This article has been accepted for publication in a future issue of this journal, but has not been

fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/TSG.2016.2550219, IEEE
Transactions on Smart Grid
1

Queuing Analysis Based PEV Load Modeling

Considering Battery Charging Behavior and
Their Impact on Distribution System Operation
Omar Hafez, Student Member, IEEE, and Kankar Bhattacharya, Senior Member, IEEE

EM Energy consumption of PEV battery per mile

Abstract This paper presents a queuing analysis based driven, kWh/mile
method for modeling the 24-hour charging load profile of a plug- E[Pchi,k] Total expected PEV charging demand at time k
in electric vehicle (PEV) charging station. The queuing model
G Conductance of feeder, p.u.
considers the arrival of PEVs as a non-homogeneous Poisson
process with different arrival rates over the day. The first PEV I Charging current, A
charging load profile assumes customer convenience as the factor IMax Charging current level, A
that influences the hourly arrival rate of vehicles at the station, ITRMax Total number of iterations used for queuing
while the second profile is developed assuming that customers model simulation
would respond to PEV charging prices and arrival rates are M1/M2/N0 Queuing model, M1 denotes PEV arrival rate
accordingly affected. One of the main contributions of the paper
(minute) / M2 denotes PEV charging time
is to model the PEV service time considering different factors
such as the state-of-charge of the vehicle battery as well as the (minute) / N0 is the number of PEVs being
effect of the battery charging behavior. The impact of PEV load charged simultaneously at a given hour
models on distribution systems is studied for a deterministic case, NCap Maximum number of PEVs that can be charged
and the impact of uncertainties is examined and compared using simultaneously at the station
the stochastic optimal power flow and the Model Predictive P(n) Probability of n
Control approaches.
Pch Active power demand from charging PEVs, p.u.
Index Terms Queuing analysis, Plug-in electric vehicle, Battery PD, QD
Min Max
Active, reactive power demand at a bus, p.u.
charging behavior, Smart grid, Distribution system, Optimal PS , PS Minimum, maximum active power limit on
power flow. substation transformer, p.u.
QSMin, QSMax Minimum, maximum reactive power limit on
I. NOMENCLATURE
substation transformer, p.u.
Sets and Indices SOC Battery state of charge
i, j Index for buses T Charging time for a PEV, minute
k Index for time Mean service time, minute
l Set of SOC intervals l {1, 2, 3,4} V Charging voltage level, V
N Total number of buses in the system VMin, VMax Minimum, maximum limit on bus voltages, p.u.
n Set of all possible options of simultaneous Yi,j Magnitude of admittance matrix element, p.u.
charging of PEVs, for a given N0 Mean of inter-arrival time, minute
n {0, 1, 2, 3, , N0} Occupation rate of PEV at charging station
s Index for stochastic scenario Angle of complex Y-bus matrix element, rad
sb Index for substation bus
y Index for PEV class Variables
PGi,k, QGi,k Active, and reactive power generation at bus i,
Parameters hour k, p.u.
CBat Total PEV battery capacity, kWh PGi,k,s, QG i,k,s Stochastic active, and reactive power generation
DD Daily driven miles by PEV, mile at bus i, hour k, scenario s, p.u.
DDMax Maximum driving distance until PEV battery is PLoss Total system losses, p.u.
fully discharged, mile E[PLoss] Expected total system losses, p.u.
EC Daily recharge energy, kWh PS , QS Active, and reactive power supplied through
i,k i,k
substation transformer at hour k, p.u.
The first author acknowledges the funding support received from Umm Al- PSi,k,s, QS i,k,s Stochastic active, and reactive power supplied
Qura University, Makkah, Saudi Arabia, to carry out this research. through substation transformer at hour k,
The authors are with the Department of Electrical and Computer
Engineering, University of Waterloo, Waterloo, Ontario, N2L 3G1, Canada
scenario s, p.u.
(email: ohafez@uwaterloo.ca; kankar@uwaterloo.ca). Vi,k Voltage magnitude at bus i, hour k, p.u.

1949-3053 (c) 2015 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/TSG.2016.2550219, IEEE
Transactions on Smart Grid
2

