epe trg 2005-06 del2
TRANSCRIPT
-
7/30/2019 EPE Trg 2005-06 Del2
1/52
Slovenian electricity marketSlovenian electricity market
openingopening -- lessons learnedlessons learned
by
Robert Golob, Ph.D.
EESTEC International Workshop
Power Engineering Today and Tomorrow
Ljubljana, Slovenia
11. December, 2003
-
7/30/2019 EPE Trg 2005-06 Del2
2/52
2
LectureLecture TopicsTopics
1. What is Deregulation All About?
2. Situation in Slovenia
3. GENCO in Deregulated Environment
4. Power Market Simulator
5. Illustration: GENCO Case Studies
-
7/30/2019 EPE Trg 2005-06 Del2
3/52
1.1. WhatWhat isis DeregulationDeregulation AllAllAboutAbout??
Regulated vs. Deregulated System
Official Goals of Deregulation
How Are This Goals Being Met?Deregulation & Prices
So What Have We Learned?
-
7/30/2019 EPE Trg 2005-06 Del2
4/52
4
WhatWhat isis DeregulationDeregulation AllAll AboutAbout??Regulated System
TRANSMITION SYSTEM
OPERATOR
(ELES)
PRODUCERS
DISTRIBUTION SYSTEM
OPERATOR(S)
CONSUMERS
Physical Power Flow
Money Flow
Vertically IntegratedPower System
-
7/30/2019 EPE Trg 2005-06 Del2
5/52
5
What is Deregulation All About?What is Deregulation All About?Deregulated System
TRANSMITION SYSTEM
OPERATOR
(ELES)
PRODUCERS
CONSUMERS
DISTRIBUTION SYSTEM
OPERATOR(S)
ORGANIZED MARKET
(BORZEN)
ELIGIBLE CONSUMERS TARIFF CONSUMERS
BROKERS
Physical Power Flow
Money Flow
Money Flow
TRANSMITION SYSTEM
OPERATOR DOES NOT
HANDLE ANY MONEY FLOW
-
7/30/2019 EPE Trg 2005-06 Del2
6/52
6
Getting rid of vertically integrated monopolies
Customer gets a right to choose
Price / tariff reduction and alignment
Entrance of new players and private capital
Improving the efficiency of operation
WhatWhat WeWe HaveHave LearnedLearned
Official Goals of Deregulation
-
7/30/2019 EPE Trg 2005-06 Del2
7/52
7
Deregulation:
(market) power has been transferred from transmission operatorsto generators
Liberalization:due to lack of transparency (liquidity), the mobility of consumersstays rather low
Privatization:large incumbents control the scene
WhatWhat WeWe HaveHave LearnedLearned
How Are They Being Met?
Efficiency:revenues are increasing and costs have been (drastically)decreased
-
7/30/2019 EPE Trg 2005-06 Del2
8/52
8
What We Have Learned
Deregulation & PricesHighly dereglated
Poorly dereglated
Price decreasePrice increase
-
7/30/2019 EPE Trg 2005-06 Del2
9/52
9
No unique solution and no guarantee that it willwork
Yet there are some extremely positive examples
Liquidity of the markets transparent priceindicators are crucial
The process is still evolving (continental Europe)
Consolidation will mark the future outcomes
WhatWhat WeWe HaveHave LearnedLearned
So What Have We Learned?
-
7/30/2019 EPE Trg 2005-06 Del2
10/52
2.2. SituationSituation inin SloveniaSlovenia
Situation in 1998
How Are This Goals Being Met?
Deregulation & PricesSo What Have We Learned?
-
7/30/2019 EPE Trg 2005-06 Del2
11/52
11
Liberalization what is it? Operation of sector with great losses (120 mill
in 1998)
Number of employees up to 40% technologicalsurpluses in individual companies
Electricity prices one of the lowest in Europeand not covering production costs
SituationSituation inin SloveniaSlovenia
Situation in 1998
-
7/30/2019 EPE Trg 2005-06 Del2
12/52
1.1. PowerPowerMarketMarket OverviewOverview
Different Market Designs
Day-Ahead Market
Auction Mechanism
-
7/30/2019 EPE Trg 2005-06 Del2
13/52
13
Two distinct trading mechanisms:
Central auction Bilateral trading
Central auction mechanism:
Simple central auctionmechanism(without economic dispatch) Highly complex auction mechanisms (based on unit commitment algorithms)
Auction types:
Double-sided auction demand side can be fully envolved
Single-sided auction onlysupplycan submitbids to the market
PowerPowerMarketMarket OverviewOverview
Different Market Designs
Transmission & ancillary services:
Electricity trading is done independetly of transmission and ancillary services The energy is traded with the transmission and ancillary service markets (usually
leads to nodal or zonal pricing market models)
-
7/30/2019 EPE Trg 2005-06 Del2
14/52
14
Hourly bids.
