A CONCEPTUAL FRAMEWORK
DESIGN FOR IMPLEMENTATION
OF VEHICLE–TO–GRID (V2G)
George Gross
Department of Electrical & Computer Engineering
University of Illinois at Urbana-Champaign
ECE Seminar Series
Carnegie
Ca
eg e Mellon
e o U
University
e s ty
November 12, 2009
© 2009 George Gross, All Rights Reserved.
1
MOTIVATION
‰ There are growing concerns around the world
about energy independence and global warming
issues
‰ The success of the Toyota Prius – with over
510,000 sold by November 2007 in the US – is a
major driver in spearheading the development of
battery vehicles (BVs) in the U.S. and Europe
© 2009 George Gross, All Rights Reserved.
2
OUTLINE
‰ Integration of BVs into the electricity grid
 BVs as a load
 BVs as a g
generation/storage
g device
 role of aggregation
‰ Development of a conceptual framework
© 2009 George Gross, All Rights Reserved.
3
OUTLINE
‰ Major implementational issues
 design of an incentive program
 metering and communication/control needs
‰ Environmental
E i
t l aspects
t
‰ Concluding remarks
© 2009 George Gross, All Rights Reserved.
4
THE ELECTRICITY GRID
‰ Not all the MWhs are equal in terms of costs and
prices
‰ The value of each MWh depends on the time of
production/consumption
‰ The integration of BVs into the grid can fully
exploit the opportunities to:
 buy when the price is low
price is high
g
 sell when the p
 provide additional services needed by the grid
© 2009 George Gross, All Rights Reserved.
5
DEMAND AND LMP
demaand (M
MW)
17,000
16,000
15,000
14 000
14,000
13,000
12,000
11,000
Source: NE ISO
0
2
4
6
8 10 12 14 16 18 20 22 24
time of day
© 2009 George Gross, All Rights Reserved.
6
DEMAND AND LMP
average
16,000 demand
demaand (M
MW)
17,000
15,000
14 000
14,000
13,000
12,000
11,000
Source: NE ISO
0
2
4
6
8 10 12 14 16 18 20 22 24
time of day
© 2009 George Gross, All Rights Reserved.
7
DEMAND AND LMP
average
16,000 demand
120
15,000
100
demaand (M
MW)
17,000
110
90
14 000
14,000
80
13,000
70
12,000
60
11,000
50
8 10 12 14 16 18 20 22 24
Source: NE ISO
0
2
4
6
price ($/MW
Wh)
130
time of day
© 2009 George Gross, All Rights Reserved.
8
DEMAND AND LMP
average
16,000 demand
120
15,000
100
demaand (M
MW)
17,000
110
90
14 000
14,000
80
13,000
average
price
12,000
11,000
Source: NE ISO
0
2
4
6
70
price ($/MW
Wh)
130
60
50
8 10 12 14 16 18 20 22 24
time of day
© 2009 George Gross, All Rights Reserved.
9
THE BV AS A “PURE LOAD”
s.o.c. (%)
100
60
6
8
time of day
18
© 2009 George Gross, All Rights Reserved.
24
10
CHARGING THE BVs
40
looad (GW
W)
36
load with 4 million
PHEVs
32
28
California
load without
BVs
24
20
6
12
18
time of day
24
Source: Lucy Sanna, “Driving the solution, the plug-in hybrid vehicle,” EPRI journal, Fall 2005
© 2009 George Gross, All Rights Reserved.
6
11
LEVELING THE LOAD
demaand (M
MW)
17,000
16,000
15,000
14 000
14,000
13,000
12,000
11,000
Source: NE ISO
0
2
4
6
8 10 12 14 16 18 20 22 24
time of day
© 2009 George Gross, All Rights Reserved.
12
LEVELING THE LOAD
loadd
impacts of the BVs
0
time of day
© 2009 George Gross, All Rights Reserved.
5
13
REGULATION SERVICE AND PRICING
1200
demaand (M
MW)
960
720
480
240
0
- 240
- 480
- 720
- 960
- 1200
Source: PJM
0
2
4
6
8 10 12 14 16 18 20 22 24
time of day
© 2009 George Gross, All Rights Reserved.
