Professional Documents
Culture Documents
18 - Congestion Management
18 - Congestion Management
Control
George Gross
Department of Electrical and Computer Engineering
University of Illinois at Urbana–Champaign
ECE 573 © 2007 – 2017 George Gross, University of Illinois at Urbana-Champaign, All Rights Reserved. 1
OUTLINE
q Review of the centralized electricity market (CEM )
q Transmission – unconstrained and constrained
markets
rights (FTR)
need to be addressed
2,000
PJM
1,500
NYISO
CAISO
million $
1,000
500
3.5
3.0
2.5
$/MWh
2.0
1.5
1.0
0.5
0.0
2000 2001 2002 2003 2004 2005 2006
Source: PJM State of the Market Reports
ECE 573 © 2007 – 2017 George Gross, University of Illinois at Urbana-Champaign, All Rights Reserved. 6
THE MARKET TIME FRAME
market decisions
ECE 573 © 2007 – 2017 George Gross, University of Illinois at Urbana-Champaign, All Rights Reserved. 7
THE CENTRALIZED ELECTRICITY
MARKET (CEM )
q We discuss the structure of the forward day-ahead
specified hour h
day–ahead CEM
CEM operator
MWh $ $ MWh $
MWh
ECE 573 © 2007 – 2017 George Gross, University of Illinois at Urbana-Champaign, All Rights Reserved. 10
CEM PARTICIPANTS
q The CEM operator is responsible for this market
marketers
C i ( p is ) = ∫ β is (ξ ) d ξ
0
p is, min p s
i
p is, max MWh/h
ECE 573 © 2007 – 2017 George Gross, University of Illinois at Urbana-Champaign, All Rights Reserved. 13
BUYER b AT NODE j BID AND
BENEFITS
p bj
B j ( p bj ) = ∫ β bj (ξ )dξ
0
$/MWh
bid curve β bj ( p bj )
the benefits of the
buyer at node j
are given by the
area under the
curve
b
0 p j MWh/h
ECE 573 © 2007 – 2017 George Gross, University of Illinois at Urbana-Champaign, All Rights Reserved. 14
THE SOCIAL WELFARE
ECE 573 © 2007 – 2017 George Gross, University of Illinois at Urbana-Champaign, All Rights Reserved. 15
MAXIMIZATION OF THE SOCIAL
WELFARE
q The objective in markets is to maximize the social
welfare, so as to determine the maximum net
benefits for society as a whole
q When we do not consider the transmission cons-
traints, the CEM operator solves the optimization
problem to determine the successful offers/bids
b s
( p ) − ∑C ( p )
N N
∑B
b s
max S = j j i i
j =1 i =1
s.t. b s
N N
∑p ∑p
b s
j
= i supply – demand balance
j =1 i =1
ECE 573 © 2007 – 2017 George Gross, University of Illinois at Urbana-Champaign, All Rights Reserved. 16
MAXIMIZATION OF THE SOCIAL
WELFARE
$/MWh demand
curve
market
clearing mar
equi ket
price social welfare libri
poin um
t
supply curve
s1 ~ ~ s3 ~ s2
bus 1 f max
ℓ
= 800 MW bus 2
lossless line
b1 b2 b3
ECE 573 © 2007 – 2017 George Gross, University of Illinois at Urbana-Champaign, All Rights Reserved. 19
OFFERS AND BIDS
$/MWh $/MWh
6 10
s1 : 5 b1 :
5
300 700 MWh/h 400 600 MWh/h
$/MWh $/MWh
8 12
b2 :
s2 : 4
200 500 MWh/h 500 MWh/h
$/MWh $/MWh
9
s3 : b3 : 7
3
700 MWh/h MWh/h
600 800
ECE 573 © 2007 – 2017 George Gross, University of Illinois at Urbana-Champaign, All Rights Reserved. 20
SOCIAL WELFARE MAXIMIZATION
IGNORING TRANSMISSION
$/MWh
500 @ 12 200 @ 7
400 @ 10
600 @ 9
S = 9,300
ρ* = 7 300 @ 8
400 @ 6 200 @ 5
300 @ 5
200 @ 4
700 @ 3
MWh/h
1,600
ECE 573 © 2007 – 2017 George Gross, University of Illinois at Urbana-Champaign, All Rights Reserved. 21
MARKET OUTCOMES WITHOUT
TRANSMISSION CONSIDERATION
s1 s2 s3
7 $/MWh 700 MWh
700 MWh 200 MWh 7 $/MWh
7 $/MWh
CEM operator
s1 ~ lossless system ~ s2
b1 CEM b2
j 0.1 operator j 0.1
bus 3
max
f 13 = 200 MW
b3 ~ s3
ECE 573 © 2007 – 2017 George Gross, University of Illinois at Urbana-Champaign, All Rights Reserved. 24
THREE – BUS SYSTEM: OFFERS AND
BIDS
$/MWh $/MWh
10 70
b1 : 50
s1 : 5
800 @ 300
200 @ 80
200 @ 70
S = 265, 600 200 @ 60
100 @ 50
λ * = 29
400 @ 20
