Document number Simulatie van het EU-wijde elektriciteitsysteem Modellering van de elektriciteitsmarkt en de hierin gebruikte elektriciteitsinfrastructuur Presentatie voor Kivi Niria Utrecht, 16 April 2013
Docum
ent
num
ber
Simulatie van het EU-wijde elektriciteitsysteem
Modellering van de elektriciteitsmarkt en de
hierin gebruikte elektriciteitsinfrastructuur
Presentatie voor Kivi Niria
Utrecht, 16 April 2013
DNV KEMA Market Modelling
Agenda
What is the European electricity system?
What kind of models are used?
What does the European Market Model look like?
What are capabilities and limitations?
Topics:
- Influence of extreme, long-lasting and EU-wide weather circumstances
in case of large RE share,
- Influence of energy storage,
- Influence of demand response.
2
DNV KEMA Market Modelling
Agenda
What is the European electricity system?
What kind of models are used?
What does the European Market Model look like?
What are capabilities and limitations?
Topics:
- Influence of extreme, long-lasting and EU-wide weather circumstances
in case of large RE share,
- Influence of energy storage,
- Influence of demand response.
3
DNV KEMA Market Modelling
What is the European electricity system?
4
DNV KEMA Market Modelling
What is the European electricity system?
The interconnected electricity system of 27 countries
Generation:
- Players, Generation portfolios
- Fuel mix, fuel prices
- CO2 prices/obligations
Transmission:
- ENTSO-E
- Reserve sharing
Markets:
- European rules, directives, regulation, ideas (market coupling, ….)
- ETS, EU energy targets
- Day-ahead, Intraday, Reserves (primary, secondary, tertiary),
5
DNV KEMA Market Modelling
Agenda
What is the European electricity system?
What kind of models are used?
What does the European Market Model look like?
What are capabilities and limitations?
Topics:
- Influence of extreme, long-lasting and EU-wide weather circumstances
in case of large RE share,
- Influence of energy storage,
- Influence of demand response.
6
DNV KEMA Market Modelling
Many models for market analysis and simulation
Dispatch models
Expansion models
Network models
Commercial models – own models
Depending on the objective a mix of models may be used
Explanation based on the DNV KEMA suite of models
What kind of models are used?
7
DNV KEMA Market Modelling
DNV KEMA use a modular portfolio of simulation tools for
comprehensive electricity market analysis and modeling,
combining several powerful simulation models.
Our models allow the simulation of …
- …commercial as well as technical market and system aspects
(Market and system operation).
- …the whole time horizon from long-term expansion planning up
to real time simulation of load flows and frequency control.
DNV KEMA modelling suite - Overall concept
8
DNV KEMA Market Modelling
Our modeling approach is based on a flexible portfolio of different
modules covering a wide range of technical and commercial aspects
PLEXOS, ProSym,
SYMBAD etc.
Short-term
market model
PLEXOS, SDDP etc.
Hydro
optimization
PSS/E, PowerFactory,
ELEKTRA etc.
Network modelling
(load flow
analyses)
KERMIT, PLEXOS
Balancing market
and frequency
control
KEEM, PLEXOS etc.
