Transcript game theory - University of Wisconsin–Madison
TRANSMISSION PLANNING AND INVESTMENT IN THE COMPETITIVE ENVIRONMENT
PS ERC Seminar Presentation by George Gross Department Of Electrical and Computer Engineering University of Illinois at Urbana – Champaign April 5, 2005 © 2005, George Gross, UIUC
OUTLINE
The changed utilization of transmission
Planning in the competitive environment
The sorry state of transmission investment
Key challenges and complexities
An analytic framework for transmission investment
Illustrative examples
Concluding remarks © 2005, George Gross, UIUC 2
OPEN ACCESS IMPACTS
Power system restructuring fosters the development of competition in wholesale electricity markets
Markets bring about changes in the way power systems are operated and planned
The vertically integrated structure is slowly disintegrating into many new parts
New structures and players have important roles and result in decentralized decision making © 2005, George Gross, UIUC 3
THE VERTICALLY INTEGRATED UTILITY INDUSTRY STRUCTURE
customers IPP © 2005, George Gross, UIUC self generation 4
THE VERTICALLY INTEGRATED UTILITY INDUSTRY STRUCTURE
IPP generation Generation self generation ion nsmiss transmission Tra ution distribution Distrib ice Serv customer service Customer customers © 2005, George Gross, UIUC 5
VERTICALLY INTEGRATED UTILITY STRUCTURE IS DISINTEGRATING
customer service distribution transmission generation 6 © 2005, George Gross, UIUC
CENTRALITY OF TRANSMISSION IN RESTRUCTURING
A common thread in the restructuring of electricity around the globe is the unbundling of transmission from the generation and the distribution of sectors
The role of transmission in evolving wholesale competition in electricity is critical
The provision of the nondiscriminatory transmission access and services to all market players under the open access transmission regime entails the establishment of independent transmission entities © 2005, George Gross, UIUC 7
PLANNING UNDER COMPETITION
Major shift in the planning paradigm
cessation of the centralized integrated planning of the past
role of regional planning under the independent grid operator
unclear responsibility for implementation under the ownership/control separation
role of decentralized decision making © 2005, George Gross, UIUC 8
PLANNING UNDER COMPETITION
Planning, to the extent it is performed in the new environment, is an asset management problem
investment under uncertainty
critical importance of effective risk management
subject to regulations in a continuous state of flux 9 © 2005, George Gross, UIUC
TRANSMISSION USAGE UNDER COMPETITION
Frequent congestion situations result whenever too many customers compete for transmission services that the grid is capable of providing
Despite the more intense utilization of the grid by the many established and new players, develop ments in transmission planning have failed to keep pace with the increases in demand © 2005, George Gross, UIUC 10
THE SORRY STATE OF TRANSMISSION INVESTMENT
As demand increases, significant additions of new generation are being made in virtually every region
The reserve margins in capacity are improving year after year
Transmission investments have failed to keep up with the increases in demand and the additions in new generation © 2005, George Gross, UIUC 11
DEMAND AND TRANSMISSION CAPACITY GROWTH
30
%
25 20 15 10 5
0
Source: EPRI 1988 – 98 1999 – 09 © 2005, George Gross, UIUC electricity demand transmission capacity expansion 12
THE NERC CAPACITY MARGIN FORECASTS 25 2002 20 15 2000 10 1999 5 1999 2001 2003 2005
year
Source: NERC Reliability Assessment, 2002
– 2011
© 2005, George Gross, UIUC 2007 2009 2001 2011 13
PROJECTED GENERATION GROWTH IN 1998 – 2007
Each percentage is with respect to the 1998 installed capacity Source: EPRI © 2005, George Gross, UIUC change in % 40 and higher 20 to 40 0 to 20 14
TRANSMISSION MAINTENANCE SPENDING
© 2005, George Gross, UIUC 16
230 kV AND ABOVE TRANSMISSION +2.2% +2.7% < .49% / yr 218.2
213.5
207.9
2003 Source: NERC 2004 2004-2008 © 2005, George Gross, UIUC 2009-2013 17
SEVERE STRESSING OF THE GRID
Large number of new and existing players Proliferation in the number of transactions
Increasing load demand Simultaneous accommodation of pool and bilateral transactions
Markedly different and more intense utilization of the grid than in the way that it was planned and designed
Low level of investment in transmission improvement © 2005, George Gross, UIUC 18
SEVERE STRESSING OF THE GRID
Severe stressing of the grid leads to frequent congestion situations with customers competing for the scarce and heavily constrained transmis sion services
The transmission-bottleneck-caused congestion situations significantly impact both the reliability and the economics of electricity supply © 2005, George Gross, UIUC 19
