Accepted Manuscript

Optimal Scheduling of Plug-in Electric Vehicles and Renewable Micro-grid in

Energy and Reserve Markets Considering Demand Response Program

Parinaz Aliasghari, Behnam Mohammadi-Ivatloo, Manijeh Alipour, Mehdi Abapour,

Kazem Zare

PII: S0959-6526(18)30715-7

DOI: 10.1016/j.jclepro.2018.03.058

Reference: JCLP 12316

To appear in: Journal of Cleaner Production

Received Date: 21 September 2017

Revised Date: 05 March 2018

Accepted Date: 06 March 2018

Please cite this article as: Parinaz Aliasghari, Behnam Mohammadi-Ivatloo, Manijeh Alipour, Mehdi
Abapour, Kazem Zare, Optimal Scheduling of Plug-in Electric Vehicles and Renewable Micro-grid
in Energy and Reserve Markets Considering Demand Response Program, Journal of Cleaner
Production (2018), doi: 10.1016/j.jclepro.2018.03.058

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to
our customers we are providing this early version of the manuscript. The manuscript will undergo
copyediting, typesetting, and review of the resulting proof before it is published in its final form.
Please note that during the production process errors may be discovered which could affect the
content, and all legal disclaimers that apply to the journal pertain.
 ACCEPTED MANUSCRIPT

Optimal Scheduling of Plug-in Electric Vehicles and Renewable Micro-grid in Energy and

Reserve Markets Considering Demand Response Program

Parinaz Aliasghari1, Behnam Mohammadi-Ivatloo1,*, Manijeh Alipour1, Mehdi Abapour1, Kazem

Zare1

1 Faculty of Electrical and Computer Engineering, University of Tabriz, Tabriz, Iran

*Corresponding author: mohammadi@ieee.org

Abstract

Plug-in electrical vehicles (PEVs) are introduced as a compatible transportation system for the

environment. Manufacturing technology of electric vehicles (EVs) could bring an opportunity to

implement them as energy storages. By expanding the use of renewable energy sources (RESs),

the role of energy storage system is highlighted to overcome power generation fluctuations.

Integrating PEVs and RESs could be profitable for both PEV and RESs owners. In this paper, a

structure of renewable energy sources based micro grid (RMG) is considered. The proposed

RMG has been equipped with a parking lot in order to control and aggregate PEVs. This paper

investigates the optimal energy management problem of the RMG with the presence of PEVs.

The objective of the RMG owner is to minimize the cost through generating power with its local

generators and trading energy with the power market considering the market price. Also, the

RMG could incentive PEV owners to take part in the demand response (DR) programs as a

flexible load. It could bring profit for both PEVs and RMG owners. The existence uncertainties

are modeled in the scenario-based framework. Three case studies are analyzed to display the

effectiveness of the proposed model. As a result, utilization of the parking lot has decreased the
 ACCEPTED MANUSCRIPT

cost of RMG about 40%. Additionally, implementing DR program during the charging process

of PEVs could bring extra profit for both RMG and PEVs owners. The results have shown the

efficiency of the proposed method.

Nomenclature

Index:

t Time index

h Time index

s Scenario index

 PEV index

 Local Generator index (MTs and FC)

Parameters:

MTU  / MDU  Minimum up/ down time of the th local Generator

R U / R D Ramp up/ down ramp of the th local Generator

E / E  Minimum/ maximum rate of charging/discharge of  th PEV

E Maximum rates of charging of  th PEV

ch /disch Charging/discharging efficiency of  th vehicle

SE  / SE  Minimum/ maximum state of charge  th PEV

et / et ,h Self/cross price elasticity of the  th PEV

Variables:

Sp t ,s Wind speed sth scenario at time t

t ,s
Pwind Wind power sth scenario at time t
 ACCEPTED MANUSCRIPT

t ,s
Pspillage,wind
Spillage wind power sth scenario at time t

Cost t ,s Operation cost of th local Generator at scenario s and time t

t ,s
PBuy Purchased power sth scenario at time t

t ,s
PSale Sold power sth scenario at time t

PRt ,s Sold reserve power sth scenario at time t

DE t ,s Discharge power of the  th PEV at scenario s and time t

SE s ,t State of charge  th PEV at scenario s and time t

fixrate
t ,s
Fix rate price for PEV

TOU
t ,s
TOU price for PEV

Cont ,s Energy consumption of the  th PEV at scenario s and time t

1. Introduction

Increasing energy demand in the world has caused reliability and security problems for power

systems. Micro grids (MGs), containing distributed generations and local loads, are introduced as

a local distributed of the electric power. A MG is able to operate not only in islanded mode but

also grid-connected [1]. This flexibility in operation has increased the efficiency and profit of the

MG. Moreover, pollution concerns and limited resources of fossil fuels have made an additional

problem for governments [2]. To deal with this problem, the authors have planned, sized and

operated of a hybrid, renewable energy based micro grid (RMG) including photovoltaic (PV)

panels and wind turbine to minimize the lifecycle cost, while considering environmental

emissions [3]. But power generation by renewable energy sources (RESs) such wind energy has
 ACCEPTED MANUSCRIPT

depended on weather and geographical conditions, which makes mismatch between supply and

demand [4]. The technology of energy storage systems (ESSs) could rebate these unbalanced

situation by charging in the surplus generating of power and discharging power into the shortage

generation period. ESS are allowing RES to behave more reliable like as traditional energy

sources [5, 6].

