Coordination of the location and time of investment of generation and transmission in a liberalised power system Jean-Michel Glachant (and Vincent Rious) Loyola de Palacio Chair & Florence School of Regulation European University Institute in Florence jean- [email protected]30 th July 2010 11th ACCC Regulatory Conference: Surfers Paradise Robert Schuman Centre for Advanced Studies Florence School of Regulation & Loyola de Palacio Chair
43
Embed
Coordination of the location and time of investment of generation and transmission in a liberalised power system Jean-Michel Glachant (and Vincent Rious)
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Coordination of the location and time of investment of generation and transmission in
a liberalised power system
Jean-Michel Glachant (and Vincent Rious)
Loyola de Palacio Chair & Florence School of Regulation
Scope for maximisation of social welfare?Model and results
Conclusion #1
FS
R a
nd
Loyola
de P
ala
cio
Ch
air o
n E
nerg
y P
olic
y
Coordination with locational signals?
• No optimal coordination with locational signals• Nodal pricing + average participation
tariff– Because of lumpiness of transmission investment– Even if improved coordination
(so locational signals are needed)– And ‘average participation’ tariff more efficient than
nodal pricing for efficient location of investment
FS
R a
nd
Loyola
de P
ala
cio
Ch
air o
n E
nerg
y P
olic
y
Coordination with locational signals?
• These signals = transmit information But limited to current grid and its current use
• And (+) many other locational constraints existing for generators: access to water, to fuel, to land, to social acceptance (NIMBY)
lasting congestions need to develop the network
FS
R a
nd
Loyola
de P
ala
cio
Ch
air o
n E
nerg
y P
olic
y
Outline
• Need for coordination between G and T
• Tool #1 = locational signals
• Tool #2 = anticipation
Scope for maximisation of social welfare?Model and results
• Conclusion
FS
R a
nd
Loyola
de P
ala
cio
Ch
air o
n E
nerg
y P
olic
y
Two alternative behaviors for the TSO
• Reactive behavior– Waits for generation connection request to study
the need of transmission investments
• Proactive behavior– Anticipates generation connection request in areas
with exploitable energy sources• Gas• Wind
– Starts first-step investment process as to get administrative green lights already agreed when generators will request for connection
25
FS
R a
nd
Loyola
de P
ala
cio
Ch
air o
n E
nerg
y P
olic
y
Pros & cons of the two alternative behaviors
• Reactive behavior
* No trsm investment cost put at risk**BUT excessive congestion if CCGTs or wind
farms connect while network is still to be upgraded
• Proactive behavior
* No excessive congestion put at risk**BUT proactive behavior is itself costly because
(if generation finally doesn’t come) TSO did already:• the study to upgrade the network • And went trough all procedures to obtain the
administrative green lights to build the line
26
FS
R a
nd
Loyola
de P
ala
cio
Ch
air o
n E
nerg
y P
olic
y
Unbundling and the need for coordination
Liberalised power system
>> Prompts investors to choose generation technologies with short
construction lead time
Right of way of powerlines facing increasing oppositions ~ 7 years to build a powerline from study to construction itself
because of administrative agreement >> 5 years!