TRANSMISSION BOTTLENECKS: WESTERN INTERCONNECTION size of transmission paths 1 GW
< 1 GW
3 GW
>
3 GW percentage of hours congested 50% and greater 40% to 49% 30% to 39% 20% to 29% 10% to 19% 20
TRANSMISSION BOTTLENECKS: EASTERN INTERCONNECTION size of transmission paths 1 GW
< 1 GW
3 GW
>
3 GW percentage of hours congested 80% and greater 60% to 79% 40% to 59% 20% to 39% 10% to 19% 21
CONGESTION IMPACTS
Decreased reliability
Reduced competition
Increased consumer prices
Creation of enhanced opportunities for market power exercise
Increased infrastructure vulnerability © 2005, George Gross, UIUC 22
CONGESTION : ECONOMIC SIGNALS
LMP
s provide short-term congestion signals
The translation of
LMP
s into long-term investment signals is complicated
LMP
s create the need for the effective integration of financial hedging instruments:
FTR
s and flowgate rights © 2005, George Gross, UIUC 23
TRANSMISSION EXPANSION
Network expansion is by its very nature a very complex multi-period and multi-objective optimi zation problem
Its nonlinear nature and the inherent uncertainty in future developments constitute major compli cations 24 © 2005, George Gross, UIUC
TRANSMISSION INVESTMENT : KEY BARRIERS
Transmission is a regulated service: tariffs are cost based and not value based
Uncertainty about the recovery of transmission investments due to
long-term revenue stream needs
lack of clarity in regulatory pricing policy © 2005, George Gross, UIUC 25
TRANSMISSION INVESTMENT : KEY BARRIERS
conflicting goals of federal and state regulators
Difficulty of recovering investment costs due to free rider problem
Organizational complexities in the new industry structure © 2005, George Gross, UIUC 26
COMPLICATIONS IN TRANSMISSION EXPANSION
Every transmission improvement impacts the transfer capabilities in the interconnected network covering a large geographic region
Each transmission investment affects market participants differently
Free rider problem creates a problem in the investment recovery
Lumpiness of transmission investments is a key complication 27 © 2005, George Gross, UIUC
COMPLICATIONS IN TRANSMISSION EXPANSION
A long-time horizon with the sequence of appropriate decisions needs to be considered
Economies of scale encourage overbuilding
Imperfect electrical markets provide opportunities for market power exercise © 2005, George Gross, UIUC 28
COMPLICATIONS IN TRANSMISSION EXPANSION
Short-run marginal costing information from the hourly
LMP
s need to be translated into long-run marginal cost for investment decisions
FTR / FGR
integration into the investment decision is needed
The explicit consideration of wide ranges of uncertainty in all aspects, including regulatory, environmental and player behavior, is required © 2005, George Gross, UIUC 29
ANALYTIC FRAMEWORK
A four-layer structure consisting of
physical
commodity market
financial
investment layers
The interrelationships between layers represen ted through information flows © 2005, George Gross, UIUC 30
THE FRAMEWORK STRUCTURE
investment layer financial market layer commodity market layer physical network layer © 2005, George Gross, UIUC 31
THE PHYSICAL LAYER
Represents the physical flows in the transmission network including real power line flows, nodal injections and physical network/operational constraints
Models congestion and allows the evaluation of congestion impacts on the transmission customers/market participants © 2005, George Gross, UIUC 32
THE COMMODITY MARKET LAYER
Models the purchases/sales in both the day ahead hourly and the bilateral transaction markets
Represents the RTO decision making process to establish feasible transmission schedules
Interacts with the physical layer and the financial layer through information transfers © 2005, George Gross, UIUC 33
THE FINANCIAL LAYER
Models the financial instruments used to provide hedging against congestion changes
Models Financial Transmission Rights (
FTR
) and flowgate rights
Represents the salient aspects of rights issuance and trading © 2005, George Gross, UIUC 34
TRANSMISSION INVESTMENT LAYER
Models the transmission investment decision making process and determines the
location
quantity
timing of the transmission assets
Evaluates the impacts of the investment decisions on the investor, system operator and the transmission customers and assesses their financial aspects 35 © 2005, George Gross, UIUC
THE INFORMATION FLOWS investment layer social welfare financial market layer
SFT
result
LMP s
feasible
FTR
commodity market layer system states market outcomes physical network layer © 2005, George Gross, UIUC topology change desired
FTR
36
RTO TRANSMISSION PLANNING PROBLEM FORMULATION
Maximize aggregate social welfare:
pool
bilateral contracts subject to:
power flow balance equations
line flow equations
generator and demand limits
line flow limits © 2005, George Gross, UIUC 37
BASIC PROBLEM FORMULATION
max S
8760
h
1
S h s.t.