Plug-in electric vehicles (PEVs) are introduced as another key solution to reduce the emission of

the greenhouse gases [8 ,7]. PEVs are mobile and uncertain consumers, which may have side

effects on the grid, especially with the high penetration in the future [9]. The authors in [10] have

anticipated that PEVs will form 20% of the U.S car market by 2030. On the other hand, PEVs

have a potential to exchange power with the grid in two states of charging (gird to vehicle) and

discharging (vehicle to gird) their batteries. Also, a MG can use PEVs as a huge number of

energy storage to provide local load, compensate the intermittent of the RESs generation or trade

energy with the grid [11]. According to vehicle to gird capability of PEVs, EV aggregators

would be able to take part in the power market and ancillary services [12, 13]. On the other hand,

EV owners would earn money or paid less money for charging their cars with respect to the

bidirectional contract deal between them and aggregators [14]. It is clear that PEV owners’

charging pattern have affected the value of the aggregators’ profit. Because of that charging and

discharging behavior have assessed in many studies. In [15], a charging algorithm based on the

specific regulation has been designed to maximize the aggregator profit. In order to formulate the

algorithm, price constraint and an optional system load are considered. The authors in [16], have

presented a model to manage energy resources based on a MG. The model consisted of practical

constraints, the uncertainty of RES generation, requirements of spinning reserve and PEVs

owner consent. This model utilized a smart parking lot. The owner of the smart parking lot
 ACCEPTED MANUSCRIPT

compares the market electricity price and desired charging/discharging prices which are received

from PEV owners for each hour. Then it is decided that the vehicle will be charged/discharged,

participated in the reserve market or stated in the idle mode. The influences of market regulation

factors have been analyzed upon the market players specially PEV owners in [17]. In this

analysis, the type of signed contract between the PEV owners and the aggregator has considered.

The authors in [18], have extracted the real time mobility behavior of PEVs in the US by

surveying the Household Travel Data in 2009. In this research, PEVs have been operated as

ESSs in the market to maximize the social benefits. Additionally, the potential of the vehicle

fleet to take part in the market of ancillary service, has been assessed. In [19] the effects of de-

regulating of the power system has been investigated upon the generating companies (Gencos).

This research has implemented the concepts of the game theory to design bidding strategy for

both the energy and reserve markets. The authors in [20], have presented a new stochastic

programming framework based on Monte Carlo simulations to carry out the optimal operation of

MG containing renewable energy generation units, ES devices and plug-in hybrid electric

vehicles (PHEVs).

With respect to the use of the vehicles, it is logical that PEVs are considered as a flexible load.

So, if they take part in demand response (DR) programs, multiple benefits are achieved for PEV

owners, MG and the grid. Few types of researches have considered DR programs and PEVs

together with traditional generators in supply side [21-23]. A new optimization framework has

been presented in [24] to optimize the bidding strategy of a smart distribution company (SDC) in

a day-ahead market. This SDC contains wind farms as stochastic generation units and PEVs as

responsive loads. To encourage PEV owners to take part in DR programs, optimal hourly prices

have determined and sent to them via the smart communication system. In the previous studies,
 ACCEPTED MANUSCRIPT

all of the mentioned options have not considered. Some of these researches have only

investigated PEVs participating in day-ahead market. Others considered both day-ahead and

reserve markets. The others, considered DR programs and EVs together on the supply side. In

the current study RMG equipped with a parking lot is considered in order to control and

aggregate PEVs. To solve the optimal scheduling problem of RMG in the presence of PEVs, the

scenario-based stochastic framework is presented over a 24-hour time horizon. It is assumed that,

the RMG owner can exchange (procure or sell) power with the grid regarding the power market

prices. Moreover, due to the ability of storing energy in the PEVs, they could take part in both

day-ahead and reserve market. The stochastic programming approach constitutes a suitable tool

to make decisions under uncertainty and reveals the fact that new information about the uncertain

data becomes known as time evolves along the planning horizon [25]. In a multi-stage stochastic

programming, the decisions made for a stage are not affected with the information arriving in

following stages [28]. In the current study, the uncertainties of the prices of day-ahead market,

wind speed, load demand, the status of being called in the reserve market are modeled through

scenarios. In this paper, the ARMA model is utilized to produce the scenarios for the

uncertainties of prices, wind speed and demand.

The rest of the present study is organized as follows. Details of the RMG components are

provided in Section 2. The problem formulation and a brief explanation of demand response

program are presented in Section 3. Moreover, the results obtained through the application of the

proposed approach upon appropriate case studies are presented in Section 4. Finally, Section 5

concludes the paper.

2. System topology
 ACCEPTED MANUSCRIPT

The topology of proposed RMG is presented in this section. The RMG consists of micro turbine

(MT) systems, fuel cell (FC) and wind turbine (WT) as shown in Fig. 1. An intelligent parking

lot is also considered in the RMG to play the role of a PEV aggregator. The RMG owner

facilitates the PEVs’ participation in the energy and reserve markets. The demand of RMG could

be served through the local generators or the grid to minimize the operation cost of the RMG.

The surplus production of electric power in the RMG could be sold in the market or stored. The

detailed modeling of generators in the RMG are presented as follows.

