See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/224205055

Planning electric-drive vehicle charging under constrained grid conditions

Conference Paper · November 2010

DOI: 10.1109/POWERCON.2010.5666620 · Source: IEEE Xplore

CITATIONS READS
147 2,306

2 authors:

Olle Sundström Carl Binding

Alpiq Cognitive Solutions and Innovation
28 PUBLICATIONS 2,581 CITATIONS 35 PUBLICATIONS 1,192 CITATIONS

SEE PROFILE SEE PROFILE

All content following this page was uploaded by Olle Sundström on 04 June 2014.

The user has requested enhancement of the downloaded file.

RZ 3785 (# Z1008-002) 08/26/2010
Computer Science 6 pages

Research Report

Planning Electric-Drive Vehicle Charging Under Constrained

Grid Conditions

O. Sundström, C. Binding

IBM Research – Zurich

8803 Rüschlikon
Switzerland

LIMITED DISTRIBUTION NOTICE

This report has been submitted for publication outside of IBM and will probably be copyrighted if accepted for publication. It has
been issued as a Research Report for early dissemination of its contents. In view of the transfer of copyright to the outside pub-
lisher, its distribution outside of IBM prior to publication should be limited to peer communications and specific requests. After
outside publication, requests should be filled only by reprints or legally obtained copies (e.g., payment of royalties). Some re-
ports are available at http://domino.watson.ibm.com/library/Cyberdig.nsf/home.

Research
Almaden • Austin • Beijing • Delhi • Haifa • T.J. Watson • Tokyo • Zurich
 Planning Electric-Drive Vehicle Charging under
Constrained Grid Conditions
Olle Sundström and Carl Binding