200 @ 10 100 @ 29
300 @ 5 200 @ 20
MWh/h
300 @ 10 1,500
ECE 573 © 2007 – 2017 George Gross, University of Illinois at Urbana-Champaign, All Rights Reserved. 26
UNCONSTRAINED SYSTEM
REVENUES AND PAYMENTS
ECE 573 © 2007 – 2017 George Gross, University of Illinois at Urbana-Champaign, All Rights Reserved. 27
BASIC MARKET PERFORMANCE
MEASURES
q The social welfare is a measure of the performance
player
m producer surplus
m consumer surplus
ECE 573 © 2007 – 2017 George Gross, University of Illinois at Urbana-Champaign, All Rights Reserved. 28
PRODUCER SURPLUS
q For a seller s at node i, the individual producer surplus
measures the difference between the revenues
received for the sale at the market clearing price and
those that would be received at the offer price
S is = ρ * ⋅ p s − C i p s
i
( ) i
i = 1, ... , N s
revenues
costs
under bid under offer
q The total producer surplus sis
N
S s
= ∑S s
i
i=1
ECE 573 © 2007 – 2017 George Gross, University of Illinois at Urbana-Champaign, All Rights Reserved. 29
CONSUMER SURPLUS
q For each buyer b at node j, the individual consumer
surplus measures the difference between the
payments for the commodity at the bid prices of
the buyer and those at the market clearing price
S b
j
= B j ( p bj ) − ρ * ⋅( p bj ) j = 1, ... , N b
actual payments
*
benefits under ρ
S s= S s
1
+S s
2
+S s
3
= 4, 400
500 @ 12
400 @ 10 200 @ 7
600 @ 9
300 @ 8
ρ* = 7 s
S s
S s S
400 @ 6 1 200 @ 5
3 2
300 @ 5
200 @ 4
700 @ 3
MWh/h
1,600
ECE 573 © 2007 – 2017 George Gross, University of Illinois at Urbana-Champaign, All Rights Reserved. 31
THE TWO – BUS SYSTEM EXAMPLE
$/MWh
b b b b
500 @ 12 S =S 1
+S 2
+S 3
= 4, 900
400 @ 10 200 @ 7
S b 600 @ 9
2 b
S 1 S b
* 3 300 @ 8
ρ =7
400 @ 6 200 @ 5
300 @ 5
200 @ 4
700 @ 3
MWh/h
1,600
ECE 573 © 2007 – 2017 George Gross, University of Illinois at Urbana-Champaign, All Rights Reserved. 32
THE TWO – BUS SYSTEM EXAMPLE
$/MWh
S = S s +S b
= 9, 300
500 @ 12
400 @ 10 200 @ 7
600 @ 9
300 @ 8
ρ* = 7
400 @ 6 200 @ 5
300 @ 5
200 @ 4
700 @ 3
MWh/h
1600
ECE 573 © 2007 – 2017 George Gross, University of Illinois at Urbana-Champaign, All Rights Reserved. 33
THREE – BUS SYSTEM: TRANSMISSION
UNCONSTRAINED EQUILIBRIUM
b
$/MWh S 2
800 @ 300 b
S 1
200 @ 80
b
S 3 200 @ 70
200 @ 60
100 @ 50
λ * = 29 s
S s S 2
400 @ 20
100 @ 29
s
1
200 @ 10
S 3
300 @ 5 200 @ 20
MWh/h
S s 300 @ 10 s 1,500
1 S 2
ECE 573 © 2007 – 2017 George Gross, University of Illinois at Urbana-Champaign, All Rights Reserved. 34
THREE – BUS SYSTEM: TRANSMISSION
UNCONSTRAINED DISPATCH
s1 ~ 600 MW ~ s2 600 MW
bus 1 bus 2
b1 CEM b2
300 MW 400 MW
lossless system
bus 3
f 13max = 200 MW
b3 ~ s3 λ *
= 29
800 MW 300 MW
ECE 573 © 2007 – 2017 George Gross, University of Illinois at Urbana-Champaign, All Rights Reserved. 35
A TRANSMISSION CONSTRAINT
s1 s3 s2
~ ~ ~
700 MW 700 MW 200 MW
bus 1 f max
ℓ
= 800 MW bus 2
ECE 573 © 2007 – 2017 George Gross, University of Illinois at Urbana-Champaign, All Rights Reserved. 36
IMPACTS OF THE TRANSMISSION
NETWORK IN THE TWO – BUS SYSTEM
To eliminate the line overload, we reduce the sale
from s 1 by 100 MW and instead use the higher –
priced energy from s 2
s1 s3 s2
~ ~ ~
700 – 200 200 + 100
700 MW = 300 MW
= 500 MW
bus 1 f ℓmax = 800 MW bus 2
b1 1,000 – 200 = 800 MW b 2 b3
!###"### $
400 MW actual flow 500 MW 700 – 100
= 600 MW
ECE 573 © 2007 – 2017 George Gross, University of Illinois at Urbana-Champaign, All Rights Reserved. 37
TWO – BUS EXAMPLE: DISCONNECTED
MARKETS AT BUSES 1 AND 2
bus 1 bus 2
$/MWh $/MWh
500 @ 12
400 @ 10 12
600 @ 9
200 @ 7
200 @ 5 300 @ 8
400 @ 6
300 @ 5 200 @ 4
3 700 @ 3
MWh/h MWh/h
