Implications of Limited Foresight on Wind Park Investments in Norway wholeSEM (28/04-2016) Arne Lind & Eva Rosenberg
Implications of Limited Foresight on Wind Park Investments in Norway
wholeSEM (28/04-2016)
Arne Lind & Eva Rosenberg
v
v
Overview
2
Introduction
• Motivation
Methodology
• TIMES-Norway
• NET-Model
Scenarios and assumptions
• Model variations
• Demand variations
• Energy prices
Model results
Concluding remarks
Part 1: Introduction
v
v
Motivation
3
Test the important of the perfect foresight assumption in the following
energy system models:
• TIMES-Norway
• NET-Model
In reality, decision makers do not act with full information about the
future
• In practice: A limited horizon is used for their decision making
• However, the future is predictable to some extent
From a market point of view, it is not reasonable to use a perfect
foresight planning horizon
• More realistic to use a planning horizon equal to the planning horizon of
investment decisions
Part 1: Introduction
v
v
Principle
4
Limited
foresight
w/ overlap
For
esig
htIn
form
atio
n
Perfect foresight
Demand, technology costs, energy prices, taxes, etc
Part 1: Introduction
Time horizon
v
v
5
Methodology
Part 2: Methodology
v
v
Modelling framework:
TIMES-Norway
6
Bottom-up, techno-economic optimisation model describing the Norwegian energy system
• Optional extension: Hard-linked version including Sweden
High time resolution
Model horizon from 2010 to 2050
Covers five Norwegian (and four Swedish) regions
Exchange of electricity between regions and neighbouring countries
Assumes perfect competition and perfect foresight and is demand driven
Energy demand is exogenous inputMore information: A. Lind et al. / Energy Policy 60 (2013) 364–377
Part 2: Methodology
v
v
Limited foresight – Implementation in
TIMES
7
Variable amount of years optimised in each solution step
The total model horizon will be solved by successive steps
• For each step the decision variables are defined only for a time subset
• In each step the periods to be optimised are advanced further in the future
• All periods before them are fixed to the solution of the previous step
The amount of overlapping years between successive steps can be
controlled by the user
• More realistic to use a planning horizon equal to the planning horizon of
investment decisions
The final (optimal) solution consists of accumulated variables
• Combination of results from each solution step
Part 2: Methodology
v
v
8
Scenarios and assumptions
Part 3: Scenarios and assumptions
v
vMain assumptions
Energy prices constant 2016-2050 (purchased
energy)
Present tax policy
• Energy taxes 2014, constant until 2050
• Biodiesel tax = fossil diesel after 2020
• Zero emission cars exempted from
purchase tax and VAT until 2020
• Purchase tax for vehicles based on CO2-
emissions,
power and weight until 2050
Other:
• Discount rate 4 % (higher for some
demand technologies)
• Enova support programs until 2020
• Common green certificate market
• Direct electric heating restricted as in TEK10
(building regulation)
9
0 100 200 300 400 500
SE2, SE3, FIN, RUS
DK
DE
SE1
NL
UK
Power export price (NOK2005/MWh)
Annaul average power price
2016-2050
2010-2014
Part 3: Scenarios and assumptions
v
v
Demand alternatives
• REF CenSES reference demand
projections
• DEM High industry activity
• PRI Increasing energy prices
Model alternatives
• Perfect foresight = No timestep
• Limited foresight I = 10 years timestep with 5
years overlap
• Limited foresight II = 5 years timestep with 2
years overlap
