Sources of experience in wind energy
technology
Joint EU/IEA workshop on Experience curves: A
tool for Energy Policy Analysis and Design
Paris, IEA, 22 – 24 January 2003
Per Dannemand Andersen
Risø National Laboratory
per.dannemand@risoe.dk
Agenda of this presentation
This presentation is based on
• Chapter 3 in the EXTOOL final report
• Paper distributed for this workshop
Contents of the presentation
• Characteristics of wind power technology and their affects on
experience curves
• Innovation theory revisited and consideration on experience
curves and Progress Ratio
Characteristics of technical changes in
wind power technology
•
•
•
•
Up-scaling of turbines
Improved efficiency
Generations for technology / turbines
Types of innovation
- Innovations as improvements of one size of machine (same
generation and same platform)
- Innovation as up-scaling on a platform (within a generation)
- Introduction of a new platform (a new generation)
èAffects on experience curve and Progress Ratio
Up-scaling of wind turbines
10,000
100
Denmark
Germany
Germany
Rotor diameter in Meter
Denmark
Generator size in kW
1,000
100
10
1980
1985
1990
1995
2000
2005
10
1980
1985
1990
1995
2000
2005
Improved technical efficiency
Partly an affect of the up-scaling (Danish data)
Full load hours in kWh/kW
3000
2500
2000
1500
1000
500
0
1975
1985
1995
2005
Generations of technology
Only 3-bladed machines from largest Danish firms
15 - 30 kW
100
Each generation's market share
45 - 75 kW
90 - 130 kW
150 - 250 kW
75
300 - 400 kW
500 - 750kW
800 - 1000 kW
50
1300 - 2000 kW
Pct of all
25
1980
1985
1990
1995
2000
Types of innovations on wind turbines
• Innovations as improvements of one size of machine (same
generation and same platform)
• Innovation as up-scaling on a platform (within a generation)
• Introduction of a new platform (a new generation)
Innovation as improvements of a machine
Micon 250 machine
10
Specific cost (DKK)/(kWh/kW)
Micon 250
Power (All Danish)
1
10
100
1,000
Cum installation in Denmark
10,000
Innovation as improvements of a machine
Micon 250 machine
35
1600
30
1400
1200
MWh - kDKK
Meter
25
20
15
10
800
600
400
5
0
1985
1000
Rotor diameter
Hub height
1990
1995
2000
200
0
1985
Annual production MWh
Price in (2000) kDKK
1990
1995
2000
Innovation as up-scaling on a platform
10
Specific cost (DKK)/(kWh/kW)
Micon 600/43
Micon 750/44
Vestas 55
Vestas 75
Power (All Danish)
1
10
100
1,000
Cum installation in Denmark
10,000
Innovation as a new platform (generation)
10
Specific cost (DKK)/(kWh/kW)
Micon 250
Micon 400
Micon 600/43
Vestas 55
Nordtank 99
Power (All Danish)
1
10
100
1,000
Cum installation in Denmark
10,000
Innovation theory revisited
Scientific challenges
• Better theoretical understanding of what experience curves
actually express
- y-axis: cost of equipment, total cost of installation, electricity
production cost, ?
• Better theoretical understanding and concepts of learning in
contemporary use of “experience curve”
- Often only focus on “learning-by-doing” or “learning-byproducing”
èOpen up the black box of industrial learning and innovation
Innovation theory revisited
Some approaches to learning
• Innovation theory (evolutionary economy)
- technological paradigms and trajectories
• Innovation theory (technology sociology)
- taxonomies of technology and innovation
• Knowledge management (more recent paradigm)
- concepts of knowledge and learning
• Science sociology
- concepts of knowledge/learning and technology/innovation
- new concepts and paradigms in science and technology (”the
new science”)
Innovation theory revisited
Some terms
• Technology: artefact and knowledge
• Technology and knowledge
- product and process (Woodward, Perrow, etc)
- concept knowledge and manufacturing or process knowledge
- learning-by-doing and learning-by-using (Rosenberg)
- design knowledge and production knowledge (Vincenti)
• Codified knowledge and tacit knowledge (Polanyi, Vincenti)
• Embodied knowledge and disembodied knowledge
(Rosenberg)
• Experience based and science based learning (Burns&Stalker,
Rosenberg, Gibbons, Nowotny)
Innovation theory revisited
A model for sources of knowledge and learning
Learning-by-doing
Learning through R&D
embodied
embodied
Learning through manufacturing
disembodied
Wright’s type of learning
Conceptual knowledge
Manufacturing knowledge
Utilisation knowledge
for developing and
designing an artefact
for manufacturing an artefact
for using an artefact
embodied
disembodied
Learning through utilisation
Learning-by-using
Sources of cost reduction (and learning)
• Total installation cost
- Ex works turbine costs (price/cost data available)
• Costs of individual components: blades, generator, tower, etc.
-
(data not available)
Additional costs (some data available – model for estimation
available)
• O&M costs (models for estimation available)
- Regular maintenance, spare parts, insurance, interest rate,
administration, land lease, etc.
• Energy production
- Efficiency (ability to extract energy from the wind – data available)
- Siting (few data available - but not used)
èCost of electricity (estimate possible)
Implications for experience curves (y-axis)
• Learning traditional only linked to changes in manufacturing
knowledge
- improvement in cost of equipment: EUR/kW
- PR = f(cost of turbines, additional costs)
• Also learning linked to changes in concept knowledge
- improvement in cost of equipment’s ability to produce electricity
-
EUR/(annual kWh)
PR = f(cost of turbines, additional costs, annual production at
standard site)
• Also learning to changes in utilisation knowledge
- improvement in cost of electricity (EUR/kWh)
- PR = f(cost of turbine, additional cost, O&M, ability to micrositing)
-
interest rates and insurance rates could be added as they reflects the financial
community’s learning on the area
Affects on experience curve and PR
Type of cost applied on
the y-axis of the
experience curves
Progress
ratio
Types of experience included
Organizational
boundary
a) Ex-works prices of
wind turbines.
91%
Experience embodied and
disembodied as cost reduction of
equipment
Manufacturing
industry
b) Total installation
costs.
90%
Experience embodied and
disembodied as cost reduction of
equipment and installations
The whole business
cluster except that
related to operations
c) Specific electricity
cost (or specific) exworks prices.
86%
Experience embodied and
disembodied as cost reduction of
equipment and as improvements
in efficiency
Manufacturing
industry
d) Levelised cost of
electricity.
83%
Experience embodied and
disembodied as cost reduction of
equipment and installation, as
improvements in efficiency and
disembodied utilization
experience
The whole business
cluster
Estimate of future cost reduction of wind
electricity
Reductions 2000 to 2004: 15%
Source
Relative share
Design improvements – weight reductions of
wind turbines (and larger rotors?)
35%
Improved conversion efficiency - aerodynamic
and electric efficiency
5%
Economy of scale – gains from steady serial
production and optimisation of logistics
50%
Other contributions – foundations, grid
connection, O&M
10%
Total
100 %
Sources:
•IEA R&D Wind, Ad Hoc Group report on Long-term R&D Needs for Wind Energy
for the time Frame 2000 tp 2020.
• BTM Consult, World Market Updata 1999.