ECE 573 © 2007 – 2017 George Gross, University of Illinois at Urbana-Champaign, All Rights Reserved. 38
TWO – BUS EXAMPLE: INTERCONNECTED
MARKETS AT BUSES 1 AND 2
f max
ℓ
= 800 MW
bus 1 bus 2
ECE 573 © 2007 – 2017 George Gross, University of Illinois at Urbana-Champaign, All Rights Reserved. 39
TWO – BUS EXAMPLE:
UNCONSTRAINED MARKET AT BUS 1
$/MWh
400 @ 10
200 @ 5 400 @ 6
300 @ 5
3
700 @ 3
MWh/h
ECE 573 © 2007 – 2017 George Gross, University of Illinois at Urbana-Champaign, All Rights Reserved. 40
TWO – BUS EXAMPLE: BUS – 2 – DEMAND
– MODIFIED MARKET AT BUS 1
$/MWh modified demand
includes bus 2
500 @ 12 demand
400 @ 10
600 @ 9
200 @ 7
400 @ 6 200 @ 5
300 @ 5
3
700 @ 3
MWh/h
ECE 573 © 2007 – 2017 George Gross, University of Illinois at Urbana-Champaign, All Rights Reserved. 41
TWO – BUS EXAMPLE: CONSTRAINED
MARKET AT BUS 1
$/MWh
500 @ 12
400 @ 10 line
300 @ 9
constraint
ρ *1 = 6 impact
200 @ 5
300 @ 5
3 400 @ 6
700 @ 3
MWh/h
1,200
ECE 573 © 2007 – 2017 George Gross, University of Illinois at Urbana-Champaign, All Rights Reserved. 42
TWO – BUS EXAMPLE:
UNCONSTRAINED MARKET AT BUS 2
$/MWh
500 @ 12
12
600 @ 9
300 @ 8 200 @ 7
200 @ 4
MWh/h
ECE 573 © 2007 – 2017 George Gross, University of Illinois at Urbana-Champaign, All Rights Reserved. 43
TWO – BUS EXAMPLE: BUS – 1 – SUPPLY
– MODIFIED MARKET AT BUS 2
$/MWh
500 @ 12
12
200 @ 7
600 @ 9
300 @ 8
400 @ 6
300 @ 5
200 @ 4 modified supply
300 @ 3
includes bus 1 supply
MWh/h
ECE 573 © 2007 – 2017 George Gross, University of Illinois at Urbana-Champaign, All Rights Reserved. 44
TWO – BUS EXAMPLE: CONSTRAINED
MARKET AT BUS 2
$/MWh
500 @ 12
12
200 @ 7
600 @ 9
ρ *2 = 8 300 @ 8
200 @ 6
300 @ 5 line constraint
200 @ 4
300 @ 3 impact
MWh/h
1,100
ECE 573 © 2007 – 2017 George Gross, University of Illinois at Urbana-Champaign, All Rights Reserved. 45
TRANSMISSION – CONSTRAINED
SETTLEMENT
quantity price revenues payments change w.r.t. the
participants
[MWh] [$/MWh] [$] [$] unconstrained case [$]
s1 500 6.0 3,000 - – 1,900
s2 300 8.0 2,400 - 1,000
s3 700 6.0 4,200 - – 700
b1 400 6.0 - 2,400 – 400
b2 500 8.0 - 4,000 500
b3 600 8.0 - 4,800 100
total 1,500 9,600 11,200 –
bus 1 bus 2
b1 IGO b2
300 MW 400 MW
lossless system
bus 3
max
f = 200 MW b3
λ* = 29
13
~ s3
800 MW 300 MW
ECE 573 © 2007 – 2017 George Gross, University of Illinois at Urbana-Champaign, All Rights Reserved. 47
THREE – BUS SYSTEM: PTDFs
1 MW 1/3 MW 1/3 MW 1 MW
1 2 1 2
3 3
1 MW 1 MW
ECE 573 © 2007 – 2017 George Gross, University of Illinois at Urbana-Champaign, All Rights Reserved. 48
THREE – BUS SYSTEM: TRANSMISSION
UNCONSTRAINED DISPATCH
s 1 ~ 600 MW ~ s 2 600 MW
b1 2 1 b2
300 MW f13 = × 300 + × 200
3 3 400 MW
violation of = 266.67
the line
constraint bus 3
since b3 λ *
= 29
~ s3
f 13 > f max
13 800 MW 300 MW
ECE 573 © 2007 – 2017 George Gross, University of Illinois at Urbana-Champaign, All Rights Reserved. 49
THREE – BUS SYSTEM: TRANSMISSION
UNCONSTRAINED DISPATCH
q The transmission unconstrained dispatch is infea-
max
sible because the line flow f 13 violates limit f 13
his load
ECE 573 © 2007 – 2017 George Gross, University of Illinois at Urbana-Champaign, All Rights Reserved. 50
SOCIAL WELFARE MAXIMIZATION
UNDER TRANSMISSION CONSTRAINTS
Nb Ns
max S = ∑ j j ∑ i i)
B ( p b
) − C ( p s
j =1 i =1
s.t. set of
g n ( p 1s, ! , p sN s ; p 1b, ! , p bN b ) = 0 ∀ node n power flow
equations
real
f ℓ ( p 1s, " , p sN s ; p 1b, " , p bN b ) ≤ f max
∀ line ℓ power
ℓ
line flow
limits
ECE 573 © 2007 – 2017 George Gross, University of Illinois at Urbana-Champaign, All Rights Reserved. 51
CONSTRAINED MARKET CLEARING
PRICES
q The equality constraints represent the balance of
binding or active :
fℓ ( p , " , p N s ; p , " , p N b ) = f
s
1
s b
1