Central auction mechanism.
Double-sided auction.
24 simultaneous auctions for next day period.
Single Market Clearing Price (MCP) for all
participants.
Billateral trade is allowed.
PowerPowerMarketMarket OverviewOverview
Day-Ahead Market
-
7/30/2019 EPE Trg 2005-06 Del2
15/52
15
PowerPowerMarketMarket OverviewOverview
Auction Design
Supply bids
Q
P
Demand bids
Q
P
Market ClearingMechanism
Q
P
Auction Results
(prices & quantities)
Physical system
contrsaints
-
7/30/2019 EPE Trg 2005-06 Del2
16/52
16
PowerPowerMarketMarket OverviewOverview
Bids
All bids are Price / Quantitybids.
Supply bids:bids to sell energyincreasing or constant
Demand bids:
bids to buy energydecreasing or constant
Q
P
Q
P
-
7/30/2019 EPE Trg 2005-06 Del2
17/52
17
PowerPowerMarketMarket OverviewOverview
Market Clearing Mechanism
Quantity [MWh/h]
Price
[/M
Wh]
Aggregated
demand bids
Aggregated
supply bids
Sold / Bought
Quantity
(Traded Volume)
Marke
tClear
ing
Price
(MCP
)
Single price discovery
Supply meets demand
-
7/30/2019 EPE Trg 2005-06 Del2
18/52
2. GENCO2. GENCO OverviewOverview
GENCO Types
Bids & Bidding Strategies
SchedulingBillateral Trade
-
7/30/2019 EPE Trg 2005-06 Del2
19/52
19
Different types of GENCOs:
Thermal (coal-fired, gas-fired, oil-fired)Combined-cycle (CHP)
HydroPumped storage
Nuclear
Alternative sources
WindSolar...
Different types of strategies & market bids.
GENCOGENCO OverviewOverview
GENCOGENCO O iO i
-
7/30/2019 EPE Trg 2005-06 Del2
20/52
20
Passive bid (strategy) must-run bid:
Bid for selling electricity by any price.
Commonly used by hydro, nuclear and some types of
thermal producers.
GENCOGENCO OverviewOverviewBasic Bid
Quantity
Price
GENCOGENCO O iO er ie
-
7/30/2019 EPE Trg 2005-06 Del2
21/52
21
Thermal GENCO:
Marginal production cost bid:
GENCOGENCO OverviewOverviewBids & Bidding Strategies
Production cost curve:
C = a + b*Q + c*Q2
Quantity
Cost
Marginal production costs:
MC = = b + 2*c*QdC
dQ
Quantity
MarginalCosts
Marginal production
cost bid
Quantity
Price
GENCOGENCO O iOverview
-
7/30/2019 EPE Trg 2005-06 Del2
22/52
22
Thermal GENCO:
Incorporating technical constraints into bid Pmin & Pmax:
GENCOGENCO OverviewOverviewBids & Bidding Strategies
Modified marginal
production cost bids
Quantity
Price
Pmin Pmax
Possibi lity for Pminconstraint violation
Possibil ity for Pmaxconstraint violation
Quantity
Price
Pmin Pmax
Quantity
Price
Pmin PmaxIncorporated
must-run bid
GENCOGENCO OverviewOverview
-
7/30/2019 EPE Trg 2005-06 Del2
23/52
23
Hydro GENCO:
Popt is optimisation problem variable
No marginal production costs
GENCOGENCO OverviewOverviewBids & Bidding Strategies
Must-run bid
Quantity
Price
Popt
Popt is hydrology dependant
Must-run bid ensures all availableenergy is sold and no spillage occurs
Quantity
Price
Popt Pmax
If unithas water storagecapacity large enough it can
incorporate gaming into its bids:
-
7/30/2019 EPE Trg 2005-06 Del2
24/52
24
Combined GENCO:
It contains more than one unit (powerplant).
According to market rules it can submita single bid for all itscapacity.
After the auction onlyuniform price and combined quantity isknown.
