Chapter 12: Introduction to

Electricity Markets
 Traditional electric utility model
Generation
• Monopoly
• Only supplier of electricity in a region
Transmission
• “service territory”
• Consumer does not have a choice of supplier
Distribution

• Vertically integrated
• A single organization performs all the technical and Retail
business functions

Consumer

© 2023 Daniel Kirschen 2

 Why vertical integration?
Generation
• Used to be considered more efficient
• Issues: Transmission
• Lack of transparency
• Poor internal communication in very large
Distribution
organizations

Retail

Consumer

© 2023 Daniel Kirschen 3

 Why monopolies?
• Building a power system is expensive
• Building competing transmission and distribution networks does not
make sense
• “Natural Monopoly”
• Until recently having several companies “share” a power system was
too complicated
• Monopoly can be:
• Private (investor-owned utility)
• Public (municipal, utility district, national)
• Size of service territories varies

© 2023 Daniel Kirschen 4

 Example 1: Seattle City Light
• Owned by the City of Seattle
• Public monopoly

© 2023 Daniel Kirschen 5

 Example 2: Puget Sound Energy
• Investor-owned utility
• Private monopoly

© 2023 Daniel Kirschen 6

 Example 3: BC Hydro
• Owned by the Province of British Columbia
• Public monopoly

© 2023 Daniel Kirschen 7

 Possible variants on vertical integration
• One entity handles generation and transmission Generation

• Other entities handle distribution and retail

Transmission
• Example from the Pacific Northwest:
• Generation and transmission:
• Bonneville Power Administration (BPA) Distribution
• Distribution and retail
• Municipal utilities
Retail
• Rural electricity cooperatives
• Public utility districts
• Example of hybrid:
• Seattle City Light is vertically integrated but buys power from BPA
Consumer
© 2023 Daniel Kirschen 8
 Types of utilities in the United States

Type Number
Investor-owned Companies 265
Municipal Systems 1,818
Public Power Districts 75
State 68
County Systems 5
Rural Electric Co-Ops 922
U.S. Government 37

© 2023 Daniel Kirschen 9

 Regulation
• A private monopoly could abuse its position
• It must be regulated by the government
• Government-owned utilities operate for the public good
• No need for outside regulation
• Elect new representatives if we are not satisfied

© 2023 Daniel Kirschen 10

 The “Regulatory Compact”
• Agreement between an Investor-Owned Utility (IOU) and a
government
• IOU receives the right to be the monopoly supplier over a certain
territory
• IOU accepts that its rates (i.e., what it can charge consumers) will be
determined by a regulator
• Example: the Washington Utilities and Transportation Commission
regulates Puget Sound Energy

© 2023 Daniel Kirschen 11

 Rate of return regulation
• Regulator sets the rates so that the IOU can:
• Recover its operating cost (fuel, personnel, etc…)
• Recover its investment costs (plants, lines, etc…)
• Pay a fair rate of return to its investors
• A monopoly utility is a low-risk investment
• No competition
• A bankrupt utility is in nobody’s interest
• The rate of return can therefore be low compared to other investments

© 2023 Daniel Kirschen 12

 Monopolies are inefficient
• No competition
• No need to be efficient to survive
• No incentive to be efficient:
• Utility earns more if it invests more
• High costs passed on to consumers as high prices of electricity
• Rates are “higher than they should be”

© 2023 Daniel Kirschen 13

 Could the regulators do better?
• Regulation is difficult
• Little basis for comparison
• Each regulator oversees a small number of utilities
• Each utility has a territory with different characteristics
• Difficult to evaluate the utilities’ decisions
• Regulator does not have as much staff as the utility
• “Information imbalance”

© 2023 Daniel Kirschen 14

 Problems with public monopolies
• A good government will run its utilities in an efficient and far-sighted
manner
• This is not always the case
• Conflicts can arise between the objectives of the government and the
objectives of the utility
• Government monopolies are inefficient too!