b max
ℓ
(500 – 3x) MW
2 1 max
×(300 − x ) + ×(200 − 2 x ) = 200 MW f 13
3 3
ECE 573 © 2007 – 2017 George Gross, University of Illinois at Urbana-Champaign, All Rights Reserved. 56
THREE – BUS SYSTEM: REDISPATCH
OF SUPPLY
q x is the amount of redispatch due to the impacts
max
of the f 13 constraint on seller s 1
q 2x is the amount of redispatch due to the impacts
max
of the f 13 constraint on seller s 2
q Redispatch calculation obtains
2 1
× (300 – x ) + × (200 – 2 x ) = 200 MW
3 3
so that
x = 50 MW
ECE 573 © 2007 – 2017 George Gross, University of Illinois at Urbana-Champaign, All Rights Reserved. 57
THREE – BUS SYSTEM: REDISPATCH
OF SUPPLY
q Then, the IGO reduces the output of seller s 1 by 50
to buyer b 3
ECE 573 © 2007 – 2017 George Gross, University of Illinois at Urbana-Champaign, All Rights Reserved. 58
THREE – BUS SYSTEM: TRANSMISSION
CONSTRAINED DISPATCH
s 1 ~ 550 MW 500 MW ~ s2
bus 1 bus 2
b1 IGO b2
300 MW 400 MW
lossless system
bus 3
max
f 13 = 200 MW b3 ~ s3
800 MW 450 MW
ECE 573 © 2007 – 2017 George Gross, University of Illinois at Urbana-Champaign, All Rights Reserved. 59
THREE – BUS SYSTEM: LMPs
s 1 ~ 550 + 1 MW 500 + 1 MW ~ s 2
bus 1 λ *1 = 10 λ *2 = 20 bus 2
b1 IGO b2
300 + 1 MW 400 + 1 MW
lossless system
bus 3
b3 ~ s3
800 MW 450 MW
ECE 573 © 2007 – 2017 George Gross, University of Illinois at Urbana-Champaign, All Rights Reserved. 60
THREE – BUS SYSTEM: LMPs
s 1 ~ 550 +Δ P1 MW 500 +Δ P2 MW ~ s 2
bus 1 λ *1 = 10 λ *2 = 20 bus 2
b1 IGO b2
300 MW 400 MW
lossless system
Δ P1 + Δ P2 = 1
bus 3 2 1
λ *3 = ? ⋅ Δ P1 + ⋅ Δ P2 = 0
b3 3 3
~ s3
800 + 1 MW 450 MW Δ P1 = − 1 Δ P2 = 2
ECE 573 © 2007 – 2017 George Gross, University of Illinois at Urbana-Champaign, All Rights Reserved. 61
THREE – BUS SYSTEM: LMPs
s 1 ~ 550 MW 500 MW ~ s2
bus 1 λ *1 = 10 λ *2 = 20 bus 2
b1 IGO b2
300 MW 400 MW
lossless system
s 1 ~ 550 MW 500 MW ~ s2
bus 1 λ *1 = 10 λ *2 = 20 bus 2
b1 IGO b2
300 MW 400 MW
lossless system
max
bus 3 λ *3 = 30
f 13 = 200 MW b3 ~ s3
800 MW 450 MW
ECE 573 © 2007 – 2017 George Gross, University of Illinois at Urbana-Champaign, All Rights Reserved. 63
SOCIAL WELFARE AND SURPLUSES
$/MWh
consumer surplus
producer surplus
MWh/h
ECE 573 © 2007 – 2017 George Gross, University of Illinois at Urbana-Champaign, All Rights Reserved. 64
EVALUATION OF METRICS
$/MWh
consumer
surplus
ρb
congestion
rents
ρs
market dead –
efficiency weight
loss loss
producer MWh/h
surplus
ECE 573 © 2007 – 2017 George Gross, University of Illinois at Urbana-Champaign, All Rights Reserved. 65
IMPACTS OF CONGESTION
q Congestion in the system leads to a change from
the single market equilibrium point to different
nodal equilibrium points, possibly one at each node
q Change in the preferred schedule for the required
generation–demand balance may lead to possible
curtailment in production or in consumption or in
both
q The individual player surpluses change from the
unconstrained market values to those of the
separate markets at each bus under the
constrained conditions
ECE 573 © 2007 – 2017 George Gross, University of Illinois at Urbana-Champaign, All Rights Reserved. 66
OTHER LONGER – TERM
CONGESTION IMPACTS
q Increase in costs arising from the delayed
various players
q Measures of congestion impacts in $
m redispatch costs
m congestion rents
Nb Ns
ˆ ˆ s ˆ
S =S +S + b
∑ρ b
⋅ p − ∑ ρi ⋅ p s
j j i
j=1 i=1
congestion rents κ
ECE 573 © 2007 – 2017 George Gross, University of Illinois at Urbana-Champaign, All Rights Reserved. 69
CONGESTION RENTS
q In the constrained case, the congestion rents are
part of the social welfare
b s
S =S +S +κ
c c c
market constrained
efficiency social welfare
loss
E =− S ( c
−S