Power Market

Anode
Electrolyte
Cathod
Fuel Cell
Micro turbine

Fig1. Structure of the studied RMG

 ACCEPTED MANUSCRIPT

1.2 Wind power

The availability of electrical power generated by RESs is depended on the availability of primary

sources such as sun or wind. Since RES based units located into a MG, have variable and

intermittent nature, the conventional operation method has encountered some challenges.

Depending on the time of the day and the availability of their primary sources, these RES based

units fluctuate significantly. This uncertainty and variability must be taken into consideration in

the scheduling of MGs. Therefore, forecasting wind speed and sun radiation play a key role in

the secure operation of MGs. In the current study, the MG has been equipped with WT. To

simulate the prediction model of hourly average wind speeds, ARMA model has been introduced

as an appropriate model [20]. ARMA models includes main basic features of wind speed data

such as non-stationary, non-Gaussian distribution and autocorrelation. In order to involve the

uncertainty of wind power generation upon the RMG scheduling problem, the ARMA models

are implemented to forecast the wind speed. In the process of the generating wind speed, the

errors of scenarios will be calculated by comparing the forecasted values with some realized

values of wind speed. These scenarios will be updated by adding the error to the forecasted

values. Afterwards, the frequency distribution of the forecasted wind speed, which is a Weibull

distribution, is generated for each section of the scheduling time horizon. As the next step, using

the scale and shape parameters of Weibull distribution, the wind power scenarios are generated

for each hour of the day. More details have been presented in [25]. Eventually, SCENRED tool

under GAMS environment is used to decrease the number of the scenarios with respect to their

probability. It is worth mentioning that, this tool is also used to reduce the number of price and
 ACCEPTED MANUSCRIPT

load scenarios. The total power generated by a WT is a function of turbine specifications and the

wind speed can be modeled by the equation (1) as proposed by [26].

0 Sp t ,s  Sp CI , Sp t ,s  Sp CO

  Sp t ,s  Sp CI 
PAt ,,wind
s
 Pwind  CI 
Sp CI  Sp t ,s  Sp R (1)
 Sp  Sp 
R

P Sp R  Sp t ,s  Sp CO
 wind

where, Sp CI , Sp CO and Sp R represent, cut-in, and cut-out and rated wind speed, respectively.

Furthermore, available and maximum wind powers are represented by PAt ,,wind
s
and Pwind ,

respectively. The power generated by the WT at time interval tth is limited by the available wind

power. Wind power spillage is also allowed. The proposed framework is responsible for deciding

about utilizing renewable generation with respect to the operational constraints and the total cost.

This restriction is enforced by the following constraint:

PAt ,,wind
s
 Pwind
t ,s
 Pspillag
t ,s
e ,wind (2)

where, t and s show the index of each hour and scenario, respectively. PAt ,,wind
s t ,s
, Pwind and

t ,s
Pspillag e ,wind refer to the available, used and spillage wind power during the scheduling horizon,

respectively.

2.2 Micro turbine and fuel cell units

Distributed generation (DG) units that are located near the consumers of electrical energy consist

different technologies such as MTs and FCs [27]. MT technology is highly flexible since some

small-scale units could be combined together to make larger systems ranged from several kWs

up to several MWs of power. Also, having low emissions is another advantage of this
 ACCEPTED MANUSCRIPT

technology. FCs have several advantages including being environmentally friendly, producing

less noise, and being highly energy efficient under varying load rates [28]. Operation data of

MTs and FC adopted from [29], is displayed in Table 2.

2.3 Parking lots

As mentioned, the studied RMG has been equipped with a parking lot in order to control and

aggregate PEVs. This situation provides a suitable opportunity for the RMG owner to store

energy in the PEVs’ batteries due to surplus generation of the electric power for the future

consumption. Moreover, the owner could take part in the energy and reserve markets on behalf

of the PEVs owners.

3. Problem Formulation

In this paper, the cost function is analyzed due to minimizing costs and maximizing benefits of

the RMG equipped with the parking lot, while providing its own demand. Moreover, PEVs

drivers are encouraged through the monetary incentive to participate in DR programs.

Mathematically, the problem and its related constraints can be drawn as follows.

3.1 Objective Function

The costs include the cost of operational and startup/shot down of the local generation units as

well as purchased power from the grid. Participating in the power market is unappealing for PEV

owners because of the more degradation of battery during the frequent charge/discharge

strategies [30]. So, the degradation cost of PEVs’ batteries is considered in the objective

function. The revenue includes the income of selling surplus power to the market and taking part

in the reserve market via PEVs.

 ACCEPTED MANUSCRIPT

Min OF  CostTotal - RevenueTotal (3)

S T
CostTotal    s . Cost LG
t ,s
 Cost market
t ,s
 Cost PEV
t ,s
 (4)
s 1 i 1

N LG
t ,s
Cost LG   Cost
 1
t ,s
 . Pt ,s (5)

t ,s
Cost market   t ,s . PBuy
t ,s
(6)

V
t
Cost PEV  Cost deg . DE t ,s (7)
 1

S  V  0.1.   Rt ,s  . Pt,R    Rt ,s  . Pt,,Rs 

Revenue t
   s .   .  Psale    n . 
t , s t , s  (8)
    fixrate  .Cont ,s 
Total
s 1  1
  

In the above formulation, CostTotal and RevenueTotal represent the total cost and revenue of the

RMG, respectively. Equations (5), (6) and (7) determine the generating cost via local generators

into the MG, cost of power purchased from the market and cost of battery degradation due to

discharging electrical [25]. Equation (8) determines earned revenue from selling surplus power

to the market, selling power to the PEV owners with a fix rate price, fixrate , as well as taking part

in the reserve via PEVs. It is supposed that the players will obtain 10% of the energy price if they

participate in the reserve market. Moreover, if they are being called in the reserve market, they

will sell the energy with the energy price which is equal to st .