© 2023 Daniel Kirschen 15

 Things bad governments do…
• Keep rates low to please the voters
• Utility does not have enough money for investments
• Keep rates high and use the surplus money for other programs
• This is an economically inefficient form of taxation
• Discourages the consumption of electricity
• Force the utility to make unnecessary investments to create jobs

© 2023 Daniel Kirschen 16

 Why introduce electricity markets?
• Market competition forces companies to make better decisions
• Be more efficient
• Introduce innovations
• Over time this should lead to lower costs to the consumers

© 2023 Daniel Kirschen 17

 Unbundling
• Building competing transmission and
distribution networks does not make
sense
• They are “natural monopolies”
• Environmental impact
• Networks must remain monopolies
• Need to separate the monopoly and
non-monopoly activities
• “Unbundling” of the vertically
integrated utilities

© 2023 Daniel Kirschen 18

 Where can we introduce competition?
• Between companies that generate electrical energy
• Wholesale market
• Between companies that buy electrical energy on the wholesale
market and sell it to consumers
• Retail market

© 2023 Daniel Kirschen 19

 Wholesale electricity markets in the US

© 2023 Daniel Kirschen 20

Fundamentals of Markets
 What are markets?
• Locations where buyers and sellers meet to trade goods
• Physical or virtual
• Venue to execute a transaction
• Mechanism for buyers and sellers to collect information
• What commodity is being traded?
• How much is being offered?
• How big is the demand?
• Gauge demand and supply for each good
• Settle at an economically efficient market equilibrium

© 2023 Daniel Kirschen 22

 Example: market for T-shirts
• Supply side
• Manufacturing T-shirts
• Selling T-shirts
• Supply function
• Demand side
• How many T-shirts shall I buy?
• How many T-shirts will people buy?
• Demand function
• Market equilibrium
• Intersection of supply and demand functions

© 2023 Daniel Kirschen 23

 Manufacturer’s perspective
• Manufacturer’s cost function:
• How much does it cost to produce T-shirts?
• Fixed costs
• Building the factory, buying machinery
• Not a function of
• Variable costs
• Raw materials, labor
• Marginal cost function:
• Ignores the fixed costs
• Cost of producing one more T-shirt

© 2023 Daniel Kirschen 24

 Typical marginal cost function
• Increases with due to extra
costs
• Overtime pay for workers
($/T-shirt)
• Additional maintenance

(T-shirts)

© 2023 Daniel Kirschen 25

 Optimal production level
• If each T-shirt can be sold for
dollars, how many should the
manufacturer sell?
($/T-shirt)
• Selling a T-shirt that costs less to
manufacture than result in a 𝜋
profit
• Selling a T-shirt that costs more
to manufacture than result in a ∗
loss 𝑞
(T-shirts)
• maximizes the overall profit

© 2023 Daniel Kirschen 26

 Inverse supply function
• Each supplier comes to the market with its own marginal cost function
• Each of these function shows how much it is willing to sell as a function
of the price
• The inverse supply function is the aggregation of each supplier’s
marginal cost function

∗ ∗ ∗ ∗
© 2023 Daniel Kirschen
𝑞1 𝑞2 𝑞1 +𝑞 2
27
 Inverse supply function
• Inverse supply function
•
• Price at which the suppliers will sell a
quantity