u )
q In economics, the market efficiency loss is also
known as deadweight loss
ECE 573 © 2007 – 2017 George Gross, University of Illinois at Urbana-Champaign, All Rights Reserved. 71
TWO-BUS SYSTEM: CONSTRAINED
CASE – BUS 1
$/MWh
consumer
500 @ 12
surplus
400 @ 10
300 @ 9
400 @ 6
ρ *1 = 6
300 @ 5 200 @ 5
700 @ 3 producer
surplus
MWh/h
ECE 573 © 2007 – 2017 George Gross, University of Illinois at Urbana-Champaign, All Rights Reserved. 72
TWO-BUS SYSTEM: CONSTRAINED
CASE – BUS 2
$/MWh
consumer
500 @ 12 surplus
200 @ 7
600 @ 9
*
ρ =8
2 300 @ 8
200 @ 6
300 @ 5
200 @ 4
300 @ 3 producer
surplus
MWh/h
ECE 573 © 2007 – 2017 George Gross, University of Illinois at Urbana-Champaign, All Rights Reserved. 73
TWO-BUS SYSTEM: CONSTRAINED
CASE
s1 300 b1 1,600
s2 800 b2 2,000
s3 2,100 b3 6000
ECE 573 © 2007 – 2017 George Gross, University of Illinois at Urbana-Champaign, All Rights Reserved. 74
TWO – BUS SYSTEM: MARKET
EFFICIENCY LOSS
q For the unconstrained case we have
S = 9, 300
u
S = 9, 000
c
E = − S ( c
−S
u ) = 300
ECE 573 © 2007 – 2017 George Gross, University of Illinois at Urbana-Champaign, All Rights Reserved. 75
THREE – BUS SYSTEM: MARKET
EFFICIENCY LOSS
q For the unconstrained case we have
S = 265, 600
u
S = 263,750
c
E =− S ( c
−S
u ) = 1, 850
ECE 573 © 2007 – 2017 George Gross, University of Illinois at Urbana-Champaign, All Rights Reserved. 76
THREE – BUS SYSTEM: CONSTRAINED
CASE
s1 1,500 b1 16,000
s2 2,000 b2 20,000
s3 2,250 b3 216,000
ECE 573 © 2007 – 2017 George Gross, University of Illinois at Urbana-Champaign, All Rights Reserved. 77
REDISPATCH COSTS
transmission constraints
ECE 573 © 2007 – 2017 George Gross, University of Illinois at Urbana-Champaign, All Rights Reserved. 78
IMPACTS OF THE TRANSMISSION
NETWORK IN THE TWO – BUS SYSTEM
To eliminate the line overload, we reduce the
sale from s 1 by 200 MW and instead use the
higher-priced energy from s 2
s1 s3 s2
~ ~ ~
700 - 200 200 + 100
700 MW
= 500 MW = 300 MW
bus 1 f max
ℓ
= 800 MW bus 2
b1 - 200 = 800 MW b b
1,000
!###"### $
2 3
300 @ 8
ρ* = 7
block must
be
300 @ 5 unloaded
200 @ 4 to prevent
400 @ 6
700 @ 3 congestion
– 200
MWh/h
1600
ECE 573 © 2007 – 2017 George Gross, University of Illinois at Urbana-Champaign, All Rights Reserved. 80
TWO – BUS SYSTEM: REDISPATCH
COSTS
$/MWh
redispatch costs = ( 8 – 6 ) ×100 = 200
300 @ 8
*
block must
ρ =7 be
dispatched
to avoid
violation of
300 @ 5 the line flow
200 @ 4 400 @ 6 limit
MWh/h
ECE 573 © 2007 – 2017 George Gross, University of Illinois at Urbana-Champaign, All Rights Reserved. 81
FACTORS CAUSING SOCIAL
WELFARE REDUCTION
q The redispatch of higher – priced blocks to
replace the output of the lower – priced generation
that could be used when the constraints are not
considered
q The decrease in market efficiency: the deadweight
loss in efficiency due to the fact that the output of
the displaced units cannot be used and sold
q The decrease in the producer surplus of some
sellers
ECE 573 © 2007 – 2017 George Gross, University of Illinois at Urbana-Champaign, All Rights Reserved. 82
FACTORS CAUSING SOCIAL
WELFARE REDUCTION
q The decrease in the consumer surplus of some
buyers
µ1 = 6 $ / MWh
ECE 573 © 2007 – 2017 George Gross, University of Illinois at Urbana-Champaign, All Rights Reserved. 86
LMPs FOR THE TWO – BUS SYSTEM
EXAMPLE
s1 s3 s2
~ ~ ~
500 MW [300 +1 ] MW
700 MW
@6 @8
µ2 = 8 $ / MWh
ECE 573 © 2007 – 2017 George Gross, University of Illinois at Urbana-Champaign, All Rights Reserved. 87
MARKET OUTCOMES INCORPORATING
TRANSMISSION CONSTRAINTS
s1 s2 s3
6 $/MWh 700 MWh
500 MWh 300 MWh 8 $/MWh
6 $/MWh
IGO
ECE 573 © 2007 – 2017 George Gross, University of Illinois at Urbana-Champaign, All Rights Reserved. 88
