al. / Energy Conversion and Management 83 (2014) 337346

C LFC T; Z; Y; ELA 

RLFC ELA
 ET; Z; Y; ELA
RPmin ELA

C LFC T; Z; Y; ELA  RLFC ELA  qhrZ; Y  af ELA  CT; ELA

X
C LFC T; Z; Y; ELA  RD  ED T; Z; Y

amounts of charged energy and the total energy of a day driving.


The EVs are assumed capable of being charged in the controllable
load pattern after 2030. They are charged in the nighttime before
2030. In addition, the control and communication equipment for
the charge control is not considered for simplicity.
Allday
X

Allday
X

EEV T; Z; Y; EV

ELA

frZ; Y 0 ; TD0 

cf T; EV  CT; EV
eff T; EV

2.4. Power output of wind power generation and PV generation


3. Analysis condition
The power output of wind power generation is assumed to be
constant at the rated power output multiplied by the capacity factor. The power output of PV generation is assumed to change with
time, supposing a sunny day, and the daily generated energy is
assumed to be the same as the integral of the rated power output
multiplied by the capacity factor in a day.

The energy system of the whole of Japan is analyzed in this


study using the advanced model presented in the previous section.

2.5. Charging patterns of EVs


Two kinds of EVs are considered in this study; normal EV and
light EV. A regular charge [25] but not a rapid charge is considered.
EVs can be charged via household wall sockets with regular charges.
They are assumed to have two charge patterns; one is the nighttime
load pattern (charge in the nighttime), and the other is the controllable load pattern (charge at any time). The charge period and
amounts of power in the controllable load pattern are variables in
the optimal problem and determined by solving it. As a result, the
controllable load pattern is the best charge pattern for the
whole energy system. The nighttime load pattern is selected as a
representative pattern and compared to the best pattern. Since
the MARKAL-based energy system model requires that electric
power distribution facilities are invested based on the increase in
electric power demand, investments in regular charge equipment
are assumed to be included in those on electric power distribution
facilities. In addition, the driving demand of EVs is assumed to be
constant all day and user convenience is not considered in this study.
The total energy for a day driving is charged during nighttime in
the nighttime load pattern. The amounts of charge power are constant all the time, as shown in (5). EEV(T, Z, Y0 , EV) is the charged
energy of EV (normal EV or light EV) at Y0 , which represents Y in
the nighttime, in Z in T; cf(T, EV) is the capacity factor of EV in T.
eff(T, EV) is the efciency of EV in T. TD represents the demand
for driving in the passenger car sector which is a factor of the
demand DM. The nighttime is dened in this study as 17 h to 6 h
in winter, 20 h to 5 h in summer, and 19 h to 5 h in spring/fall, so
as to have the same time allocation as Japanese MARKAL [2].

Fig. 1. Upper bound of CO2 emissions (normalized by the amount in 1990).

Allday
X
qhrZ; Y 0

EEV T; Z; Y 0 ; EV Nighttime
frZ; Y 0 ; TD0
X
Y
qhrZ; Y 0
Y0

cf T; EV  CT; EV

eff T; EV

In Japan, the average driving distance of a car is short, and


almost all the cars are parked. For example, the daily driving distances of as many as 50% of cars are less than 30 km [26]. Thus,
the nighttime load pattern is assumed to be feasible by shifting
the charging start time of EVs, although the charge period of each
EV is expected to be only a few hours.
The charge period is changeable in the controllable load pattern,
as long as the total energy for a day driving is equal to the total
charged energy in a day. However, the maximum charged energy
in one hour is limited to 20% of the total energy in a day so as to
avoid a great disturbance caused by the simultaneous charge of
EVs. Eq. (6) is the constraint equation of the balance between the

Fig. 2. Assumed energy demand (normalized by each demand in 2010).

340

T. Masuta et al. / Energy Conversion and Management 83 (2014) 337346

Fig. 3. Daily change of demand for light and appliances.

Fig. 4. Daily change of demand for cooling.

T. Masuta et al. / Energy Conversion and Management 83 (2014) 337346

341

Fig. 5. Daily change of demand for space heating.