Vi,k,s Stochastic voltage magnitude at bus i, hour k, impacts of PEV charging on a real low voltage distribution
scenario s, p.u. network for various PEV penetrations is discussed in [10], it is
i,k Voltage angle at bus i, hour k, rad shown that PEV charging can have negative impacts in terms
i,k,s Stochastic voltage angle at bus i, hour k, of increased peak load, increase of power losses, overload of
scenario s, rad transformers and lines, voltage drop and increased voltage
asymmetry. Bidirectional dispatch coordination of PEVs in a
II. INTRODUCTION power grid is proposed in [11] using a robust optimization
The global demand for energy has been increasing rapidly, model; it is concluded that coordinating PEV
which imposes a large burden on the existing energy resources, charging/discharging with thermal generators and the ability of
and adversely impacts environmental pollution and global PEV aggregators to provide regulation reserve service are
warming. As all governments around the world move toward a beneficial to power grid operation. In [12], an optimization
green energy economy, the Plug-in Electric Vehicles (PEVs) model is developed to determine the optimal and maximum
have an increasingly important role to play, because of their penetration of PHEVs in the transport sector of Ontario
contribution to emissions reduction from the transportation considering the grid limitations.
sector. In Canada, almost 35% of the total energy demand is The travel patterns of light-duty vehicles in the U.S.
from the transport sector and it is the second largest source of obtained from the 2009 National Household Travel Survey
greenhouse gas (GHG) emissions [1]. The Government of (NHTS) is used in [13] to estimate the electric energy and
Ontario, Canada, has chalked out a path to move toward a power consumption of PEVs for two uncontrolled charging
green energy economy, aimed towards increased penetration of scenarios.
renewable energy sources and PEVs [2]. Penetration of PEVs The charging demand of PEVs is affected by different
into the market is expected to be large in the near future, and uncertain factors, such as the number of PEVs being charged
with their complex charging behavior, many technical simultaneously, their charging levels, battery capacity, and
problems related to their impact on the power grid need be charging duration. Monte Carlo Simulation (MCS) is used in
investigated. Since the charging behavior of PEVs is dependent [14] to generate virtual trip distances which consider driving
on a number of factors, their overall charging demand tends to habits, different vehicle models, etc., and hence formulates an
be uncertain. annual energy consumption model of light duty fleet of PEVs.
Several studies show that the power distribution grid can be In [15], a max-weight PEV dispatch algorithm, based on a
significantly impacted by high penetration levels of PEVs [3- queuing formulation integrated with renewable energy sources
5]. Electric vehicle (EV) charging will likely coincide with the is used to control the PEV charging in order to avoid costly
system peak demand and thus, in order to avoid overloading of distribution system infrastructure upgrades. A probabilistic
the distribution feeders, adequate load management schemes constrained load flow problem with wind generation and EV
need be in place [3]. In [4], a comprehensive study to assess demand is presented in [16] where the charging and
the implications arising from adoption of PEVs in Ontario, discharging processes are considered using M/M/ queuing
from the technical, consumer, policy, regulatory and market model. The spatial and temporal distribution of demand, based
points of view are presented and specific measures and on fluid dynamic traffic model and queuing theory is
approaches, and policy initiatives relevant to the Ontario developed in [17] to determine the EV charging demand for a
context is discussed. Different charging scenario studies rapid charging station. In [18], an EV demand model suitable
reported in [5] conclude that PEVs significantly increase for load flow studies is proposed wherein the EV demand is
demand side uncertainties and can potentially reduce the represented as a PQ buses with stochastic characteristics based
distribution transformer insulation life. In [6], the impact of EV on the concept of queuing theory. In [19], four different types
charging on a medium-voltage network and the Local of PHEVs are considered and factors that affect their charging
Distribution Companys (LDCs) benefits arising from the use behavior, e.g., differences in battery capacity and charging
of smart charging schemes is discussed. In [7] PEV charging level, is discussed. A single PHEV charging demand model is
control strategies are developed to mitigate distribution formulated and queuing theory is used to describe the behavior
transformer ageing that could result from load peaks caused by of multiple PHEVs.
PEV charging. It is noted that grid stability, need for In recent research, the Model Predictive Control (MPC)
infrastructure investment, and other challenges can be approach has been applied to various operational and control
mitigated by the LDC by adopting smart charging strategies. problems in the context of smart grids to consider the effect of
The impact of uncoordinated Plug-in Hybrid Electric uncertainties. In [20] a prediction based real-time charging
Vehicle (PHEV) charging on system peak load, losses, voltage method is proposed that considers the effect of future vehicles
and system load factor are discussed in [8] and is noted to have penetrating into the grid. In [21] an MPC based approach is
adversely affected the efficiency of the distribution grid. It is proposed for a minimum cost dispatch based operation while
shown in [9] that a 10% penetration of PEV may cause including the effect of PEV charging loads.
Most of the works on PEV charging demand modeling that
unacceptable variations in voltage profiles if there is no
use queuing analysis [16-19], consider the arrival rates as a
regulation on PEV charging, while coordinated charging can
Poisson process, which is a stochastic process that assumes the
reduce system peak load, losses, and mitigate the impacts of PEVs have a constant arrival rate. Only a few have modeled
uncoordinated PEV charging in the distribution system. The

1949-3053 (c) 2015 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/TSG.2016.2550219, IEEE
Transactions on Smart Grid
the arrival rate as a non-homogeneous Poisson process [22].
Moreover, the charging time is typically modeled by
exponential distributions with given upper and lower limits,
which are randomly assigned to each PEV, and the waiting
time is assumed to be infinite. Furthermore, it should be noted
that during the fast charging process, the charging power
typically starts at a high rate, and drops off as the battery State
of Charge (SOC) approaches its full capacity, as per the battery
charging behavior (BCB) of PEVs, discussed in [23]. This
affects the charging time and needs to be considered in the
PEV load model, which has not been considered so far.
Furthermore, in a fast charging station an infinite waiting time
is not a reasonable assumption. In view of the above, the main
objectives of this paper are:
Estimate the distribution of vehicles which are on the road
during the day using mobility statistics from the Waterloo
Region Transportation Tomorrow Survey (TTS) [24] and
construct the PEV arrival rate profile at a charging station
based on customer convenience. Also, develop PEV arrival
rate profile based on customers response to PEV charging
price.
Using the developed non-homogeneous Poisson process type
arrival rate profiles, the service time of PEVs modeled using
their BCBs and waiting times, develop appropriate queuing
models to determine the 24-hour charging demand profile at
a PEV charging station.
Integrate the developed PEV charging load model within a
distribution system operation framework to study the impact
of PEV charging on distribution system operation
considering optimal power flow (OPF), stochastic analysis,
and the Model Predictive Control (MPC) approaches.
The novel features and contributions of this paper are as
follows:
The arrival of customers to charge their PEVs at the charging
station has been modeled as continuous and random, non-
homogeneous Poisson processes, considering two different
arrival patterns- customers convenience, and customers
response to PEV charging price. The first arrival rate profile
makes use of the large database of mobility statistics
available from the Waterloo Region TTS [24]. The second
arrival rate profile considers customers response to PEV
charging price, where the price data is obtained from the
winter Time-of-Use (TOU) prices applicable in Ontario,
Canada [25].
The queuing model proposed in this paper considers for the
first time, a detailed representation of the BCB of PEVs and
the SOC, and the time taken to charge PEVs, including a
finite waiting time. Such detailed models provide accurate
information of the charging load on the distribution system,
taking into account different classes of PEVs.
These features render the proposed queuing model more
realistic and accurate and also more generic and universal,
than the existing models. For example, homogenous Poisson
process models are a special case of non-homogeneous
Poisson process models, and similarly, the modeling of in
this work considers the SOC, the BCB, and the time of
charging including a finite waiting time- which is the most
generic form of model for M2.
1949-3053 (c) 2015 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
3
A comprehensive set of studies considering load flow and
OPF analysis is carried out to examine the impact on LDC
operation, and to determine the optimal strategies for the
LDC to improve its operational performance. Thereafter,
considering the PEV charging load model may have
uncertainties, therefore to study the impact of such load
model on LDC operation in the presence of uncertainties,
stochastic load flow, stochastic OPF analysis, and MPC
based analysis are carried out.
The rest of this paper is organized as follows. The basics of
the queuing model, the PEV BCB model, the PEV charging
model and the distribution system operations model are
presented in Section III. In Section IV, the system description
pertaining to the case study carried out is presented. The results
and discussions are presented in Section V and the concluding
remarks are presented in Section VI.
III. MATHEMATICAL MODELING
A. PEV Queuing Model
Queuing analysis is applied to estimate the total charging
power of PEVs. The PEV customers are considered to be
served using M1/M2/N0 queue model at a PEV charging station,
where M1 denotes the arrival rate which varies from hour to
hour of the day and is modeled as a non-homogeneous Poisson
process, the service time denoted by M2 includes the waiting
time and the charging time. The service time is modeled in this
paper considering the PEV BCB. Poisson process is a
continuous-time stochastic process that counts the number of
arrivals in a given time interval where the time between each
pair of consecutive arrivals has an exponential distribution with
(mean of inter-arrival time) and each of these inter-arrival
times are assumed to be independent of other inter-arrival
times. It is useful for modeling arrival that occur independently
from each other [26]. Since the arrival of PEVs at the charging
station is a continuous-time stochastic process, Poisson process
has been considered in this work. In accordance to M1/M2/N0
queuing analysis [27], the probability of the number of PEVs
charging simultaneously at an hour is modeled as a discrete
distribution, as follows:
pk (n ) =
(N 0 )n pk (0 ) n = 1,2,3,...., N 0 (1)
n!
where,
1
N 0 1
(N 0 ) m + (N 0 ) N 0 1
pk (0) =
m! N 0! (1 )
(2)
m=0
and is the occupation rate of the PEV charging station and is
calculated as follows:
1
k
= (3)
1
N0
k
The expected number of customers waiting at the charging
station, at hour k is given by Littles Law in queuing theory as
follows [27]:
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/TSG.2016.2550219, IEEE
Transactions on Smart Grid
4