THREE – BUS SYSTEM: LMPs
s 1 ~ 550 MW 500 MW ~ s2
bus 1 λ *1 = 10 λ *2 = 20 bus 2
b1 IGO b2
300 MW 400 MW
lossless system
max
bus 3 λ *3 = 30
f 13 = 200 MW b3 ~ s3
800 MW 450 MW
ECE 573 © 2007 – 2017 George Gross, University of Illinois at Urbana-Champaign, All Rights Reserved. 89
SENSITIVITY STUDY IN THE TWO –
BUS SYSTEM
q We investigate the impacts of changing the offer
500 @ 12
200 @ 7
600 @ 9
* * 9
ρρρρ ====8.6
*
2 2* 8.4
8.2
8
300 @
300 @
300
98.6
@ 8.2
8.4
22 300 @ 8
200 @ 6
300 @ 5
200 @ 4
300 @ 3
MWh/h
ECE 573 © 2007 – 2017 George Gross, University of Illinois at Urbana-Champaign, All Rights Reserved. 91
SENSITIVITY STUDY IN THE TWO –
BUS SYSTEM
$/MWh
500 @ 12
200 @ 7
600 @ 9
ρ =9
*
2 300 @ 9
200 @ 6
300 @ 5
200 @ 4
300 @ 3
MWh/h
ECE 573 © 2007 – 2017 George Gross, University of Illinois at Urbana-Champaign, All Rights Reserved. 92
SENSITIVITY STUDY IN THE TWO –
BUS SYSTEM
limiting price
3000 S
$ S
Series1
2
2500
B
2000 S
Series2
2
1500
B
S
Series3
3
1000
congestion
500 rents
Series4
0 market
8 8.2 8.4 8.6 8.8 9 9.2 efficiency
Series5
$/MWh loss
ECE 573 © 2007 – 2017 George Gross, University of Illinois at Urbana-Champaign, All Rights Reserved. 93
SENSITIVITY STUDY IN THE
THREE – BUS SYSTEM
q We investigate the impacts of changing the offer
60 350
50 300
40 250
$/MWh
200
10 3 $
30
150
20 interval where
λ *3 increasing price 100
10 does not impact the
S s 50
3 market outcomes
0 0
30 70 110 150 190 230 270 310
offer for the last block of seller s3 in $/MWh
ECE 573 © 2007 – 2017 George Gross, University of Illinois at Urbana-Champaign, All Rights Reserved. 95
THREE – BUS SYSTEM: PRODUCER
SURPLUS
90
80
s
70 S 2
60
10 3 $
50
40 s
S 3
30
20 S s
1
10
200 b
S 3
150
10 3 $
100
b b
S S
50 2 1
0
30 70 110 150 190 230 270 310
offer for the last block of seller s3 in $/MWh
ECE 573 © 2007 – 2017 George Gross, University of Illinois at Urbana-Champaign, All Rights Reserved. 97
THREE – BUS SYSTEM: MARKET
PERFORMANCE MEASURES
100
90
80 congestion rents
70
60
market
10 3 $
50
efficiency losses
40
30
20
10
0
30 70 110 150 190 230 270 310
offer for the last block of seller s3 in $/MWh
ECE 573 © 2007 – 2017 George Gross, University of Illinois at Urbana-Champaign, All Rights Reserved. 98
LOCAL MARKET POWER
q Market power is the ability of a firm to profitably
raise the price of a product
q The exercise of market power may be carried out
by:
m the physical withholding of units
m the financial withholding of units
q Transmission constraints may create locational
market power since they may set up area markets
with limited importing capability
ECE 573 © 2007 – 2017 George Gross, University of Illinois at Urbana-Champaign, All Rights Reserved. 99
SIMULATION STUDIES
3 6 7
~ ~
ECE 573 © 2007 – 2017 George Gross, University of Illinois at Urbana-Champaign, All Rights Reserved. 101
SEVEN – BUS SYSTEM: PRODUCER
SURPLUS
100
impacts of
changes in the
80 flow directions
S 3S
60
10 3 $
S 4S
40 S
S7
20
S1 S S 5S
0
0 100 200 300 400 500 600
offer for the last block of seller s3 in $/MWh
ECE 573 © 2007 – 2017 George Gross, University of Illinois at Urbana-Champaign, All Rights Reserved. 102
SEVEN – BUS SYSTEM: CONSUMER
SURPLUS
350
300
250
10 3 $
200
150 b
S 3
100
50
0
300
congestion rents
250
200 market
10 3 $
150 efficiency
losses
100
50
~ 16
2
12
~ 1
17
~
3 ~
4
15
14
18
5
45
44
46 13
~
6 19 11
38
47
20
48 49 50
21
39 57
42 43 51
22 37
10
26
23
40 56 41
36
27 24
35
28 25 9
~
7 34
29 30
31
~ 32
8
52 53 54 33 55
ECE 573 © 2007 – 2017 George Gross, University of Illinois at Urbana-Champaign, All Rights Reserved. 105
THE IEEE 57 – BUS SYSTEM:
CONGESTION RENTS
10000
!
!
!