S
h
N
n b
(
p n b
)
n s
(
p n s
)
W
w p 0 s
p 0 b
p 0
p s
p b
p
b 0 T
B
B A d
f
max
0
Note: all parameters and variables are hourly quantities © 2005, George Gross, UIUC 38
EVALUATION OF METRICS
$
/
MWh
consumer surplus
B
S
producer surplus
congestion rents
market efficie ncy loss dead weight loss
MWh/h
© 2005, George Gross, UIUC 39
APPROPRIATE METRICS FOR TRANSMISSION INVESTMENT
RTO metrics:
social welfare: aggregated value
loss of efficiency: decrease in social welfare due to transmission constraints
congestion rents: money collected by the system operator because of congestion © 2005, George Gross, UIUC 40
APPROPRIATE METRICS FOR TRANSMISSION INVESTMENT
Producer metrics:
producer surplus: difference between what the producer collects from the system and the real costs
redispatch costs: difference in the produ cers’ costs with and without congestion © 2005, George Gross, UIUC 41
APPROPRIATE METRICS FOR TRANSMISSION INVESTMENT
Consumer metrics :
consumer surplus: difference between the demand bids and the demand payments
load payment costs: difference in demand payments with and without congestion © 2005, George Gross, UIUC 42
THREE – BUS SYSTEM EXAMPLE
One-hour horizon
Lossless network
Quadratic functions for the costs and benefits
No bilateral transactions © 2005, George Gross, UIUC 43
1
S 1
~
B 1
NETWORK TOPOLOGY
~
S 2
2
B 2 lossless system B 3
3 ~
S 3
© 2005, George Gross, UIUC 44
NETWORK DESCRIPTION
line
l
= (i, j) with
i
1
j
2 1 2 3 3
x
l
( p.u.) 0.1
0.1
0.1
f max
l
( MW ) 300 300 300 © 2005, George Gross, UIUC 45
OFFER REPRESENTATION
Cost function:
C s i
s i P s i
0.5
s i s i
2
Offer function:
s i s i
s i
s i P s i
© 2005, George Gross, UIUC 46
OFFER DATA
i
1 2 3
s i
( $/MWh ) 3.0
4.5
4.0
s i
[( $/MWh ) 2 h] 0.001
0.005
0.003
(
p s i
)
max
( MWh/h ) 1000 1000 1000 © 2005, George Gross, UIUC 47
OFFER PARAMETERS $/MWh generator offer
s i
s i MWh/h
© 2005, George Gross, UIUC 48
BID REPRESENTATION
Benefit function:
B b j P b j
b j P b j
0.5
b j P b j
2
Bid function:
v b j P b j
b j
b j P b j
© 2005, George Gross, UIUC 49
BID DATA
i
1 2 3
b j
( $/MWh ) 13 23 16
b j
[( $/MWh ) 2 h] 0.0150
0.0200
0.0150
( p
b j
)
max
( MWh/h ) 1000 1000 1000 © 2005, George Gross, UIUC 50
$/MWh
b j
BID PARAMETERS
b j
demand bid © 2005, George Gross, UIUC
MWh/h
51
PRE
–
EXPANSION RESULTS
metric total producer surplus total consumer surplus congestion rents social welfare value in
$
761.98
6632.01
520.67
7914.66
total production = 1056.57 MW © 2005, George Gross, UIUC 52
POST
–
EXPANSION RESULTS
metric total producer surplus total consumer surplus congestion rents social welfare total production = MW 1092.60
© 2005, George Gross, UIUC value in
$
880.24
7150.03
163.83
8194.10
53
PRE – AND POST – COMPARISON
pre-expansion power generated producer 1 2 (MW) 593.75
142.31
3 320.51
pre-expansion consumer power demanded (MW) 1 2 3 293.75
426.92
335.90
surplus ($) 352.54
101.26
308.19
surplus ($) 1294.34
3645.27
1692.41
© 2005, George Gross, UIUC post-expansion power generated (MW) 898.44
90.00
440.00
104.17
379.17
post-expansion power demanded surplus (MW) 273.44
440.00
379.17
surplus ($) 273.44
($) 1121.52
3872.00
2156.51
54
PRE – AND POST – COMPARISON
metric pre-expansion post-expansion total producer surplus (
$
) 761.98
total consumer surplus (
$
) 6632.01
congestion rents (
$
) 520.67
social welfare (
$
) 7914.66
total production (
MW
) 1056.57
© 2005, George Gross, UIUC 880.24
7150.03
163.83
8194.10
1092.60
55
MULTI – PERIOD ANALYSIS
topology change investment layer social welfare market outcomes financial market layer
. . .
social welfare market outcomes financial market layer
LMP s SFT
feasible
FTR LMP s
feasible
FTR
commodity market
. . .
commodity market system states market outcomes physical network system
. . .
states market outcomes physical network desired
FTR
operational period 1
. . .
