Available via license: CC BY-NC-ND 4.0
Content may be subject to copyright.
FACTA UNIVERSITATIS
Series: Electronics and Energetics Vol. 33, No 4, December 2020, pp. 583-603
https://doi.org/10.2298/FUEE2004583B
© 2020 by University of Niš, Serbia | Creative Commons License: CC BY-NC-ND
RISK MANAGEMENT AND PARTICIPATION OF ELECTRIC
VEHICLE CONSIDERING TRANSMISSION LINE CONGESTION
IN THE SMART GRIDS FOR DEMAND RESPONSE
Cyrous Beyzaee, Sara Karimi Marvi, Mahdi Zarif
Department of Electrical Engineering, Mashhad Branch, Islamic Azad University,
Mashhad, Iran
Abstract. Demand response (DR) could serve as an effective tool to further balance the
electricity demand and supply in smart grids. It is also defined as the changes in
normal electricity usage by end-use customers in response to pricing and incentive
payments. Electric cars (EVs) are potentially distributed energy sources, which support
the grid-to-vehicle (G2V) and vehicle-to-grid (V2G) modes, and their participation in
time-based (e.g., time of use) and incentive-based (e.g., regulation services) DR
programs helps improve the stability and reduce the potential risks to the grid.
Moreover, the smart scheduling of EV charging and discharging activities supports the
high penetration of renewable energies with volatile energy generation. This article
was focused on DR in the presence of EVs to assess the effects of transmission line
congestion on a 33-bit grid. A random model from the standpoint of an independent
system operator was used to manage the risk and participation of EVs in the DR of
smart grids. The main risk factors were those caused by the uncertainties in renewable
energies (e.g., wind and solar), imbalance between demand and renewable energy
sources, and transmission line congestion. The effectiveness of the model in a 33-bit
grid in response to various settings (e.g., penetration rate of EVs and risk level) was
evaluated based on the transmission line congestion and system exploitation costs.
According to the results, the use of services such as time-based DR programs was
effective in the reduction of the electricity costs for independent system operators and
aggregators. In addition, the results demonstrated that the participation of EVs in
incentive-based DR programs with the park model was particularly effective in this
regard.
Key words: Electric Vehicles, Smart Grid, V2G, G2V, GAMS, Loss Function,
Demand Response
Received March 3, 2020; received in revised form August 21, 2020
Corresponding author: Cyrous Beyzaee
Department of Electrical Engineering, Mashhad Branch, Islamic Azad University, Mashhad, Iran
E-mail: mahaniranian1@gmail.com
584 C. BEYZAEE, S. KARIMI MARVI, M. ZARIF
1. INTRODUCTION
Electric vehicle (EV) sales are growing rapidly worldwide [1,2], with the amount
exceeding one million. Several factors have been involved in this growing trend in the
past few years, including the ability to replace fossil fuel vehicles with EVs, which results
in the preservation of natural reservoirs. However, the increased number of EVs leads to
increased grid demand. With the growth of domestic, industrial, and commercial demands,
the power network must be capable of responding to all types of demands. The current
power grids used in most countries are unable to fully respond to the large volume of
EVs. In this regard, the simplest solution is to increase transmission lines and various
power plants to supply the electricity required by the grid. Nonetheless, this solution
requires unjustified large operating and economic costs. As such, the proper management
of various parameters such as EVs, wind and solar power plants and new energies, programs
to reduce consumption, increased grid sustainability and customer satisfaction, and
operational costs of the system is of paramount importance.
In this context, one of the important topics is the transmission line congestion and
management of grid demand response (DR) using EVs since the lack of management of
EV charging may lead to issues such as increased grid demand, power loss, and voltage
fluctuations [3].This article was focused on the management and participation of EVs in
smart grid DR considering the impact of transmission line congestion in the form of time-
based and incentive-based programs. To obtain our goals, we have first introduced EVs,
their types, and DR in this field.
2. OVERVIEW OF EVS AND RENEWABLE ENERGY SOURCES AND DR
2.1. Evs [4-9]
There are different types of EVs, some of which use the electronic grid to supply their
required energy, which increases the energy received from the grid, thereby causing more
problems for the grid. In general, EVs are able to operate in frequency regulation, voltage
regulation, spinning and non-spinning reserves, subsidiary services, and demand profile
adjustment [4]. Compared to common vehicles (e.g., fossil fuel cars), EVs have a different
propellant. The electric power required for EVs is provided by three main sources,
including power plants, generators, and energy savers; however, most EVs are of the third
type. In recent years, special attention has been paid to plug-in EVs (PEVs; especially
battery EVs and hybrid PEVs) in the industrial and university sectors.