In fully liberalised market environment there is no centraldispatching.
The problem of optimal unitcomitment (UC) and economicdispatch (ED) remains within combined GENCO.
Different approaches to UC and ED can be used:
Classical optimization and hydro-thermal coordination techniquesGenetic algorithms, evolutionary computation, ....
GENCOGENCO OverviewOverview
Scheduling
GENCOGENCO OverviewOverview
-
7/30/2019 EPE Trg 2005-06 Del2
25/52
25
Combined GENCO - example:
Contains several different type units:
GENCOGENCO OverviewOverviewScheduling
Quantity
Price
Pmax250 MW
Unit type Pmin [MW] Pmax [MW] MC [/MWh]
Hydro 0 50 0
30
60
Thermal
(coal-fired)70 130
Thermal
(gas-fired) 20 70
Submitted combined bid forsingle hour:
Sum of all available
capacity within GENCO
MCP
Final economic dispatchdone by GENCO:
Psold230 MW
Unit type P [MW]
Hydro 50
Thermal
(coal-fired)130
Thermal
(gas-fired)50
In this simple example the ED
is done by dispatching units
in MC merit order.
-
7/30/2019 EPE Trg 2005-06 Del2
26/52
3. GENCO in3. GENCO in DeregulatedDeregulatedEnvironmentEnvironment
General Overview
Power Market Simulator as Optimization Tool
Market Power
-
7/30/2019 EPE Trg 2005-06 Del2
27/52
27
Significant shift in optimization criteria:
GENCO inGENCO in DeregulatedDeregulated EnvironmentEnvironment
General Overview
REGULATED
Optimization criteria:operating costs minimization
DEREGULATED
Optimization criteria:profits maximization
The need for new
optimization approachesand tools.
Power Market
Simulator
-
7/30/2019 EPE Trg 2005-06 Del2
28/52
28
Bidding strategies optimization:
Power market simulator can be on of the tools used in optimalbidding strategy development for GENCO
With simulation of different scenarios and approaches the
optimalbidding strategy can be found:- by simulating GENCOs own different approaches and
tactics
- by modeling expected actions of the competition andreviewing impacts a suitable counter-action can be formed
- by forecasting different trends (market prices, scheduledplant decomissions or openings, ...)
GENCO inGENCO in DeregulatedDeregulated EnvironmentEnvironment
Power Market Simulator as Optimization Tool
GENCO i D l d E i
-
7/30/2019 EPE Trg 2005-06 Del2
29/52
29
GENCO inGENCO in DeregulatedDeregulated EnvironmentEnvironment
Market Power
MARKET POWER the ability of a single or a group ofparticipants to raise prices above competit ive levels.
Increases in
market prices
Increases in
profits
Due to the GENCOs focus on profitmaximization ahigh probability of market abuse exists.
The posibility (probability) of market abuse is usuallycorrelated with the GENCOs market power.
GENCO iGENCO i D l dD l t d E iE i t
-
7/30/2019 EPE Trg 2005-06 Del2
30/52
30
GENCO inGENCO in DeregulatedDeregulated EnvironmentEnvironment
Market Power - Detection
=
=N
i
iSHHI
1
210000
Common indices:
Hirschmann-Herfindahl (HHI) Index:
Lerner Index:
Si market share of producer i
N number of producers in the system
( )%100
=
P
MCP
P market priceMC short-term marginal costs ofproduction
- Static index
- Gives the level of concentration of production
- Lacks the information about possibility ofmarket power abuse
- Much more dynamic
- Troubles with obtaining trueproduction
short-term marginal costs
All indices share the low ability to detectparticipants abusive behavior patterns.
GENCO iGENCO i D l t dD l t d E i tE i t
-
7/30/2019 EPE Trg 2005-06 Del2
31/52
31
Power Market Simulator:
Ability to simulate reference scenario.Possibilty of dynamic observations of prices, market shares,quantities, profits...
Ability to forecastmarket trends and impacts of differentevents(planned plant maintainance outage, plantdecomission or opening,...)
Simulation tool can be extremely efficient in detecting possible
participants abusive behavior patterns.
GENCO inGENCO in DeregulatedDeregulated EnvironmentEnvironment
Market Power - Detection
Power market simulator already used as a monitoring tool at
Agency for energy a market monitoring body in Slovenia.
