Evaluating the Potential Impact of Transmission Constraints on the Operation of a Competitive Electricity Market in Illinois

prepared for

Illinois Commerce Commission

June 2006

Richard R. Cirillo

Decision and Information Sciences Division

phone: 630-252-5629 email: cirillor@anl.gov

The Study Was a Joint Effort:Argonne National LaboratoryUniversity of Illinois at Urbana-Champaign

Work commenced July 2002

Initial draft report submitted December 2003

Revised draft report with additional analysis September 2004

Briefing to Transmission Advisory Group May 2005

Final TAG comments received January 2006

Letter response to TAG comments sent March 2006

Final report with editorial changes April 2006

The Study Was Commissioned to Address Some Important Questions Regarding the Electricity Market in Illinois

Is the transmission system adequate to support market operation?

Can competition keep prices in check?

Is there the potential for market power to be exerted?– Market power: The ability to raise prices and increase profitability by

unilateral action

The Study Was Designed to Be:

– Exploratory, not predictive of specific outcomes

– Initial investigation

– Issue-oriented, not regulatory-oriented

Basic Assumptions Used in the Study

Analysis year : 2007

Projections for 2007 based on data from NERC, FERC, EIA, IEPA

A single day-ahead pool market operating in the State– All buyers and sellers of electricity bid into a common market every day

No bilateral contracts

Out-of-state generation and load participate in the market

Groups of cases– Case Study Assumptions – with generator forced outages– Conservative Assumptions – no forced outages

In the Pool Market, a Locational Marginal Price (LMP) Settlement Process Was Assumed to Be Operative

The cost of providing the next MW of power at each point in the network (i.e., the LMP) is calculated

In the absence of transmission limitations:– The lowest cost generators are used first– Higher cost generators are used only as

needed– The LMP at every point in the network is

the same

With transmission limitations:– It is not always possible to use the lowest

cost generators– The LMPs vary, sometimes considerably,

across the networkTransmission

Nodes

Cases Were Constructed as “Electronic Experiments”to Study Market Behavior

Production Cost •Generation companies bid production cost•Gives the lowest system cost

Physical Withholding •Intentionally take generators out of service•Prices rise

Economic Withholding •Generation companies bid above production cost•Prices rise

•Experiments moved from very simple to more complex strategies

•Production Cost Case used as a benchmark

•Not intended to imply that any company would attempt to exercise market power

•Only an initial mapping of possible market bidding

In the Production Cost Case for the Analysis Year 2007, the State Is a Net Exporter of Electricity

0

2,000

4,000

6,000

8,000

10,000

12,000

14,000

16,000

18,000

Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec

In-State Generation

(GWh)Net ExportIn-State Demand

Approximately 6% of the electricity generated is exported

Coal and Nuclear Capture the Largest Market Share

0

2

4

6

8

10

12

14

16

18

Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec

Generation(TWh)

Renewable

Oil

Nat Gas

Coal

Nuclear

Approximately 2% of the electricity is from natural gas and other sources

In the Production Cost Case, 4 out of 22 Companies Account for About 95% of the Market

Out-of-state suppliers do not gain significant market share

Generation Company

Market Share Exelon Nuclear43.1%

Ameren20.1%

Midwest Generation

16.8%

DynegyMidwest

Generation14%

Dominion Energy3.4%

Other 1.5%

City of Springfield1.0%

0

200

400

600

800

1,000

1,200

1,400

1,600

Exelon N

uclear

Ameren

Midwest G

enera

tion LLC

Dyneg

y Midwes

t Gen

eratio

n Inc.

Dominion Energy

Exelon N

uclear/

Midameri

can Energ

y

City of S

pringfie

ld

NRG Energy

Relian

t Energ

yCalp

ine

Duke Energ

y

Dyneg

y/NRG Energ

y

Constella

tion Power

Southern Ill

inois Power

Coop

MidAmerica

n Energy C

o

Alleghen

y Power

Aquila Energ

yPPL

Power Energ

y Part

ners

Soyland Power

Coop Inc

Calumet

Energy L

LC

Southwester

n Electri

c Coop.

Revenuesand

Costs($ million)

RevenuesCosts

Only 5 of 22 Generation Companies Show an Operating Profit In the Production Cost Case. Others Are Not Dispatched Enough to Recover Fixed Operating Costs.

LMPs at Transmission Nodes Increase In the Peak Load Summer Months and Vary Across the State, Even Under Production Cost Conditions

January February March April JuneMay

July August September October November December

HigherLowerLMP Price and Duration:

Four Downstate Zones in April (Low Load Month)LMPs Are Low and Close Together

0

20

40

60

80

100

120

140

160

180

200

SunMonTueWedThu Fri SatSunMonTueWedThu Fri SatSunMonTueWedThu Fri SatSunMonTueWedThu Fri SatSunMon

LMP($/MWh)

Little Transmission Congestion

Four Downstate Zones in July (High Load Month)LMPs Are a Little Higher But Still Close Together

0

20

40

60

80

100

120

140

160

180

200

SunMonTueWedThu Fri SatSunMonTueWedThu Fri SatSunMonTueWedThu Fri SatSunMonTueWedThu Fri SatSunMon

LMP($/MWh)

Little Transmission Congestion

Six Northern Zones in April (Low Load Month)LMPs Are a Little Higher than Downstate But Close Together

0

20

40

60

80

100

120

140

160

180

200

SunMonTueWedThu Fri SatSunMonTueWedThu Fri SatSunMonTueWedThu Fri SatSunMonTueWedThu Fri SatSunMonTue

LMP($/MWh)

Little Transmission Congestion

Six Northern Zones in July (High Load Month)LMPs Are High and with a Significant Spread