Abstract—This paper presents a novel method of planning charging should of course be automated, and could essentially
the charging of electric-drive vehicles that takes electricity grid be performed by the vehicle’s system if the price information
constraints into account. The method computes an individual is available. Several concepts have been proposed for charging
charging plan for each vehicle while minimizing the cost of
electricity, avoiding distribution grid congestion, and satisfying management using price-based methods. In [9], the charging
the individual vehicle owner’s requirements. The underlying and discharging of eight plug-in vehicles is optimized based
algorithm is explained through a simple example and tested on on the day-ahead electricity price. The optimization scheme
a simulated electricity grid. The method is shown to significantly used is based on particle swarm optimization, i.e., a variation
reduce the overloading in the electricity grid compared to of a randomized search algorithm. The study also investigates
charging schemes that do not consider grid constraints.
Index Terms—Balancing Power, Demand Management, Distri- the impact of grid faults on smart charging parking lots. The
bution Networks, Electric Vehicles, , Electricity Grid, Intermit- price-based charging scheme is also described in [10], where
tent Energy Sources, Netflow Algorithms, Smart Charging time-based or price-threshold charging is suggested. A time-
shifted charging, in which different geographical regions have
I. I NTRODUCTION different charging hours, is proposed in [11]. In general, price-
For reasons of CO2 reduction, independence from fossil based charging schemes depend on that the resulting reduction
fuels, and potential provision of balancing energy in grids in charging cost is a sufficiently strong incentive for the EV
with high proportions of intermittent power generation, plug- owner to choose price-based charging. Additional studies on
in hybrid and fully electrical vehicle technology has attracted forecasting and managing EV charging can be found in [12],
renewed attention. It is clear that electric vehicles (EVs) will [13].
be available on the market because most manufacturers are More interestingly, distribution grid operators also have
developing EVs. The impact of the EV fleet on the power an interest to manage the charging because for them it is
grid is therefore inevitable. The impact of EVs on the grid was important to incorporate a large number of EVs without
studied as early as 1983 [1]. However, the time and extent of massivly reinforcing the distribution grid. The impact of EVs
this impact are largely dependent on the success of EVs. Ungar on the electricity grid is studied by Letendre et al. [14], where
et al. [2] argue that the impact will in fact be a major concern the focus is on the Vermont power grid. They assume a dual-
to the power grid operators, in particular the distribution grid tariff, nightly charging scheme, and conclude that enough
operator. transport capacities are available in the power grid. Grid
constraints are only considered for the transport and high-
Several new concepts have been proposed on how to use
voltage distribution grid, whereas the low-voltage distribution
the grid-connected EVs for grid services, also known as
grid is not considered.
vehicle-to-grid (V2G). These concepts usually involve both
More details on potential impact on a low-voltage distri-
discharging and charging of EVs to help the grid to level out
bution grid are given in [15]. Smart charging behavior is
peaks in overall consumption. A literature overview of the
considered to maximize the density of EV deployment into
V2G field is given in Jenkins et al. [3]. Jenkins et al. focus on
the grid, i.e., to reach the maximally tolerable number of EVs
the United States power grid and argue that several challenges
while maintaining grid constraints. The charging goals were
such as managing and dispatching power remain to be solved.
set to achieve full charge for all plugged-in EVs. Hence this
Additional V2G concepts have been studied in [4], [5], [6], [7],
approach does not consider flexible charging schemes with
[8]. However, there are challenges with V2G services because
variable charging goals. In [16], the impact of a large fleet
it is not clear how much EV battery ageing, and therefore also
of plug-in hybrid electric vehicles on the Virginia-Carolinas
the associated warranty, are affected during V2G operation.
electric grid is studied. The focus is on the supply of electricity
It is clear that there are several entities interested in man-
and not on the transmission and distribution capabilities.
aging the charging of the electric vehicles. The vehicle owner
However, it is concluded that a large fleets of plug-in vehicles
has of course an interest in minimizing the cost of charging.
will impact the electricity grid.
The recent trend to move to hourly electricity pricing can
This paper considers the power grid on the Danish island of
help the owner to shift EV charging request from high-price
Bornholm, similarly as in [17], where the isolated main island
hours to hours with less demand. This management of the
of the Azores is used to study the impact of EVs and the
O. Sundström and C. Binding are with IBM Research, Zurich, 8803 potential profit to be made on grid services. The focus of this
Rüschlikon, Switzerland. (email: {osu, cbd}@zurich.ibm.com) paper is not on estimating the impact of the EVs, but rather
on proposing a method of planning the individual charging on customers, their billing information, as well as the
schedules of a large EV fleet while respecting the constraints metering of the EV-specific power consumption and feed-
in the low-voltage distribution grid. The method has been backs into the grid. A user-facing client GUI needs to be
tested in a simulation environment in which the movement provided to enable users to manage their data and to let
and charging of individual EVs are simulated simultaneously. the VPP operator interact with customer data.
A load-flow simulation of the electric grid is performed to • Communication: To gather data from the various entities
assess the impact of the large fleet of EVs. (EVs, transport and distribution grid, power generation,
The paper is structured as follows: the basic functions of markets) an appropriate communication infrastructure
the fleet operator are described in Section II. The simulator must be deployed.
is described in Section III with the focus on the electric As the goal of this paper is to propose and evaluate a novel
vehicle model and the topology of the simulated grid. The algorithm for performing grid-aware charging of large EV
optimization of the charging schedules for an unconstrained fleets, the focus is on the Optimization block. It is assumed
grid is described in Section IV and the proposed method of that the EV trips for the next day are perfectly predictable, i.e.,
optimizing the schedules for a constrained grid is shown in the location and time of connection and disconnection, as well
Section V. The results of using the method on a simulated as the required energy are known. This assumption is further
large-scale distribution grid are given in Section VI. Finally emphasized in the discussion and conclusions in Section VII.
Section VII discusses and concludes the paper.
III. S IMULATOR
II. F LEET O PERATOR
To evaluate the proposed algorithm of coordinating the
There are several ideas on how electric vehicles will be charging of EVs, a simulator is used. The simulator is a
integrated into society. Waraich et al. [18] envision two hybrid simulator including both discrete and continuous state
charging schemes: a per-EV, decentralized, decision process variables. It simulates a large fleet of EVs and the electricity