-
7/30/2019 EPE Trg 2005-06 Del2
32/52
4.4. PowerPowerMarket SimulatorMarket Simulator
General Overview
Technical & Financial Parameters
Bids
Bidding Strategies
Bilateral Contracts
Clearing Mechanism
Results
PP M k t Si l tM k t Si l t
-
7/30/2019 EPE Trg 2005-06 Del2
33/52
33
SUPPLY
PowerPowerMarket SimulatorMarket Simulator
General Overview Modell ing of Day-Ahead Market
Technical & Financial
Parameters
Bids
Bidding Strategies
DEMAND
Technical & Financial
Parameters
Bids
Bidding Strategies
Clearing Mechanism
System Parameters
Results:Quantities & Prices
Bilateral Contracts Bilateral Contracts
Billateral contracts
are alsoconsidered.
PPower M k tMarket Si l tSimulator
-
7/30/2019 EPE Trg 2005-06 Del2
34/52
34
SUPPLY:
Installed Capacity [MW]
Start-up Costs (thermal units)
Operating Costs (thermal units)
Hydrology reservoir inflow (hydro units)
DEMAND:
Load curve
Load forecast (short & long-term)
PowerPowerMarketMarket SimulatorSimulator
Technical & Financial Parameters
PowerPower MarketMarket SimulatorSimulator
-
7/30/2019 EPE Trg 2005-06 Del2
35/52
35
Bids are made of blocks of quantities and prices.
PowerPowerMarketMarket SimulatorSimulator
Bids
SUPPLY bid:(bid for sale)
DEMAND bid:(bid for purchase)
PowerPower MarketMarket SimulatorSimulator
-
7/30/2019 EPE Trg 2005-06 Del2
36/52
36
Bidding strategy is basically a series of appropriate one-hour bids for
corresponding number of simultaneous auctions.
PowerPowerMarketMarket SimulatorSimulator
Bidding Strategies
PowerPower MarketMarket SimulatorSimulator
-
7/30/2019 EPE Trg 2005-06 Del2
37/52
37
Bilateral contracts are entered in the form of their load over time.
The actual negotiated prices in bilateral contracts are unknown.
PowerPowerMarketMarket SimulatorSimulator
Bilateral Contracts
PowerPower MarketMarket SimulatorSimulator
-
7/30/2019 EPE Trg 2005-06 Del2
38/52
38
Clearing mechanismmimics the actual clearing process on day-aheadpower market.
For each trading hour the Market Clearing Price (MCP) and quantitiesof bought/sold electricity are determined according to market rules.
PowerPowerMarketMarket SimulatorSimulator
Clearing Mechanism
PowerPower MarketMarket SimulatorSimulator
-
7/30/2019 EPE Trg 2005-06 Del2
39/52
39
Price levels:
PowerPowerMarketMarket SimulatorSimulator
ResultsQuantities:
Profits:
Market shares:
-
7/30/2019 EPE Trg 2005-06 Del2
40/52
5.5. CaseCase StudiesStudies
Abusive Behavior Illustration
Scheduled Outages Planning
Pumped-storage Plant Operation
CaseCase StudiesStudies
-
7/30/2019 EPE Trg 2005-06 Del2
41/52
41
Test systemcharacteristics:
CaseCase StudiesStudies
Abusive Behavior IllustrationProducer Unit
Installed capacity
[MW]
Units marginal cost
[c/kWh]
1 300 1.74
2 50 3.041 200 2.392 150 2.833 50 5.221 300 1.522 100 5.65
3 50 6.96
C
B
A
500
600
700
800
900
1000
1100
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
Time (h)
Loa
d(MW
)
Test system24-hour inelasticdemand diagram:
CaseCase StudiesStudies
-
7/30/2019 EPE Trg 2005-06 Del2
42/52
42
DemandQuantity
Price
MCP
Extra profit for
producer A
Example of possible price manipulation by producer A inparticular trading hour:
CaseCase StudiesStudies
Abusive Behavior Illustration Witholding Strategy
MCP
DemandQuantity
Price
Producer A plays
unilateral witholding
strategy
Producer A
Costs
Profit
Producer B
CostsProfit
Producer C
CostsProfit
Producers A
unit 2 to be
witheld from
market
Price mark-up in
particular trading hour
CaseCase StudiesStudies
-
7/30/2019 EPE Trg 2005-06 Del2
43/52
43
CaseCase StudiesStudies
Abusive Behavior Illustration Witholding Strategy
Producer A Producer B Producer C
Reference scenario
A view fromthe simulator - 6th hour of the 24-hour period:
Price mark-up in
particular trading hour
Producer A is witholding
Producers A uni t
2 is to be witheld
CaseCase StudiesStudies
-
7/30/2019 EPE Trg 2005-06 Del2
44/52
44
0
50
100
150
200
Producer A Producer B Producer C
Pro
fit[*1000]
Reference scenario Producer A withholding unit 2
Results:
CaseCase StudiesStudiesAbusive Behavior Illustration Witholding Strategy
2
3
4
5
6
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
Time [h]
MCP[c/kWh]
Reference scenario Producer A withholding unit 2
MCP levels duringobserved period.
