Electric Power Systems

Challenges in the Nordic Power

system
NordREG meeting, Arlanda, June 25, 2014
Lennart Sder
Professor in Electric Power Systems, KTH

Smart
Transmission
Systems

Power System
Operation and
Planning

Mehrdad
Robert (pd)
Mohamadreza
Amin
Mohammad
Na.
Marina
Afshin
Harold
Omar

Luigi
Raphael (pd)
Rujiroj (pd)
Almas
Tetiana
Yuwa
Wei
Vedran
Doan Tu
Jan

Lennart
Lars (pd)
Pia
Camille
Yalin
Angela
German
Ilan
Desta

Mikael
Ebrahim
Ilias
Richard

Smart
Transmission
Systems

Power System
Operation and
Planning

Electricity Market
analysis

Some Nordic Power System Challenges

Electric Power Systems

Power System
Dynamics and
Control

Power System
Dynamics and
Control

Electricity
Market analysis
Mohammad
Hector (pd)
Ekatrina
Mahir
Richard
Yaser
Kristina
Yelena
Ezgi

A. Pricing challenges
B. High share of
variable renewables
C. Smart-grids: Consumer interaction
D. Covering Peak load

Discussion:
1. DSO role and responsibility
2. Information exchange TSO role and
responsibility, access between
hubs/information-exchange systems
3. Market rules for demand response
4. Market rules energy services
5. Market rules micro-production
6. Market rules for charging electrical
vehicles
7. Unbundling etc

Pricing challenges in the future

Nordic Power Market with large
amounts of renewable low marginal
cost units.

Nordic hydro power

In 2012 the Nordic hydro power
production was 237 TWh.
This corresponds to an inflow of
around 4,5 TWh/week
A heavy rain every second
week implies 9 TWh

Nordic Power Supply

TWh 2012
Country

Nuclear Hydro Fossil

Wind

Bio

Total

Norway
Finland
Sweden
Denmark

0
22,1
61,2
0

1,6
0,5
7,1
10,2

0
9,9
10,8
2,3

147,8
67,7
161,6
29,4

142,9
16,6
77,7
0,02

2,4
17,9
4,6
16,3

New Nordic interconnections. ENTSO-E: Ten-Year

Network Development Plan 2012

20122016

2017-2022

Renewable energy
systems
Energy is produced where the resource is
The energy has to be transported to
consumption center
The energy inflow varies, which requires
storage and/or flexible system solutions

Example
Nordic hydro power (inflow) can
vary 86 TWh between different
years (2001 to 1996)
Transport from NV to SE +
continent
Energy balancing with thermal
power in i Dk+F+Ge+Pl+NL+Ee

This is valid for hydro power, wind power,

and solar power

Wind power gives the same

variations/uncertainties (and
solutions) as hydro power.
But: time perspective is much
shorter!

Transmission capacity plans

for Sweden

Pricing in power
systems

Sweden-neighbours: ca 10100 MW
(continuously 88 TWh/year)
Nordel-neighbours: ca 5500 MW (DC!)
Plans:
Jrpstrmmen-Nea, S-N, ~1000 MW
(South-West link, S-N, 2x600 MW cancelled)
Nordbalt, S-L, ~600 MW
New line to Gotland
Strengthening North to Finland
New cut 2 line, ev. DC
Sweden-neighbors: +~1600 MW

Thermal power systems: Price is set by

marginal cost
Hydro power: Price is set by the water
value = the expected marginal cost in the
future to which the water could be stored.
Wind power: Price is set by marginal cost
= negative subsidy, since subsidy is only
obtained at production (e.g. -2 Euro-cent if
certificate price is 2 Euro-cent.)

Pricing in power
systems:
Norway
Nearly only hydro power
(97%)
Price is set by the water value
= the expected marginal cost
in the future to which the
water could be stored.

Price is not set in

Norway!

Pricing in power
systems:
Denmark

Pricing in power
systems:
Sweden
Hydro + Nuclear + wind (90%)
Large part of the rest is CHP
(industrial and distr. heat)
Price is set by the water value =
the expected marginal cost in the
future to which the water could
be stored.

Price is not set in Sweden!

Pricing in power
systems:
Finland

2020: High wind power

(50%)
A part of the rest is CHP
(industrial and distr. heat)
When it is windy, then the
prices will be low

Nuclear + hydro + wind (58%now)

CHP + more nuclear in the
future
At wind and low demand, then
the prices will be low

High prices are often not

set in Denmark!