The objective function is set to minimize the total system cost. The
discount rate is set to 5%. The analysis condition is presented in
this section. The technical characteristics of other technologies
than are offered in this section are based on the data of Japanese
MARKAL [2] in the analysis. The analysis period is from 1990 to
2050. The parameters from 1990 to 2010 are tuned to match the
actual data.

Pumped storage power plants are presumed to pump from 1 to


6 h and generate from 11 to 16 h (before 2020) or 14 to 19 h (after
2020). The generating period varies from before 2020 to after 2020
because the output of PV generation around noon increases with
its increasing installed capacity of it after 2020.

3.1. Setting of CO2 emission levels

Fig. 6 shows the installation potential (upper bound) of nuclear


power plants, wind power generation, and PV generation. Nuclear
power plants are assumed to be phased out in 40 years. The potentials of wind power generation and PV generation are obtained
from [29,30], respectively. Table 1 shows the data on wind power
generation and PV generation used in the analysis. Table 2 presents
the data on other electric power facilities. The construction cost
and maintenance cost are determined in reference to [31].

Fig. 1 shows the upper bound of the CO2 emission (normalized


by the amount in 1990) imposed in the analysis. It is reduced by
15% in 2030 and 40% in 2050 from the amount of the CO2 emission
in 1990. Carbon dioxide capture and storage is not considered in
this study.

3.3. Data on electric power facilities

3.2. Data on energy demand


3.4. Data on EDC and LFC
In the same manner as MARKAL, the data on demand
(transportation, cooling, space heating, hot water demand, etc.)
is set, and the effective energy demand supplied by the energy
devices (e.g., electric and diesel trains, gas, oil, and gaswater,
and electric-water heaters, and so on) is determined by solving
the optimal problem in the advanced model. Fig. 2 shows the
energy demand assumed in the analysis, which is obtained from
[27] for reference. The plotted demands are normalized by each
demand in 2010.
FR(Z, Y, DM), which is the proportion of the demand DM used
during time Y in season Z to a year, is determined in 24 time periods in 9 patterns (3 daily patterns in 3 seasons) by reference to
[28]. Figs. 35 show the change of daily demand for light and appliances, cooling, and space heating, respectively, which are
normalized by the peak demand. The demand for industry and
transportation is assumed to be constant in this study.

The model used in the analysis considers the EDC and LFC of
thermal and hydro power plants as described in Section 2. Table 3

Fig. 6. Installation potential.

342

T. Masuta et al. / Energy Conversion and Management 83 (2014) 337346

Table 1
Data on wind power generation and PV generation.

2010

2020

2030

2040

2050

Wind (onshore)

Construction cost (10 JPY/kW)


Annual maintenance cost (103JPY/kW)
Capacity factor (%)

275
15
24

188
11
27

177
11
30

173
11
30

170
11
30

Wind (offshore)

Construction cost (103JPY/kW)


Annual maintenance cost (103JPY/kW)
Capacity factor (%)

492
27
31

357
21
36

260
17
40

257
17
40

254
17
40

PV (residential)

Construction cost (103JPY/kW)


Annual maintenance cost (103JPY/kW)
Capacity factor (%)

515
8
12

230
4
12

1S9
3
12

156
2
12

128
2
12

PV (non-residential)

Construction cost (103JPY/kW)


Annual maintenance cost (103JPY/kW)
Capacity factor (%)

450
16
16

185
7
16

158
6
16

135
5
16

116
4
16

2010

2020

2030

2040

2050

Table 2
Data on other electric power facilities.
Type
Nuclear
Nuclear (LWR)
Hydro
Hydro
Hydro (pumped storage)
Hydro (mixed pumped storage)
Coal
Coal (PC)
Coal (IGCC)
Coal (AF)
Oil
Oil
Oil (private)
Oil (privatev cogen.)
Oil (cogen.)
Gas
Gas turbine
Gas engine
LNG
Blast FNC
LNG (CC)
Gas fuel cell (cogen.)
Gas fuel cell
Gas engine (cogen.)
Gas turbine (cogen.)
Other
Hydrogen fuel cell
Geohydrothermal
Biomass

Cost (103JPY/kW)
Construction cost
Annual maintenance cost

350
25

350
25

350
25

350
25

350
25

Construction cost
Annual maintenance cost
Construction cost
Annual maintenance cost
Construction cost
Annual maintenance cost