3.2 Constraints

In this subsection, the operational and technical constraints of the RMG’s scheduling problem

including the constraints of power balance between supply and demand, PEV, MT and FC

systems are presented individually.

 ACCEPTED MANUSCRIPT

3.2.1 Power balance constraint

It is necessary that each scenario and hour the summation of generated power (purchased, sold

and generated) and consumption power must be zero. The power balance between power

consumption and generation within the RMG is guaranteed in the below equation:

N N LG N
t ,s
Pwind  PBuy
t ,s
   n . DE t ,s    PLG
t ,s
,s  PSale  PL
t ,s t ,s
   n . E t ,s  t , s (9)
 1 l 1  1

3.2.2 PEV’s constraints:

The parking lot into the RMG would behave as a virtual power plant and inject the electrical

energy to the grid through the participation in the energy and reserve markets. The parking lot

could sell power in the reserve market if it is being called by ISO. This can be described by the

following equation:

Pt,R . call t ,s  Pt,,Rs (10)

where, call t ,s represents a binary stochastic process which call t ,s  1 means that ISO will call

the parking lot owner to deliver energy and vice versa. Equation (11) describes the discharged

energy during plugged-in time interval

Pt,,sale
s
 Pt,,Rs  DE t ,s (11)

In which, the amount of discharging power of PEV’s battery, DE t ,s , is equal to the amount of

discharging power during the day-ahead market, Pt,,sale

s
, and called power for the reserve market,

Pt,,Rs .The following equations are used to update the state of charge for each PEV at hour t . Is is

drawn from the charging power, the power sold in the market, the power used for driving Cont ,s ,

and its state of energy at the previous hour:

 ACCEPTED MANUSCRIPT

SE t ,s  SE t 1,s  ch E t ,s  Cont ,s disch DE s ,t  (12)

SE   SE t ,s  SE  (13)

E   E t ,s  E  (14)

Pt,,Rs  SE t ,s  SE  (15)

E t ,s , Pt,,Rs , Pt,Sale
,s
0 t ,  Con t ,s
  0 (16)

Pt,,Rs , Pt,Sale
,s
0 t ,  E t ,s  0 (17)

where, SE t ,s calculated by (12) depicts the state of charge vehicle th at scenario s and hour t .

The state of charge PEV’s battery is limited by (13). In constrain (14), charging rate, E t ,s , is

restricted by the maximum and minimum charging rates depicted by E  and E  , respectively.

The amount of reserve power, Pt,,Rs , is limited by (15). Constraint (16) states that PEV can only

be charged or discharged during plugged times and constraint (17) states the battery cannot be

charged and discharged simultaneously.

3.2.3 Constraints of local generation units

In order to force the minimum up /down time of the th local generator of RMG including MT1,

MT2 and FC, Equations (20-23) are used. Two auxiliary variables  t and  t are utilized which

are connected to each other according to (21).

U t Pt,min
,s
 Pt ,s  U t Pt,max
,s
(18)

 t   t  U t U t 1 (19)

t  MTU  1

 t
U t   t . MTU  (t  1, ,T  MTU   1) (20)
 ACCEPTED MANUSCRIPT

U
t
t
   t . T  t  1 (t  T  MTU   2, ,T ) (21)

t  MTD 

 1 U    . T
t
t

t
  t  1 (t  1, ,T  MTD   1) (22)

 1 U    . T
t
t

t
  t  1 (t  T  MTD   2, ,T ) (23)

The restriction of minimum up time for each unit is forced by (20) and (21). Moreover, to

observe the restriction of minimum down time, Equations (22) and (23) should be considered for

each unit. The binary variable U t is equal to 1 when unit th is on, otherwise is 0.The Ramp up

and down of the th local generator of RMG are implemented by (24) and (25), respectively. :

Pt ,s .U t  Pt 1,s .U t 1  R U (24)

Pt 1,s .U t 1  Pt ,s .U t  R D (25)

3.3 Process of scenario generation

The rendered stochastic programming model is applied to solve the optimal scheduling problem

of RMG in the presence of PEVs over a 24-hour time period. It is assumed that the framework is

bidirectional and the RMG owner could exchange power with the grid regarding the power

market prices. The stochastic programming procedure could be utilized as a suitable tool for

deciding under uncertainty and appearing new information along the planning horizon [25]. In

the multi-stage programming with a stochastic framework, decisions in the current stage are

independently made from the information of the foreword stage [31]. The ARIMA model is

employed to produce the scenarios for the uncertainties of prices, wind speed and RMG demand.
 ACCEPTED MANUSCRIPT

An appropriate MATLAB function is implemented in order to calculate the parameters of

ARIMA model. The process of ARIMA parameters calculation and its equations have been

presented by [25]. Additionally, the calling state scenarios for the reserve market are produced

by utilizing the uniform distribution. The process of scenario generation for ISO calling reserve

market is indicated in the following algorithm as shown in Fig. 2:

 ACCEPTED MANUSCRIPT

Start

Set the scenario counter sc = 0

Set the time counter tc = 0

Extract the probability of being call

and set parameter probcall,t

Produce the random number Calls,t

(between (0,1))

Calls,t <= probcall,t

Calls,t = 0 Calls,t = 1

Yes
tc < T tc=tc+1

No
Yes
sc=sc+1 sc < Ns

No

End

Fig. 1. Scenario generation for reserve market calling

In the algorithm, a random number, call t ,s , is produced by the uniform distribution for both each

hour and scenario. Then it is compared with the probability of being called by the ISO depicted
 ACCEPTED MANUSCRIPT

t
by probcall . The RMG owner will be called if the produced number is bigger than the probability

of being called.