2.1.1. Battery EVs
Battery EVs encompass three parts, including an electronic engine, a battery, and a
controller. The electronic motor uses the battery as the driving force. The two-input
controller is only able to manage the power provided to the electric motor, which provides
the driving force for the vehicle to move backward or forward. Simultaneously, a four-input
controller supports the brake as well. Another important part of battery EVs is the power
inverter, which is responsible for the conversion of the stored electrical energy in the
battery from the DC into the AC mode. This is mainly due to the fact that most EVs have
an AC engine, which has a simple, low-cost structure.
Risk Management and Participation of Electric Vehicle Considering Transmission Line Congestion... 585
2.1.2. Hybrid PEVs
Hybrid EVs are classified into three categories of parallel, series, and two-part hybrids
based on their engine type. The first category is recognized as the most common engines of
such vehicles. PEVs often have two electronic and internal combustion engines as the
propellant, which enables the vehicle to move in the no-charge and full-charge modes.
Hybrid PEVs supply their propulsion energy from batteries. When the battery power levels
are lower than a certain amount in the no-charge mode, the vehicle changes its status and
switches to the use of the internal combustion engine as the propellant. In the full-charge
mode, the vehicle uses a combination of electric engine and internal combustion engine for
maximum efficiency in propulsion. Simultaneously, the controller controls the battery
charge level and maintains it at a certain level.
2.2. Renewable energy sources
The increased awareness of environmental crises and reduction of fossil fuel use are
leading to new directions for energy production and consumption. One of these issues is
renewable energy sources with eco-friendly features, including the wind energy and solar
energy.
2.3. Response demand
The necessity to define new electronic energy sources with quick response ability in the
emergency situations of power network is ever-increasing due to the growing load of power
networks, especially the increased loads sensitive to the changes in the power supply
parameters by the network. Therefore, it is essential to address consumer management
issues. Structural changes in the electricity industry have led to the emergence of new
paradigms alongside consumer management. DR is one of these paradigms, which
encompasses the consumer management methods that lead to changes in the consumption
level of costumers caused by the changes in electricity prices in the market. According to
the United States Department of Energy, DR is defined as the empowerment of industrial,
commercial, and residential users to improve electronic energy consumption, so that
appropriate costs could be established and the network exploitation conditions could be
improved [10]. In other words, DR could change the form of electronic energy consumption,
so that the maximum system demand would reduce and consumptions would be transferred
to non-peak hours. The US Energy Regulatory Commission divides DR programs into two
main groups of motivation-based and time-based DR [11]. In each classification, the DR
programs are divided into several subcategories, which have been discussed in the
following section [12]:
Incentive-based DR programs
1) Direct demand control programs
2) Demand reduction/cessation programs
3) Repurchase/demand sales programs
4) Emergency DR programs
5) Market capacity programs
6) Subsidiary service market programs
586 C. BEYZAEE, S. KARIMI MARVI, M. ZARIF
Time-based DR programs
1) Application-time pricing plans
2) Actual-time pricing plans
3) Critical peak-time pricing plans
3. PROBLEM STATEMENT AND MODEL PRESENTATION
With the increased prevalence of EVs and their use worldwide, there has been growing
demand for attention and planning to exploit these vehicles. Owing to their numerous
benefits, fuel fossil vehicles are being rapidly replaced by EVs. However, the increased
number of EVs has resulted in higher demands in this regard. On the other hand, the
electricity network must be able to respond to all types of demands with the ever-increasing
growth of housing, industrial, and commercial demands. To this end, the simplest solution
is to increase transmission lines and various power plants to supply the required electricity
level. Nonetheless, this solution requires substantial operating and economic costs, which
may not be economical. Therefore, the proper management of various network parameters
such as EVs, wind and solar power plants, and new energies, various programs to reduce
consumption, increased network stability, customer satisfaction, and system operating costs
is of paramount importance.
One of the key topics in this regard is the discussion of transmission line congestion,
grid demand response, and management using EVs. As such, the present study aimed to
evaluate the management and participation of EVs in the demand response of smart grids,
while considering transmission line congestion and its impact in time-based and motivation-
based programs.
3.1. Time-based programs
These programs involve the use of global networks by consumers and grid demands. By
pricing electricity at different hours (load peak, mean load, and low load), the consumption
peak is divided into non-peak hours and may reduce. Therefore, there would be no
transmission line congestion, and electricity purchase level would decrease significantly.
3.2. Motivation-based programs
Focusing on regulatory services and supplying the reserve amount are essential to the
states of G2V and V2G, leading to demand-response balance and reduction of global
network costs. Moreover, it results in increased profitability for customers and higher use
of EVs. In this program, EVs are connected to the grid in the two states of G2V and V2G,
experiencing smart discharging in addition to smart charging [13-15].