Long-term
expansion
planning
DNV KEMA modelling suite - models
9
DNV KEMA Market Modelling
Our models cover the whole range from many years ahead to real time
– and generation as well as network and system operation
PSS/E,
PowerFactory,
ELEKTRA
Many
years
1
Month
1
Day
1
Hour
15
Minutes
1
Minute
1
Sec
1
Cycle
1
ms
1
Year A few
years
Network
modelling
DC
PTDF
AC
Balancing market
and frequency control
Wholesale
market
Strategic
bidding
KERMIT
KEEM
KEEM KERMIT
SDDP
PLEXOS Cost-based
optimisation
Hydro
optimisation
NTC
SDDP
Time horizon and scope of simulation
10
Symbad
DNV KEMA Market Modelling
A suite of specialized tools with well-developed interfaces enables a
comprehensive analysis of all relevant issues
Network
modelling (load
flow analyses)
Short-term
market model
Balancing and
frequency control
Long-term
generation
expansion
Hydro
optimization
Installed
capacities
Production schedules
Production schedules
Balancing offers
Water- value
Adjusted network model
Adjusted reserve demand
Adjusted network model
Adjusted reserve demand
Interaction between models
11
DNV KEMA Market Modelling
Symbad
PLEXOS covers the most important issues and timeframe for market
simulation – if necessary we can use other models for further analysis
PSS/E,
PowerFactory,
ELEKTRA
Many
years
1
Month
1
Day
1
Hour
15
Minutes
1
Minute
1
Sec
1
Cycle
1
ms
1
Year A few
years
Network
modelling
DC
PTDF
AC
Balancing market
and frequency control
Wholesale
market
Strategic
bidding
KERMIT
KEEM
KEEM KERMIT
SDDP
Cost-based
optimisation
Hydro
optimisation
NTC
SDDP
Time horizon and scope of simulation
PLEXOS
12
DNV KEMA Market Modelling
Agenda
What is the European electricity system?
What kind of models are used?
What does the European Market Model look like?
What are capabilities and limitations?
Topics:
- Influence of extreme, long-lasting and EU-wide weather circumstances
in case of large RE share,
- Influence of energy storage,
- Influence of demand response.
13
DNV KEMA Market Modelling
PLEXOS provides for an integrated approach from long-term
expansion planning to spot and real-time markets
• Load data
• Reserve margin
• Exist. generation structure
• Network constraints
• New built options
• Fuel prices and availability
• CO2 prices & emission cap
• (Spin.) reserve requirements
• etc.
Input Data
• Generation expansion
• Transmission expansion
Optimal expansion plan
• Emission targets
• Fuel contracts / usage
• Hydro optimization
Optimal resource
allocation
• Unit commitment & dispatch
• Power & Reserve prices
• Fuel consumption &
emissions
• Load flows / Exchanges
Optimal dispatch
Long-Term Schedule
(capacity expansion)
Medium-Term Schedule
(inter-temporal constraints)
Short-Term Schedule
(unit commitment)
PLEXOS: Integrated optimization of capacity
expansion & dispatch
General market model based on PLEXOS
14
DNV KEMA Market Modelling
Market model based on least-cost unit
commitment and dispatch
Representation of the regional market,
model set-up adjusted to actual project
requirements
Aggregation of generation and
transmission on a regional level with
detailed technical and commercial
parameters
Use of NTC or PTDF
Co-optimization of energy and various
types of reserves
Regional data for wind generation (incl.
correlation of forecast errors)
ES
FR
IT
DE
BeNl
GB
SEE
CHAT
EE
PT
GB SC
GB E&WNL
BE
FR-1
FR-5
FR-3
FR-2
DE-3
DE-5
DE-6
ATCH
IT
DK-W
NO SW FI
CZ-W
DE-1
DE-2
DE-4
DE-7
FR-7
FR-4
FR-6
CZ-E
PL
DK-E
Simplified modeled areas
Planned interconnectors
GB SC
GB E&WNL
BE
FR-1
FR-5
FR-3
FR-2
DE-3
DE-5
DE-6
ATCH
IT
DK-W
NO SW FI
CZ-W
DE-1
DE-2
DE-4
DE-7
FR-7
FR-4
FR-6
CZ-E
PL
DK-E
Simplified modeled areas
Planned interconnectors
Stu
dy f
oc
us
:
Ma
rke
t P
ric
es
Stu
dy f
oc
us
:
Re
gio
na
l e
xc
ha
ng
es
European market model – features
15
DNV KEMA Market Modelling
Due to our extensive project experience in the NW-European region, most of the
required data is available to us, including up-to-date information on:
- Fuel and CO2 prices
- Installed capacity by fuel/technology
- Electricity demand and load profile
- Trans-border transport capacities
- etc.