and a centrally controlled charging scheme. Each method has grid to which the EVs connect for charging. In this section,
benefits and drawbacks. The decentralized approach assumed the electric vehicle model and the electricity grid simulation
that the EV itself optimizes its charging behavior based on, methods are explained.
for example, a price signal. The drawback with this approach
is that the EV needs to collect and store trip history and, if A. Electric vehicle fleet overview
the EVs need to coordinate their charging, for example to For the fleet operator that optimizes the charging schedules
include grid constraints, the need for communication is high. the electric vehicles are mobile energy buffers. The energy
The centralized approach on the other hand assumes that the buffers in this study are batteries, which are simulated using
centralized unit optimizes the charging and that the resulting a non-linear model. Each battery is modeled as an equivalent
charging schedules are communicated to the EVs. electric circuit containing a voltage source in series with a
In this study, it is assumed that there is a centralized resistor. Both the voltage source and the internal resistance
EV aggregator that can act on the power market and use depend on the state-of-energy ζ ∈ [0, 1] of the battery. The
the available electricity products and financial instruments to, battery model also depends on the specific cell charateristics
for example, minimize the cost of charging the EVs. In the and the size of each battery pack. The dynamic state variable
remainder of this paper, the EV aggregator is referred to as for a single electric vehicle is the state-of-energy
the fleet operator (FO). It has the following basic modules:
Pint (ζ, Pb )
• Data storage: The FO needs to gather considerable ζ̇ = f (ζ, Pb ) = . (1)
Eint0
amounts of data to perform its EV-fleet management
tasks. In particular, historical trip data is needed to predict where Eint0 is the maximum stored energy in the battery
future EV usage. In addition, end customer information and Pint (ζ, Pb ) is the internal power of the battery. The
may need to be stored, as well as billing information. detailed battery model is described in [19]. Because of battery
• Trip forecasting: The forecasting of the anticipated energy limitations, the charging power is limited to
requirements for EV usage is essential to minimize the Pb ∈ [Pb,min (ζ), Pb,max (ζ)] (2)
driver interactions. The FO has to estimate how much
energy has to be fed into an EV while it is connected to and the state-of-energy is constrained to
the power grid. The connection location also plays a role
ζ ∈ [0.2, 0.8] (3)
when handling potential grid congestions. This module
evidently depends on the data storage subsystems. to avoid premature aging.
• Optimization: This module computes an optimal EV
charging plan, taking into account estimated energy pro- B. Electricity grid overview
duction, required energy needs, expected durations of The electricity grid is simulated using a conventional load-
charging periods and potential grid constraints. flow simulation. In the load-flow simulation, each component
• Customer relationship and billing information: This is is modeled as a two-port element. The electricity grid model,
traditional IT infrastructure for maintaining information i.e., the admittance matrix, is based on the grid on the Danish
island of Bornholm. The parameters for the grid model are index for the time slot contained in one plan duration. The
constructed using both real-world and synthetic data, where decision variable pb then has m · n elements. The cost vector
no real-world data is available. c comprises of the cost associated for each vehicle in each
The grid contains three voltage levels 60 kV, 10 kV, and time slot ci,j . The charging power vector pb comprises the
400 V. The 60 kV network is meshed whereas the 10 kV charging power for each vehicle and time slot pb,i,j .
and the 400 V networks are radial trees extending from the
c = [c1,1 , c1,2 , . . . , c1,n , . . . ,
60/10 kV substations. In total, the grid contains roughly 12000
T
400 V end nodes. The ratio of cables to end nodes is roughly cm,1 , cm,2 , . . . , cm,n ] (6)
1:1. There are several types of consumers in the grid, such as
households, farms, and bakeries. These consumers vary in size pb = [pb,1,1 , pb,1,2 , . . . , pb,1,n , . . . ,
and are spread out through the grid. The load curves are based T
pb,m,1 , pb,m,2 , . . . , pb,m,n ] . (7)
on German VDEW consumption profiles. On the generation
side, the grid simulator includes a single power plant and Because of battery or charging spot limitations, the charging
several wind turbines spread out across the island. The wind power is limited to p̄b,i,j . The charging power is then limited
speed, which is an input to the turbines, is based on historical to pb ∈ [0, p̄b ], i.e., bl = 0 and bu = p̄b in (5). Note that
measurements of actual wind speeds in Denmark. The grid all those slots in which vehicles are not connected can be
is artificially dimensioned to handle twice the peak load of eliminated prior to solving the optimization problem.
the base consumption. After the grid has been dimensioned, V. O PTIMIZATION IN A CONSTRAINED GRID
each 400 V end node is equipped with a charging spot with
The goal of the optimization in this section is to derive
a maximum power of 16 kW. Each charging spot can handle
a charging schedule for each vehicle that ensures sufficient
two EVs simultanously.
energy for the predicted trips while respecting the grid capacity
Thus, for an outlet node, u, in the electricity grid, the base
and minimizing the total cost of electricity for the fleet. The
load, pf (u), and the EV load, pv (u), give the total load at
grid can be considered by including each edge, i.e., line and
each end node:
transformer load, as an additional variable. The additional
p(u) = pf (u) + pv (u) (4) constraints are then the load balance in each node in the
network. The upper bound of each network-load variable is the
at each time instance. The grid is simulated using a load-flow actual maximum tolerable load. The new linear optimziation
simulation and is evaluated every 15 minutes. problem can be formulated using a single large linear program.
IV. O PTIMIZATION IN AN UNCONSTRAINED GRID The number of variables in this problem is
The goal of the optimization in an unconstrained grid is nvariables ≈ nslots · nvehicles + nslots · nedges (8)
to derive a charging schedule for each vehicle that ensures
and the number of constraints is
sufficient energy for the predicted trips, while, for example,
minimizing the total cost of the electricity used for the fleet. nconstraints ≈ 2 · nstopovers/vehicle · nvehicles +
The charging schedules are divided into time slots for the given nslots + nslots · nnodes . (9)
planning period. In this paper the planning period, which is the
next day, is divided into 96 slots of 15 minutes each. An EV If assuming 15-minute slots, 50’000 managed electric vehicles,
charging schedule therefore contains a charging power level and a small city with a distribution grid that contains 200’000
for each of the 15-minute slots during the day of operation. edges and 100’000 nodes, the total number of variables is
Establishing charging schedules for an EV fleet can be done by on the order of 25 · 106 and the number of constraints is 15 ·
solving an optimization problem [19], which can be formulated 106 . However, even though this problem could be solved using
as the linear program appropriate hardware and software, the size of the problem is
even larger for a larger city.
min ts cT pb (5a) Also, it is not clear whether the information regarding the
pb