Prices are then often not set

in Finland!

Pricing in future Nordic

power systems:
Much more often: Prices are not
set by Nordic power plants.
At wind and low demand, then
the prices can be really low
There is then a challenge to get
prices that are high enough to
finance all power plant.
Enough transmission to high MC
areas essential

Results from new Swedish studies

on larger amounts of wind power
Lennart Sder, KTH
IEA Wind, Task 25, Meeting
Golden/Denver/NREL, April 9, 2014

Identified wind power

projects in Sweden:
Identified wind power projects:
45000 MW ( 100 TWh/year)
Today capacities:
Hydro Power: 16000 MW ( 65
TWh)
Nuclear power: 9000 MW ( 65
TWh)
total of 25000 MW

Swedish production: Total: 145,6 TWh

(same as 2011)

Current (2011) Swedish Power System

Source
Hydro
Nuclear

TWh 2011
66,0
58,0

Energy % 2011
44,9
39,5

MW-capacity 2011
16197
9363

Wind
Solar
CHP-Ind
CHP-distr.
Condens

6,1
0
6,4
9,4
1,01

4,2
0
4,4
6,4
0,7

2899
0
1240
3551
3197

Total

146,9

100

36447

Studied Swedish Power System

Source

TWh

Energy %

MW-max

Hydro
Nuclear
Wind
Solar
CHP-Ind
CHP-distr.

64,9
0
46,7
12,6
6,4
13,9

44,5
0
32,1
8,6
4,4
9,5

12951
0
15633
9849
1240
4126

Other
Total

1,3
139,9

0,9
100

5081
48180

Hydro power: Duration curve

Deficit situation
4

x 10
2.5

MWh/h

Consumption
Hydro power
Wind power
Solar power
CHP

1.5

0.5

Min level: 1875 MW: Needed during 860 hours

Max level: 12951 MW: Needed during 765 hours

50

100
150
200
250
300
Consumption from 14 January to 30 January

High wind decrease in CHP

350

400

Surplus situation (August)

Deficit situation (yearly basis)

Assumed need of OCGT
5000
16000

Consumption
Hydro power
Wind power
Solar power
CHP

Max level: 5081 MW

4000

Number of hours with need: 765 h

3500

Energy: 1.259 TWh

14000
12000
10000

3000
MWh/h

Energy level [MWh/h]

4500

2500
2000

8000
6000

1500
4000

1000
2000

500
0

500
1000
Number of hours with need of more production

20

1500

Cost for this: 2 re/kWh = 0,2 Eurocent/kWh

40

MWh/h

10000
8000
6000
4000

Max level: 9510 MW

8000

Number of surplus hours: 860 h

7000
6000
5000
4000
3000
2000

2000
0

9000

Energy level [MWh/h]

12000

220

Surplus during a year

Consumption
Hydro power
Wind power
Solar power
CHP

14000

200

Not OK: 83% limit, min-hydro, min-CHP

Surplus situation (August)

16000

60
80
100
120
140
160
180
Consumption from 1 August to 10 August

20

40

60
80
100
120
140
160
180
Consumption from 1 August to 10 August

Now OK: 83% limit, min-hydro, min-CHP

200

220

1000

Energy volume: 1.63 TWh

0

200

400
600
800
Number of hours with surplus/possible export

1000

1200

General transmission challenge

A. Voltage stability limits between

areas
B. Q-control important
C. More transmission required, but
low utilization time
D. Challenge to identify future
transmission capacity with less
nuclear
E. Detailed hydro simulation takes
10 minutes per week.

Surplus situation (August 1-10)

Surplus situation (August 1-10)

Wind Power production

Surplus situation (August 1-10)

Transmission situation (Jan 21 Feb 1)

Transmission situation (Jan 21 Feb 1)

Wind Power production

Transmission: Yearly duration : today 7000 MW

On transmission needs

A. Increase production in receiving

end (= thermal, currently OCGT)
B. Capacity is available, small
energy increase for first GW.
C. Since limit is voltage stability,
SVC may be enough
D. Discussion on exchange of AC to
DC
E. Optimization approach may be
interesting

Pricing in power systems

Pricing in presence of
variable sources (e.g. wind)