14
1200
80
960
70
720
60
480
240
50
0
40
- 240
30
- 480
20
- 720
10
- 960
- 1200
Source: PJM
price ($/MW
W/h)
demaand (M
MW)
REGULATION SERVICE AND PRICING
0
2
4
6
0
8 10 12 14 16 18 20 22 24
time of day
© 2009 George Gross, All Rights Reserved.
15
REGULATION SERVICE AND PRICING
average
price
demaand (M
MW)
960
720
70
60
480
240
50
0
40
- 240
30
- 480
20
- 720
10
- 960
- 1200
Source: PJM
80
price ($/MW
W/h)
1200
0
2
4
6
0
8 10 12 14 16 18 20 22 24
time of day
© 2009 George Gross, All Rights Reserved.
16
ROLE OF BVs IN FREQUENCY
REGULATION
‰ A basic objective of the system operator is to
ensure that the supply – demand equilibrium is
maintained around the clock
‰ Imbalances lead to frequency fluctuations that
need
eed to be regulated
egu ated
‰ The supply-demand imbalance is checked every 2
to 4 s
© 2009 George Gross, All Rights Reserved.
17
ROLE OF BVs IN FREQUENCY
REGULATION
loadd/generration
generation
load
time of day
© 2009 George Gross, All Rights Reserved.
18
OFF – PEAK REGULATION
‰ Compliance with the unit commitment schedules
becomes difficult during low load conditions that
characterize the off – peak periods
‰ While, the operator may not wish to turn off units,
there may be no choice
‰ Wind integration further exacerbates these
conditions
g
prices
p
are,, typically,
yp
y, the highest
g
as
‰ The regulation
many units need to reduce their outputs
© 2009 George Gross, All Rights Reserved.
19
pricees > 2550 $/M
MW/h
num
mber off days w
with
PEAK AND OFF – PEAK REGULATION
time of day
Source: CAISO
© 2009 George Gross, All Rights Reserved.
20
BVs AND FREQUENCY REGULATION
‰ Batteries have the ability to both absorb and
discharge energy
‰ The regulation capacity quantity provided by a
battery is relatively small
‰ Batteries have very short response times (of the
order of ms))
‰ The frequent switching of batteries may, however,
severely impact their life expectancies
© 2009 George Gross, All Rights Reserved.
21
THE BV AS A “SUPPLY-SIDE
RESOURCE”
s.o.c. (%)
100
60
6
8
time of day
18
© 2009 George Gross, All Rights Reserved.
24
22
BATTERY ISSUES
‰ The battery capacity of each BV is small in terms
of kWh storage
‰ This
Thi capacity
it limitation
li it ti restricts
t i t consequently
tl the
th
“supply-side
supply-side resource
resource” capability of each BV
‰ A key
y requirement for g
grid integration
g
is the
aggregation of BVs into a collection with
capability to impact the grid
© 2009 George Gross, All Rights Reserved.
23
V2G CONCEPTUAL FRAMEWORK
‰ Load aggregation
‰ Resource aggregation
‰ Explicit representation of uncertainty
‰ Communications/control layer construction
‰ Development of incentives
© 2009 George Gross, All Rights Reserved.
24
PRINCIPAL PLAYERS IN THE V2G
INTEGRATION
‰ BV
‰ Aggregator
‰ Aggregated
gg g
BVs
‰ ESP
‰ ISO/RTO
© 2009 George Gross, All Rights Reserved.
25
THE INTEGRATION FRAMEWORK
aggregator
…
…
…
resource aggregation
load aggregation
ESP
© 2009 George Gross, All Rights Reserved.
ISO/RTO
26
THE ROLE OF AGGREGATION
‰ The storage capacity C for a typical BV is in the
1 – 30 kWh range
‰ If we consider the total discharge of the full
battery over 5 h, the output is in the 0.2 – 6 kW
range
‰ The aggregator, who gathers together “many”
BVs to result in a nontrivial aggregated output
pp y
and load, interfaces with the ISO/RTO on supplyside issues and the ESP on demand-side issues
© 2009 George Gross, All Rights Reserved.
27
FLOWS IN V2G INTEGRATION
individual
BV owner
ESP
ISO/RTO
6
…
…
aggregated BVs as a resource
parking
facility
battery
supplier
aggregated BVs as a load
capacity/energy flows
dollar flows
© 2009 George Gross, All Rights Reserved.