3.3. Studied system
The system assessed in the present study is illustrated in Figure 1 [16].The independent
system operator plays a pivotal role in this system, managing the market by collecting and
exporting information among the market members, such as power plants, demand centers,
and EV aggregators. The independent system operator aims to reduce the operational costs
of the system. However, the balance between supply and demand remains constant at all
times. The power system encompasses the energy distribution of various manufacturing
Risk Management and Participation of Electric Vehicle Considering Transmission Line Congestion... 587
units, such as conventional power generators and renewable energy systems (wind and solar
power). Considering the limited capacity of electric car batteries, the contribution of each
battery separately to the grid is negligible. Previous studies have indicated the inefficiency
of planning for small-scale consumer participation in wholesale electricity market [11].
Therefore, it is essential to control the charging and discharging of numerous EVs by an
aggregator to participate in tenders and coordinate the charging and discharging activities of
EVs. Notably, the aggregator units cover both V2G and G2V models. Vehicle owners
announce their battery capacity and traffic route to the aggregator units by considering the
additional time and distance and possible parameters for the proper and accurate planning
of EVs. On the other hand, the aggregator units inform the independent system operator on
the available and anticipated capacity of the state of charge (SOC) in order to participate in
the demand level and frequency tuning services.
Fig. 1 An Overview of studied system [16]
The aggregators support both the time-based and motivation-based states in demand
response programs. In this article, the time of use was selected as the time-based program
to supply the demand service provision. The vehicles participating in the program were
required with different costs at various times (e.g., load peak or low load) [17-20].
The aggregators often participate in the motivation-based programs of the demand
response to supply the required V2G and G2V for regulatory services. These services
have two classifications in terms of the costs for the independent system operator, which
involve paying the reserving capacity costs and energy costs to the aggregators [21-
22].The reserve capacity costs are equal to the maximum capacity supplied by each
aggregator during the contract. The energy costs are associated with the costs of energy
transfer from the V2G state to the G2V state. In addition, a specific number of EVs is
required for the rapid responding to the demand, as well as saving the excess energy or
compensating for its shortage. The level of emergency storage must also be set correctly.
Moreover, the decision-makings in this regard are mainly focused on the charging and
discharging of electric cars, and the plan of producing electricity from various sources
often has to be precise. Decisions should be made by considering various risk factors for
588 C. BEYZAEE, S. KARIMI MARVI, M. ZARIF
the possible future scenarios. In this regard, the risk in the mentioned conditions is the
possible imbalance between demand and power supplies (supply and demand). In the
current research, the model presented for the management of the participation of EVs in
demand response programs was based on the DC OPF model.
(1)
(2)
(3)
(4)
(5)
Equations 1-5 show the main formula of OPF, which minimizes the costs associated
with various generating units and the current load in terms of the technical limitations of the
electricity grid. In addition, Equations 2 and 3 demonstrate the load balance per bus and
power flux per line. The heat flux limit and generator capacity are shown in Equations 4
and 5. The OPF model presented above has been corrected for the integration of dynamic
issues into our model. In this respect, the main goal was to manage the level of necessary
reservation for the V2G and G2V states, as well as the anticipated costs in the future system
vision. The modified model was presented for the management of the cooperation of EVs in
Equations 5 and 6. The first part of the target function shown in Equation 6 is related to the
reservation capacity costs of EVs to conclude the V2G and G2V service contracts.
The second part includes the expected operating costs of the independent system
operators for the actual energy payments sent for regulatory services. The next part of the
production costs of conventional generators is the costs of the current load and decreased
costs of renewable energies. In addition, the energy outage costs are considered in the
model because when the surplus energy is generated by renewable sources, ISO should
be allocated to others to receive the additional energy [23].
Equation 7 is similar to Equation 2 in terms of showing the power balance per bus.
The overall energy flux to the bus (generating electricity) through conventional
generators, renewable energy sources, and energy discharge from the aggregators is equal
to the total overall energy output from the bus (base demand, charged energy of the
aggregators, reduction of renewable energies).
The SOC of each aggregator is presented in Equation 8, which changes based on the
charge/discharge status and battery efficiency. Initially, the aggregators must supply the
charge of the vehicles that immediately leave the place and need charge. The SOC of the
input and output vehicles often affects the overall charge of the aggregators. Equation 9
+,
2+,
2
=1
=1 +
=1 ,,+,,
=1
=1
+
=1 ,,+,,+,,
=1
=1
(6)
(7)
Risk Management and Participation of Electric Vehicle Considering Transmission Line Congestion... 589
shows the charge of the remaining battery capacity of the aggregators. The improvement
of the charge/discharge pattern affects the remaining battery capacity (RBC) of an EV
after arriving at the parking lot. Moreover, the RBC is affected by the SOC of the input
and output vehicles.