Weekly demand

2500
2000
1500
1000
500
0
10

19

28

37 46

55 64

73

82

91 100 109 118 127 136 145 154 163

120
100
80
60
40
20

3000

MWh/h

2500

Eu ro /
M Wh

13 25 37 49 61 73 85 97 109 121 133 145 157 169

3500
3000
2500
2000

Weekly
net
demand

1500
1000
500
0
-500 1

13

25 37 49 61

73 85 97 109 121 133 145 157 169

Solutions and competition

3500

Note: This is
independent of fixed
price etc

Weekly
demand
+ wind

13 25 37 49 61 73 85 97 109 121 133 145 157 169

Pricing in presence of
variable sources
Other units have to
cover the net load =
demand wind
The other units
production is controlled
by price!
more volatile prices

W Denmark 10/1-17/1 2005

0
1

4000
3500
3000
2500
2000
1500
1000
500
0

Weekly
net
demand

2000
1500
1000
500

Assume a system with large price

variation:
Three types of business
opportunities

100
80
Eu ro /
MWh

Pricing

Wind power has a

marginal cost zero
The production level is
depending on wind
speed
It is not easy to make
good long term (hours)
forecasts
Other units have to
cover the net load =
demand - wind

MWh/h

3500
3000

M W h /h

4000

Euro/
MWh

With an
assumption of
perfect competition:
Prices are based on
production marginal
costs
Low costs units are
used first
Higher load higher
prices:

60
40
20
0
1

13 25 37 49 61 73 85 97 109 121 133 145 157 169

0
-500 1

13

25 37 49 61

73 85 97 109 121 133 145 157 169

More trading with neighbors

Demand side management

Flexible plants

100
80
60

Thermal
pricing

40
20
0
1

13 25 37 49 61 73 85 97 109 121 133 145 157 169

There is a competition between these methods.

Much transmission reduces price changes less interest in DSM

Peak capacity
responsibilities
Norway: TSO-Statnet is responsible
for enough capacity
Finland: TSO-Fingrid is NOT
responsible for enough capacity
Sweden: TSO-Svenska Kraftnt is
NOT responsible for enough
capacity. But: up to 2000 MW
Denmark: TSO-Energinet.dk is
responsible for enough capacity

Peak capacity responsibilities

example 1
1. Assume that there is a capacity problem in
South Sweden and Denmark exports 1000 MW
to Sweden.
2. Assume that there is an outage in Denmark so
they have to decrease consumption.
3. According to EU legislation nondiscrimination Denmark cannot prioritize
Danish consumers before Swedish ones.
4. Does this has as a consequence that Denmark
is also responsible for Sweden?

Peak capacity responsibilities

example - 2

MWh / h

1. There are discussions of capacity payments to

a rather large volume in UK
2. Probably this then leads to comparatively low
energy prices compared to a case with no cap.
payments
3. Both Norway and Denmark plan new cables to
UK.
4. Does this mean that Denmark and Norway can
import and only pay the energy price?

Peak loads in Sweden

1992-2011

Year 1992 - 2011

High load reserves in Sweden

Selective capacity market
TSO responsible to purchase up to 2000 MW of reserves
for peak load situations.
There is a bidding process where the cheapest offers are
accepted.
Pricing:
The bids are placed on Nordpool spot. They are only used if all
other bids are accepted.
The Net Regulation Price should not be allowed to exceed
5,000 Euro/MWh.
TSO can immediately impose a Disconnection Price in The
event of Critical Power Shortage of 20 000 SEK/MWh 2300
Euro/MWh
Australia: Max price 12000 AUD 9000 Euro/MWh

Summary of (some)
Nordic market challenges
Risk for prices so low so power plants cannot
be financed
Large amounts of renewables often very
low prices
But still other units are needed
need of either (very) high prices or some
kind of capacity payment mechanism.
Large amount of transmission is one part
solution, but perhaps also large amounts of
solar/wind power on the other end?

Reserves in Sweden 2012-13

Consumers accepted to
reduce consumption
Company
Stora Enso AB
Hgans Sweden AB
Rottneros Bruk AB
Befesa Scandust AB
Vattenfall AB
Gteborg Energi AB
AV Reserveffekt
TOTAL

Area
3-4
4
3
4
3-4
3
3-4

MW
210
25
27
18
92
25
+ 67
464

Idea to market solution to last unit

There should be a (renewable) unit (biogas?) with MC as

High operation
cost (or bid
price) essential

Call it a market maker unit reduced need of cap. payment.

If DSM is cheaper then it will be used instead
As low LOLP as requested can be obtained (= size of unit)

Comments to
Nordic market challenges
How high costs should we allow to
make market work:
Costs for market-making plants
Costs for new lines are needed to
increase the number of participants and
to decrease risk of use of market power
Costs for IT solutions for consumer
flexibility since this is essential to make
the market work.