• Supply function

• Quantity that the suppliers will sell at

a price

• Describes the supply side of the

market: the aggregated willingness
to sell
© 2023 Daniel Kirschen 28
 Consumer’s perspective
• Shall I buy a T-shirt?
• What factors do I consider when making that decision?
• How expensive is the T-shirt?
• How much enjoyment/utility will I get from this T-shirt?
• What else could I buy with this money?
• Different people  different decisions
• Depending on the price:
• Some people will buy a T-shirt, other people will not
• Depends on their marginal utility
• Marginal utility > price  buy
• Marginal utility < price  don’t buy
© 2023 Daniel Kirschen 29
 Demand and inverse demand functions
• Aggregation of the marginal utility of all
consumers
• Demand function
•
• Quantity that consumers will buy as a function
of the price
• Inverse demand function
•
• Price at which consumers will buy a quantity
• Describes the demand side of the market:
the aggregated willingness to buy
© 2023 Daniel Kirschen 30
 Market equilibrium
• Willingness to sell meets willingness to buy
• Intersection of the demand and supply 𝜋
functions: , Demand function
• At , as a group consumers will not:
Supply function
• Buy more than because the utility of the extra
quantity would be less than the price
• Buy less than because they could get more utility for
that price
∗
• At , as a group suppliers will not: 𝑞
• Sell more than because they would sell for less than
the marginal cost of production
• Sell less than because they would forgo a profit
opportunity

© 2023 Daniel Kirschen 31

 Market equilibrium
• At the market equilibrium:
• Market price = suppliers’ marginal cost
𝜋
• Market price = consumers’ marginal utility
Demand function
• Marginal cost = marginal utility
• “Marginal pricing” Supply function

• Valid only if there is perfect competition

• Perfect competition requires
• Many buyers 𝑞
∗

• Many sellers
• If some buyers or sellers have too big a market share,
they have market power and can influence the
market price

© 2023 Daniel Kirschen 32

 Example 12.1: Market equilibrium
• Inverse supply function $/T-shirt
• Inverse demand function $/T-shirt
• Intersection of these two functions:
•
• T-shirts
• $/T-shirt

© 2023 Daniel Kirschen 33

 Structure of a wholesale electricity market
• Gencos
• Generating companies Genco 1 Genco 2 Genco N

• Own and operate power plants

• Sell energy on the wholesale Wholesale Market and
TSO
Transmission Network
market
• LSEs
• Load Serving Entities LSE A LSE B LSE M
• Buy energy on the wholesale
market
• Sell this energy to the Consumers Consumers Consumers
consumers in their monopoly
service area

© 2023 Daniel Kirschen 34

 Trading vs. physics
• Commodity being traded is energy (MWh)
• Power system delivers power (MW) on a continuous basis
• Define trading periods
• Typically, one hour, 15 minutes, 5 minutes
• Trading assumes constant delivery of power over this trading period
• Example
• Genco sells 20 MW from 2:15 to 2:30 pm
• It has sold 5 MWh during this trading period

© 2023 Daniel Kirschen 35

 Trading vs. physics
• All MWs get merged in the transmission network
• Cannot direct energy sold by a given Genco to a given LSE
• All MW are the same (“fungible”)
• Keep track of the transactions using meters at both ends

© 2023 Daniel Kirschen 36

 Types of electricity trading
• Bilateral trading
• Conventional form of trading
• Buyer and seller agree on price and quantity without intermediary
• However, TSO may limit trades to make sure that they are compatible with
the reliable operation of the system
• Common form of trading in European markets and the Pacific Northwest
• Centralized trading
• Specific to electricity markets
• TSO takes on the role of market operator
• Common form of trading in North America:
• PJM, CAISO, ISO-NE, ERCOT, MISO, NYISO, SPP

© 2023 Daniel Kirschen 37

 Centralized trading: supply side
• Gencos submit offers to sell for each trading period
• Quantity/price pairs
• Example: 100 MW at 23 $/MWh
• TSO ranks the offers in increasing order of price
•  Creates a supply curve

© 2023 Daniel Kirschen 38

 Example: supply curve in a centralized
market
Offers to Sell $/MWh 200
Quantity Price
Genco
(MW) ($/MWh) 180

100 0.00 160

A 100 30.00
140
50 50.00
100 15.00 120

B 100 40.00 100 B

50 90.00
100 0.00 80
C
C 100 25.00 60 A
50 65.00 D
B
40 A
50 10.00 D
C
B
D 50 20.00 20 D
C
100 35.00 0
A

0 100 200 300 400 500 600 700 800 900 1000
MW
Supply Curve

© 2023 Daniel Kirschen 39

 Centralized trading: demand side
• LSEs submit bids to buy for each trading period
• Quantity/price pairs
• Example: 100 MW at 50 $/MWh
• TSO ranks the offers in decreasing order of price
•  Creates a demand curve