⌣!⌢
8000
6000
$
4000
2000
0
0 0.5 1 1.5 2 2.5 3
7000
6000
5000
4000
$
3000
2000
1000
0
0 0.5 1 1.5 2 2.5 3
offer of seller s2 in $/MWh
ECE 573 © 2007 – 2017 George Gross, University of Illinois at Urbana-Champaign, All Rights Reserved. 107
SIMULATION RESULTS
q Congestion situations produce, typically,
bilateral trades
scheduling
W ! {ω 1 , ω 2 ,!, ωW }
ECE 573 © 2007 – 2017 George Gross, University of Illinois at Urbana-Champaign, All Rights Reserved. 113
SOCIAL WELFARE MAXIMIZATION
UNDER TRANSACTIONS
q We incorporate the basic transactions into the
transmission-constrained social welfare
optimization problem statement
Nb Ns
max S = ∑ B j ( pB ) − ∑ C i ( pS )
j i
j=1 i=1
s.t.
max S = ∑ B j ( pB ) − ∑ C i ( pS )
j i
j=1 i=1
s.t.
bus 1 bus 2
100 MW
50 MW ω 2 = {2,1,50 MW } 50 MW
bus 1 bus 2
50 MW
ECE 573 © 2007 – 2017 George Gross, University of Illinois at Urbana-Champaign, All Rights Reserved. 117
r
EVALUATION OF f IN THE ℓ
TWO – BUS EXAMPLE
f ℓmax = 800 MW
bus 1 bus 2
100 MW
50 MW
s1 s3 s2
~ ~ ~
500 - 50 300 + 50
700 MW = 350 MW
= 450 MW
f ℓr =750 MW
bus 1 bus 2
800 – 50 = 750 MW
b1 b2 b3
400 MW 500 MW 600 MW
ECE 573 © 2007 – 2017 George Gross, University of Illinois at Urbana-Champaign, All Rights Reserved. 119
MARKET OUTCOMES
INCORPORATING TRANSACTIONS
s1 s2 s3
6 $/MWh 700 MWh
450 MWh 350 MWh 8 $/MWh
6 $/MWh
IGO
S = 8, 950
ECE 573 © 2007 – 2017 George Gross, University of Illinois at Urbana-Champaign, All Rights Reserved. 120
IMPACTS OF TRANSACTIONS
settlement time
ECE 573 © 2007 – 2017 George Gross, University of Illinois at Urbana-Champaign, All Rights Reserved. 123
CONGESTION CHARGES IN THE TWO
– BUS SYSTEM EXAMPLE
q The line is congested in the base case and
IGO pays
m ω 2 = {2,1,50 MW } :
$ 100 to ω 2
(µ 1
− µ 2 ) t2 = (6 - 8) i 50 = − $ 100
ECE 573 © 2007 – 2017 George Gross, University of Illinois at Urbana-Champaign, All Rights Reserved. 124
BILATERAL TRANSACTIONS
h-year
seller
MWh
$ bilateral
transactions h-month
buyer
CEM seller
bilateral
seller
IGO
bilateral
buyer
CEM buyer
transmission constraints
ECE 573 © 2007 – 2017 George Gross, University of Illinois at Urbana-Champaign, All Rights Reserved. 126
A NEW SCHEDULING TOOL
q We develop a comprehensive formulation for the
transmission service provision that we call the
transmission allocation problem (TAP)
q In the formulation, we
m explicitly represent the contributions of each
bilateral transaction to the social welfare
m consider the provision of transmission
services to all the customers on a consistent
basis in conformance with the willingness to
pay of each customer
ECE 573 © 2007 – 2017 George Gross, University of Illinois at Urbana-Champaign, All Rights Reserved. 127
TAP FORMULATION
welfare
of bilateral transactions
ECE 573 © 2007 – 2017 George Gross, University of Illinois at Urbana-Champaign, All Rights Reserved. 128
TRANSMISSION REQUESTS OF THE
BILATERAL TRANSACTIONS
seller at
ω ! m ,n ,t
w
{ w w w
} node m w
IGO tw
α w ( t w ) ≥ 0 for buyer at
node n w
t w ∈ ⎡⎣ 0, t w ⎤⎦
bilateral
willingness to transaction w
pay
ECE 573 © 2007 – 2017 George Gross, University of Illinois at Urbana-Champaign, All Rights Reserved. 129
FORMULATION OF THE TAP
q We modify the social welfare expression to include
the contributions of the bilateral transactions
s.t. p 0s − p 0b + pW0 = b T0 θ ↔ µ0
p s − p b + pW = B θ ↔ µ
max
Bd A θ ≤ f
s s
impacts of the 0≤ p ≤ p n n
n = 0,1, 2, ... , N
bilateral
0 ≤ pnb ≤ pnb n = 0,1, 2, ... , N
transactions
0 ≤ tw ≤ t w w = 1, 2, ... ,W
ECE 573 © 2007 – 2017 George Gross, University of Illinois at Urbana-Champaign, All Rights Reserved. 131
THE TAP SOLUTION
IGO
µ*j p s*j buyer at
(µ *
n −µ *
m )t w*
node m
p b*j
pool buyer
at node j congestion
charges
ECE 573 © 2007 – 2017 George Gross, University of Illinois at Urbana-Champaign, All Rights Reserved. 132
TAP : KEY ADVANTAGES
q Transmission services are allocated in accordance
with each transaction’s willingness to pay: the
limited transfer capabilities are used so as to
35
30
25
20
15
10
5
0
1 2 3 4 5 6 7 8 9 10 11 12 13 14
node n
ECE 573 © 2007 – 2017 George Gross, University of Illinois at Urbana-Champaign, All Rights Reserved. 135
LMP DEVIATION AS A FUNCTION OF
TRANSACTION VOLUME
1.4 µ n* − µ n*
TAP TSP