The equations 10-12 show a method similar to the equations 3-5. However, the flux
constraint was not presented for lines with error ().
Equations 10-12 show a method similar to Equations 3-5. However, the flux constraint
was not presented for the lines with error ().
In addition, Equation 13 guarantees the use of the generator in the allowed range. On
the other hand, Equation 14 indicates the risk coefficient required for the operator.
Accordingly, the probability of any mismatch between the power source and demand
would be less than or equal to the specified error limit ). In addition, Equations 15 and
16 guarantee the operation of the aggregator only in one of the V2G or G2V states at any
moment.
Moreover, Equation 17 demonstrates that discharge is limited by the available
energy, while Equation 18 guarantees that the level of charge does not exceed the empty
capacity available to the accumulators.
(8)
(9)
(10)
(11)
(12)
(13)
(14)
(15)
(16)
(17)
(18)
590 C. BEYZAEE, S. KARIMI MARVI, M. ZARIF
(24)
Nonetheless, Equations 19 and 22 are boundary constraints. The non-provided load
and decreased energy are limited by the actual load and renewable energy available in
Equations 19 and 20. The required storage was determined in the aggregators’ contract
and limited to their capacities to support the V2G and G2V services, while constraint 21
shows the limit of this capacity in the G2V state. Similarly, the discharge energy of the
aggregators is limited by the maximum storage defined for the V2G state in their
contracts.
Equation 23 shows the G2V reservation storage. Moreover, the maximum period
guarantees the level of G2V reservation required when the generated energy is higher
than the system’s demand. In such case, EVs are charged, and the G2V reservation level
is estimated based on their participation in the use of the surplus energy. Equation 24
shows that the G2V service provided by each aggregator cannot exceed its charge
amount, and the range of changes in the variables is shown in Equation 25.
The aforementioned model is a nonlinear complex number programming problem,
which could be converted into a linear complex number programming problem. To
establish linearity, Equation 14 is replaced by Equations 26 and 27. Moreover, the
binary variable is equal to one if there is incompatibility between the energy sources and
existing demand; otherwise, it would be zero. In order to make Equation 23 linear, we
used Equation 28 through Equation 31 to cover all the possible cases.
,,
=1 =,,
+
=,0,,,
=1 +,,
=1 ,,
=1 ,,
,,,
=1 +,,
=1 ,,
=1 1, ,
(19)
(20)
(21)
(22)
(25)
(23)
(26)
(27)
(28)
Risk Management and Participation of Electric Vehicle Considering Transmission Line Congestion... 591
4. MODEL IMPLEMENTATION AND SIMULATION
In order to evaluate the proposed models, we applied a one-day program on the
standard 33 bus grid as the case study, the characteristics of which are presented in Table
1, along with the base load. The maximum generating capacity was 700 kw, and the
minimum production value for the conventional generators was not set. The transmission
capacity of each 2 MW line with equal susceptance risk was estimated at 10 p.u. The
charging of EVs imposes an additional load to the system, which does not include the
base load.
The Parking pattern in Figure 2 was considered for the evaluation of the number of
the EV inputs and outputs each day. Each parking region had the maximum capacity of
200 vehicles and was managed by an aggregator. Therefore, it was assumed that each EV
has a battery with a 24 kwh capacity and 99% charge/discharge efficiency. In addition, it
is expected that 35% of the parked vehicles are EVs. In general, EVs enter the parking
with 30% charge and prefer to leave the parking with 90% battery charge. Figure 3 shows
the generated energy by the wind and solar power plants as selected based on the data of
California ISO wind and solar power plants [24].
The cost related to renewable energy decreased, and the reduced load was assumed as
1.5 and 5$/kw, respectively as shown in Equation 23. Furthermore, the cost related to the
generation of emergency electricity by a conventional generator was presented as
0.20 $/kwh. The aggregators cost 0.02 $/kwh for the available capacity to provide the
V2G and G2V services. The aggregators could benefit from 100% discount if they charge
when there is the need for energy reduction. In addition, the independent system operator
deals with the aggregators, high costs of regulation services, and other services. However,
the different services had various costs, which mostly depend on the electricity market
cost. For instance, 0.01 $/kwh was considered as the base cost of electricity.