© 2023 Daniel Kirschen 40

 Example: demand curve in a centralized market
Bids to Buy $/MWh
200
Q S
Quantity Price P R
LSE
(MW) ($/MWh) 180
P R Q S
100 200.00 160 P
P 100 175.00
140
50 150.00
100 200.00 120

Q 50 175.00 100
50 10.00
80
100 200.00
R 100 175.00 60

50 15.00 40
100 200.00
20 R
S 50 175.00 Q
S
50 5.00 0
0 100 200 300 400 500 600 700 800 900 1000
MW
Demand Curve

© 2023 Daniel Kirschen 41

 Centralized trading: clearing
• TSO identifies the intersection of the supply and demand curves
• Determines the market clearing price and the quantity transacted
• Genco offers with price less than the market price are accepted
• LSE bids with price higher than the market price are accepted
• Other bids and offers are rejected
• Gencos and LSEs whose bids and offers are accepted are committed
to inject or extract from the system the corresponding amount of
power during this trading period

© 2023 Daniel Kirschen 42

 Example: centralized market clearing
$/MWh
Accepted bids
200
P R Q S

180
P R Q S
160 P
Market price: 40 $/MWh
140

120 Quantity traded: 750 MW

100 B

80
C
Accepted offers Market equilibrium
60 A
B
40 D
A
C
D
B R
20 D Q
S
A C
0
0 100 200 300 400 500 600 700 800 900 1000
MW
Supply Curve Demand Curve

© 2023 Daniel Kirschen 43

 Centralized trading: settlement
• All energy traded is bought/sold at the market price
• All Gencos collect the market price for the accepted offers
• All LSEs pay the market price for the accepted bids
• Market price reflects the marginal cost of generation

© 2023 Daniel Kirschen 44

 Example: centralized market settlement
$/MWh

200 Genco Accepted offers = quantity sold Revenue

P R Q S

180 A 100+100 = 200MW 200 x 40 = $8,000

P R Q S B 100+50 = 150 MW 150 x 40 = $6,000
160 P C 100 + 100 = 200 MW 200 x 40 = $8,000
140 D 50 + 50 + 100 = 200 MW 200 x 40 = $8,000
Total 750 MW $30,000
120

100 B

80
C
Market equilibrium
60 A
B
40 D
A
C
D
B R
20 D Q
S
A C
0
0 100 200 300 400 500 600 700 800 900 1000
MW
Supply Curve Demand Curve

© 2023 Daniel Kirschen 45

 Example: centralized market settlement
$/MWh

200 Accepted bids = quantity

P R Q S LSE Expenditure
purchased
180
R Q S
P 100+100+50 = 250MW 250 x 40 = $10,000
P
160 P
Q 100+50 = 150 MW 150 x 40 = $6,000
R 100 + 100 = 200 MW 200 x 40 = $8,000
140
S 100 + 50 = 150 MW 150 x 40 = $6,000
120 Total 750 MW $30,000

100 B

80
C
Market equilibrium
60 A
B
40 D
A
C
D
B R
20 D Q
S
A C
0
0 100 200 300 400 500 600 700 800 900 1000
MW
Supply Curve Demand Curve

© 2023 Daniel Kirschen 46

 Bidding in a centralized market
• Sufficiently large number of market participant  perfect competition
• Gencos optimal bidding strategy: bid marginal cost of generation
• Offer accepted at market price
Profit equal to the difference between the market price and the offer price
• Offer rejected
Genco avoids producing at a loss
• Offer at higher than marginal cost
 profitable offer risks being rejected
• Offer at lower than marginal cost
 non-profitable offer risks being accepted

© 2023 Daniel Kirschen 47

 Bidding in centralized markets
• Less competitive market
• Example: demand is high and most of the available generation capacity is needed
• Gencos can offer at substantially more than their marginal cost
• Try to drive up the market price to recover their fixed costs