Raised MCP levels as a
result of Producer Aabusive behavior.
Raised profit levels as a
result of Producer Aabusive behavior.
CaseCase StudiesStudies
-
7/30/2019 EPE Trg 2005-06 Del2
45/52
45
Results fromPower Market Simulator:
CaseCase StudiesStudiesAbusive Behavior Illustration Witholding Strategy
Rised MCP levels as a result
of Producer A abusive
behavior.
Reference scenario
Producer A is witholding
CaseCase StudiesStudies
-
7/30/2019 EPE Trg 2005-06 Del2
46/52
46
Test system:
Supply:
CaseCase StudiesStudies
Scheduled Outages Planning
Producer type No. of producersCombined installed capacity
[MW]
Thermal (coal-fired) 7
11
Hydro 3 81212
Thermal (gas-fired)
857
386337Nuclear
2392Total:
Demand:Peak demand: 1891 MW
Base load: 692 MWAnnual consumption: 11.3 TWh
Neighbouring systems:2 neighbouring systems
Maximumtransfer capacity: 300 MW
Simulated period:
1 year
Nuclear producer is
planning a 1 month
maintenance outage.
CaseCase StudiesStudies
-
7/30/2019 EPE Trg 2005-06 Del2
47/52
47
CaseCase StudiesStudies
Scheduled Outages Planning
Changes in prices:
Moderate price spikes as a
result of outage in May.
Daily average prices
No scheduled outage
Daily average prices
Scheduled outage in May
Av. price
levels
No outage
[c/kWh]
Outage
[c/kWh]
In May 2.40
3.78
3.87 (+61%)
Yearly 3.91 (+3%)
Outage planned inMay
CaseCase StudiesStudies
-
7/30/2019 EPE Trg 2005-06 Del2
48/52
48
CaseCase StudiesStudiesScheduled Outages Planning Uncoordinated Planning
Changes in prices:
Daily average prices
No scheduled outage
Daily average prices
Scheduled outage in August
Av. price
levels
No
outage
[c/kWh]
Outage
[c/kWh]
In Aug. 2.63
3.78
7.47 (+184%)
Yearly 4.20 (+11%)
Outage planned inAugust
Special event: in August also one ofthe biggest thermal power plants
consideres 1-month scheduled
outage.!
CaseCase StudiesStudies
-
7/30/2019 EPE Trg 2005-06 Del2
49/52
49
CaseCase StudiesStudies
Pumped-storage Plant OperationPumped-storage plant characteristics:
- Maximumproduction capacity: 50 MW
- Maximumconsumption capacity: 50 MW- Upper reservoir capacity: unlimited
Minimal di fference between sell price
and buy price needed to cover the
energy transformation losses.
Bids submitted each trading hour:
Quantity Quantity
Price
Price
Buy bid Sell bid
CaseCase StudiesStudies
-
7/30/2019 EPE Trg 2005-06 Del2
50/52
50
CaseCase StudiesStud es
Pumped-storage Plant OperationResulting price levels during observed period:
During work-day peak hours the
electricity prices are high.During work-day night hours and
during the weekend the electricity
prices are lower.
Pumped-storageplant is taking
advantage of suchperiodic pricefluctuations.
CaseCase StudiesStudies
-
7/30/2019 EPE Trg 2005-06 Del2
51/52
51
Case Studies
Pumped-storage Plant OperationSimulation results (1 week period):
Bought hourly quantities: Sold hourly quantit ies:
During work-day peak hours theenergy is produced and sold to the
market.
During work-day night hours and
during the weekend the energy is
consumed for pumping water into
upper reservoir.
-
7/30/2019 EPE Trg 2005-06 Del2
52/52
THANK YOU FOR YOUR
ATTENTION