850
23
122
9
381
10

850
23
122
9
381
10

850
23
122
9
381
10

850
23
122
9
381
10

850
23
122
9
381
10

Construction cost
Annual maintenance cost
Construction cost
Annual maintenance cost
Construction cost
Annual maintenance cost

230
12

239
13

230
12

239
13

230
12
288
12
239
13

230
12
288
12
239
13

230
12
288
12
239
13

Construction cost
Annual maintenance
Construction cost
Annual maintenance
Construction cost
Annual maintenance
Construction cost
Annual maintenance

190
10
218
11
238
12
130
10

190
10
218
11
238
12
130
10

190
10
218
11
238
12
130
10

190
10
218
11
238
12
130
10

190
10
218
11
238
12
130
10

cost

107
6
171
9
603
38
249
21
120
7
2776
108
2103
82
120
11
120
11

107
6
171
9
603
38
249
21
120
7
860
8
587
5
120
11
120
11

107
6
171
9
603
38
249
21
120
7
490
4
441
4
120
11
120
11

107
6
171
9
603
38
249
21
120
7
280
4
289
4
120
11
120
11

107
6
171
9
603
38
249
21
120
7
160
3
193
4
120
11
120
11

Construction cost
Annual maintenance cost
Construction cost
Annual maintenance cost
Construction cost
Annual maintenance cost

1924
75
800
46
356
19

624
5
800
46
356
19

382
3
800
46
356
19

218
3
800
46
356
19

125
2
800
46
356
19

Construction cost
Annual maintenance
Construction cost
Annual maintenance
Construction cost
Annual maintenance
Construction cost
Annual maintenance
Construction cost
Annual maintenance
Construction cost
Annual maintenance
Construction cost
Annual maintenance
Construction cost
Annual maintenance
Construction cost
Annual maintenance

cost
cost
cost
cost

cost
cost
cost
cost
cost
cost
cost
cost

T. Masuta et al. / Energy Conversion and Management 83 (2014) 337346


Table 3
Data on thermal and hydro power plants.
Minimum output (%)

Coal
LNG-CC
Oil, Gas
Hydro

30
30
30
30

Rate of change (%/h)


Up

Down

31
82
100
100

58
75
100
100

LFC capacity (%)

5.0
5.0
5.0
5.0

presents the minimum output, the rate of change, and the LFC
capacity of thermal and hydro power plants (percentage to the
rated capacity) in the analysis [6,32].
3.5. Data on next generation vehicles
Eleven types of vehicles are considered as the nal demand
technologies in the passenger car sector: gasoline internal combustion engine vehicle (ICEV), gasoline ICEV (light vehicle), diesel
ICEV, liqueed petroleum gas (LPG) ICEV, liqueed natural gas

Fig. 7. Expected vehicle price ratio to gasoline ICEV.

Fig. 8. Expected vehicle efciency.

Fig. 9. Assumed upper bound of total share of next generation vehicle in Japan.