N0 where IMax, and V are dependent on the charging level and

1 k k k
Ek [N w ] = pk (0) (4) hence fixed. In (9), EC y is the daily recharge energy of a PEV,
( N 0 1) ! k ( N 0 k k ) 2
and is obtained as follows:
And consequently, the average waiting time of an arrival at
hour k is given by: C Bat y
if DD y DDMax
EC y = (10)
E [N ]
E My DD y if DD y < DDMax
Wk = k w (5)
k where E M y is the energy consumption by a PEV of class y, per
B. PEV Battery Charging Behavior (BCB) Model mile. The maximum driving distance of a PEV, DDMax y is
The charging service time is affected by different factors, calculated as follows:
such as the charging level, battery capacity, battery SOC, and C Bat y
PEV BCB. One of the main objectives of PEV customers is to DDMax y = (11)
have fast charging at a charging station i.e., minimum service EM y
time. In order to achieve this, the BCB of each class of PEVs Therefore, the total charging power for N0 number of PEVs
are considered; for example, the battery of a typical Compact being charged simultaneously at time k is given as follows:
PEV during fast charging attains an SOC of 50% in 10 N0
minutes, 75% in 15 minutes, beyond which there is a drop in Pchi, k = I m, k , yV (12)
charging rate, as shown in Fig.1 [23]. Considering a maximum m =1
required SOC for the PEVs to be 85%, it can be noted from The total expected PEV charging demand at time k for all
Fig.1 that this is attained in 22 minutes, for Compact PEVs, possible values N0 (where N0 1 to Ncap) is given as follows:
and therefore, the service time depends on the BCB of the PEV N cap
class. E[ Pchi , k ] = P (N
N0
k 0 )Pchi, k (13)
100 %
100
90 90 %
95 %
The total PEV charging demand obtained from (13) is used
80 75 % in the OPF model discussed next, to examine the impact of
70 PEV charging on the distribution system performance.
60
SOC %

50
50 %
C. OPF Model for System Operation Including PEV Load
40 Once the PEV charging load is estimated using the queuing
30
20
model, the impact on system operation is examined by
10 10 Min 5 Min 7 Min 12 Min 25 Min formulating the following OPF model, with the objective of
0 minimization of feeder losses, as given below:
0 10 15 25 35 60
1
24 N N

( ( ))
Minutes 2 2
PLoss = G i , j V i , k + V j , k 2Vi , k V j , k cos j , k i , k (14)
Fig.1: Typical BCB of a Compact PEV during fast charging [23] 2 k =1 i =1 j =1

The SOC of the battery can be obtained as follows: The PEV charging load (Pchi,k) is included in the active
EC y power balance at bus i where the charging station is located.
SOC y = 1 y (6) The demand-supply balance for both active and reactive power
C Bat y
is given by the load flow equations as follows.
and the following conditions are imposed, from the above BCB N
of the PEV in Fig.1:
PGi ,k PDi ,k Pchi ,k = Vi ,k V j ,k Yi , j cos i , j + j ,k i ,k ( (15) )
0.2 if SOC y 0.2 j =1
N
SOC y = SOC y if 0.2 < SOC y 0.85

(7)
(
QGi , k QDi , k = Vi , kV j , k Yi , j sin i , j + j , k i , k )
(16)
0 . 85 if SOC y > 0 . 85 j =1

Once the SOC of a PEV is known, the required charging (or The voltage magnitudes at each bus at hour k are constrained
service) time of a PEV, given by T, is obtained from the BCB by their respective upper and lower limits.
(Fig.1) using the piece-wise linear relationship: Vi Min Vi , k Vi Max i Load buses (17)
SOCl y bl
T= l {1, 2, 3, 4} , y (8) The slack bus voltage magnitude and voltage angle, which is
al the substation bus, are fixed, as follows.
where al and bl are the slope and intercept of the linear equation Vsb, k = 1 p.u., sb, k = 0
obtained from the BCB and depend on the SOC of interval l. The substation capacity limit determines the maximum and
The charging current drawn by a PEV over the charging minimum active and reactive power transfer capacity over the
period T, at time k for PEV class y is given as follows: substation transformer.
EC y
I k , y = min( , I Max ) (9) PSiMin PSi , k PSiMax i = sb (18)
VTk

1949-3053 (c) 2015 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/TSG.2016.2550219, IEEE
Transactions on Smart Grid
5

QSiMin QSi , k QSiMax i = sb (19) E. Model Predictive Control (MPC)