We cross-check our information against public and private data sources, covering
European market model – market data
16
Public data sources: Private data sources:
• ENTSO-E
• TSOs
• Eurelectric (Europrog)
• Ministries / Regulators
• Generation companies
• Platts
• DNV KEMA’s power plant database
• DNV KEMA’s renewable generation
database (modelling volatility)
• DNV KEMA‘s network of local experts
• DNV KEMA’s executed projects
DNV KEMA Market Modelling
We collect and use data from more than one source and cross-check and
reconfirm the data accuracy.
We undertake a comparative analysis and filter-out the best data quality for
market modelling. Our selection criteria are:
- Consistency and plausibility,
- Cross-checks,
- Source reliability,
- Local expert knowledge
- In house knowledge of the power industry(1)
From our experience, this approach significantly minimises data gaps and the
need to make data estimations
We summarize the information in a concise way to facilitate decision making on
executive level
(1) DNV KEMA have many experts on generation (GT, CCGT, CHP, coal plant, nuclear and
renewable energy like wind, solar and biomass).
European market model – market data
17
DNV KEMA Market Modelling
Information heat rate curve; another option is
through load points and associated heat rates
Variable operations and maintenance costs
Units 7 -
Max Capacity 400 MW
Min Stable Level 160 MW
Heat Rate Base 562.2005 GJ/hr
Heat Rate Incr 4.61 GJ/MWh
Heat Rate Incr2 2.40E-08 GJ/MWh²
VO&M Charge 1.4 EUR/MWh
Start Cost 3600 EUR
Min Up Time 2 hrs
Min Down Time 5 hrs
Max Ramp Up 28 MW/min.
Max Ramp Down 28 MW/min.
Maintenance Rate 7.3 %
Forced Outage Rate 3.2 %
Mean Time to Repair 24 hrs
Min Time To Repair 2 hrs
Max Time To Repair 48 hrs
Information for random distribution of forced
outages
Power plant inputs
Other required inputs:
- Load data
- Representative wind and solar irradiation profiles
- Interconnector values
- Fuel and CO2 prices
Possibilities are there to also
incorporate modeling of:
- Reserve markets
- Various types of regulation
- Steam/ heat supply requirements
- Congestion within certain areas
(requiring detailed load and network
data)
A scenario approach
is adopted to assess
the impact of
changes in particular
parameters
Market model – typical model inputs
18
DNV KEMA Market Modelling
Category Potential model results Granularity
Generation - Dispatch schedules
Hourly,
by power plant
- Reserve provision
- Dispatch of hydro stations
- Costs (O&M, fuel, CO2)
- Revenues (for energy and reserves)
- Emissions
Grid - Load flows between potential market
areas Hourly, by line
- Congestion rent
Electricity prices
and commercial
exchanges
- Cost reflective electricity prices for
each market area
Hourly, by market
area
- Electricity im-/exports between and
neighbouring countries Hourly, by border
Reserves - Reserve provision Hourly for each
reserve type - Costs and prices for reserves
Market model – exemplary model outputs
19
DNV KEMA Market Modelling
Agenda
What is the European electricity system?
What kind of models are used?
What does the European Market Model look like?
What are capabilities and limitations?
Topics:
- Influence of extreme, long-lasting and EU-wide weather circumstances
in case of large RE share,
- Influence of energy storage,
- Influence of demand response.
20
DNV KEMA Market Modelling
What are capabilities and limitations?
Capabilities required for Roadmap-like studies
Future fuel mix (capacity and energy)
Network expansion
Costs of scenarios (based on targets)
Detailed dispatch
Reliability (will the system work?)