subject to grid topology and capacity is available to the EV fleet operator

that optimizes the charging schedules. For example, if the
As pb ≥ bs (5b) distribution system operator is separated from the FO, there
Ag pb ≤ bg (5c) must be an exchange of information with the districution
Ab pb ≤ bb (5d) system operator to be able to consider grid constraints. The
bl ≤ pb ≤ bu , (5e) method proposed in this paper uses a clear separation between
evaluating the grid congestion and optimizing the charging
with the cost vector c, the charging power vector pb , the schedules. This separation can be used when the grid infor-
stop-over inequality constraints (As , bs ), the generation in- mation is not available to the FO.
equality constraints (Ag , bg ), the battery inequality constraints In the unconstrained case, the planning can be done by
(Ab , bb ), and the upper and lower bounds (bu , bl ). Assume solving the optimziation problem in (5). In the constrained,
i = 1, 2..., m is the index of the vehicle, j = 1, 2, ..., n the case the same optimization problem is solved and iteratively
 F = {(V, E), s, t ∈ V, c : V × V → R+ }, where
(V, E) is a directed graph with edges E and nodes
1 Forecast generation and base consumption
for each time slot and EV energy needs V containing a source s ∈ V , a target t ∈ V ,
and a capacity c(u, v) ∈ R+ ∀ (u, v) ∈ E and
c(u, v) = 0 ∀ (u, v) 6∈ E. Let the flow in a flow
2 Formulate and solve charging the network be a function f : V × V → R where
schedule optimization problem

f (u, v) ≤ c(u, v) ∀ (u, v) ∈ V × V (10a)

3 For each time slot, generate the flow
network including the charging schedules f (u, v) = −f (v, u) ∀ (u, v) ∈ V × V (10b)
X
4 For each time slot, solve the maximum flow f (u, v) = 0 ∀ v ∈ V \ {s, t}. (10c)
problem based on the flow network
u∈V