The above non-linear programming (NLP) model is solved
using the MINOS5.1 solver in the GAMS environment.
At each time step, the total charging load is estimated using
the proposed queuing model, and then applied to the OPF. The
flowchart describing the queuing process is shown in Fig.2. A
large number of queuing simulation runs are needed, for every
simulation run ITR1 ( 1,, ITRMax), the hourly arrival rate is
input, and N0 is randomly selected in the range [1, NCap].
D. Stochastic OPF Including PEV Load
In the stochastic OPF the output of the queuing model is
used to examine the impact of PEV charging loads on
distribution system operation, with the objective of
minimization of expected feeder losses, considering the
scenarios of different queuing models, as given below:
N N
1 17 24
Pk (s ) i =1 j =1
(
Gi , j Vi 2,k ,s + V j2,k ,s
E[ PLoss ] =
2 S =1 k =1
2Vi ,k ,sV j ,k ,s cos j ,k ,s
Where s represents the scenario, which in this paper,
represents N0 i.e., the number of PEVs charging
simultaneously. The PEV charging load for each scenario
(Pchi,k,s) is included in the active power balance at bus i where
the charging station is located. The demand-supply balance for
both active and reactive power is given by the load flow
equations as follows.
The MPC approach is applied in this paper to examine the
impact of PEV charging loads on LDCs operation in the
presence of uncertainties. The MPC approach determines a
series of optimal operations over a finite horizon, wherein the
decision for the next time step is obtained by solving the
problem using the current state of the system as the initial state
[20], [21], [29]. The MPC approach works on an iterative,
finite-horizon optimization time-frame [k:k+t] where at the
current time step k the OPF is solved using (14) - (19). This
provides the optimal set of decisions at the current point of
time, for the next 24 hours, based on the estimated PEV
charging demand. Only the first sample from the set of optimal
decisions is implemented. In the next iteration, the new state of
the system is considered, the optimization horizon is shifted
forward and the OPF is solved again; MPC repeats until the
last time step [k+t] is reached.
) (20)
Set ITR1 = 1, ., ITRMax
(
i ,k ,s )
Input arrival rate depending on the time of
the day and customers objective
Select N0 = uniform [1,Ncap]
Set ITR2 = 1, 2, .., N0
Is Arrival rate > NCap Yes
Add wait time to No
No
charging time

N No Execute queuing model (1)-(13)

PGi,k ,s PDi,k Pchi,k ,s = (

Vi,k ,sV j ,k ,s Yi, j cos i, j + j ,k ,s i,k ,s ) (21)
j =1 No Is ITR2 = N0 ?

N Yes
QGi,k ,s QDi,k = V (
i,k ,sV j ,k ,s Yi, j sin i, j + j ,k ,s i,k ,s ) (22) Calculate the expected power from (13)
j =1

All values of N0 considered ?

The power flow equations (21)-(22) shows that fluctuations
in PEV charging load in each scenario (Pchi,k,s) leads to Yes
ITR1 = ITR1 + 1
variations in the system bus voltages (Vi,k,s) and angles (i,k,s) Determine the PEV load of specific hour

across the network. The following constraint is enforced on the

expected voltage at every bus i and time interval k: Simulation completed for 24-hours ?
17
ViMin Pk ( s) Vi,k ,s ViMax i N (23) Yes
s =1
Is ITR1 = ITRMax ? No
The voltage magnitude and angle at the slack bus, which is
the substation bus, are fixed for all scenarios, as follows: Yes
Vsb, k , s = 1 p.u., sb, k , s = 0 Execute OPF
The substation capacity limit determines the maximum and
minimum active and reactive power transfer capacity over the Display and compare results

substation transformer for all scenarios.

Fig.2: The flowchart of the queuing simulation process
PS iMin PS i,k , s PS iMax i = sb (24)

QSiMin QSi,k , s QSiMax i = sb (25) IV. CASE STUDY

The above non-linear programming (NLP) model is solved A. Distribution System and Mobility Data
using the SNOPT solver in the GAMS environment. There are
The analysis reported in this paper is carried out considering
several extensive research works on stochastic OPF with
the IEEE 69-bus radial distribution system, whose single line-
comprehensive detail on the solution of such problems [28].
diagram is given in Fig. 3 [30]. The distribution system is
supplied through the substation at bus-1. It is assumed that a

1949-3053 (c) 2015 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/TSG.2016.2550219, IEEE
Transactions on Smart Grid
PEV charging station is arbitrarily located at bus-59. In this
paper, Level-3 charging is considered since high power level
charging is preferred at PEV charging stations, and thus IMax =
63 amps and V = 400 volts.
Distribution systems are generally balanced by using various
load balancing schemes, and hence can be represented by
single phase equivalents [31]. The unbalanced nature of
distribution system is more prevalent at the end-user level
(residential customer level) but since this work considers a
PEV charging station load model, it is assumed to be connected
at 12.66 kV feeder level; and at this voltage level, the loads are
assumed to be balanced three-phase, and all line segments are
three-phase, and perfectly transposed [31].
With these assumptions, a single line-to-neutral equivalent
circuit for the feeder has been used, and a three phase
distribution system is represented by a single-phase equivalent.
Fig.3: 69-Bus radial distribution system [30]
Waterloo Region TTS data and Ontario, Canada, TOU
winter tariff rates are used in order to obtain realistic results.
The Waterloo Region TTS is a comprehensive travel survey
conducted every five years in the region wherein 5% sample of
households are reached by telephone. In this work the 2011
TTS for Waterloo Region is used which considers of 43,165
unique trips. Fig. 4 presents the winter TOU rates of Ontario
over a day. The distribution of vehicles on the road over a 24
hour period is calculated using the same TTS data and shown
in Fig. 4.
Fig.4: Distribution of vehicles on the road and Ontario winter TOU
1949-3053 (c) 2015 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
6
B. Modeling of Daily Driven Miles and PEV Arrival Rate, M1
Fig. 5 shows the distribution of daily miles driven on all
vehicle driving days based on the TTS data. It is noted that the
best fit for the TTS data (using EasyFit software [32]) is
obtained with a lognormal distribution as evident from Fig.6.
So, the daily driven miles by the PEVs, DD, is modeled as a
lognormal distribution in this work, and is given by:
DD y = e ( M + M f ) (26)
where M and M are the mean and the variance of the
lognormal distribution, respectively.
14
Percent of Vehicles %
12
10
8
6
4
2
0
Daily driven miles
Fig.5: Distribution of daily driven distance per vehicle as per TTS
Fig.6: Distribution of daily driven distance per vehicle using EasyFit [32]
In this paper, four classes of PEVs are considered- Compact,
Economy, Mid-Size, and Light Truck/SUV, to present a realistic
picture of the PEV charging station load. The queuing
algorithm is initiated by randomly generating N0. The PEV
arrival rate M1 depends on the hour of the day and customers
behavior pattern. Under a rational behavior assumption, two
M1 profiles are modeled as follows:
Scenario-1: considers that the PEV arrival rate depends on
customer convenience, i.e., the number of vehicles on the
road. When the number of vehicles on the road is high the
arrival rate is high, irrespective of the price or LDCs
operational constraints. In this paper, using TTS data, a
relationship between vehicles on the road and PEV arrival
rate at the charging station is assumed. As shown in Fig.7, if
the percent of vehicles on the road is up to 4%, at any hour, a
uniformly distributed PEV arrival in the range of 1 to 4
PEVs/hour is assumed, and similarly 5 to 11 PEVs are
assumed to arrive if 4-7% of vehicles are on the road, and so
on. For example, at hour 7, about 7% of the vehicles are on
the road (see Fig.4) and consequently, 5 to 11 PEVs may
arrive for charging (as per Fig.7), on the other hand, at hour
17, 9.3% of the vehicles are on the road, and it is assumed
that 12 to 17 PEVs may arrive for charging.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/TSG.2016.2550219, IEEE
Transactions on Smart Grid
7