• Generators that mostly care about producing energy

• Nuclear units that cannot shutdown if their bid is not accepted
• Renewable generation whose marginal cost is close to zero
• Offer at zero
• Let the offers of the other gencos set the market price
© 2023 Daniel Kirschen 48
 Example: extreme offers
$/MWh
200
P R Q S

180
P R Q S
160 P

140

120

100 B Less competitive generators

that run only when the
80
C demand is high
60 A
B
40 D
A
C
D
B R
Generators that bid zero 20 D Q
S
A C
to make sure that they get 0
0 100 200 300 400 500 600 700 800 900 1000
to produce. Typically nukes MW
Supply Curve Demand Curve
and renewables.
© 2023 Daniel Kirschen 49
 Example: demand side biddingSince loads tend to be inflexible,
most demand side will bid high to ensure
$/MWh that they get selected
200
P R Q S

180
P R Q S
160
Some consumers are flexible and
P
willing to curtail or shift their load
140
if the price is too high
120

100 B

80
C

60 A
B
40 D
A
D
C Some loads are highly flexible and
B
20 D
R
Q
S
wait for a very low price e.g.,
C
0
A
battery charging
0 100 200 300 400 500 600 700 800 900 1000
MW
Supply Curve Demand Curve

© 2023 Daniel Kirschen 50

 Replacing the demand curve by a load forecast
• Some markets neglect
$/MWh 200
the effect of the price on
the demand 180

160

140

• The demand curve is 120

then replaced by a 100

vertical line at the
forecast value of the load
80

60

40

20

0
0 100 200 300 400 500 600 700 800 900 1000
MW
© 2023 Daniel Kirschen 51
 Variation of the load over a day
50,000

45,000

40,000

35,000

30,000
Load (MW)

25,000

20,000

15,000

10,000

5,000

0
0 3 6 9 12 15 18 21 24
Hour
© 2023 Daniel Kirschen 52
 Discretized variation of the load over a day
110000

100000

90000
Forecast Load (MW)

80000

70000

60000

50000
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

Hour

Assume that the load is constant over each one-hour trading period
© 2023 Daniel Kirschen 53
 Market clearing for each hour
• Replace the demand curve $/MWh 200

by the load forecast for 180

Hour 5 Hour 23 Hour 10 Hour 19
each trading hour 160

• Offers from the Gencos 140

are valid for the whole day 120

• Clear the market for each 100

hour 80

60
• Market price increases
with the load
40

20

0
0 100 200 300 400 500 600 700 800 900 1000
MW

© 2023 Daniel Kirschen 54

 Example: time-varying market prices
$/MWh 200

180
Hour 5 Hour 23 Hour 10 Hour 19
160

Hour 1 2 3 4 5 6 140
Load 390 375 300 275 225 300
120
Hour 13 14 15 16 17 18
Load 810 775 750 750 825 875 100

Hour 7 8 9 10 11 12 80
Load 425 550 650 750 775 825
60
Hour 19 20 21 22 23 24
Load 925 825 750 650 450 425 40

20

0
0 100 200 300 400 500 600 700 800 900 1000
MW

© 2023 Daniel Kirschen 55

 Example: time-varying market prices

100
$/MWh
90
Hour 1 2 3 4 5 6 80

Load 390 375 300 275 225 300 70

Hour 13 14 15 16 17 18 60

Load 810 775 750 750 825 875 50

Hour 7 8 9 10 11 12 40

Load 425 550 650 750 775 825 30

Hour 19 20 21 22 23 24 20

Load 925 825 750 650 450 425 10

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
Hour

© 2023 Daniel Kirschen 56

 Example: Price-duration curve for ISO-NE in 2022
400
$/MWh

350

300

250

200

150

100

50

0
0 10 20 30 40 50 60 70 80 90 100
Percentage of hours
© 2023 Daniel Kirschen 57
 Effect of transmission capacity limits
• Assumption so far: supply and demand curves reflect the bids and
offers of all the Gencos and LSEs
• Market price applied uniformly to all transactions
• Problem: pattern of generation and demand may result in violations
of operating limits of the transmission network
• Must modify the results of the market clearing to remove these
violations