The capacitites of the interior edges in the flow network

7 For each subnetwork in each congested time slot,
add one constraint to the optimization problem
{c(u, v) | u, v ∈ V \ {s, t}} (11)
6 For each congested time slot, determine are the maximum power that can be transmitted though
disjoint the subnetworks under the minimum cut
the two ports. The flow network includes additional
yes
edges from a virtual source node s to all nodes repre-
5 Is there congestion senting generation units. The capacities of these edges
in any time slot?
are the generated power g(u) for each generation unit
no u. Similarly, the flow network also includes additional
egdes from all nodes representing outlets to a virtual
target node t. The capacities of these edges are the total
Figure 1. Flowchart of the algorithm of optimizing the charging schedules
under constrained grid conditions. power demand of each outlet p(u) in (4).
4) For each time slot, solve the maximum flow problem
updated to include the constraints in the electricity grid. A based on the flow network: The problem of maxi-
flow chart of the proposed algorithm is shown in Fig. 1. The mizing the flow from the source to the target can be
details of each element are presented in the following list: formulated as a linear program
1) Forecast generation and base consumption for each X
max f (s, u) (12)
time slot and EV energy needs: The available power
u∈V
for the EV fleet can, for example, be the predicted
subject to (10a), (10b), (10c),
available wind power. The base load is determined
using generalized consumption profiles for a variety of which in this paper is solved using IBM ILOG CPLEX.
consumer types, e.g. farms, households, etc. The EV The maximum flow problem can, of course, also be
energy needs are predicted using the trip history of each solved using other techniques such as the Edmond-
vehicle. The EV trip prediction contains the location Karp algorithm or the Goldberg-Tarjan algorithm. The
and duration of the connection events and the minimum choice of using an LP solver was made because of the
required state-of-energy after each charging interval. straightforward formulation and the simple possibility of
2) Formulate and solve the charging schedule optimiza- warm starting the solver.
tion problem: Formulate the optimization problem in 5) Is there congestion in any time slot? If the capacities
(5) based on the given generation, consumption, and EV of the flows to the target are equal to the actual flows,
charging need forcasts. If this is not the first iteration, the requested power levels at the outlets are feasible and
include the grid constraints calculated in the preceding can be delivered through the network. The criteria can
iterations. The optimization problem (5) is solved using be checked as follows:
IBM ILOG CPLEX1 library [20].
3) For each time slot, generate the flow network includ- f (u, t) = c(u, t) ∀ u ∈ V. (13)
ing the charging schedules: To determine whether there If (13) is not true, there is a congestion in the grid and
is a congestion in the electricity grid, a flow network is the optimization problem needs to be updated and solved
constructed for each time slot j based on the electricity again in the next iteration. If (13) is true the flow is not
grid model, which in this paper is a network of two congested and the algorithm terminates.
ports. A flow network is defined as follows. 6) For each congested time slot, determine the disjoint
Definition Let a flow network be subnetworks under the maximal source minimum
1 IBM, ILOG and CPLEX are trademarks of International Business Ma-
cut: This step starts by determining the maximal source
chines Corporation in the United States, other countries, or both. Other product minimum cut. This cut is calculated after the maximum
and service names might be trademarks of other companies. flow by finding the nodes in the residual graph that
 can reach the target node. After the cut with partition 6

price [cents/kWh]
(S, T ) has been determined, the disjoint subnetworks
are determined by finding the reachable nodes in T from 5
each edge in the cut. In this paper a subnetwork is
defined as follows.
Definition Let a subnetwork NT of a constrained flow 4
0 6 12 18 24
network F , i.e., with partition (S, T ) from the maximal time [h]
source minimum cut C, be
Figure 2. Artificial day-ahead price of electricity used for optimzing the
NT = (V ′ , E ′ ), (14a) charging schedules.

V ′ ⊆ T \ {t} (14b)
50
′ ′ price-based charging
E = {(u, v) | u, v ∈ V ∩ (u, v) ∈ E} . (14c)
The subnetworks N ′ and N ′′ are disjoint if V ′ ∩ V ′′ = ∅ 40 eager charging
and E ′ ∩ E ′′ = ∅.