C. Simulation of the Queuing Process

Once the arrival rate scenario is selected, for every value of
N0, the following steps are repeated:
Assume the PEV classes to be uniformly distributed over
the sample set N0, randomly select a PEV class from
amongst the four different classes.
The battery capacity for each PEV class is uniformly
Fig.7: Relationship between vehicles on road and arrival rate of PEVs distributed between their upper and lower limits.
Calculate battery SOC for each PEV from its daily recharge
Scenario-2: considers that the PEV arrival rate depends on
energy (6), (7), which depends on different factors such as
the charging price, i.e., more PEVs will charge when the
the daily driven miles, and battery capacity.
price is low, and vice versa. In this scenario a relationship
Determine the time required for charging for each PEV,
between Ontarios winter TOU price and PEV arrival rate at using the BCB given in (8).
the charging station is assumed. As shown in Fig.8, if the Determine the total charging power arising at the charging
charging price ranges between 7.5 and 11.2 cents, at any station, for the total N0 PEVs using (9) - (13).
hour, a uniformly distributed PEV arrival in the range of 5 to Following are the parameters used in this paper to simulate
11 PEVs/hour is assumed, and similarly 12 to 17 PEVs are the queuing process of PEVs charging at a charging station:
assumed to arrive if the charging price is in the range of 11.2 TABLE I
to 13.5 cents. For example, at hour 7, the charging price is PARAMETERS FOR SIMULATION OF QUEUING MODEL
13.5 cents (see Fig.4) and consequently, only 1 to 4 PEVs ITRMax 2000
may arrive for charging (as per Fig.8). NCap 17
k Non-homogeneous Poisson Process (Fig.7 & 9)
Non-homogeneous Poisson Process, includes BCB
k
and SOC (Fig.8 & 9)
Light Truck
Class Compact Economy Mid-size
/ SUV
CBat, kWh 8 - 12 10 - 14 14 - 18 19 - 23
PEV EM, kWh/mile 0.2- 0.3 0.25-0.35 0.35-0.45 0.48-0.58
Lognormal Distribution (Fig.5 & 6)
DD, miles
Fig.8: Relationship between TOU tariff and arrival rate of PEVs M = 40 miles, M = 20 miles

It is important to point out that, the arrival rates modeled in
this paper are based on some assumptions pertaining to
vehicles on the road, charging price and how PEVs arrive for
charging at the charging station. Such assumptions are
necessary in order to understand the impact of PEV charging
on the distribution grid but, need be validated with realistic
data from ground level surveys.
Moreover, as per Scenario-2, the arrival rate would be high
at night since the PEV charging price is low at these hours.
However, considering that the probability of charging during
night is low, because of customer inconvenience, the arrival
rate is modified appropriately, as shown in Fig. 9, where the
removed arrival data of early hours are indicated. Also to be
noted that since home charging has been ignored in this work,
there will be no effect on the early morning arrival rates. The
two arrival rate profiles, as discussed above, are modeled as
non-homogeneous Poisson processes with mean k which is the
time dependent number of expected car arrivals at a charging
station throughout the day.
Fig.9: Number of PEVs arriving per hour at the charging station (Arrival rate)
V. RESULTS AND DISCUSSIONS
A. PEV Charging Load Using Queuing Analysis
In this section the effect of PEV charging on the distribution
system performance is examined. Queuing analysis is used to
model the 24-hour PEV charging demand at the charging
station. The objective is to determine the optimal distribution
system operation considering PEV charging demand while
minimizing the system losses using the OPF model. The
probability distribution of N0 is obtained from (1) and shown in
Fig.10 which is used as an input to the M1/M2/N0 queuing
model [27]. Different NCap values are examined within the
proposed queuing model and it is noted that when NCap is high,
the probability of simultaneously charging of NCap number of
PEVs, i.e., P(NCap) is very low, as seen from Fig.10. Because
of this low probability, there is insignificant impact on the total
expected charging demand for high values of NCap. By various
trial runs it is noted that beyond NCap =17, the effect on
expected charging demand does not change significantly and
hence, NCap =17 was chosen for the paper.
It is noted from Fig.11 that the expected PEV charging
demand is low for low arrival rates (M1 = 40, i.e., a PEV
arriving every 40 minutes) and the discrete distribution pattern
of the demand as a function of N0 is skewed normal with a low
mean value of N0. As the arrival rate increases, i.e., M1 = 25 (a
PEV arriving every 25 minutes), and then for M1 = 10, the
expected PEV charging demand increases and the distribution