0

20

40

60

80

100

120

140

160

180

200

SunMonTueWedThu Fri SatSunMonTueWedThu Fri SatSunMonTueWedThu Fri SatSunMonTueWedThu Fri SatSunMonTue

LMP($/MWh)

Noticeable Transmission Congestion

Physical Withholding Cases Were Studied

Withhold generator units from the market

– Reduce supply to cause prices to rise

– Increase profitability of other units owned by the company

Cases studied:

– Withhold one unit at a time: Does not increase profitability

– Use a profitability screen: Does not significantly increase profitability

– System reserve screen: Can increase profitability

Units Withheld Using the System Reserve Screen Could Potentially Increase Prices and Profitability

•Impact of physical withholding depends on:– Capacity withheld– Location on the network– Availability of replacement capacity

•Increase in company peak day operating profit: $2-14 million

•Increase in peak day consumer costs: $45-230 million

•Conservative assumptions do not change the basic result

•This assumes there is no market oversight or prohibition on withholding•There are some physical constraints on the ability to withhold large units•There is no implication that any company would do this

Economic Withholding Cases Were Studied

Generator unit prices are increased– Units are not taken out of service

– Prices increased relative to production cost

Cases studied– Single unit price increases: Does not increase profitability

– Companywide price increases in all hours: Profitability increase but loss of market share

– Companywide price increases in peak hours: Profitability increase with little loss of market share

For Some Companies, Increasing Prices All Day Leads to a Loss of Generation Market Share But Can Increase Profitability

0

2

4

6

8

10

12

14

16

18

20

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20

Companywide Price Increase (Multiple of Production Cost)

Peak Day Operating

Profit ($million)

0

50

100

150

200

250Peak Day

Generation (GWh)

Profit

Generation

All Day Price Increases as a Multiple of Production Cost

Peak Day Generation(GWh)

Peak Day Operating Profit($million)

Company rate of generation drops as prices are increased, then levels off when units are needed to meet peak load

Company profitability increases only slightly until units are needed to meet load, then increase significantly

For Some Companies, Increasing Prices Only During Peak Hours Minimizes the Loss of Market Share

0

2

4

6

8

10

12

14

16

18

20

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20

Companywide Price Increase (Multiple of Production Cost)

Peak Day Operating

Profit ($million)

0

50

100

150

200

250Peak Day

Generation (GWh)

Profit

Generation

Peak Hour Price Increases as a Multiple of Production Cost

Peak Day Generation(GWh)

Peak Day Operating Profit($million)

Company rate of generation drops only slightly as prices are increased

Company profitability increases steadily

Overall Observations Under the Assumed Conditions

The State has an abundance of generation capacity. The State is a net exporter of electricity. Virtually all the generation is supplied by coal and nuclear; only a small amount by natural gas.

Using several indices, the generation market in the State is highly concentrated.

The transmission system is adequate for most hours but will show signs of congestion about 5% of the time. During high load periods, transmission congestion causes prices to rise at different rates across the State, even with no exerting of market power.

There is the potential for some companies to exercise market power (i.e., raise prices and increase profitability by unilateral action) and raise consumer costs under selected conditions, particularly when there is transmission congestion.

The Conservative Assumptions (no forced outages) do not change the basic results.

Additional Work Is Needed to Expand the Understanding of the Electricity Market

Improved representation of out-of-state generation and load

Sensitivity analyses

Effect of bilateral contracts

Effect of consumer price responsiveness

Addition of generation and transmission resources in key places

Effect of market rules

Comments Received on the Study Made Several Major Points

The study does not reflect the current operating practices of the PJM/MISO markets– Agree – Modeling of alternative market rules was originally proposed as a

later phase of the study.

The data and information used in the report have been superseded– Agree– The study was not intended to be a one-time effort. The plan was to

have ICC staff use the model and data in-house for updating and further analysis.

The results have no relevance to the current situation– Strongly disagree

Examples of Generator Bids in PJM for July 26, 2005Illustrate Bidding That Goes Beyond Production Cost

8X2D

0

200

400

600

800

1000

0 10 20 30 40 50 60Amount (MW)

Pric

e ($

/MW

)

2O

0

200

400

600

800

1,000

1,200

0 2,000 4,000 6,000 8,000 10,000Amount (MW)

Pric

e ($

/MW

)

On a companywide basis, GenCo 2O bid almost 9,000 MW into the market. Approximately the last 1,000 MW of the company’s capacity were bid at prices greater than 600 $/MW.

GenCo 9O bid unit 8X2D of 48 MW at 800-900 $/MWh

•These are actual bids, not model results•It is not possible to determine if the bids were accepted•Market monitor actions in response to the bids are unknown

PJM Data on LMPs for July 26, 2005 Show the Effects of Generator Bids and Transmission Congestion

Example Individual Bus LMPs

0

100

200

300

400

500

600

700

800

900

1000

0 200 400 600 800 1000 1200 1400 1600 1800 2000 2200 2400

Hour of the Day

LMP($/MW)

BIXBYMOUNDHOCKINGHOCKINGMONTGOMEHILLVALLBAKER

Bus

ComEd Bus and Zone LMPs

0

20

40

60

80

100

120

140

160

180

0 200 400 600 800 1000 1200 1400 1600 1800 2000 2200 2400Hour of the Day

LMP($/MW)

Lowest Bus - 176 STILHigest Bus - 389 E ROComEd Zone

Selected ComEd LMPsSelected PJM LMPs

There Is Strong Agreement with Several Comments

“The most important finding of the study is the rigorous demonstration that the transmission system is a vital part of an energy market.”

- Ameren Electric

“…the necessity for RTOs to continue to refine the independence of their transmission system operations, market designs, market rules andoversight of the RTO market by market monitors.”

- PJM