© 2005, George Gross, UIUC operational period
H
56
IEEE RTS SEVEN – BUS NETWORK EXAMPLE
Study horizon of one year; typical week day and week end day for each of four seasons
Lossless network
Quadratic functions representation for costs and benefits
No bilateral transactions
Hourly computations © 2005, George Gross, UIUC 57
STUDY SCENARIOS
Reference scenario: the pre-expansion system
Scenario 1 : addition of line ( 3 , 4 )
Scenario 2 : addition of line ( 5 , 6 )
Scenario 3 : addition of lines ( 3 , 4 ) and ( 5 , 6 ) © 2005, George Gross, UIUC 58
NETWORK TOPOLOGY
S
1
~ B
1 bus 1 bus 2
B
2 bus 3
B
3
~ S
2
B
4
~ S
3 bus 4 bus 5
B
5
~ S
4 bus 6
B
6 © 2005, George Gross, UIUC bus 7
B
7
~ S
5 59
NETWORK DESCRIPTION 5 6 3 4 5 1 2 3
line
l
= ( i, j )
with i
1 2
j
3 4 4
x
l
(
p.u.
0.0576
0.0920
0.0586
0.1008
6 5 6 0.1720
0.0625
0.1610
7 7 0.0850
© 2005, George Gross, UIUC 0.0856
)
f max
l
(
p.u.
) 300 200 300 150 300 300 300 300 200 60
i
1 2 3 4 5 OFFER DATA
s i
3.5
s i
0.002
5.0
4.5
3.8
0.005
0.003
0.004
3.8
0.004
© 2005, George Gross, UIUC ( p )
max
1000 1000 1000 1000 1000 61
5 6 7
i
1 2 3 4 BID DATA
b j
20 21 50 20 28 20 27 © 2005, George Gross, UIUC
b j
0.015
0.018
0.022
0.010
0.017
0.016
0.015
( p )
max
1000 1000 1000 1000 1000 1000 1000 62
ANNUAL RTO METRICS
scenario
reference
1 2 3 social welfare 305,101.73
loss of efficiency (
k$
) congestion rents 6,679.58
7,664.69
308,204.19
305,975.03
3,577.12
5,806.28
308,799.57
2,981.74
© 2005, George Gross, UIUC 8,715.52
4,939.40
5,179.23
63
ANNUAL PRODUCER AND CONSUMER METRICS
scenario producer surplus consumer surplus (
k$
)
reference
1 2 27,363.09
27,503.96
28,706.49
27,0073.95
27,1984.71
27,2329.14
3 30,005.20
© 2005, George Gross, UIUC 27,3615.14
64
$ $
AGGREGATE METRICS FOR A SUMMER WEEKDAY
$ $
© 2005, George Gross, UIUC 65
NODAL PRICES FOR A SUMMER WEEKDAY
nodal prices, reference scenario
nodal prices, scenario 1 nodal prices, scenario 2 nodal prices, scenario 3 © 2005, George Gross, UIUC 66
NODAL PRICE DIFFERENCES FOR A SUMMER WEEKDAY
nodal price differences, scenario 1
nodal price differences, reference scenario
nodal price differences, scenario 2 nodal price differences, scenario 3 © 2005, George Gross, UIUC 67
SEVEN – BUS SYSTEM RESULTS
Best overall solution is scenario 3 with the lines ( 3, 4 ) and ( 5, 6 ) added
Scenario 1 results in the highest congestion results
Scenarios 2 and 3 are characterized by flat nodal price differences and lower average
LMP
s than in the reference scenario and scenario 1 © 2005, George Gross, UIUC 68
CONCLUDING REMARKS
Multi-layer analytic framework for transmission expansion planning
Framework capability to deal with the complex issues in transmission investment
Appropriate metrics to determine the best investment policy
Scenario analysis allows the identification of optimal strategy and investigation of
what if
questions © 2005, George Gross, UIUC 69
FUTURE WORK
Transmission service pricing on a value rather than cost basis
Formulation of effective incentives for transmis sion investment
The formulation and solution of the individual investor problem © 2005, George Gross, UIUC 70