1949-3053 (c) 2015 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/TSG.2016.2550219, IEEE
Transactions on Smart Grid
8

pattern becomes a standard normal distribution with a high

mean value of N0.
Using the chronological arrival rate profiles for each
scenario (Fig.9) and the probability distribution associated with
a specific arrival rate, the expected PEV charging demand at a
specific hour can be obtained, as shown in Figs. 12-14. It is
seen that when the arrival rate is high, the expected load is
high, and the discrete distribution pattern of the PEV charging
demand as a function of N0 is normally distributed.
Fig.13: Expected PEV charging demand at hour-22 for different N0

Fig.10: Probability distribution of N0 as input to M1/M2/N0 queuing model

Fig.14: Expected PEV charging demand at hour-10 for different N0
For example, at hour-8 the PEV arrival rate is high for
Scenario-1 (Fig.12) while at hour-22 it is high for Scenario-2
(Fig.13), and the discrete distribution patterns are accordingly
normally distributed at these hours, for the respective
scenarios. When PEV charging takes place at hour 10 (Fig.14),
which is not the most convenient hour and neither the cheapest
hour for customers to charge their vehicles, both scenarios
have almost similar distributions.

Fig.15: Total PEV expected charging demand for all queuing model

B. Impact of BCB on Service Time and Charging Load

Since minimizing the charging time and wait time for PEVs
at the charging station is one of the main objectives, the PEV
charging time (i.e., the service time) is modeled considering
the BCB as discussed earlier. In classical queuing analysis
based approaches [16-19], charging time is typically modeled
using exponential distributions with upper and lower limits
Fig.11: Expected PEV charging demand for some typical arrival rates
randomly assigned to each PEV. It is seen that from Fig.16 that
when BCB is considered, all the PEVs are served within the
The overall expected PEV charging demand obtained from hour they arrive at the station, there is no overflow across the
(13) for both Scenarios, presented in Fig.15, shows that the hour, and there is no waiting time.
PEV charging demand increases in both scenarios as compared
to the Base Case. In Scenario-1 the charging demand appears
during the peak price hours since these hours are more
convenient for customers, while in Scenario-2 the increase is
significant during off-peak price hours. It is also noted that as
N0 increases, the expected charging demand will gradually
merge with the Base Case load profile as the probability of a
large N0 is low.

Fig.16: PEVs charging, overflow, waiting: considering BCB

On the other hand, when BCB is not considered, and the

charging time is selected randomly, there will be service
overflows if the selected charging time is more than 60
minutes. These service overflows will be transferred to the next
Fig.12: Expected PEV charging demand at hour-08 for different N0 hour, which may lead to waiting times if the total number of

1949-3053 (c) 2015 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/TSG.2016.2550219, IEEE
Transactions on Smart Grid
9

PEVs to be served, exceeds NCap (Fig.17). The effect of

considering the BCB on the PEV demand profile is quite
noticeable (Fig.18), when the BCB model is not considered the
charging demand is higher at certain hours because of the
service overflow and shifting of demand taking place, as
discussed earlier.
In order to introduce a service overflow and waiting time for
PEVs when considering BCB of the PEVs, it is now assumed
that the arrival rate is greater than the station capacity, i.e., >
NCap. As can be seen from Figs.19 and 20, the service overflow Fig.21: Effect of BCB on average waiting time, > N
Cap
and PEVs waiting are significantly increased when BCB is not
considered. It is also noted that the average waiting time is
significantly reduced when BCB is considered as compared to
the case without BCB (Fig. 21) which is in line with the
preferences of customers for fast charging. Finally, Fig.22
shows the effect of BCB on total charging demand when the
arrival rate exceeds the station capacity. Comparing this profile
with Fig.18, it is noted that the charging demand is
significantly affected by the BCB, and also when > NCap.

Fig.22: Effect of BCB on total charging demand, > NCap

C. Impact of PEV Charging on Distribution System Operation

Fig.17: PEVs charging, overflow, waiting: not considering BCB
Fig.18: Effect of BCB on total charging demand
Fig.19: PEVs charging, overflow, waiting: considering BCB, > NCap
The Base Case is the case when no PEVs are present in the
system. Analyses are then carried out to examine the impact of
PEV charging loads considering:
a. Uncontrolled operation of distribution system- in this case
power flow analysis is carried out to examine the impact of
PEV charging loads appearing on the distribution feeder
while the LDC takes no operational or control actions to
manage the system voltages.
b. Optimal operation of distribution system- this case
considers the OPF model and examines the impact of
LDCs optimal operation actions and how the system
voltages are controlled to ensure system security.
It is noted from Fig.23 that the expected system losses for
the two scenarios of PEV charging with optimal operation of
the distribution system are significantly higher than the Base
Case, with no PEVs. It is noted that the expected loss is
maximum for N0 = 3, i.e., when three PEVs are charging
simultaneously, because M1/M2/3 results in the highest
expected system load. It is also noted from Fig.23 that for N0
9, i.e., for nine or more PEVs charging simultaneously, the
expected system loss in both scenarios are almost the same as
that in the Base Case.

Fig.20: PEVs charging, overflow, waiting: not considering BCB, > NCap Fig.23: Impact of PEV charging on losses during optimal operation of system

1949-3053 (c) 2015 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/TSG.2016.2550219, IEEE
Transactions on Smart Grid
10

The expected bus voltage profiles are also affected by PEV

charging. For example, at Bus-65, which is a remote bus and
located close to the PEV charging station, significant drop in
the expected voltage profile takes place at various hours,
depending on the scenario considered (Fig.24 - Fig.27). Figs.
24 and 25 presents the expected voltage profiles at bus-65
considering uncontrolled operation and optimal operation of
the distribution system, respectively; the voltage drop is
significant in uncontrolled operation (Fig.24), while much
improved with optimal operation (Fig.25) because of the
imposition of voltage constraints by the LDC. Fig.25 Fig.26: Expected voltage profile at Bus 65, stochastic uncontrolled operation
demonstrates that the LDC can indeed accommodate PEV
charging loads by appropriate control and operational decisions
to maintain the system voltages within limits.
In order to capture the uncertainties, stochastic power flow
studies are presented with p.d.f. of the PEV charging load and
compared with stochastic OPF, to examine the LDCs
operational impacts. As expected, the stochastic OPF (Fig.27)
ensures that the bus voltage at bus-65 is within the specified
limit at all hours, while the stochastic power flow (Fig.26)
gives an idea of the expected worsening of the bus voltage due
to PEV charging.
It is noted from Figs. 24 27 that the voltage drops coincide
with the appearance of PEV charging loads; for example, the Fig.27: Expected voltage profile at Bus 65, stochastic optimal operation
voltage drop is significant during peak hours in Scenario-1,
because customers charge their PEVs as per their convenience; The profiles of expected active power transfer from the
while in Scenario-2 customers opt to charge during low price external grid over the substation transformer are also affected
hours and the voltage drop is most significant during the off- by PEV charging (Figs. 28 - 31). When the optimal operation
peak hours. of the distribution system, is considered (Figs. 29 and 31), the
expected active power transfer is reduced as compared to the
uncontrolled operation (Figs. 28 and 30).