343

(LNG) ICEV, methanol ICEV, hydrogen ICEV, gasoline hybrid vehicle, hydrogen FCV, EV, and EV (light vehicle). LPG ICEV, LNG ICEV,
methanol ICEV, hydrogen ICEV, gasoline hybrid vehicle, hydrogen
FCV, EV, and EV (light vehicle) are assumed as next generation
vehicles. The expected vehicle price ratio to gasoline ICEV and
vehicle efciency is shown in Figs. 7 and 8, respectively. The prices
of gasoline hybrid vehicles, hydrogen FCVs, and EVs are determined to be the same as that of gasoline ICEVs after 2030. The
upper bound of the total share of next generation vehicle in Japan
is determined with reference to the governments target [33] as
shown in Fig. 9. The cost of hydrogen fueling station is based on
[22].
4. Analysis results
Analysis results are shown in this section under the condition
presented in the previous section. Two cases of analysis are implemented. In Case 1, EVs are charged in the nighttime load pattern,
and in Case 2, they are charged in the controllable load pattern.
4.1. Electricity generation mix
The total system cost in Case 2 is smaller by 0.034% than that in
Case 1 because EVs can be charged to minimize the total system
cost in the former case. Figs. 10 and 11 show the transition of
the installed capacity of electric power facilities and that of annual
electricity generation in Japan from 2010 to 2050, respectively.
In Fig. 10, the ratio of the installed capacities of nuclear and coal
power plants decreases and that of gas power plants, wind power
generation, and PV generation increases as the years go by. Both
onshore and offshore wind power generation and residential PV
generation, but not non-residential PV generation, are installed
up to the upper bound in 2050 in the both cases. The trends of
the transition of the installed capacity are almost the same in the
both cases. In 2040, the total capacity in Case 1 is, however, smaller
than that in Case 2 because the installed capacity of PV generation
in the former is smaller than that in the latter. Actually, the
installed capacity of the residential PV generation reaches the
upper bound after 2045 in Case 1, whereas it is attained after
2040 in Case 2. As described hereinafter, the electric power
demand in the daytime is small in Case 1 because all the EVs are
charged in the nighttime, and the power supply from PV generation in the daytime has less of an advantage in Case 1. Therefore,
the age of the PV installation to the upper bound in Case 1 is later
than that in Case 2.
Similar to Fig. 10, the trends of the transition of the generation
are almost the same in the both cases before 2040 in Fig. 11. However, the total generation amount in Case 1 is smaller than that in
Case 2 after 2040 because, as will be discussed in detail, of the
energy breakdown consumed in the passenger car sector.
Figs. 1214 respectively show the daily generation curve, the
daily load curve, and the marginal generation cost on a weekday
in summer in 2050. Different colors are used for each generation
in Fig. 12 and for EVs and the others in Fig. 13. In Fig. 12, which
is assumed to represent 2050, the installed capacity of the coalred power plants is zero, as shown in Figs. 10 and 11. The EVs
are charged at constant power in the nighttime in Case 1 in
Fig. 13(a), whereas they are charged at about 4 h and midday, for
both load leveling and the efcient use of PV generation in Case
2 in Fig. 13(b). The charge period at midday corresponds to the
time zone of the maximum PV power output, as shown in
Fig. 12. The marginal generation cost in Fig. 14 means the incremental system cost per unit of electricity power when the total
system cost is minimal. It is high from 17 to 21 h in Case 1 in
Fig. 14. This result indicates that the total cost for electricity generation increases because the EVs are charged in this time zone.

344

T. Masuta et al. / Energy Conversion and Management 83 (2014) 337346

Fig. 10. Transition of installed capacity of electric power facilities in Japan.

Fig. 11. Transition of annual electricity generation in Japan.

However, the marginal costs in all periods are the same as in Case 2
in Fig. 14. The generation costs are leveled by the charge control of
EVs in Case 2.
4.2. Impact on technology competitiveness of next generation vehicles
The MARKAL-based energy system model considers the total
energy consumption but not the number of EVs as their diffusion.
The energy consumption is determined as a result of the analysis.
Fig. 15 shows the energy breakdown in the passenger car sector.
Hydrogen FCVs have competitiveness against EVs in Case 1 after
2040, whereas they do not have in Case 2. In Fig. 11, electricity generation in Case 1 is smaller than that in Case 2 after 2040, as mentioned in Section 4.1, since the electric power demand also
decreases, given that the energy consumed by EVs in Case 1 is
smaller than that in Case 2 after 2040. In Figs. 12 and 13, electricity
generation and demand in Case 1 are particular larger around 20 h
compared to Case 2. In Case 1, the generation from other around
this time mainly hydrogen fuel cells, whose generation cost is
expensive is larger than that in Case 2 in Fig. 12. The generation
ratio of others around this time becomes higher, if the total electricity generation is greater than 160 GW, in answer to the increase

in the charging demand of EVs. The electricity generation cost is


usually considered to be low during the nighttime. However, a
large electricity consumption of EVs requires the operation of more
expensive power plants, and the generation mix becomes costly
when all EVs are charged in the nighttime as shown in Figs. 12
and 14, which explains why EVs are less competitive than hydrogen FCVs in Case 1. Since the objective function of the optimization
is to minimize cost, the result that the FCVs running on hydrogen
are more popular than the EVs running on electricity means that
hydrogen as a car fuel is superior to electricity in terms of the total
system cost in Case 1.
Thus, the economic operation of the power system differs from
the charge patterns of EVs. Moreover, the charge patterns of EVs
have a great impact on the technology competitiveness of next
generation vehicles. Although little data on EVs is currently available, it is important to examine how and when EVs are charged
in the future with a large-scale diffusion of such vehicles.
As indicated in Section 4.1, the optimal charge period of EVs is
during around 4 h (for load leveling) and midday (for efcient use
of PV generation), and the total system cost in Case 2 is smaller
than that in Case 1. Therefore, under the analysis condition
assumed in this study, it is better to diffuse more EVs than

T. Masuta et al. / Energy Conversion and Management 83 (2014) 337346

345

Fig. 14. Marginal generation cost (weekday in summer, 2050).