εµ =* bus 35
n
µ n*
1.2 TAP
bus 7
1
bus 54
IEEE 57 – bus system
0.8
εµ *
n
0.6 bus 12
bus 3
0.4 bus 28
bus 43
0.2
0
0 10 20 30 40 50 60 70
percentage of total load supplied by transactions
ECE 573 © 2007 – 2017 George Gross, University of Illinois at Urbana-Champaign, All Rights Reserved. 136
SOCIAL WELFARE DEVIATION AS A
FUNCTION OF TRANSACTION VOLUME
* *
S −S
TAP TSP
εS * = *
S
TAP
εS *
0 10 20 30 40 50 60
percentage of total load supplied by transactions
ECE 573 © 2007 – 2017 George Gross, University of Illinois at Urbana-Champaign, All Rights Reserved. 137
THE TAP
transmission
bilateral constraints energy sale
transmission quantities
requests for pool
customers
sellers’
offers
TAP LMPs
transmission
buyers’ services
bids for bilateral
transactions
ECE 573 © 2007 – 2017 George Gross, University of Illinois at Urbana-Champaign, All Rights Reserved. 138
UNCERTAINTY IN THE CONGESTION
CHARGES FROM TRANSACTIONS
h-year
congestion
charges
100
80
60
40
20
0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
-20
χ ! µ n* − µ m* γ s
( )
χ FTR holder
receives from
γ IGO
* *
µ −µn m
FTR holder
pays to IGO
ECE 573 © 2007 – 2017 George Gross, University of Illinois at Urbana-Champaign, All Rights Reserved. 142
THE FTR MARKETS
h-year
FTR issuance
h-month
simultaneous
IGO feasibility
test
Γ1 =
{m1 , n1 , γ 1 , ρ1 }
ΓK γ K ρK
γ 1 ρ1 Γk
γ k ρk
ECE 573 © 2007 – 2017 George Gross, University of Illinois at Urbana-Champaign, All Rights Reserved. 144
THE SIMULTANEOUS FEASIBILITY
TEST (SFT)
FTR Γ1 ... Γk ... ΓK
fictitious
ω! 1 ω! k ! {mk , nk , γ k } ω! K
transactions
w" 1 w" k
PTDFs { }ϕℓ { } ϕ ℓ { } "
ϕ ℓwK
K
w" k
≤ f ℓmax yes
∑γ k ϕ ℓ feasible
k=1
∀ ℓ∈ L
no
infeasible
ECE 573 © 2007 – 2017 George Gross, University of Illinois at Urbana-Champaign, All Rights Reserved. 145
BASIC REQUIREMENTS FOR THE
FRAMEWORK
q Simplicity : applications to the electricity market
analysis requires the models in the framework to
be simple and easy to understand
q Comprehensiveness : close interactions among
the various elements of the problem requires
detailed representation of all of the three aspects
q Flexibility : the lack of a standard market design
and the evolution of the electricity markets
requires the framework to be able to accommodate
different market designs
ECE 573 © 2007 – 2017 George Gross, University of Illinois at Urbana-Champaign, All Rights Reserved. 146
CONSTRUCTION OF THE FRAMEWORK
ECE 573 © 2007 – 2017 George Gross, University of Illinois at Urbana-Champaign, All Rights Reserved. 147
CONSTRUCTION OF THE
FRAMEWORK
IGO
MWh $ $ MWh $
MWh
ECE 573 © 2007 – 2017 George Gross, University of Illinois at Urbana-Champaign, All Rights Reserved. 148
CONSTRUCTION OF THE FRAMEWORK
ECE 573 © 2007 – 2017 George Gross, University of Illinois at Urbana-Champaign, All Rights Reserved. 149
CONSTRUCTION OF THE
FRAMEWORK
ECE 573 © 2007 – 2017 George Gross, University of Illinois at Urbana-Champaign, All Rights Reserved. 150
THE INFORMATION FLOWS
ECE 573 © 2007 – 2017 George Gross, University of Illinois at Urbana-Champaign, All Rights Reserved. 151
COMPUTATIONAL ISSUES
q To provide computational simplicity, we adopt the
DC power flow assumptions and use the
distribution factors in the implementation
q Such approximation may result in errors on the
outcomes of various markets
m these factors are linear approximation of the
actual sensitivities
m the factors calculated for the base case are
often used to evaluate impacts in changed
system conditions
ECE 573 © 2007 – 2017 George Gross, University of Illinois at Urbana-Champaign, All Rights Reserved. 152
ILLUSTRATIVE APPLICATIONS OF
THE FRAMEWORK
q Assurance of the IGO revenue adequacy
q Design of FTR markets with enhanced liquidity
m new trading rules based on payoff parity
generation
ECE 573 © 2007 – 2017 George Gross, University of Illinois at Urbana-Champaign, All Rights Reserved. 153
ASSURANCE OF IGO REVENUE
ADEQUACY
pool
customers A = merchandising surplus
of the pool market
FTR
IGO holders
C = FTR
payoffs
bilateral B = congestion
transactions charges
ECE 573 © 2007 – 2017 George Gross, University of Illinois at Urbana-Champaign, All Rights Reserved. 154
IGO REVENUE ADEQUACY THEOREM
q Given that
q The inequality
A+ B ≥ C
REFERENCES