Fig.24: Expected voltage profile at Bus 65 for uncontrolled operation

Fig.28: Expected active power transfer, uncontrolled operation

Fig.25: Expected voltage profile at Bus 65 for optimal operation Fig.29: Expected active power transfer, optimal operation

1949-3053 (c) 2015 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/TSG.2016.2550219, IEEE
Transactions on Smart Grid
11

Fig.30: Expected active power transfer, stochastic uncontrolled operation

Fig.33: Comparison of optimal operation versus MPC for Scenario-2

Fig.31: Expected active power transfer, stochastic optimal operation

As discussed earlier, the MPC approach is considered in this

paper to examine the impact of the PEV charging load on the
LDCs operations in the presence of uncertainties, and these
are compared with the optimal operation of the distribution
system. The expected voltage profiles at bus-65 for these cases Fig.34: Comparison of optimal operation versus MPC for Scenario-1
and the two scenarios are presented in Figs.32 and 33,
respectively. The profiles of expected active power transfer
from the external grid are presented in Figs.34 and 35,
respectively. It is noted that when MPC is considered, the
expected voltage profile is improved and the active power
transfer is reduced as compared to the corresponding optimal
operations profiles (Figs. 32-35).
Figs. 36 and 37 present the hourly expected system losses
considering the MPC and the OPF models for both scenarios. It
is noted that the system losses are reduced by 17.95% and
16.08% in Scenario-1 and 2, respectively with the MPC based
approach as compared to the OPF. Therefore, the overall
operational performance of the distribution system considering
the uncertainty of PEV charging demand, is improved when
Fig.35: Comparison of optimal operation versus MPC for Scenario-2
the MPC based approach is considered for LDC operations.

Fig.32: Comparison of optimal operation versus MPC for Scenario-1 Fig.36: Comparison of optimal operation versus MPC for Scenario-1