Some Nordic Power System Challenges

A. Pricing challenges
B. High share of
variable renewables
C. Smart-grids: Consumer interaction
D. Covering Peak load

Discussion:
1. DSO role and responsibility
2. Information exchange TSO role and
responsibility, access between
hubs/information-exchange systems
3. Market rules for demand response
4. Market rules energy services
5. Market rules micro-production
6. Market rules for charging electrical
vehicles
7. Unbundling

Comments to
Nordic market challenges
How high costs should we allow to
make market work:
Costs for market-making plants
Costs for new lines are needed to
increase the number of participants and
to decrease risk of use of market power
Costs for IT solutions for consumer
flexibility since this is essential to make
the market work.
(But what is the alternative?)

Comments to some
Nordic market challenges - 1
1. DSO role and responsibility
- More distributed power production,
- PV voltage control (rules or market?)
- Supply voltage to feeding grid
- Coordination with retailer concerning DSM
- Grid planning? (if DG pays connection, then why plan?)
2. Information exchange TSO role and responsibility,
access between hubs/information-exchange systems
- DSO:s possibility to help TSO
- Peak capacity: Market? Tender? Rely on neighbours?

Comments to some
Nordic market challenges - 2
3. Market rules for demand response
- Contract with DSO and/or retailer?
- React on prices and/or externally controlled?
- Time frame: minutes/hours
- How long time in advance are one informed?
4. Market rules energy services
- DSO:s possibility/responsibility to help TSO (Q-support)
- DG ancillary services requires ICT technology

Comments to some
Nordic market challenges - 4
7. Unbundling etc
- Not so rational if TSO should plan for all possible
expansion of wind power
- Off-shore grids
- How much should Swedish grid-users pay for power
transferred through Sweden?
- More detailed DSO-regulation?
- How to allow TSO:s to take risks for long-term
investments?
- How to make neighbor TSO:s to think about total social
surplus

Comments to some
Nordic market challenges - 3
5. Market rules micro-production
- How to handle voltage control at large scale?
- React on prices and/or externally controlled?
- Time frame: minutes/hours
- 100% access to grid or less? (cheaper)
6. Market rules for charging electrical vehicles
- Possibility to have same price everywhere?
- Different price/taxes special meter ?
- Externally controllable? If so: How?

Reserve slides
Lennart Sder, KTH
Arlanda, June 25, 2014

What is a good market? - 2

What is a good market? - 1

Static allocation efficiency (=are available

resources used as efficient as possible? E.g. in
deficit situations)
Management efficiency (= is the administrative
organization efficient?)
Plant operation efficiency (= is each plant
operated in an efficient and reliable manner?)
Production optimization (= correct merit order = is
cheapest possible operation, including externalities,
applied?)

Transaction cost efficiency (= the amount

of transactions costs, as measurements,
spread of information and contracts)
Dynamic investment efficiency (= are the
correct investments done at the right time?)
Risk management efficiency (= are risks
and uncertainties handled in an efficient
way?)
System reliability (= is it on a correct
level?)

C2: Surplus at 55 TWh wind+solar

Different assumptions on wind/solar 70-90% of consumption
(no nuclear)
10000
Exempel
Exempel
Exempel
Exempel
Exempel
Exempel

9000
8000

Effektniv [MW]

7000
6000

2a
2a
2b
2b
2c
2c

ej fjrrvrme
med fjrrvrme
ej fjrrvrme
med fjrrvrme
ej fjrrvrme
med fjrrvrme

Max ca 90 percent of
consumption
- No distr. heat.
- surplus:
- 0,61 TWh/year

5000
4000
3000
2000

Solution:
- Export
- District heating
- Other head
- El. vehicles
- Etc
- Inertia?

1000
0

500
1000
Antal timmar som verskottsproduktion i Exempel 2a-2c verskrids

1500

Model of CHP and district heating

a. In Part 2 there is the assumption that one can
use surplus electricity (MC=0) in the district
heating instead of spillage.
b. In Part 3 there is the assumption that one can
use surplus electricity (MC=0) to replace CHP
instead of spillage.
c. Reality is probably both with some limitations.