28
V2G CONCEPTUAL FRAMEWORK
all services
$
6
aggregated
BVs as a
resource
BV
…
$
aggregated
BVs as a
load
BV
…
BV
MWh
$
MW
BV
MWh
$
MWh
h
ESP 1
ISO/RTO
…
ESP
n
parking
services
batteries
$
battery
supplier
A
…
battery
supplier
Z
$
parking
facility D
© 2009 George Gross, All Rights Reserved.
…
parking
facility ]
29
REPRESENTATION OF UNCERTAINTY
SOURCES
‰ We take into account various sources of
uncertainty, including:
 time of arrival
 parking time
 state of charge
 storage of the vehicle
 demand
‰ For the aggregated BVs,
BVs we use the Central Limit
Theorem (N > 30) to justify the representation of
th various
the
i
random
d
variables
i bl by
b normall
distributions
© 2009 George Gross, All Rights Reserved.
30
DAILY COMMUTE DISTANCES
100
cumulativee percen
ntage
of veh
hicles
90
80
70
60
50
40
average commute distance is
22 miles = 35 km
30
20
10
0
30
60
90
120
150
average daily travel distance (miles)
Source: Lucy Sanna,“Driving the solution, the plug-in hybrid vehicle,” EPRI journal, Fall 2005
© 2009 George Gross, All Rights Reserved.
31
PARKING LOT UTILIZATION AS A
FRACTION OF ITS CAPACITY
1
frraction
0.8
day 1
0.6
0.4
0.2
0
6
8
10
12
14
time of day
16
© 2009 George Gross, All Rights Reserved.
18
20
32
PARKING LOT UTILIZATION AS A
FRACTION OF ITS CAPACITY
1
day 2
frraction
0.8
day 1
0.6
0.4
0.2
0
6
8
10
12
14
time of day
16
© 2009 George Gross, All Rights Reserved.
18
20
33
PARKING LOT UTILIZATION AS A
FRACTION OF ITS CAPACITY
day 3
day 2
day 4
1
frraction
0.8
day 1
0.6
0.4
0.2
0
6
8
10
12
14
16
18
20
time of day
© 2009 George Gross, All Rights Reserved.
34
APPROXIMATION OF PARKING
UTILIZATION
1
fr
fraction
0.8
0.6
0.4
normal
approximation
0.2
0
6
8
10
12
14
16
18
20
time of day
© 2009 George Gross, All Rights Reserved.
35
MODEL OF PARKING UTILIZATION
1
fr
fraction
0 .8
0 .6
0 .4
normal
approximation
0 .2
0
6
8
10
12
14
16
18
20
time of day
© 2009 George Gross, All Rights Reserved.
36
s.o.c. OF THE BV BATTERY
‰ The role of the s.o.c. is
key to the effective
power output (kW)
management of the
C
5
aggregated BV
60
integration into the
grid
y is a function of
battery
its storage capability
s.o.c.
100
‰ The output of a
BV acts as
supply-side
resource
(%)
C
5
BV acts as
BV acts as either
demand-side demand- or supplyresource
side resource
© 2009 George Gross, All Rights Reserved.
37
BVs PROVIDE IMPORTANT SERVICE
‰ The aggregated BVs constitute a very important
supply-side resource to the grid
presence of the BVs p
provide the ISO/RTO with
‰ The p
considerable flexibility in the scheduling of units
‰ As a result, the start-up of cycling and peaking
units may be delayed or avoided; the availability
of reserves is improved and, during off-peak, the
need for reserves is reduced
© 2009 George Gross, All Rights Reserved.
38
CYCLING UNITS WITHOUT V2G
peaking units
demaand (M
MW)
17,000
16,000
15,000
cycling
i
units
start-up the
cycling units
14 000
14,000
13,000
12,000
11,000
Source: NE ISO
0
2
4
6
8 10 12 14 16 18 20 22 24
time of day
© 2009 George Gross, All Rights Reserved.
39
CYCLING UNITS WITH V2G
peaking units
demaand (M
MW)
17,000
16,000
15,000
cycling
units
i
start-up the
cycling units
14 000
14,000
13,000
12,000
11,000
Source: NE ISO
delay
0
2
4
6
8 10 12 14 16 18 20 22 24
time of day
© 2009 George Gross, All Rights Reserved.
40
AGGREGATED BV POWER/CAPACITY
450
400
poweer (MW
W)
350
300
250
200
150
100
50
0
5
10
15
time of day
© 2009 George Gross, All Rights Reserved.