Generation technology
Time to build (year)
Notional size (MW)
CCGT 2 200-800
Wind onshore 2 25
offshore 2 100
Coal 4-5 200-1600
Nuclear 5-7 1600
Congestion while the network is not upgraded
27
Need for coordination between generation and transmission
investments
FS
R a
nd
Loyola
de P
ala
cio
Ch
air o
n E
nerg
y P
olic
y
Outline
• Need for coordination between G and T
• Tool #1 = locational signals
• Tool #2 = anticipation
Scope for maximisation of social welfare?Model and results
• Conclusion
FS
R a
nd
Loyola
de P
ala
cio
Ch
air o
n E
nerg
y P
olic
y
Problem
• Efficiency of anticipating generation connection for the TSO in terms of minimization of her total network costs?– If no anticipation congestion for quite long
period while network must be upgraded
– If TSO anticipes the ‘’first-stage costs’’ costly if anticipated generators do not connect
One must arbitrate between these two costs• weighed by a probability for the connection of
generation (been evaluated by an expert panel « à dire d’expert »)
29
FS
R a
nd
Loyola
de P
ala
cio
Ch
air o
n E
nerg
y P
olic
y
Construction
Generation Investment
Year
CU(d)CW(d)
Transmission Investment
d
0
Study, admin. proced. construction
Investment sequencing with reactive TSO 30
Excessive congestion Optimal value for network capacity
FS
R a
nd
Loyola
de P
ala
cio
Ch
air o
n E
nerg
y P
olic
y
Construction
Year
Study, admin. proced. construction
Generation and Network Investments
CU(0)
0
Investment sequencing with proactive TSO31
FS
R a
nd
Loyola
de P
ala
cio
Ch
air o
n E
nerg
y P
olic
y“Probability limit”
and condition for proactive TSO with known first-stage costs32
First-stage costs = 10% of transmission investment costs
FS
R a
nd
Loyola
de P
ala
cio
Ch
air o
n E
nerg
y P
olic
y
With first-stage costs = 50% of transmission investment costs
Still efficient for the TSO to be proactive when 40% probability for the connection of a new plant
“Probability limit” and condition for a proactive TSO
with unknown first-stage costs
FS
R a
nd
Loyola
de P
ala
cio
Ch
air o
n E
nerg
y P
olic
y
Outline
• Need for coordination between G and T
• Tool #1 = locational signals
• Tool #2 = anticipation
Scope for maximisation of social welfare?Model and results
• Conclusions
FS
R a
nd
Loyola
de P
ala
cio
Ch
air o
n E
nerg
y P
olic
y
Conclusion (1/4)
• A model to evaluate the efficiency of anticipating plants investment to minimize the total network costs
• Illustration on simple realistic examples (like CCGT or wind farms)– Efficient to anticipate the connection of power
plants for the TSO– Planning in advance network reinforcement– Reduce congestion costs
35
FS
R a
nd
Loyola
de P
ala
cio
Ch
air o
n E
nerg
y P
olic
y
Conclusion (2/4)
– Efficient for TSO to anticipate the connection of new generator
• All the more that (a/)– Proactive behavior favours dialog between
TSO and market participants about planning assumptions
• Facilitates coordination through sharing information
• Facilitates dialog and acceptability from local populations
36
FS
R a
nd
Loyola
de P
ala
cio
Ch
air o
n E
nerg
y P
olic
y
Conclusion (3/4)
• All the more that (b/)– Possibility to send potential locational signals
related to the anticipated network• Volume signals: new generation capacity that will
not constrain the grid• Tariff and price signals: potential levels of
locational network access fees and energy prices for different network and generation scenarios
37
FS
R a
nd
Loyola
de P
ala
cio
Ch
air o
n E
nerg
y P
olic
y
Conclusion (4/4)
• All the more that (c/)– Possible to incentivise TSOs to be proactive
• Make them bear a part of congestion costs• Example in France: the TSO compensates the
wind farms when they have a ‘curtailment obligation’ for more than 3 years
38
FS
R a
nd
Loyola
de P
ala
cio
Ch
air o
n E
nerg
y P
olic
y
Future research on network anticipation
• Feedback effects with locational signals
• Effects of (milestones payment) in connection tariffs to create increasing locational commitment from generator
39
FS
R a
nd
Loyola
de P
ala
cio
Ch
air o
n E
nerg
y P
olic
y
Conclusion of conclusions
• Tool 1- Coordination of generation and transmission investment with signals because of intrinsic lumpy network cost structure– information sharing with Generation – Necessary but not sufficient locational
signal when generation quicker to develop than the network
FS
R a
nd
Loyola
de P
ala
cio
Ch
air o
n E
nerg
y P
olic
y
Conclusion of conclusions
• Tool 2- Importance of information / consultation platform for network planning– Information sharing building of common
knowledge about possible future(s) more certainty for investors
– All the more needed to integrate new generation technologies new location, new network usage
FS
R a
nd
Loyola
de P
ala
cio
Ch
air o
n E
nerg
y P
olic
y
That’s it: Even in a Surfers (non)Paradise more coordination between Gen. and
Transmission is not necessarily a luxury…
Jean-Michel Glachant
Loyola de Palacio Chair & Florence School of Regulation