Model of district heating - 1

Model of district heating - 2

4000

12000

3500

10000

januari

februari

mars

april

maj

juni

juli

augusti

september

oktober

november

december

3000

8000
Effektniv [MW]

[MW]

2500
2000
1500

6000

4000

1000
500

2000
0

1000

2000

3000

4000
5000
Stockholm fjrrvremeleverens under 2012

6000

7000

8000

9000

a. Total district heating in Stockholm during 2012

per hour (City-Sder, including Sderenergi and
Nordvstra including E.ON Jrflla)

1000

2000

3000
4000
5000
6000
Timme p ret med niv ver 75 procent av frbrukning

7000

8000

9000

a. Amount of excess of wind+solar > 75% of

deman.

Model of district heating - 3

Model of district heating - 4

18000
16000
14000
12000

[MW]

Assumptions:
a. Swedish yearly disstrict heating same profile as
in Stockholm
b. Same level of fuel waste as today (18%) which
is not replace (negative MC)
c. Heat spillage or rkgaskondensering not
replaced with electricity.

10000
8000
6000
4000
2000
0

Spillvrme
0

1000

2000

Avfallsfrbrnning
3000

4000
5000
Vrmebehov under 2012

6000

7000

a. Result for the potential of heat replacement

8000

9000

Model of CHP- 1

Model of district heating - 5

DataKraftvrme:1Januari 31December
4500,00

12000

4000,00

10000

3500,00
8000

[MW]

3000,00
6000

2500,00
4000

2000,00

2000

1500,00

1000,00
0

1000

2000

3000
4000
5000
6000
Mjlig elfrbrukning (trappstegskurva) och tillgnglig elproduktion

7000

8000

a. Result for the replacement: 1.2 TWh used (out of

3,0 TWh available surplus)

9000

500,00
0,00

a. Yearly CHP production in 2011

Model of CHP- 3

Model of CHP - 2
4

x 10
2.5

MWh/h

Model:
a. 50% more CHP than today (from Profu report)
b. It is possible to decrease CHP down to 25% of
its original value during each hour
c. But the total amount of hydro + CHP muast be
17% of production (synchronous machine
requirement)
d. Excel sheet available to change these data.

Elfrbrukning
Vattenkraft
Vindkraft
Solkraft
Vrmekraft

1.5

0.5

50

100

150
200
250
Frbrukning frn 14 januari till 30 januari

300

a. Example of resulting CHP decrease:

350

400

Model of CHP- 4

Model of CHP- 5

Resultatkraftvrmeminskning:1Januari 31December
Resultatkraftvrmeminskning:Varaktighetskurva;Max=2267MW;
Energi=0,79TWh

2500,00

2500
2000,00
2000

1500,00

1500

1000
1000,00
500
500,00
0
0

1000

2000

3000

4000

5000

6000

7000

8000

0,00

a. Resulting CHP decrease: 0,79 TWh = 5,4%

Questions on CHP + District heating

a.
b.
c.
d.
e.

Amount of waste etc (MC <0) in future?

How much can one decrease CHP?
How fast can one decrease CHP?
Electric boilers in district heating?
Uncertainties on amount of CHP/district heating
in future?

a. Resulting CHP decrease: 0,79 TWh = 5,4%

Simulation method (Sweden isolated)

2: Detailed simulation

a. Simulate one week with details of Swedish

hydro system
b. Assume wind+solar+CHP from preliminary
simulation as fixed
c. Formulate a linear optimization problem
d. Problem has 359147 equations and 438985
variables

Preliminary simulation for a January week

Simulation method (Sweden isolated)
2: Detailed simulation

x 10
2

a. Objective function:

1.8
1.6

M Wh/h

1.4
1.2
1
Elfrbrukning
Vattenkraft
Vindkraft
Solkraft
Vrmekraft

0.8

b. where Hspill(k)= hydro spillage during hour k, Vspill(k)

= Wind spillage during hour k, EXP(k) = extra export
during hour k, IMP(k) = extra import during hour k,
K = number of hours in the studied period (=168).

0.6
0.4
0.2
0

20

Result after detailed simulation (normal

inflow)

140

160

x 10

1.5

1.5
MWh/h

MWh/h

120

Frbrukning
Vattenkraft
Vind + sol
vrig produktion

0.5

20

40

60
80
100
Timme under studerad vecka

120

140

160

Frbrukning
Vattenkraft
Vind + sol
vrig produktion

0.5

60
80
100
Frbrukning frn 1 november till 7 november

Result after detailed simulation (160% of

normal inflow)

x 10

40

20

40

60
80
100
Timme under studerad vecka

120

140

160