© 2023 Daniel Kirschen 58

 Locational marginal pricing
Modify the market clearing to respect transmission constraints
 Reject some cheaper offers in the “wrong” place
 Replace them with more expensive offers in the “right” place
Market price depends on the location
Locational marginal pricing

© 2023 Daniel Kirschen 59

 Example: locational marginal pricing
• Same generators and same load as in the previous centralized market
example
• Generators and loads separated in two regions connected by a single
transmission line

Western Eastern

A B

D C

225 MW 525 MW

© 2023 Daniel Kirschen 60

 Example: locational marginal pricing
• Market clearing from previous example
• Ignores transmission network
• Violates the 50 MW rating of the transmission line

Western Eastern
200 MW 150 MW
A B
175 MW

200 MW 200 MW
D C

225 MW 525 MW

© 2023 Daniel Kirschen 61

 Example: locational marginal pricing
Western Eastern
• To remove the overload: 200 MW 150 MW
A B
• Reduce the generation in the 175 MW

Western region by 125 MW 200 MW 200 MW

D C
• Increase the generation in the
Eastern region by 125 MW 225 MW 525 MW

• A & D compete to supply 275 MW Western Eastern

A B
50 MW

• B & C compete to supply 475 MW D C

225 MW 525 MW

© 2023 Daniel Kirschen 62

 Example: locational marginal pricing
Western Eastern
• To remove the overload: 200 MW 150 MW
A B
• Reduce the generation in the 175 MW

Western region by 125 MW 200 MW 200 MW

D C
• Increase the generation in the
Eastern region by 125 MW 225 MW 525 MW

Western Eastern
175 MW 225 MW
A B
50 MW

100 MW 250 MW
D C

225 MW 525 MW

© 2023 Daniel Kirschen 63

 Example: locational marginal pricing
Western region Eastern region
$/MWh 100 $/MWh 100
B
90 90
80 80
70 70 C

60 60
A
50 50
B
40 D 40
A
30 30 C
D
20 20 B
D
10 A 10 C
0 0
0 50 100 150 200 250 300 350 400 450 500 0 50 100 150 200 250 300 350 400 450 500
MW MW

Market price in the Western region: 30 $/MWh Market price in the Eastern region: 90 $/MWh

© 2023 Daniel Kirschen 64

 Locational marginal pricing (LMP)
• Generators are paid the LMP at the bus where they are connected
• LSEs pay the LMP at the bus where they are connected
• Prices will vary depending on which constraints are binding

• Example with two zones and one transmission line

 Easy to calculate the locational marginal prices
Larger, meshed transmission networks
Solve a linearized optimal power flow
 Prices are the Lagrange multipliers of the nodal power balance constraints

© 2023 Daniel Kirschen 65

 Examples of locational marginal prices
• CAISO
• ERCOT
• MISO
• ISO-NE
• NYISO
• SPP
• PJM

© 2023 Daniel Kirschen 66

 Day-ahead market
• Benefits of unit commitment:
• Amortize startup costs over enough trading period
• Satisfy operating constraints on the generating units
• Minimum up- and down-time, ramp rates
• Clear the market one day-ahead using a unit commitment

© 2023 Daniel Kirschen 67

 Day-ahead market based on unit commitment
• Clear the market on day D-1 for day D
• Gencos submit offers for each unit
• Offers are valid for all trading periods of day D
• Offers must include:
• Price/quantity pairs
• Startup costs
• Operating constraints
• LSEs submit their demand curve for each trading period
• TSO runs a unit commitment combined with an OPF
• Offer prices treated like the marginal costs in a conventional UC
• Offer price of the most expensive unit scheduled at each trading period
sets the market price
© 2023 Daniel Kirschen 68
 Problem with day-ahead market
• Things never turn out as expected…
• Error in the load forecast
• Some generators do not produce what they were scheduled to produce
• Outages of large units, wind and solar generation
• How can we maintain the load/generation balance?
• Procure some reserve on the day-ahead market
• Provide insurance against outages of large units
• Implement a balancing market
• Economically efficient way of handling the imbalances