20
41
REGULATION SERVICE AND PRICING
average
price
demaand (M
MW)
960
720
70
60
480
240
50
0
40
- 240
30
- 480
20
- 720
10
- 960
- 1200
Source: PJM
80
price ($/MW
W/h)
1200
0
2
4
6
0
8 10 12 14 16 18 20 22 24
time of day
© 2009 George Gross, All Rights Reserved.
42
DAY – TIME REGULATION SERVICE
PROVISION BY 100,000
,
BVs
40
regulatioon (MW)
30
20
time of
day
10
0
-10
8
10
12
14
16
18
-20
-30
-40
© 2009 George Gross, All Rights Reserved.
43
PERCENTAGE OF BVS PROVIDING
THE REGULATION SERVICE
perrcentage off BVs provid
iding servicce
10
9
8
7
6
5
4
3
2
1
0
8
9
10
11
12
13
14
15
16
17
18
time of day
© 2009 George Gross, All Rights Reserved.
44
PROVISION OF LOAD SHAVING
SERVICE IN ADDITION TO REGULATION
‰ The number of BVs providing regulation service
remains rather low
low, with fewer than 10% of the
BVs in the aggregation providing service at each
point in time
‰ We consider the provision of load shaving service
in addition to the regulation service
‰ We show that the Aggregator can provide 100
MWh of load shaving service at a constant power
output between 9:00 and 9:30 a.m for an
aggregation of 100,000 BVs
© 2009 George Gross, All Rights Reserved.
45
PERCENTAGE OF BVS PROVIDING LOAD
SHAVING AND REGULATION SERVICE
percen
ntage of BV
Vs providin
ng service
60
50
40
30
20
10
0
8
9
10
11
12
13 14
time of day
15
© 2009 George Gross, All Rights Reserved.
16
17
18
46
maxim
mum consttant powerr output fo
or
the 9--9:30 perio
od in addiition to thee
provvision of 3
30-MW reg
egulation
serviice (MW)
ENERGY PROVIDED IN ADDITION TO
THE REGULATION SERVICE
200
150
100
50
0
40
50
60
70
80
90
100
size of BV aggregation in thousands
© 2009 George Gross, All Rights Reserved.
47
REGULATION FOR OFF-PEAK
CONDITIONS
r
regulati
ion (M
MW)
1000
500
0
1
2
3
4
5
6
1
2
3
4
5
6
- 500
time of
day
- 10 00
0
© 2009 George Gross, All Rights Reserved.
48
REGULATION FOR OFF-PEAK
CONDITIONS
r
regulati
ion (M
MW)
1000
500
0
1
2
3
4
5
6
2
3
4
5
6
- 500
- 10 00
time of
day
10,000 BVs
0
1
© 2009 George Gross, All Rights Reserved.
49
REGULATION FOR OFF-PEAK
CONDITIONS
r
regulati
ion (M
MW)
1000
500
0
1
2
3
4
5
6
- 500
- 10 00
10,000 BVs
0
1
time of
day
50,000 BVs
2
3
4
© 2009 George Gross, All Rights Reserved.
5
6
50
REGULATION FOR OFF-PEAK
CONDITIONS
r
regulati
ion (M
MW)
1000
100,000 BVs
500
0
1
2
3
4
5
6
- 500
- 10 00
10,000 BVs
0
1
time of
day
50,000 BVs
2
3
4
© 2009 George Gross, All Rights Reserved.
5
6
51
REGULATION FOR OFF-PEAK
CONDITIONS
r
regulati
ion (M
MW)
1000
100,000 BVs
500
500,000 BVs
0
- 500
- 10 00
10,000 BVs
0
1
50,000 BVs
2
3
4
time of day
© 2009 George Gross, All Rights Reserved.
5
6
52
KEY IMPLEMENTATIONAL ISSUES
‰ Aggregation implementation
‰ Information layer construction
‰ Incentive development
‰ Realization of environmental benefits
© 2009 George Gross, All Rights Reserved.
53
V2G COMMUNICATION AND
METERING
charging
station
stat
o
Aggregator
6
distribution
ggrid
.
ZigBee
i
transceiver
ISO/RTO
ISO/RTO
.
ESP
© 2009 George Gross, All Rights Reserved.