power [MW]
7) For each subnetwork in each congested time slot, add 30
one constraint to the optimization problem: For each
disjoint subnetwork NT = (V ′ , E ′ ) in the constraint
20
flow network F with the maximal source minimum cut
base load
C,
X X X 10 grid-aware price-based charging
pv (u) ≤ f (u, v) − pf (u), (15)
(u,t)∈E ′ (u,v)∈C (u,t)∈E ′
v∈V ′ 0
0 6 12 18 24
where pf (u) is the base load used to set up the flow time [h]
network in step 3) of the algorithm. Constraint (15)
limits the sum of the EV loads pv (u) of all EVs in Figure 3. The base load excluding EV charging and the total load including
EV charging during the day of operation. The total load is shown for the three
the disjoint subnetwork to the difference between the different levels of charging management.
maximum inflow to the subnetwork given by the cut
and the total base load of the outlets in the subnetwork. charging, all vehicles charge at the low-priced slots during the
day. In Fig. 3 these peaks in charging, can be seen at 00:45 and
VI. S IMULATION R ESULTS
02:30 in the morning. When using our approach of grid-aware
This section shows the results when applying the proposed price-based charging these peaks are reduced and charging is
algorithm to plan the charging of an EV fleet with 3’500 moved to more expensive slots to avoid grid congestion.
commuter vehicles driving over the simulated grid. To evaluate To assess the benefits of using the proposed method, the
the proposed algorithm, three different charging schemes have actual loading of the grid must be further analyzed. Figure 4
been tested: shows the distribution of the average loading during the day
• Eager charging: All EVs charge their battery fully when of each two-port element in the grid. The differences between
connecting to the charging spot. the three charging schemes are not significant. The majority of
• Price-based charging: The charging schedules are deter- the loading is in fact below 50%, which is the effect of having
mined by solving the unconstrained optimization problem our grid dimensioned to handle twice the base load peak.
in (5). More interestingly is the loading during the EV charging
• Grid-aware price-based charging: The charging is done peaks. Figure 5 shows the average loading distribution, simi-
using the algorithm proposed in Section V. larly to Fig. 4, for the four slots with the highest EV charging
For both the price-based and grid-aware charging, the infor- power. This means that for eager charging the average is
mation of the price of electricity is needed. Figure 2 shows the taken during the morning and the afternoon peaks. For price-
artificial price curve that has been used in the optimization. It based and grid-aware charging, the average is taken during the
is assumed that the price is not influenced by the charging of nightly charging peaks. Figure 5 shows that both for eager and
the electric vehicles. pure price-based charging, the grid is significantly overloaded.
The resulting total load in the system is shown in Fig. 3 for In fact, for the four considered slots eager charging overloads
the three charging schemes. Because the fleet only contains 1.8% of the grid by more than 10%. For pure price-based
commuter vehicles, the eager charging is done when the charging, 4.1% of the grid is overloaded by more than 10%.
vehicles arrive at work in the morning and also after returning Using the proposed method of including the grid constraints
home in the afternoon. The charging peaks therefore overlap in the optimization actually reduces the overload. This can
with the existing peaks in the base load. This effect should be seen in Fig. 5 by the peak in loading at 100% of the
be avoided as much as possible. For the pure price-based grid capacity and the reduced loading above 100%. Using the
 share of grid [-]
in the optimization. There is therefore a need to study other
0.04
objectives and business models for the FO that will enable
eager charging
smooth integration of large EV fleets.
0.02
grid-aware price-based charging
price-based charging
R EFERENCES
0 [1] M. M. Collins and G. H. Mader, “The timing of EV recharging and its
0 0.5 1 1.5 2 effect on utilities,” IEEE Transactions on Vehicular Technology, vol. 32,
loading [-] no. 1, pp. 90–97, 1983.
[2] E. Ungar and K. Fell, “Plug in, turn on, and load up,” IEEE Power &
Figure 4. The grid loading distribution at the peak charging slots during the Energy Magazine, pp. 30–35, May/June 2010.
day for the three different levels of charging management. [3] S. D. Jenkins, J. R. Rossmaier, and M. Ferdowsi, “Utilization and effect
of plug-in hybrid electric vehicles in the United States power grid,”
in Proceedings of the IEEE Vehicle Power and Propulsion Conference,
0.01 Harbin, China, 3–5 September 2008, pp. 1–5.
[4] W. Kempton and S. E. Letendre, “Electric vehicles as a new power
grid-aware price-based charging source for electric utilities,” Transportation Research, vol. 2, no. 3, pp.
0.008 157–175, 1997.
[5] S. E. Letendre and W. Kempton, “The V2G concept: A new for
model power? Connecting utility infrastructure and automobiles,” Public
share of grid [-]