© 2023 Daniel Kirschen 69

 Balancing market
• 5-minute or 15-minute trading period
• Operate shortly before real-time
• Bids and offers:
• Gencos with flexible generating units submit offers
• Flexible loads
• Storage
• TSO determines which bids and offers are needed to keep the system
in balance
• Most expensive bid or offer sets the real-time price

© 2023 Daniel Kirschen 70

 Two-settlement market
• Day-ahead transactions
• Settled at the day-ahead price
• As if powers produced and consumed were exactly as scheduled
• Imbalances settled at the real-time price
• Generator produced less than scheduled
 pays for the difference at the real-time price
• Load consumes more than scheduled
 pays for the difference at the real-time price
• Generator produced more than scheduled
 paid for the difference at the real-time price
• Load consumes less than scheduled
 paid for the difference at the real-time price

© 2023 Daniel Kirschen 71

 Example: Two-settlement market
• Day-ahead market • Real-time market
• June 11, hourly trading period 8 • June 11, hourly trading period 8
• Bus Patagonia • Bus Patagonia
• Day-ahead price: 23 $/MWh • What they actually did:
• Genco Blue scheduled to inject Trading Blue Orange RT price
interval (MW) (MW) ($/MWh)
100 MW at that bus 7:00 – 7:15 92 50 25.00
7:15 – 7:30 104 46 20.00
• LSE Orange scheduled to extract 7:30 – 7:45 100 54 23.00
50 MW at that bus 7:45 – 8:00 20 42 30.00

© 2023 Daniel Kirschen 72

 Example: Two-settlement market
Settlement for Genco Blue:

Trading period MWh Price Amount

DA: 7:00 – 8:00 100 23.00 $2300
RT: 7:00 – 7:15 25.00 ($50)
RT: 7:15 – 7:30 20.00 $20
RT: 7:30 – 7:45 23.00 $0
RT: 7:45 – 8:00 30.00 ($600)
Total: $1670

The factors ¼ convert MW imbalances over a 15-minute period to a MWh value.

© 2023 Daniel Kirschen 73

 Example: Two-settlement market
Settlement for LSE Orange:

Trading period MWh Price Amount

DA: 7:00 – 8:00 50 23.00 ($1150)

RT: 7:00 – 7:15 25.00 $0

RT: 7:15 – 7:30 20.00 $20

RT: 7:30 – 7:45 23.00 ($23)

RT: 7:45 – 8:00 30.00 $60

Total: ($1093)

The factors ¼ convert MW imbalances over a 15-minute period to a MWh value.

© 2023 Daniel Kirschen 74

 Ancillary services
• TSO responsible for maintaining the stability of the system
• Needs resources to do that
• In an unbundled system, the TSO does not own these resources
• These resources are owned by market participants
• These market participants expect to be paid for these resources
 market for ancillary services

© 2023 Daniel Kirschen 75

 Ancillary services
• Contingency reserve
• Compensate providers for the opportunity cost of not providing energy
• Opportunity for batteries
• Frequency control
• Constant minor adjustments to power injections/extractions
• Zero energy on average
• Another opportunity for batteries
• Reactive power
• Providing reactive power causes losses and reduces opportunity to provide
active power
• Black start capability
• Need to be paid to be available
© 2023 Daniel Kirschen 76
 Retail market
Genco 1 Genco 2 Genco N

Wholesale Market and

TSO
Transmission Network

Retailer A Retailer B Retailer M

Retail Market and

Discos
Distribution Networks

Consumer Consumer Consumer

© 2023 Daniel Kirschen 77