First three items are output of Least Cost Expansion analysis and
show the future electricity supply system
Last two can be investigated with short term commitment and
dispatch analysis and are used to check whether the model is a
reasonable representation of the real system
21
Capacity expansion
System operation
DNV KEMA Market Modelling
What are capabilities and limitations? - Capabilities
Example Least Cost Expansion analysis (Asian country)
22
0
5,000
10,000
15,000
20,000
25,000
30,000
201
0
201
1
201
2
201
3
201
4
201
5
201
6
201
7
201
8
201
9
202
0
202
1
202
2
202
3
202
4
202
5
Ge
ne
rati
ng
ca
pa
cit
y [
MW
]
New OCGT
New CHP
Wind
Refurbish
New Coal
Existing
Hydro
Peak load
DNV KEMA Market Modelling
14.71%
11.48%
0.00%
13.88%
2.30%
57.63%
Total Generation 2030 in NL: 140.01 TWh
CCGT
OCGT+Gas Engine
Gas-fired steam turbine
Coal-fired power plant
Uranium
RES
23
Example future fuel mix
Netherlands
What are capabilities and limitations? - Capabilities
DNV KEMA Market Modelling
What are capabilities and limitations? – Capabilities
0
2000
4000
6000
8000
10000
12000
0
10
20
30
40
50
60
70
80
90
0 24 48 72 96 120 144
Win
d p
rod
uc
tio
n [
MW
]
Ele
ctr
icit
y p
ric
e [€/M
Wh
]
Hours of one week
Wind (12 GW) Base Sce 1c Sce 1d
24
Example hourly electricity price
Netherlands 2020
DNV KEMA Market Modelling
What are capabilities and limitations? - Limitations
Major limitations:
A model is merely a model (what you may discover only later)
- Garbage in – garbage out (commercial info)
- Simulations often cost based (actual prices mostly higher, more volatile)
- Real market behaviour (and competition) difficult to predict and model
- Too much detail will blow the model
The world in 2050 is hard to predict
- Electricity market
- Fuel and CO2 prices
- Cost of technology
The present energy only market will most probably fail
- Market prices and contribution conventional plants drop (investment incentives?)
25
DNV KEMA Market Modelling
What are capabilities and limitations? - Limitations
Overcoming limitations:
Deviation in prices => Tuning with historical price information
Future fuel and CO2 prices
Cost of technology
New market model => Introduce capacity markets/mechanisms?
Often relative outcome is more important than absolute values:
What technologies will help us to meet energy targets?
Will the system work
The right choice of input and scenarios is a major challenge for
modellers
26
Sensitivity scenarios
DNV KEMA Market Modelling
Agenda
What is the European electricity system?
What kind of models are used?
What does the European Market Model look like?
What are capabilities and limitations?
Topics:
- Influence of extreme, long-lasting and EU-wide weather circumstances
in case of large RE share,
- Influence of energy storage,
- Influence of demand response.
27
DNV KEMA Market Modelling
Agenda
What is the European electricity system?
What kind of models are used?
What does the European Market Model look like?
What are capabilities and limitations?
Topics:
- Influence of extreme, long-lasting and EU-wide weather circumstances
in case of large RE share,
- Influence of energy storage,
- Influence of demand response.
28
DNV KEMA Market Modelling
Topics - Influence of weather circumstances
Influence of extreme, long-lasting and EU-wide weather circumstances
in case of large RE share,
- Security of supply
- Extra reserves for back up during times with low wind / solar
Simulation based on correlation between weather systems
Clear dependence of wind in NW Europe
Large differences possible for solar
Stochastic input or sensitivity analysis
29
DNV KEMA Market Modelling
Topics - Influence of weather circumstances
ECF Roadmap 2050 result
- Extra capacity (53% in 2030, 83% in 2050)
- Application of demand response
30
DNV KEMA Market Modelling
Topics - Influence of weather circumstances
Influence of extreme, long-lasting and EU-wide weather circumstances
in case of large RE share,
- Security of supply
- Extra reserves for back up during times with low wind / solar
Capacity credit of wind is limited
Much more total capacity with high RE share
Contribution Demand Response
Large extension of interconnection capacity
31
DNV KEMA Market Modelling
Agenda
What is the European electricity system?
What kind of models are used?
What does the European Market Model look like?
What are capabilities and limitations?
Topics:
- Influence of extreme, long-lasting and EU-wide weather circumstances
in case of large RE share,
- Influence of energy storage,
- Influence of demand response.