3.4 Demand response model

Each consumer is capable to manage its consumption independently from the others, according

to its characteristics. Different kinds of consumers respond differently to the same price. The

behavior of each consumer can be modeled through its utility or benefit function. Each consumer

adjusts its power consumption in order to maximize its own welfare. Moreover, electrical loads

can be classified into two categories: shiftable and non-shiftable loads. The first category

includes the loads that can be operated at any time throughout a day or a particular time window.

The latter includes loads that cannot be shifted in a day. A discussion has been presented in [32]

about the process of modeling and formulating how the electricity demand is affected by the time

of use program.

Usually, vehicles only about 5% of time are driven on the road and for most of the time they

remain stationary [33]. Thus, the charging pattern of PEVs is the most likely to change in

accordance with hourly price signals. The RMG owner can design a proper DR program for PEV

drivers to manage their charging pattern. It could earn profit for both PEVs and RMG owners.

The DR program based on time of use (TOU) price could encourage customers to manage their

consumption according to the received price signal. Therefore, the customers prefer to reduce

their consumption during the high price periods or shift it into the low price periods. Based on

load characteristics including demand profile and price elasticity, the authors of [24] have driven

a comprehensive economic model of responsive demand. The proposed model is based on TOU

price. In this study, PEVs are considered as flexible load and PEVs owners are quite willing to

take part in the DR program to reduce their bill. So the economic model of [24, 34] is modified
 ACCEPTED MANUSCRIPT

for PEV’s energy consumption depicted by Cont ,s . The following equation presents the final

responsive economic demand model of PEVs in time interval t :

 e t .   t ,s    24 t ,h  TOU
h ,s
 fixrate  
   TOU fixrate 
Con t ,s
Con 1 
t
  e  t  h (26)
 fixrate h 1 fixrate 
 h t 

In order to protect consumers against high energy prices, the price of the TOU program, TOU
t ,s
,

for each level is determined by the mean value of the prices into the related TOU interval. Mixed

integer non-linear programming (MINLP) using DICOPT solver under GAMS is utilized to

solve the optimal scheduling of PEVs and RMG model.

1. Numerical results

The rendered stochastic programming model is used for solving the optimal scheduling problem

of energy management of a RMG in the presence of PEVs in Fig. 1. The ARIMA model is

employed to produce scenarios for the uncertain behaviors of prices, speed of wind and demand.

The types of PEVs are selected by considering the information of Table 1 based on the number

of selling brands in year 2017. According to [31], ten scenarios are generated from the three

different driving patterns of PEV. Figure 2 portrayed different states of PEV through the day and

night. The amount of consumed energy and the used periods in the various driving patterns are

indicated by the bar graph in Fig. 3.

Table 1. Bestselling PEV brands

PEV Brand Capacity (kWh) Percentage (% ) Pch-max(kW)
Nissan Leaf 40 25.52 11.5
Tesla Model S 100 21.81 17.2
Tesla Model X 100 18.64 17.2
Renault Zoe 41 15.00 20
Alliance other PEV 25 19.03 12.5
 ACCEPTED MANUSCRIPT

Daily energy(kW)
3

20
0

15
1
2
3
4
5
10
Time (h)
6
7 5
PEV's number 8
9
10

Fig. 3. Ten different driving patterns of PEVs

It is assuming that, the capacity of the parking lot is 200 vehicles. The number of PEVs in

different driving pattern scenarios is calculated by the probability of scenarios in which shown in

Table 2. The Operation data of WT, MTs and FC adopted from [25, 29, 35], are displayed in

Tables 3 and 4, respectively. Additionally, the expected value of the generation scenarios of the

market price, TOU price and demand are illustrated in Fig. 4.

Table 2. Probability of number of EV types

PEV type Probability
1 0.09
2 0.05
3 0.057
4 0.06
5 0.077
6 0.12
7 0.13
8 0.083
9 0.183
10 0.15
 ACCEPTED MANUSCRIPT

4
x 10
100 2
Market Price
TOU Price
RMG Demand
Price(MWh$)

Load (MW)
50 1.5

0 1
0 5 10 15 20 25
Time (h)

Fig. 4. The expected value of price, TOU price and demand

Table 3. Operational data of WT

Rated power Cut-in speed Rated speed Cut-out speed
Type (kW) (m/s) (m/s) (m/s)

WT 500 3 12 13

Table 4. Operational data of local generators

Min Power Max Power Bid Start-up/Shut- Ram up/down
Type
(kW) (kW) ($ /kWh) down cost ($) (kW)