1949-3053 (c) 2015 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/TSG.2016.2550219, IEEE
Transactions on Smart Grid
Fig.37: Comparison of optimal operation versus MPC for Scenario-2
VI. CONCLUSIONS
In this paper, a queuing analysis based methodology for
modeling the 24-hour charging demand of PEVs at a charging
station was presented. Four different PEV classes with their
appropriate parameters that determined the charging behavior
were taken into consideration. The proposed queuing model
considered the arrival of PEVs as a non-homogeneous Poisson
process with different arrival rates at different times of the day
and the PEV BCB was also considered. Different arrival rate
patterns were considered for different groups of customers- one
based on customer convenience and the other based on PEV
charging price which were estimated from Waterloo Region
TTS data and Ontario TOU rates respectively. A novel feature
of the proposed PEV charging load model was that a piece-
wise linear function of the SOC was used to represent the BCB
of PEVs, and hence determine the charging time which then
integrated with the M1/M2/N0 queuing model. PEV SOC and
BCB were found to be a sufficient method in order to
determine appropriate PEV charging time. The developed load
model of PEVs was then incorporated in deterministic and
stochastic analysis frameworks of the distribution system,
through power flow and OPF models as well as MPC based
method, to study their impact on the distribution system, and
examine how the LDC can accommodate PEV charging loads
while maintaining the system constraints.
Charging during overnight hours is not a convenient time
although the price is low during these hours. In the future,
combinations of both scenarios need be studied and multi-
objective analysis may be considered. Furthermore, a rational
driver will always select the PEV charging station considering
traffic factors, such as the location and distance so optimal
siting, sizing of charging station need also be studied.
REFERENCES
[1] Transport Canada. [Online]. Available:
http://www.tc.gc.ca/environment/menu.htm#climatechange
[2] Ontario Power Authority, Ontarios Integrated Power System Plan:
Scope and Overview, June 2006 [online]. Available:
http://www.powerauthority.on.ca/Storage/24/1922_OPA__IPSP_Scope
_and_Overview.pdf
[3] G. T. Heydt, The impact of electric vehicle deployment on load
management strategies, IEEE Transactions on Power Apparatus and
Systems, vol. 1, no. 144, pp. 12531259, May 1983.
[4] C. Canizares, J. Nathwani, K. Bhattacharya, M. Fowler, M. Kazerani, R.
Fraser, I. Rowlands, H. Gabbar, Towards an Ontario action plan for
1949-3053 (c) 2015 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
12
pluginelectric vehicles (PEVs), Waterloo Institute for Sustainable
Energy, University of Waterloo, May 17, 2010.
[5] Q. Gong, S. Mohler, V. Marano, and G. Rizzoni, Study of PEV
charging on residential distribution transformer life, IEEE
Transactions on Smart Grid, vol. 3, no. 1, Mar 2012.
[6] J. A. P. Lopes, F. J. Soares and P. M. R. Almeida, Integration of
electric vehicles in the electric power system, Proceedings of the IEEE,
vol. 99, No. 1, pp. 168-183, 2011.
[7] Q. Gong, S. Midlam-Mohler, E. Serra, V. Marano, and G. Rizzoni,
PEV charging control considering transformer life and experimental
validation of a 25 kVA distribution transformer, IEEE Transactions on
Smart Grid, vol. 6, no. 2, Mar 2015.
[8] K. Clement-Nyns, E. Haesen, and J. Driesen, The impact of charging
plug-in hybrid electric vehicles on a residential distribution grid, IEEE
Transactions on Power Systems, vol. 25, no.1, pp. 371380, Feb 2010.
[9] E. Sortomme, M. M. Hindi, S. D. MacPherson, and S. S. Venkata,
Coordinated charging of plug-in hybrid electric vehicles to minimize
distribution system losses, IEEE Transactions on Smart Grid, vol. 2,
no. 1, Mar 2011.
[10] A. Bosovic, M. Music, S. Sadovic, Analysis of the impacts of plug-in
electric vehicle charging on the part of a real low voltage distribution
network, PowerTech, Eindhoven, 2015.
[11] X. Bai and W. Qiao, Robust optimization for bidirectional dispatch
coordination of large-scale V2G, IEEE Transactions on Smart Grid,
vol. 6, no. 4, July 2015.
[12] A. Hajimiragha, C. A. Caizares, M.W. Fowler, and A. Elkamel,
Optimal transition to plug-in hybrid electric vehicles in Ontario,
Canada, considering the electricity-grid limitations, IEEE Transactions
on Industrial Electronics, vol. 57, pp. 690701, Feb. 2010.
[13] D. Wu, D. C. Aliprantis, and K. Gkritza, Electric energy and power
consumption by light-duty plug-in electric vehicles, IEEE
Transactions on Power Systems, vol. 26, no. 2, pp. 738-746, May 2011.
[14] M. F. Shaaban, Y. M. Atwa, and E. F. El-Saadany, PEVs modeling
and impacts mitigation in distribution networks, IEEE Transactions on
Power Systems, vol. 28, no. 2, May 2013.
[15] Q. Li, R. Negi, and M. D. Ilic, "A queueing based scheduling approach
to plug-in electric vehicle dispatch in distribution systems", [Online].
Available: http://arxiv.org/pdf/1203.5449v1.pdf.
[16] J. G. Vlachogiannis, Probabilistic constrained load flow considering
integration of wind power generation and electric vehicles, IEEE
Transactions on Power Systems, vol. 24, no. 4, pp. 18081817, Nov
2009.
[17] S. Bae, and A. Kwasinski, Spatial and temporal model of electric
vehicle charging demand, IEEE Transactions on Smart Grid, vol. 3,
no. 1, Mar 2012.
[18] R. Garcia-Valle and J. G. Vlachogiannis, Electric vehicle demand
model for load flow studies, IEEE Transactions on Electric Power
Components and Systems, vol. 37, no. 5, pp. 577582, May 2009.
[19] G. Li and X. Zhang, Modeling of plug-in hybrid electric vehicle
charging demand in probabilistic power flow calculations, IEEE
Transactions on Smart Grid, vol. 3, no. 1, Mar 2012.
[20] Z. Li, Q. Guo, H. Sun, S. Xin, and J. Wang, "A new real-time smart-
charging method considering expected electric vehicle fleet
connections," IEEE Trans. Power Syst, vol. 29, no. 6, pp. 3114-3115,
Nov. 2014.
[21] W. Su, J. Wang, K. Zhang, and A. Q. Huang, "Model predictive
control-based power dispatch for distribution system considering plug-
in electric vehicle uncertainty," Electr Pow Syst Res, vol. 106, pp. 29-
35, 2014.
[22] L. Bremermann, M. Matos, J. Peas Lopes, M. Rosa, Electric vehicle
models for evaluating the security of supply, Electric Power Systems
Research, vol.111, pp.32-39, Fevereiro, 2014.
[23] S. Dhameja, Electric Vehicle Battery Systems, Newnes, 2001, ISBN
0750699167.
[24] The 2011 Transportation Tomorrow Survey [Online]. Available:
http://www.dmg.utoronto.ca/reports/ttsreports.html.
[25] Ontario hydro Time of use pricing rates [Online]. Available:
http://www.ontario-hydro.com/index.php?page=current_rates.
[26] H. C. Tijms, A First Course in Stochastic Models, John Wiley & Sons
Ltd, The Netherlands, 2003.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/TSG.2016.2550219, IEEE
Transactions on Smart Grid
13

[27] Adan, I., and Resing, J., Queueing theory, available at:
http://www.win.tue.nl/_iadan/queueing. pdf, Department of
Mathematics and Computing Science, Eindhoven University of
Technology, The Netherlands, 2002.
[28] Y. Cao, Y. Tan, C. Li, and C. Rehtanz, "Chance-constrained
optimization-based unbalanced optimal power flow for radial
distribution networks," IEEE Trans. Power Del, vol. 28, no. 3, July
2013.
[29] F. Allgwer, R. Findeisen, and Z. Nagy, Nonlinear model predictive
control: from theory to application, J. Chin. Inst. Chem. Engrs., Vol.
35, No. 3, 299-315, 2004.
[30] M. E. Baran and F. F. Wu, Optimal sizing of capacitor placed on ra-
dial distribution systems, IEEE Trans. Power Del. , vol. 4, no. 1, pp.
735743, Jan. 1989.
[31] W. H. Kersting, Distribution System Modeling and Analysis, CRC
Press, New Mexico, 2002.
[32] EasyFit Software [Online].Available:
http://www.mathwave.com/en/home.html.

ACKNOWLEDGMENTS
The authors would like to thank P. Fisher, and G. Keyworth, Regional
Municipality of Waterloo and the Data Management Group, Department of
Civil Engineering, University of Toronto for providing the data to perform this
study.

Omar Hafez (S11) received the BSc degree from

Umm Al-Qura University, Makkah, Saudi Arabia, in
2004 in Electrical and Computer Engineering, and
the MASc degree in Electrical and Computer
Engineering from the University of Waterloo,
Waterloo, ON, Canada, in 2011 where he is
currently pursuing the Ph.D. degree. His research
interests include micro-grids, renewables based
distributed generation, distribution system operation
and planning, electric vehicles, and smart grids.

Kankar Bhattacharya (M95SM01) received the

Ph.D. degree in Electrical Engineering from the
Indian Institute of Technology, New Delhi, in 1993.
He was in the faculty of Indira Gandhi Institute of
Development Research, Mumbai, India (19931998)
and then the Department of Electric Power
Engineering, Chalmers University of Technology
Gteborg, Sweden (19982002). He has been with
the Department of Electrical and Computer
Engineering, University of Waterloo, Waterloo, ON,
Canada, since 2003 and is currently a full Professor. His research interests are
in power system economics and operational aspects.

1949-3053 (c) 2015 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.