32
DNV KEMA Market Modelling
Topics - Influence energy storage
Storage at
Transmission
Distribution
End user
Electric
Thermal
33
Source: DG ENER
DNV KEMA Market Modelling
Topics - Influence energy storage
Large bulk energy (GW):
- Thermal storage, pumped hydro;
- Compressed Air Energy Storage (CAES);
- Chemical storage (e.g. hydrogen - large scale >100MW, up to weeks and
months)
Grid storage systems (MW) able to provide:
- Power: super-capacitors, Superconducting Magnetic Energy Storage (SMES),
flywheels,
- Energy : batteries such as Lead Acid , Li-ion, NaS & Flow batteries Energy &
Power: LA & Li-ion batteries
- Hydrogen Energy Storage / CAES / Pumped Hydro Energy Storage (PHES)
(small scale, 10MW< P > 100MW, hours to days)
End-user storage systems (kW):
- Power: super-capacitors, flywheels
- Energy: batteries such as Lead acid and Li-ion
- Energy & Power: Li-ion batteries
34
DNV KEMA Market Modelling
Topics - Influence energy storage
Roles of energy storage at transmission level:
Balancing demand and supply
- Seasonal and weekly variation
- Strong variability of solar and wind energy
Grid management
- Participation in balancing market
- Voltage and frequency regulation
Energy efficiency
- Arbitrage
- Reduction of RE curtailment
- Enabling base load plants (coal, nuclear)
35
DNV KEMA Market Modelling
Topics - Influence energy storage
Influence of storage on wind curtailment (example NL)
Storage helps but cannot prevent all curtailment in a cost effective way
36
0
1000
2000
3000
4000
5000
6000
7000
8000
9000
0 500 1000
Hours of the year
Win
d c
urt
ailm
en
t (M
W) No LSES
2000MW
4000MW
DNV KEMA Market Modelling
Topics - Influence energy storage
General conclusions from several investigations for the Netherlands:
Energy storage facility adds flexibility and available reserve capacity to the system.
As a result:
- Existing coal increases its generation (more hours at full load, so more efficient)
- Less renewable energy curtailment
- Lower total generation costs
Benefit of storage facility is mostly analysed from a system perspective. Different
operational strategy affects impact of storage:
- Storage to compensate (RES) imbalance in own portfolio.
- To increase company profits by applying arbitrage: storing energy during hours of low
electricity price, and generating power at hours of a high electricity price. (profit maximization.)
- A different operational strategy may change the observed benefit for the ‘B.V. Nederland’.
37
DNV KEMA Market Modelling
Agenda
What is the European electricity system?
What kind of models are used?
What does the European Market Model look like?
What are capabilities and limitations?
Topics:
- Influence of extreme, long-lasting and EU-wide weather circumstances
in case of large RE share,
- Influence of energy storage,
- Influence of demand response.
38
DNV KEMA Market Modelling
Topics - Influence of demand response
Demand Response: Complicated but necessary?
39
Source: IEA DSM 2012
DNV KEMA Market Modelling
Topics - Influence of demand response
Example: Influence of Impact of integration of 5 million EVs and 400.000
heat pumps on the electricity demand curve of Nordel for the year 2030
40
Source: EC DG-TREN IRENE-40
DNV KEMA Market Modelling
Topics - Influence of demand response
Example Germany: Possible influence of DR on demand
41
Source: Pöyry
DNV KEMA Market Modelling
Topics - Influence of demand response
Influence of DR on wholesale prices
Less volatile especially at high RE shares
Competition with storage
42
Source: Pöyry
DNV KEMA Market Modelling
Topics - Influence of demand response
Influence of DR can be investigated using Market Models
Cost of DR however is complicated and simultaneous optimisation is a
high burden on run time
Normally DR is examined by adjusting the load based on specific DR
investigations
DR seems indispensable for integration of large amounts of RE
43
DNV KEMA Market Modelling
Too much to investigate …..
What would we do without models?
44
Source: Rethinking
DNV KEMA Market Modelling
www.dnvkema.com
www.dnv.com