,,
=1 +,,
=1 ,,
=1 ,,
+
=1, ,
,,
=1 ,,
=1 +,,
=1 ,,
=1 1,, ,
,,
=1 ,,
+
=1 ,, ,
(29)
(30)
(31)
592 C. BEYZAEE, S. KARIMI MARVI, M. ZARIF
Table 1 Characteristics of a standard 33 bus grid
Br.No
Rc.Nd
Sn.Nd
r(ohm)
x(ohm)
PL(KW)
1
0
1
0.0922
0.47
100
2
1
2
0.493
0.2511
90
3
2
3
0.366
0.1864
120
4
3
4
0.3811
0.1941
60
5
4
5
0.819
0.707
60
6
5
6
0.1872
0.6188
200
7
6
7
0.7114
0.2351
200
8
7
8
1.03
0.74
60
9
8
9
1.044
0.74
60
10
9
10
0.1966
0.065
45
11
10
11
0.3744
0.1238
60
12
11
12
1.468
1.155
60
13
12
13
0.5416
0.7129
120
14
13
14
0.591
0.526
60
15
14
15
0.7463
0.545
60
16
15
16
1.289
1.721
60
17
16
17
0.732
0.574
90
18
1
18
0.164
0.1565
90
19
18
19
1.5042
1.3554
90
20
19
20
0.4095
0.4784
90
21
20
21
0.7089
0.9373
90
22
2
22
0.4512
0.3083
90
23
22
23
0.898
0.7091
420
24
23
24
0.896
0.7011
420
25
5
25
0.203
0.1034
60
26
25
26
0.2842
0.1447
60
27
26
27
1.059
0.9337
60
28
27
28
0.8042
0.7006
120
29
28
29
0.5075
0.2585
200
30
29
30
0.9744
0.963
150
31
30
31
0.3105
0.3619
210
32
31
32
0.341
0.5302
60
Risk Management and Participation of Electric Vehicle Considering Transmission Line Congestion... 593
Fig. 2 Parking pattern
Fig. 3 Pattern of electricity generation by wind and solar sources
This model was developed in MATLAB software and solved by the CPLEX solver in
the definitive and randomized forms. It is notable that the definitive cases were
considered as the base case, and no risk range was considered for the definitive cases.
The model was solved after adjusting the random parameters for their expected values.
The results regarding the load levels in the definitive cases are shown in Figure 4, where
the collaboration of EVs was observed to be effective in correcting the available load and
using renewable energies, while transferring the load charge to off-peak periods. Figure 5
illustrates the results on the G2V and V2G services in the definitive cases. In this regard,
0
20
40
60
80
100
120
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
Parking Utilization(%)
Time(hour)
0
200
400
600
800
1000
1200
1400
1600
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
POWER GENERATION(KW)
TIME(HOUR)
WT PV
594 C. BEYZAEE, S. KARIMI MARVI, M. ZARIF
EVs provided the G2V reserve at hours by generating more renewable energy and
insufficient base load. In addition, the EVs were discharged to provide V2G services at
the load peak. Conventional generators are applied to generate the necessary electricity in
periods when the sum of renewable energies and emitted energy by EVs is insufficient to
reach the base load.
Fig. 4 Results of EV participation in base state
Fig. 5 Results of EV participation to provide reservation services in base state
The capacity of the lines also reduced to observe the effect of transmission line
congestion on the cost function in the definitive form. The maximum capacity of
transmission lines was 2 MW, and decreased congestion constraint to 1 MW led to the
0
500
1000
1500
2000
2500
3000
3500
4000
4500
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
hourly-load EV-load RES
-1000
-500
0
500
1000
1500
2000
12345678910 11 12 13 14 15 16 17 18 19 20 21 22 23 24
RG2V RV2G
Risk Management and Participation of Electric Vehicle Considering Transmission Line Congestion... 595
congestion of the lines. As a result, the cost of system operation increased. Figure 6
illustrates the results of decreased transmission line congestion and the effects on the V2G
and G2V states. As is observed, the reservation amount decreased in the G2V state with the
transmission line congestion. In contrast, the reservation amount increased in the V2G state.
Fig. 6 Effect of transmission line congestion on reservation plans in base state
In general, the reservation level in the G2V and V2G states increased, which in turn
led to the increased system operating costs.
4.1. Charging method
This section illustrates the effects of charging the EVs on the operating costs of the
power systems. The definitive base model was run with three different charging models
4917.322909
4385.643281
4100
4200
4300
4400
4500
4600
4700
4800
4900
5000
F_max(l)=2MW F_max(l)=1MW
Reserving in G2V mode
2713.14846
3592.512
0
500
1000
1500
2000
2500
3000
3500
4000
F_max(l)=2MW F_max(l)=1MW
Reserving in V2G mode
596 C. BEYZAEE, S. KARIMI MARVI, M. ZARIF
and policies. In the first policy, it was assumed that the EVs do not participate in recharge
programs and are charged once when they arrive in the parking lot. The second policy
showed that the EVs participated in the time-based program of the demand response,
which led to the planning of EV charging by the aggregators to reduce the electricity
costs and eliminate the load peak. When the aggregators attended the time-based programs
of the demand response, the independent system operator only responded to the charging
patterns by minimizing its operational costs. Table 2 shows the time spent to manage
consumer recharge. The total charging cost of the aggregators participating in the time of
use program was calculated using the
equation.