54
ESSENTIAL COMMUNICATION /
CONTROL SYSTEM REQUIREMENTS
‰ Speed: signals need to be sent every 1 to 2 s
‰ Range:
R
every BV in
i a parking
ki lot
l t mustt be
b on the
th
communication network
‰ Measurement: metering must be installed to
enable payment for services
‰ Reliability: full utilization of all parked aggregated
BVs
‰ Security: BVs may be used to hack the network
© 2009 George Gross, All Rights Reserved.
55
ESSENTIAL COMMUNICATION /
CONTROL SYSTEM REQUIREMENTS
‰ Costs: each BV has an implanted device and the
costs per unit must be low for the large collection
of aggregated BVs
‰ Extendibility: the communication layer must allow
g
of additional BVs
the integration
‰ Interoperability: a non-restrictive, flexible standard
needs to be introduced and implemented
© 2009 George Gross, All Rights Reserved.
56
COMPUTER / COMMUNICATION / CONTROL
NETWORK: INFORMATION FLOWS
‰ ID of each BV
‰ Preferences/constraints of each BV
‰ Parking
P ki status
t t off each
h BV
‰ Storage capability of the BV battery
‰ The BV battery s.o.c.
‰ Power flows from BV battery to the grid
‰ Measured value of metered quantities
© 2009 George Gross, All Rights Reserved.
57
THE ROLES OF THE AGGREGATOR
‰ Development of the parking infrastructure
‰ Maintenance of the batteries and the network
‰ Creation of relationships with the BV
manufacturers
f t
‰ Interface with ISO/RTO
© 2009 George Gross, All Rights Reserved.
58
VALUE ADDED BY THE
AGGREGATOR
‰ Provides a “package deal” to the aggregated BVs
in terms of:
 parking facilities
 service
i acquisition
i iti
and
d provision
i i
 charging of BVs
 battery service
p shopping”
pp g for potential
p
BV
‰ Allows “one-stop
participants
‰ Acts as a provider of environmental benefits for
reduced emissions
© 2009 George Gross, All Rights Reserved.
59
NET GENERATION BY ENERGY
SOURCE UP TO 6/31/2009
petroleum 1.15
1.15%
%
nuclear
20.8%
hydro 7.6
7.6%
%
natural gas
21.4%
21 4%
21.4
4%
other renewable
3.6%
3.6
%
coal 45%
other
th gases 0.5%
0 5%
0.5
5%
industry total = 4,115 Twh
Source : Ameren IP
© 2009 George Gross, All Rights Reserved.
60
CO2 EMISSION BY PLANT TYPE
1000
900
800
600
500
400
300
200
100
© 2009 George Gross, All Rights Reserved.
co
al
na
tu
ra
lg
as
pe
tr
ol
eu
m
so
la
r
nu
cl
ea
r
hy
dr
o
0
w
in
d
g/kWh
h
700
61
ENERGY CONSUMPTION UNDER
MIXED CONDITIONS
0,30
.
kW
Wh/km
m
0,25
.
0,20
.
0,15
.
0,10
.
0,05
,
.
0,00
Miles EV
Volvo C30
Recharge
Tesla roadster
.
© 2009 George Gross, All Rights Reserved.
Chevrolet LUV
(Chile 2000)
.
62
VEHICLE CO2 EMISSIONS
350
300
g
g/km
250
200
150
100
50
0
EV France
EV Canada
EV USA
EV Germany
© 2009 George Gross, All Rights Reserved.
internal
combustion
63
FUTURE WORK
‰ Improvement of the BV selection for the provision
of higher energy and regulation performance
‰ Design and implementation of a secure and
economic communication/control architecture
‰ Design of an effective incentive program for high
BV participation and retention
© 2009 George Gross, All Rights Reserved.
64
CONCLUDING REMARKS
‰ Integration of BVs helps the grid both as loads
and as generation sources
S sa
and
d ISO/RTOs
SO/ Os will have
a e a new
e player
p aye to do
‰ ESPs
business with
‰ Aggregators are key for the implementation of
V2G concept to be successful
‰ The BV aggregator has the potential for making
sizeable benefits
© 2009 George Gross, All Rights Reserved.
65
REFERENCES
‰ C. Guille and G. Gross, "A Conceptual Framework
for the Vehicle-to-Grid (V2G) Implementation,"
accepted for publication in Energy Policy, 2009
© 2009 George Gross, All Rights Reserved.
66