Micro turbine 1 60 300 0.0054 0.0173 60

Micro turbine 2 90 450 0.0054 0.0173 90
Fuel cell 3 30 0.0035 0.0165 -

To show the effectiveness of the proposed method three case studies are considered:

Case study 1: Energy management of the renewable micro grid without considering parking lot

Case study 2: Energy management of the renewable micro grid with considering parking lot
 ACCEPTED MANUSCRIPT

Case study 3: Energy management of the renewable micro grid with considering parking lot and

implementing DR program for PEVs

In Table 5, the difference of the case studies has reflected. Table 6 has summarized the

comparison of the RMG’s cost for different number of scenarios in different cases. The objective

function’s value has been stabled after solving the problem for 200 scenarios. The difference

between the results at 200 and 150 scenarios is equal to 0.22%. It means that, when the number

of scenarios is increased, there is no-significant change in the results. Moreover, increasing the

number of the scenarios increases the burden of the computation. Therefore, 200 scenarios have

been implemented for each stochastic value.

Table 5. Characteristics of case studies

Day-ahead market Reserve market Parking lot DR program
Case 1 
Case 2   
Case 3    

Table 6. Simulation results for different case studies

Scenario numbers Case 1 Case 2 Case 3
50 871.753 527.190 502.429
100 882.960 533.121 506.323
150 873.579 527.306 502.689
200 874.379 530.627 503.804

Case study 1:

In this case, the RMG is not equipped with the parking lot. It can reduce its cost by producing

power with local generators in high price and buying power in low price of energy. Since in this

case there are no PEVs and ESSs, the RGM owner can adjusts the value of generation and

purchased power during the day from the local generations and the energy market to reduce its

cost. Because of the lower price of the production power than the purchased power, FC is

generating the power all over the day with its high capacity. MT2 is operated between hours 7 to

20 when the price of electricity is higher than its operation cost as well as MT1 is operated 2
 ACCEPTED MANUSCRIPT

hours less than MT2. The RMG has to buy power from the market to provide the demand while

it schedules to reduce its cost. The generation power of the WT is also providing the part of the

demand with respect to its available power. Fig. 5 has depicted the expected value of dispatching

power in this case study.

1200
Pwind
1000 PMT2
PMT1
PFC
800
Psale
Pbuy
600
Power(kW)

400

200

-200

-400

-600

-800
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
Time(h)

Fig. 5. Expected power dispatching in Case 1.

Case study 2:

The RMG is equipped with the parking lot in this case. The RMG owner can use the capacity of

PEVs’ batteries to store the lower-cost production power, which are producing by the local

generators and WT as well as purchasing power from the market. The scenarios of PEVs’ type

and driving pattern are generated through the Mont Carlo simulation with respect to their

probability. The capacity and charging rate of each PEV are identified by the producing scenario

as shown in Fig. 6. The efficiency of charging/discharging of PEV is considered to be 100%. The

process of degradation cost calculation is adopted from [31]. As depicted in Fig. 7 in the
 ACCEPTED MANUSCRIPT

presence of PEVs, the owner is able to participate in

effective than Case 1. The owner is going to purchase power in low price and store the surplus

power into the PEVs’ batteries. He can consume or sell the stored power in the high price of

energy according to availability of PEVs in the parking lot. The RMG owner charges the PEVs

with constant price. In order to protect both the RMG and PEV owners, the fix rate is equal to

the mean value of the day-ahead price. Based on Table 6

it means that the operator can reduce its cost about 343.75$ compared with the first case for 200

scenarios.
 ACCEPTED MANUSCRIPT

Start A

Receive the information from Fig. 2 and

Set the numbers of PEV Nev = 0
Table 2 and generate the probability of
each driving pattern (prob_s)

Receive the information from Table 1 and

generate the probability of each type Produce the random number Mev
(prob_t) (between (0,1))

Produce the random number Tev Yes

(between (0,1)) Mev <= prob_s1 Mev_s1=1

. .
.. No ..
Yes Yes
Type_ev=1 Mev <= prob_s9 Mev_s9=9
Tev <= prob_t1
Pch=1
. .
.. No .. No
Yes Type_ev=3 Mev_s10
Tev <= prob_t3
Pch=3

No
Type_ev=4
Pch=4
Calculate the level of SOC

No
Nev=Nev+1 Nev<200

Yes

A End

Fig. 6. Mont Carlo simulation for generating scenario of PEVs’ type and driving pattern

As shown in the flowchart, Fig. 6, T ev and M ev are random numbers generated between (0,1),

which are compared with the probability of each type of PEVs ( prob _ t ) and driving pattern (

prob  s ), respectively.
 ACCEPTED MANUSCRIPT

Psale
800
4000 Pwind
PMT2
PMT1 600
PFC
Pres 400
2000 Pbuy

21
Power(kW)

-2000

-4000

-6000
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
Time(h)

Fig. 7. Expected power dispatching in Case 2

According to Fig. 7, the amount of sold power significantly is increased, in the presence of

PEVs, especially during the high price periods. Actually, the RMG owner plans to buy and sell

power from the market during the low and peak demand hours. The MT units are operated by

satisfying the technical constraints from hours 7 and 8 to 18 and 18, respectively. The FC is

operated whole day with its maximum capacity. The PEVs are called by ISO with low

probability in reserve market. Entirely, the total amount of energy sold in reserve market is about

100 kWh.