Fig. 12. Daily generation curve (weekday in summer, 2050).

Fig. 15. Energy breakdown in the passenger car sector.

5. Conclusion

Fig. 13. Daily load curve (weekday in summer, 2050).

hydrogen FCVs in terms of the whole energy system. However, the


ideal condition is assumed in Case 2. The cost of charge control will
increase from a practical application perspective because communication and control equipment is necessary for charge control.
Moreover, the controllable period of EVs is limited because of
EVs user convenience. The application of V2G to power system
operation is important also for best mix energy in the transportation sector.

In this paper, the impact of charge patterns of EVs on the electricity generation mix and technology competitiveness of next generation vehicles has been evaluated by an energy system analysis.
The analysis results reveal that hydrogen FCVs have competitiveness against EVs, if the EVs are charged in the nighttime, when
the electricity generation cost is expensive. However, under the
analysis condition assumed in this study, the optimal charge period is around 4 h and midday, when the electric power demand
is the smallest, and the PV power output is the largest, respectively; therefore, hydrogen FCVs are not competitive against EVs.
We will compile all the detailed analysis conditions and results,
including those shown in this paper, into a standalone report.
The development of an estimation method of EV use (drive/
charge) pattern and a V2G method that considers user convenience
is necessary because the EVs cannot always be practically charged
in the optimal period for the system. In future works, we will
develop a statistical estimation method of EV charge patterns
and a new V2G method, which considers both charge and

346

T. Masuta et al. / Energy Conversion and Management 83 (2014) 337346

discharge, based on the estimation method. The impact of V2G


method on the competitiveness of EVs will also be evaluated. In
addition, we will improve the energy system model by considering
the number of EVs; the different characteristics of each EV user,
such as driving periods and driving distances; and other specications of EV, such as battery capacity and charging period.
Moreover, we will discuss business models for the V2G method,
including electricity price structure and charging and communication infrastructure.
Acknowledgement
This work was supported by JSPS KAKENHI Grant No.
24860074.
References
[1] Mazi N, Assoumou E. Electricity generation and renewables under carbon
mitigation policies. IEEE Energy 2030 Conf, Atlanta, U.S.; 2008..
[2] Endo E. Effects of the low nuclear policy on technology competitiveness among
next generation vehicles in Japan. Cape Town, South Africa: International
Energy Workshop (IEA); 2012.
[3] Chu CM, Jong TL. A novel direct air-conditioning load control method. IEEE
Trans Power Syst 2008;23:135663.
[4] Yao L, Lu H-R. A two-way direct control of central air-conditioning load via the
internet. IEEE Trans Power Delivery 2009;24:2408.
[5] Short JA, Ineld DG, Freris LL. Stabilization of grid frequency through dynamic
demand control. IEEE Trans Power Syst 2007;22:128493.
[6] Takagi M, Iwafune Y, Yamamoto H, Yamaji K, Okano K, Hiwatari R, et al.
Economic value of LFC substitution by charge control for plug-in hybrid
vehicles. IEE Jpn Trans Power Energy 2010;130:20313 [in Japanese].
[7] Han S, Han S, Sezaki K. Development of an optimal vehicle-to-grid aggregator
for frequency regulation. IEEE Trans Smart Grid 2011;1:6572.
[8] Masuta T, Yokoyama A. Supplementary load frequency control by use of a
number of both electric vehicles and heat pump water heaters. IEEE Trans
Smart Grid 2012;3:125362.
[9] Shimizu K, Masuta T, Ota Y, Yokoyama A. A new load frequency control method
in power system using vehicle-to-grid system considering users
convenience. Stockholm, Sweden: 17th Power Syst Comput Conf; 2011.
[10] Galus MD, Koch S, Andersson G. Provision of load frequency control by PHEV,
controllable loads, and a cogeneration unit. IEEE Trans Ind Electron
2011;58:45826.
[11] Fazelpour F, Vafaeipour M, Rahbari O, Rosen MA. Intelligent optimization to
integrate a plug-in hybrid electric vehicle smart parking lot with renewable
energy resources and enhance grid characteristics. Energy Convers Manage.
2014;77:25061.
[12] Zakariazadeh A, Jadid S, Siano P. Multi-objective scheduling of electric vehicles
in smart distribution system. Energy Convers Manage 2014;79:4353.