0.006 Utilities Fortnightly, pp. 16–26, 15 February 2002.

[6] W. Kempton and J. Tomic, “Vehicle-to-grid power fundamentals: Cal-
culating capacity and net revenue,” Journal of Power Sources, vol. 144,
0.004 no. 1, pp. 268–279, June 2005.
[7] ——, “Vehicle-to-grid power implementation: From stabilizing the grid
to supporting large-scale renewable energy,” Journal of Power Sources,
vol. 144, no. 1, pp. 280–294, June 2005.
0.002
eager charging [8] J. Tomic and W. Kempton, “Using fleets of electric-drive vehicles for
price-based charging
grid support,” Journal of Power Sources, vol. 168, no. 2, pp. 459–468,
June 2007.
0 [9] G. K. Venayagamoorthy, P. Mitra, K. Corzine, and C. Huston, “Real-
0.5 1 1.5 2
time modeling of distributed plug-in vehicles for V2G transactions,” in
loading [-] Proceedings of the IEEE Energy Conversion Congress and Exposition,
San Jose, CA, 20–24 September 2009, pp. 3937–3941.
Figure 5. The grid loading distribution at the four peak charging slots during [10] E. Larsen, D. K. Chandrashekhara, and J. Oestergaard, “Electric vehicles
the day for the three different levels of charging management. for improved operation of power systems with high wind power pene-
tration,” in Proceedings of the IEEE Energy 2030 Conference, Atlanta,
proposed method, only 0.04% of the grid is overloaded by GA, 17–18 November 2008, pp. 1–6.
more than 10%. [11] F. Koyanagi and Y. Uriu, “A strategy of load leveling by charging and
discharging time control of electric vehicles,” IEEE Transactions on
As the grid-aware charging shifts some charging to more Power Systems, vol. 13, no. 3, pp. 1179–1184, 1998.
expensive slots during the day, the total cost of charging the [12] S. Rahman and Y. Teklu, “Role of the electric vehicle as a distributed
fleet is increased. However, based on the assumed price curve resource,” in Proceedings of the IEEE Power Engineering Society Winter
Meeting, vol. 1, Singapore, 23–27 January 2000, pp. 528–533.
the actual cost per kWh charged to the fleet is only increased [13] M. D. Galus and G. Andersson, “Demand management of grid connected
by 0.2%. plug-in hybrid electric vehicles (PHEV),” in Proceedings of the IEEE
Energy 2030 Conference, Atlanta, GA, 17–18 November 2008, pp. 1–8.
VII. C ONCLUSIONS AND F UTURE W ORK [14] S. Letendre and R. A. Watts, “Effects of plug-in hybrid electric ve-
hicles on the vermont electric transmission system,” Presented at the
The proposed method of considering the electricity grid Transportation Research Board Annual Meeting, Washington D.C., 11–
15 January 2009.
when planning the charging of large EV fleets has been [15] J. A. P. Lopes, F. J. Soares, and R. M. R. Almeida, “Identifying
tested in a simulation environment. The simulation results management procedures to deal with connection of electric vehicles in
show clearly that using the proposed algorithm the overloading the grid,” in Proceedings of the IEEE Bucharest Power Tech Conference,
Bucharest, Romania, 28 June–2 July 2009, pp. 1–8.
in the grid is significantly reduced. The simulated grid is [16] S. W. Hadley, “Evaluating the impact of plug-in hybrid electric vehicles
dimensioned to handle twice the peak in base load. If the on regional electricity supplies,” in Proceedings of the iREP Symposium
grid is more constrained, the overloading will be even more - Bulk Power System Dynamics and Control - VII, Revitalizing Opera-
tional Reliability, Charleston, SC, 19–24 August 2007, pp. 1–12.
significant for all types of charging. Also, this paper considers [17] P. Kadurek, C. Ioakimidis, and P. Ferrao, “Electric vehicles and their
only one type of EV and charging spots rated to 16kW. It impact to the electric grid in isolated systems,” in Proceedings of the
is therefore interesting to consider a mix of vehicle types International Conference on Power Engineering, Energy and Electrical
Drives, Lisbon, Portugal, 18–20 March 2009, pp. 49–54.
and a mix of charging spot ratings, such as low-power home [18] R. A. Waraich, M. Galus, C. Dobler, M. Balmer, G. Andersson,
charging and fast charging spots in central locations. and K. Axhausen, “Plug-in hybrid electric vehicles and smart grid:
In this paper perfect EV trip prediction is assumed. Future Investigations based on a micro-simulation,” ETH Zurich, Switzerland,
Institute for Transport Planning and Systems, Tech. Rep. 10.3929/ethz-
work will, therfore, involve studies on the impact of trip a-005916811, 2009.
prediction errors on the results. If the EV charging does [19] O. Sundström and C. Binding, “Optimization methods to plan the
not impact the price of electricity price-based charging will charging of electric vehicle fleets,” in Proceedings of the International
Conference on Control, Communication and Power Engineering, Chen-
actually syncronize the charging of the vehicles at the low nai, India, 28-29 July 2010, accepted for publication.
price hours during the day. As shown in this paper, the [20] IBM ILOG CPLEX V12.1: User’s Manual for CPLEX, International
peaks are still present even if grid constraints are included Business Machines Corporation, 2009.

View publication stats