Table 2 Hourly electricity cost
Hour
Price($)
Hour
Price($)2
1
0.05
13
0.19
2
0.05
14
0.19
3
0.05
15
0.19
4
0.05
16
0.12
5
0.05
17
0.12
6
0.05
18
0.12
7
0.05
19
0.19
8
0.12
20
0.19
9
0.12
21
0.19
10
0.12
22
0.12
11
0.19
23
0.12
12
0.19
24
0.05
In the third policy, the participation of the EVs in the motivation-based program of
the demand response was assumed, and the vehicles were motivated to participate in the
G2V and V2G states. The overall energy cost of the aggregators in this policy was
estimated using the equation below:
In the equation above, the negative values indicated that not only the aggregators did
not pay the costs, but they also inspire revenue generation in most cases. The results of
the charging strategy are presented in Table 3.
Table 3 Results of three charging policies of EVs
DR charging policy
ISO reserve
cost ($)
ISO operation cost
($)
Aggregator's
energy payment ($)
Generation
(KWh)
No participation
0
44500.572
261.791
15812
time-based
0
37312.752
116.64
14756
incentive-based
417.89
25804.761
-179.741
13590
,,
+,,,,
+,
2,
2
=1
=1
(32)
Risk Management and Participation of Electric Vehicle Considering Transmission Line Congestion... 597
The conventional power generation and charging patterns of the three policies are
shown in figures 7 and 8. Participation in the demand response programs decreased the
costs of the aggregators and independent system operator. Compared to the time-based
program, the motivation-based program provided more saving in the costs of the
independent system operator, which was mainly due to the fact that the use of EVs for the
management of the V2G and G2V states could reduce the costs related to the lost load
and energy reduction. Participation in the motivation-based programs is often associated
with positive income generation for the aggregators.
As is depicted in Figure 7, unplanned charging forced the conventional power systems
to generate more power during the peak times when the system experienced higher load.
Motivational programs often cover the need for routine energy generation by entirely
using renewable sources. The main goal of demand response programs is to decrease the
load peak. According to the obtained results, participation in the time-based and
motivation-based programs led to 48% and 51% decrease in the load peak, respectively.
As is shown in Figure 8, the participation of the aggregators in the demand response
programs created the motivation for the lack of charging at the peak hours, thereby
increasing the desire to charge at the non-peak hours.
Fig. 7 Conventional power plant production rates in three different charging policies
Fig. 8 Charging activity of EVs in three different charging policies
0
500
1000
1500
2000
2500
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
Power(KW)
Time(hour)
No Participation Time_Based Incentive_Based
0
500
1000
1500
2000
12345678910 11 12 13 14 15 16 17 18 19 20 21 22 23 24
No participation(sen1) Time-based(sen2) Incentive-based(sen3)
598 C. BEYZAEE, S. KARIMI MARVI, M. ZARIF
Figure 9 shows the effect of transmission line congestion on the production of
conventional power plants in the three charging policies. According to the results, the
energy produced by conventional power plants significantly reduced in case of
congestion in the transmission lines. Considering that conventional power plants are used
to supply part of the system load that is not responsive to renewable energies and electric
vehicles, the network cannot supply that part of the system load.
Fig. 9 Effect of transmission line congestion on production of conventional power plants
0
5000
10000
15000
20000
F_max(l)=2MW F_max(l)=1MW
Conventional Genertions
Senario1
0
5000
10000
15000
20000
F_max(l)=2MW F_max(l)=1MW
Conventional
generations
Senario2
10500
11000
11500
12000
12500
13000
13500
14000
F_max(l)=2MW F_max(l)=1MW
Conventional generations
senario3
Risk Management and Participation of Electric Vehicle Considering Transmission Line Congestion... 599
According to the simulation results, the costs of the independent system operator
increased with the decreased transmission line congestion. Figure 10 depicts the results in
the three charging policies.