Case study 3:

In this case, PEV owners take part in DR program, which are implemented by the RMG owner,

which is profitable for both of them. Three levels on-peak (11:00-17:00), mid-peak (7:00-11:00,

17:00-19:00) and off-peak (19:00-7:00) have been considered for TOU program. The value of

elasticity of Canadian’s power system are adopted from [34] and presented in Table 7.
 ACCEPTED MANUSCRIPT

Table 7. Characteristics of the elasticity

On-peak Mid-peak Off-peak

On-peak -0.150 0.080 0.070

Mid-peak 0.080 -0.140 0.050
Off-peak 0.070 0.050 -0.120

The energy consumption of Each PEV during the day is calculated by (21). The amounts of the

TOU prices for three levels are considered the mean value of the forecasted prices on that

interval. Therefore, drivers prefer to shift their consumption from high price to low price of the

TOU program and reduce their consumption with respect to their cross elasticity and self-

elasticity, respectively. The created changes in the driving patterns can reduce the cost of both

PEVs and the RMG owners. As shown in Table 6 the cost of RMG has been reduced about 5.1%

against Case 2. The value of improvement could be increased by enlarging the capacity of the

parking lot.

2. Summery and Conclusion

In the current study renewable energy sources based MG was considered. The proposed RMG

has been equipped with a parking lot in order to control and aggregate PEVs. The optimal energy

management problem of the RMG in the presence of PEVs was formulated in the stochastic

framework. The ARIMA model was used to generate the scenarios of the market price, wind

speed and load demand of RMG. It is assumed that

the grid. Moreover, due to the ability of storing energy in the PEVs’ batteries, it can take part in

the energy and reserve market more flexible. Additionally, the RMG owner has encouraged PEV

drivers to participate in the DR program. To evaluate the efficiency of the proposed model, three

case studies were designed, including scheduling of RMG without and with considering parking

lot as well as charging PEVs with respect to the designed DR program. The PEVs owners have
 ACCEPTED MANUSCRIPT

motivation to participate in DR program in order to reduce their bill. The result show that the

equipped RMG to parking lot could take advantage of PEVs’ batteries to store surplus generated

power based RESs which could be utilized or sold at the high price electricity in the market.

Moreover, the owner could participate in the reserve market with more flexible scheduling. The

positive role of parking lot in the reduction of RMG’s total cost by selling electricity to the

PEVs, participating in the energy and reserve market with respect to the price of electricity was

portrayed in the second case study. According to the results, there is 39.3% reduction of total

cost in Case2 against Case1. Finally, in Case 3, utilized DR program has improved the trend of

PEVs’ charging and decreased the cost of RMG about 5.1% against Case 2. This amount can

increased by enlarging the capacity of the parking lot.

References

[1] M. E. Khodayar, M. Barati, and M. Shahidehpour, "Integration of high reliability distribution system in
microgrid operation," IEEE Transactions on Smart Grid, vol. 3, no. 4, pp. 1997-2006, 2012.
[2] M. S. Kumar and S. T. Revankar, "Development scheme and key technology of an electric vehicle: An
overview," Renewable and Sustainable Energy Reviews, vol. 70, pp. 1266-1285, 2017.
[3] O. Hafez and K. Bhattacharya, "Optimal planning and design of a renewable energy based supply system
for microgrids," Renewable Energy, vol. 45, pp. 7-15, 2012.
[4] J. Wang et al., "Wind power forecasting uncertainty and unit commitment," Applied Energy, vol. 88, no.
11, pp. 4014-4023, 2011.
[5] D. Rastler, Electricity energy storage technology options: a white paper primer on applications, costs and
benefits. Electric Power Research Institute, 2010.
[6] X. Fang, N. Kutkut, J. Shen, and I. Batarseh, "Analysis of generalized parallel-series ultracapacitor shift
circuits for energy storage systems," Renewable Energy, vol. 36, no. 10, pp. 2599-2604, 2011.
[7] M. Hannan, F. Azidin, and A. Mohamed, "Multi-sources model and control algorithm of an energy
management system for light electric vehicles," Energy Conversion and Management, vol. 62, pp. 123-130,
2012.
[8] K. Zhang et al., "Optimal decentralized valley-filling charging strategy for electric vehicles," Energy
Conversion and Management, vol. 78, pp. 537-550, 2014.
[9] M. Noori and O. Tatari, "Development of an agent-based model for regional market penetration projections
of electric vehicles in the United States," Energy, vol. 96, pp. 215-230, 2016.
[10] H. Lee and G. Lovellette, "Will electric cars transform the US market?," 2011.
[11] X. Tan, Q. Li, and H. Wang, "Advances and trends of energy storage technology in Microgrid,"
International Journal of Electrical Power & Energy Systems, vol. 44, no. 1, pp. 179-191, 2013.
[12] J. Druitt and W.-G. Früh, "Simulation of demand management and grid balancing with electric vehicles,"
Journal of Power Sources, vol. 216, pp. 104-116, 2012.
[13] E. Sortomme and M. A. El-Sharkawi, "Optimal scheduling of vehicle-to-grid energy and ancillary
services," IEEE Transactions on Smart Grid, vol. 3, no. 1, pp. 351-359, 2012.
 ACCEPTED MANUSCRIPT