[13] Jian L, Zhu X, Shao Z, Niu S, Chan CC. A scenario of vehicle-to-grid


implementation and its double-layer optimal charging strategy for
minimizing load variance within regional smart grids. Energy Convers
Manage 2014;78:50817.
[14] Zhang K, Xu L, Ouyang M, Wang H, Lu L, Li J, et al. Optimal decentralized
valley-lling charging strategy for electric vehicles. Energy Convers Manage
2014;78:53750.
[15] Next Generation Vehicle Promotion Center. Available at: <https://www.cevpc.or.jp/index.html> [in Japanese].
[16] Masuta T, Murata A, Endo E. Evaluation of relationship between demand side
management and future generation mix by energy system analysis. In: The
international conference on electrical engineering, Xiamen, China; 2013.
[17] IEA ETSAP MARKAL. Available: <http://www.iea-etsap.org/web/Markal.asp>.
[18] Hemmati R, Hooshmand RA, Khodabakhshian A. Reliability constrained
generation expansion planning with consideration of wind farms
uncertainties in deregulated electricity market. Energy Convers Manage
2013;76:51726.
[19] Li YP, Huang GH. Electric-power systems planning and greenhouse-gas
emission management under uncertainty. Energy Convers Manage
2012;57:17382.
[20] Aminloei RT, Ghaderi SF. Generation planning in Iranian power plants with
fuzzy hierarchical production planning. Energy Convers Manage
2010;51:123041.
[21] Fishbone LG, Giesen G, Goldstein G, Hymmen HA, Stocks KJ, Vos H, et al. Users
guide for MARKAL (BNL/KFA version 2.0). A multi-period, linear-programming
model for energy systems analysis. Juelich, Germany: Brookhaven National
Lab. and Kernforschungsanlage; 1983.
[22] Endo E. Market penetration analysis of fuel cell vehicles in Japan by using the
energy system model MARKAL. Int J Hydrogen Energy 2007;32:134754.
[23] Okken PA, Lako P, Gerbers D, Kram T, Ybema JR. CO2 removal in competition
with other options for reducing CO2 emissions. Energy Convers Manage
1992;33:73745.
[24] Morris SC, Lee J, Goldstein G, Steinberg M. Comparison of pre- and postcombustion carbon removal. Energy Convers Manage 1992;33:74754.
[25] Nissan LEAF electric car charging. Available: <http://www.nissanusa.com/
electric-cars/leaf/charging-range/charging/>.
[26] Ministry of land in Japan. Road trafc census 2005. Tokyo, Japan: Japan Society
of Trafc Engineers; 2005 [in Japanese].
[27] The Energy Data and Modeling Center (EDMC). Handbook of energy &
economic statics in Japan. Tokyo, Japan: The Energy Conservation Center;
2012.
[28] Japan Industrial Technology Association (JITA). Trend survey on residence in
the future. Tokyo, Japan: JITA; 1990 [in Japanese].
[29] Japan Wind Power Association (JWPA). Wind power potential and long term
target along with the road map ver. 1.1. Tokyo, Japan: JWPA; 2010 [in
Japanese].
[30] National Economic Development Ofce (NEDO). PV road map toward 2030
(PV2030+). Kawasaki, Japan: NEDO; 2009.
[31] National Policy unit in Japan. Cost verication committee report. Tokyo,
Japan: Cost Verication Committee; 2011 [in Japanese].
[32] IEE Japan, demand and supply control technology in power system, IEE Japan
technical report, No. 2302, IEEJ, Tokyo, Japan; 1989 [in Japanese].
[33] Ministry of economy, trade, and industry in Japan. Next-generation vehicle
strategy 2010. Tokyo, Japan: METI; 2010.

You might also like