Fig. 10 Effect of transmission line congestion on costs of independent system operator
44500.572
67941.623
0
10000
20000
30000
40000
50000
60000
70000
80000
F_max(l)=2MW F_max(l)=1MW
Cost of ISO($)
Senario1
37312.752
60353.122
0
10000
20000
30000
40000
50000
60000
70000
F_max(l)=2MW F_max(l)=1MW
Cost of ISO($)
Senario2
25804.761
39323.319
0
5000
10000
15000
20000
25000
30000
35000
40000
45000
F_max(l)=2MW F_max(l)=1MW
Cost of ISO($)
Senario3
600 C. BEYZAEE, S. KARIMI MARVI, M. ZARIF
4.2. Risk Perspective and Random Solutions
In this section, the model is solved in the random form by the predefined risk level of
0.01, which indicated that the possibility of mismatch between the load and source must
be less than 1%. Therefore, it was assumed that the load, renewable energy production,
behavior of the EV owners, SOC input and output of the aggregators, and line errors were
uncertain. To reduce the computational time of the random model, the reduction scenario
presented in was used to construct a tree scenario with 10 scenarios [25-26]. In the
random model, a higher reserve level was required compared to the definitive status due
to the uncertainty and risk level parameters. The random model was also solved for
various risk thresholds, including 0.01, 0.1, and 1. As can be seen in Figure 11, the higher
risk threshold tolerated the higher probability of mismatch between the source and load,
thereby requiring less storage.
Fig. 11 Effect of imbalance between energy source and demand on reservation programs
with various risk factors
1800
1900
2000
2100
2200
2300
2400
2500
2600
stochastic33(0.01)-G2V stochastic33(0.1)-G2V stochastic33(1.00)-G2V
Regulation down(KW)
Risk telorance
2000
2050
2100
2150
2200
2250
2300
stochastic33(0.01)-V2G stochastic33(0.1)-V2G stochastic33(1.00)-V2G
Regulation up(KW)
Risk telorance
Risk Management and Participation of Electric Vehicle Considering Transmission Line Congestion... 601
Similar to the definitive form, the reservation level increased in the V2G and G2V
states by applying line congestion in the random form, which led to the increased cost of
system operation. However, the amount was lower compared to the definitive form,
which was due to the presence of a risk coefficient and possible disproportion between
the load and energy source. The results for 1% risk coefficient are shown in Figure 12.
Fig. 12 Effect of transmission line congestion on reservation level of demand respond
programs in random form
2490.466
2503.466
2480
2485
2490
2495
2500
2505
F_max(l)=2MW F_max(l)=1.8MW
Reserving in G2V mode(KW)
stochastic model
2260.198
2287.32
2245
2250
2255
2260
2265
2270
2275
2280
2285
2290
F_max(l)=2MW F_max(l)=1.8MW
Reserving in V2G mode(KW)
stochastic model
602 C. BEYZAEE, S. KARIMI MARVI, M. ZARIF
5. CONCLUSION
In the present study, we applied a new EV participation plan in demand response
programs and their timing in a smart grid. In addition, we evaluated the effects of
transmission line congestion on the cost of system operation and level of reservation in the
definitive and random forms. The applied system was a standard 33-bus system exposed to
the possible risk of various load levels due to the uncertainty of EVs, production of
renewable energies, transmission line congestion, and behavior of the EV owners in a group
manner. We also assessed the participation of EVs in demand response programs in time-
based and motivation-based areas, observing that the participation could be extensively
effective in the response to the load of the smart grid, thereby providing considerable load
and reducing the load peak, which led to the reduction of the operational costs of the
system, aggregators, and EV owners, as well as monetization in some cases. The random
model enables users to determine the level of risk and costs and their profits considering the
available factors. The model evaluated in this thesis could be used to improve the storage
levels required by an independent system operator by considering the profits of the
aggregators. The independent system operator could reduce operational costs by improving
the conventional production schedule and renewable energies, as well as the participation of
EVs. Moreover, the aggregators attempted to reduce the electricity costs by optimizing the
charge/discharge schedule of EVs in order to receive the maximum discount and revenue
from participation in the demand response. The definitive and random cases were assessed
to demonstrate the effects of parameters such as charging policy, level of risk, penetration
of renewable energies, and residential load pattern. According to the results, services such
as time-based programs affected the reduction of electricity costs for the independent
system operator and aggregators. In addition, the participation of EVs in the motivation-
based programs by the park model had a significant impact in this regard.
REFERENCES
[1] Lutsey N. Global milestone: the first million electric vehicles. 2015.
[2] Wilson L. The EV wedge: how electric vehicle fuel savings vary by country
(and car). 2015.
[3] Z. Yang, K. Li, A . Foley, “Computational scheduling methods for integrating plugin electric vehicles
with power systems: a review”, Renew Sustain Energy Rev, vol. 51, pp. 396–416, 2015.