10
ScenariosPart 3: Scenarios and assumptions
Model altern
atives
Demand alternatives
Time step
REF REF REF REF DEM DEM DEM DEM PRI PRI PRI PRI
No
Global rate 4% 10% 4% 4% 4% 10% 4% 4% 4% 10% 4% 4%
Certificate price
NOK/MWh 165 165 200 120 165 165 200 120 165 165 200 120
10 yrs
Global rate 4% 10% 4% 4% 4% 10% 4% 4% 4% 10% 4% 4%
Certificate price
NOK/MWh 165 165 200 120 165 165 200 120 165 165 200 120
5 yrs
Global rate 4% 10% 4% 4% 4% 10% 4% 4% 4% 10% 4% 4%
Certificate price
NOK/MWh 165 165 200 120 165 165 200 120 165 165 200 120
v
v
11
Analyses
Part 4: Analyses
v
v
Total power production (2030)
12
140
145
150
155
160
165
170
175
180
185
190
REF DEM PRI
[TW
h]
PF TS10_5O TS5_2O 4% discount rate
Certificate price = 165
NOK/MWh
Part 4: Analyses
2.7 TWh
5.1 TWh
11 TWh
v
v
13
140
145
150
155
160
165
170
175
180
185
190
REF DEM PRI
[TW
h]
PF TS10_5O TS5_2O
Total power production (2030)
Part 4: Analyses
10% discount rate
Certificate price = 165
NOK/MWh
0.1 TWh0.2 TWh
6.3 TWh
v
v
Wind power production (2030)
14
0
5
10
15
20
25
30
35
REF DEM PRI REF DEM PRI REF DEM PRI REF DEM PRI
165 NOK/MWh 165 NOK/MWh 200 NOK/MWh 120 NOK/MWh
4% rate 10 % rate 4% rate 4% rate
[TW
h]
PF TS10_5O TS5_2O
Part 4: Analyses
v
v
Wind power production (2030)
15
0
5
10
15
20
25
30
35
REF DEM PRI REF DEM PRI REF DEM PRI REF DEM PRI
165 NOK/MWh 165 NOK/MWh 200 NOK/MWh 120 NOK/MWh
4% rate 10 % rate 4% rate 4% rate
[TW
h]
PF TS10_5O TS5_2O
Part 4: Analyses
v
v
Wind power production (2030)
16
0
5
10
15
20
25
30
35
REF DEM PRI REF DEM PRI REF DEM PRI REF DEM PRI
165 NOK/MWh 165 NOK/MWh 200 NOK/MWh 120 NOK/MWh
4% rate 10 % rate 4% rate 4% rate
[TW
h]
PF TS10_5O TS5_2O
Part 4: Analyses
v
v
Wind power production (2030)
17
0
5
10
15
20
25
30
35
REF DEM PRI REF DEM PRI REF DEM PRI REF DEM PRI
165 NOK/MWh 165 NOK/MWh 200 NOK/MWh 120 NOK/MWh
4% rate 10 % rate 4% rate 4% rate
[TW
h]
PF TS10_5O TS5_2O
Part 4: Analyses
v
v
Electricity consumption
18
-
20
40
60
80
100
120
140
160
1975 2000 2025 2050
Elec
tric
ity
use
(TW
h/y
ear)
Statistics
REF
DEM
PRI
Part 4: Analyses
v
v
Electricity consumption
19
100
105
110
115
120
125
130
135
140
145
150
1975 2000 2025 2050
Elec
tric
ity
use
(TW
h/y
ear)
Statistics
REF
DEM
PRI
Part 4: Analyses
v
v
Net power export (2030)
20
Part 4: Analyses
-10
0
10
20
30
40
50
REF DEM PRI REF DEM PRI REF DEM PRI REF DEM PRI
165 NOK/MWh 165 NOK/MWh 200 NOK/MWh 120 NOK/MWh
4% rate 10 % rate 4% rate 4% rate
[TW
h]
PF TS10_5O TS5_2O
v
v
21
Concluding remarks
Part 5: Conclusion
v
v
Concluding remarks
22
For predictive scenarios (forecasts):• Small effects of using limited foresight in TIMES-Norway
For explorative (external) or predictive (what if) scenarios:• Occasionally large effects of using limited foresight in TIMES-Norway
The effects of limited foresight is reduced with a higher discount rate
The model results show that limited foresight had larger effect on the production side than on the end-use side of the energy system:• Typical features of the production side:
Large, expensive investments
Long technological lifetime
Examples: Wind power, hydropower and export connections
• Typical features of the end-use side: Small, “inexpensive” investments
Shorter technological lifetime
Constant demand -> Change of technology (and energy carrier) possible
Significant reduced computational time• Over 80% for certain scenarios
Part 5: Conclusion
v
v
23
Thank you!
Arne Lind, PhD
Senior research scientist
Institute for Energy Technology
2027 Kjeller, Norway
e-mail: [email protected]
Part 5: Conclusion