[14] M. Honarmand, A. Zakariazadeh, and 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.
[15] E. Sortomme and M. A. El-Sharkawi, "Optimal charging strategies for unidirectional vehicle-to-grid,"
IEEE Transactions on Smart Grid, vol. 2, no. 1, pp. 131-138, 2011.
[16] M. Honarmand, A. Zakariazadeh, and S. Jadid, "Integrated scheduling of renewable generation and electric
vehicles parking lot in a smart microgrid," Energy Conversion and Management, vol. 86, pp. 745-755,
2014.
[17] M. Shafie-khah, M. P. Moghaddam, M. K. Sheikh-El-Eslami, and M. Rahmani-Andebili, "Modeling of
interactions between market regulations and behavior of plug-in electric vehicle aggregators in a virtual
power market environment," Energy, vol. 40, no. 1, pp. 139-150, 2012.
[18] N. H. Tehrani, G. Shrestha, and P. Wang, "Vehicle-to-grid service potential with price based PEV
charging/discharging," in Power and Energy Society General Meeting (PES), 2013 IEEE, 2013, pp. 1-5:
IEEE.
[19] S. Soleymani, A. Ranjbar, and A. Shirani, "New approach for strategic bidding of Gencos in energy and
spinning reserve markets," Energy conversion and management, vol. 48, no. 7, pp. 2044-2052, 2007.
[20] H. Kamankesh, V. G. Agelidis, and A. Kavousi-Fard, "Optimal scheduling of renewable micro-grids
considering plug-in hybrid electric vehicle charging demand," Energy, vol. 100, pp. 285-297, 2016.
[21] C. Ziogou et al., "Automation infrastructure and operation control strategy in a stand-alone power system
based on renewable energy sources," Journal of Power Sources, vol. 196, no. 22, pp. 9488-9499, 2011.
[22] A. Y. Saber and G. K. Venayagamoorthy, "Intelligent unit commitment with vehicle-to-grid—A cost-
emission optimization," Journal of Power Sources, vol. 195, no. 3, pp. 898-911, 2010.
[23] S. Goleijani, T. Ghanbarzadeh, F. S. Nikoo, and M. P. Moghaddam, "Reliability constrained unit
commitment in smart grid environment," Electric power systems research, vol. 97, pp. 100-108, 2013.
[24] A. Ghasemi, S. S. Mortazavi, and E. Mashhour, "Hourly demand response and battery energy storage for
imbalance reduction of smart distribution company embedded with electric vehicles and wind farms,"
Renewable Energy, vol. 85, pp. 124-136, 2016.
[25] M. Alipour, B. Mohammadi-Ivatloo, and K. Zare, "Stochastic scheduling of renewable and CHP-based
microgrids," IEEE Transactions on Industrial Informatics, vol. 11, no. 5, pp. 1049-1058, 2015.
[26] M. Abbaspour, M. Satkin, B. Mohammadi-Ivatloo, F. H. Lotfi, and Y. Noorollahi, "Optimal operation
scheduling of wind power integrated with compressed air energy storage (CAES)," Renewable Energy, vol.
51, pp. 53-59, 2013.
[27] W. Gu et al., "Modeling, planning and optimal energy management of combined cooling, heating and
power microgrid: A review," International Journal of Electrical Power & Energy Systems, vol. 54, pp. 26-
37, 2014.
[28] M. Matos et al., "Probabilistic evaluation of reserve requirements of generating systems with renewable
power sources: The Portuguese and Spanish cases," International Journal of Electrical Power & Energy
Systems, vol. 31, no. 9, pp. 562-569, 2009.
[29] S. Tabatabaee, S. S. Mortazavi, and T. Niknam, "Stochastic scheduling of local distribution systems
considering high penetration of plug-in electric vehicles and renewable energy sources," Energy, vol. 121,
pp. 480-490, 2017.
[30] A. Ostadi and M. Kazerani, "A comparative analysis of optimal sizing of battery-only, ultracapacitor-only,
and battery–ultracapacitor hybrid energy storage systems for a city bus," IEEE Transactions on Vehicular
Technology, vol. 64, no. 10, pp. 4449-4460, 2015.
[31] M. Alipour, B. Mohammadi-Ivatloo, M. Moradi-Dalvand, and K. Zare, "Stochastic scheduling of
aggregators of plug-in electric vehicles for participation in energy and ancillary service markets," Energy,
vol. 118, pp. 1168-1179, 2017.
[32] H. Aalami, M. P. Moghaddam, and G. Yousefi, "Demand response modeling considering
interruptible/curtailable loads and capacity market programs," Applied Energy, vol. 87, no. 1, pp. 243-250,
2010.
[33] S. Huang, L. Wu, D. Infield, and T. Zhang, "Using electric vehicle fleet as responsive demand for power
system frequency support," in Vehicle Power and Propulsion Conference (VPPC), 2013 IEEE, 2013, pp. 1-
5: IEEE.
[34] M. Doostizadeh and H. Ghasemi, "A day-ahead electricity pricing model based on smart metering and
demand-side management," Energy, vol. 46, no. 1, pp. 221-230, 2012.
 ACCEPTED MANUSCRIPT

[35] S. Mohammadi, S. Soleymani, and B. Mozafari, "Scenario-based stochastic operation management of

microgrid including wind, photovoltaic, micro-turbine, fuel cell and energy storage devices," International
Journal of Electrical Power & Energy Systems, vol. 54, pp. 525-535, 2014.