[4] G. Cardoso, M. Stadler, M.C. Bozchalui, R. Sharma, C. Marnay, A. Barbosa-Povoa,
et al., “Optimal
investment and scheduling of distributed energy resources with
uncertainty in electric vehicle driving
schedules”, Energy, vol. 64, pp. 17–30, 2014.
[5] K. Mets, R. D'hulst, C. Develder, “Comparison of intelligent charging algorithms
for electric vehicles to
reduce peak load and demand variability in a distribution grid”, Commun Netw J, vol. 14, no. 6, pp. 672–
81, 2012.
[6] M. Musio, P. Lombardi, A. Damiano, “Vehicles to grid (V2G) concept applied to a
virtual power plant
structure”, In Proceedings of XIX international conference
on electrical machines - ICEM 2010, Rome,
Italy, 2010
.
[7] C. White, K. Zhang, “Using vehicle-to-grid technology for frequency regulation
and peak-load
reduction”, J Power Sources; vol. 196, no. 8, pp. 392–398, 2011
.
[8] M. Honarmand, A. Zakariazadeh, S. Jadid, “Optimal scheduling of electric vehicles in an intelligent
parking lot considering vehicle-to-grid concept and
battery condition”, Energy, vol. 65, pp. 572–579,
2014
.
[9] S. Shao, M. Pipattanasomporn, S. Rahman, “Grid integration of electric vehicles
and demand response
with customer choice”, IEEE Trans Smart Grid, vol. 3, no. 1, pp. 543–50, 2012.
Risk Management and Participation of Electric Vehicle Considering Transmission Line Congestion... 603
[10] A. Khazali, M. Kalantar, “A stochasticeprobabilistic energy and reserve market
clearing scheme for
smart power systems with plug-in electrical vehicles”,
Energy Convers Manag, vol. 105, pp. 1046–105,
2015
.
[11] E. Sortomme, M.A. El-Sharkawi, “Optimal combined bidding of vehicle-to-grid
ancillary services”,
IEEE Trans Smart Grid, vol. 3, no. 1, pp. 70–79, 2012.
[12] A. Damiano, G. Gatto, I. Marongiu, M. Porru, A. Serpi, “Vehicle-to-Grid technology: state-of-the-art and
future scenarios”, J Energy Power Eng, vol. 8, no. 1, p. 152, 2014.
[13] R. Sioshansi, “OR Forum-Modeling the impacts of electricity tariffs on plug-in
hybrid electric vehicle
charging, costs, and emissions”, Oper Res, vol. 60, no. 3, pp.
506–516, 2012.
[14] S. Habib, M. Kamran, U. Rashid, “Impact analysis of vehicle-to-grid technology
and charging strategies
of electric vehicles on distribution networksea review”, J Power Sources, vol. 277, pp. 205–214, 2015
.
[15] M. Yilmaz, P.T. Krein, “Review of the impact of vehicle-to-grid technologies on
distribution systems
and utility interfaces”, IEEE Trans Power Electron,
lov
28, no. 12, pp. 5673–5689, 2013.
[16] N. Nezamoddini, Y. Wang, Risk management and participation planning of electric vehicles in smart
grids for demand response.
[17] L. Zhang, Z. Zhao, J. Chai and Z. Kan, “Risk Identification and Analysis for PPP Projects of Electric
Vehicle Charging Infrastructure Based on 2-Tuple and the DEMATEL Model”, World Electric Vehicle
Journal, vol. 10, no. 4, 2019.
[18] Risk Management and Public Participation (GEOG90020), Handbook,the university of melbourne, 2019.
[19] Y. Wang, L. Li, “Time-of-use based electricity demand response for sustainable manufacturing systems”,
Energy, vol. 63, pp. 233–244, 2013.
[20] Y. Wang, L. Li, “Time-of-use electricity pricing for industrial customers: a survey of US utilities”, Appl
Energy, vol. 149, pp. 89–103, 2015.
[21] A. Khalid, N. Javaid, A. Mateen, M. Ilahi, T. Saba and A. Rehman. “Enhanced Time-of-Use Electricity
Price Rate Using Game Theory”.
[22] W. Kempton, J. Tomic, “Vehicle-to-grid power fundamentals: calculating capacity and net revenue”, J
Power Sources, vol. 144, no. 1, pp. 268–279, 2005.
[23] D. Howarth, B. Monsen, “Renewable energy faces daytime curtailment in California”, 2015.
[24] CAISO renewable watch data.
[25] N. Growe-Kuska, H. Heitsch, W. Romisch, “Scenario reduction and scenario tree € construction for
power management problems”, In Proceedings of IEEE power tech conference, Bologna, Italy, 2003.
[26] F. Thangaiyan, Risk Management in renewable energy and sustainability in india., project. July 2019.
