Meta-Analysis of Net Energy Return for Wind Power Systems

Abstract

This analysis reviews and synthesizes the literature on the net energy return for electric power generation by wind turbines. Energy return on investment (EROI) is the ratio of energy delivered to energy costs. We examine 119 wind turbines from 50 different analyses, ranging in publication date from 1977 to 2007. We extend on previous work by including additional and more recent analyses, distinguishing between important assumptions about system boundaries and methodological approaches, and viewing the EROI as function of power rating. Our survey shows an average EROI for all studies (operational and conceptual) of 25.2 (n = 114; std. dev = 22.3). The average EROI for just the operational studies is 19.8 (n = 60; std. dev = 13.7). This places wind in a favorable position relative to fossil fuels, nuclear, and solar power generation technologies in terms of EROI.

Full text

Meta-analysis of net energy return for wind power systems
Ida Kubiszewski
a
,
*
, Cutler J. Cleveland
b
, Peter K. Endres
c
a
Gund Institute for Ecological Economics, University of Vermont, 617 Main Street, Burlington, VT 05405, USA
b
Department of Geography and Environment, Boston University, 675 Commonwealth Avenue, Boston, MA 02215, USA
c
JW Great Lakes Wind, LLC, 1900 Superior Avenue, Suite 333, Cleveland, OH 44114-4420, USA
article info
Article history:
Received 10 December 2008
Accepted 29 January 2009
Available online 25 February 2009
Keywords:
Energy return on investment (EROI)
Wind energy
Net energy
Input/output analysis
abstract
This analysis reviews and synthesizes the literature on the net energy return for electric power gener-
ation by wind turbines. Energy return on investment (EROI) is the ratio of energy delivered to energy
costs. We examine 119 wind turbines from 50 different analyses, ranging in publication date from 1977 to
2007. We extend on previous work by including additional and more recent analyses, distinguishing
between important assumptions about system boundaries and methodological approaches, and viewing
the EROI as function of power rating. Our survey shows an average EROI for all studies (operational and
conceptual) of 25.2 (n¼114; std. dev ¼22.3). The average EROI for just the operational studies is 19.8
(n¼60; std. dev ¼13.7). This places wind in a favorable position relative to fossil fuels, nuclear, and solar
power generation technologies in terms of EROI.
!2009 Elsevier Ltd. All rights reserved.
1. Introduction
Wind energy is one of the fastest growing energy systems in the
world. Global installed annual wind capacity grew by more than 31
percent from 1997 to 2007 as seen in the global annual installed
wind power capacity graph created by the Global Wind Energy
Council (Fig. 1), and will approach 100,000 MW by the end of 2008
[1]. The surge in wind energy is due to a combination of factors,
including reduction in the cost of wind turbines, volatile and high
prices for conventional forms of energy, the demand for non-
carbon forms of energy to mitigate the effects of climate change,
and favorable policies such as feed-in tariffs in Europe and
renewable portfolio standards in the United States. Despite the
impressive growth, wind energy still accounts for a small fraction of
total installed power generation.
Global electricity use is projected to double from 2005 to 2030,
with its share of final energy consumption rising from 17 to 22
percent [2]. How will the increase in demand be met, particularly in
light to the increasing pressure to reduce carbon emissions? A
variety of alternatives are proposed, including wind, biomass,
various forms of solar power, nuclear, fossil fuel systems with
carbon sequestration, among others. A variety of techniques are
available to compare power generation systems, including life cycle
analysis (LCA), learning or experience curves, and various forms of
economic and financial analysis.
Another technique for evaluating energy systems is net energy
analysis, which seeks to compare the amount of energy delivered to
society by a technology to the total energy required to find, extract,
process, deliver, and otherwise upgrade that energy to a socially
useful form. Energy return on investment (EROI) is the ratio of
energy delivered to energy costs [3]. In the case of electricity
generation, the EROI entails the comparison of the electricity
generated to the amount of primary energy used in the manufac-
ture, transport, construction, operation, decommissioning, and
other stages of facility’s life cycle (Fig. 2). Comparing cumulative
energy requirements with the amount of electricity the technology
produces over its lifetime yields a simple ratio for energy return on
investment (EROI):
EROI [ðcumulative electricity generatedÞ=
ðcumulative primary energy requiredÞ
This analysis reviews 119 wind turbines from 50 different
analyses, ranging in publication date from 1977 to 2007. We extend
the work of Lenzen and Munksgaard [4] by including additional and
more recent analyses, distinguishing between important assump-
tions about system boundaries and methodological approaches,
and viewing the EROI as function of power rating. Our survey shows
average EROI for all studies (operational and conceptual) of 25.2
(n¼114; std. dev ¼22.3). The average EROI for just the operational
studies is 19.8 (n¼60; std. dev ¼13.7). This places wind in
*Corresponding author. Tel.: þ1 860 729 1126; fax: þ1 802 656 2995.
E-mail addresses: ida.kub@gmail.com (I. Kubiszewski), cutler@bu.edu
(C.J. Cleveland), peterkendres@gmail.com (P.K. Endres).
Contents lists available at ScienceDirect
Renewable Energy
j o u rn a l h o me p a g e : ww w . e l s e v i e r . c o m / l o c at e / r e n e n e
0960-1481/$ – see front matter !2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.renene.2009.01.012
Renewable Energy 35 (2010) 218–225

a favorable position relative to fossil fuels, nuclear, and solar power
generation technologies in terms of EROI.
2. Importance of net energy
Economies with access to energy sources with a large energy
surplus have greater potential for economic expansion and/or
diversification than those with access to lower quality fuels [5]. The
history of the expansion of human civilization and its material
standard of living is directly linked to successive access to and
development of fuel sources with increasingly greater energy
surpluses [6]. The transitions from animate energy sources such as
plant, biomass, and draft animals, to wind and water power, to
fossil fuels and electricity enabled increases in per capita output
due to increases in the quantity of fuel available to produce non-
energy goods. The transition to higher surplus fuels also enabled
social and economic diversification as decreasing amounts of
energy were used in the energy securing process, meaning more
fuel was available to support non-extractive activities.
An EROI ¼1 is an absolute cutoff point for an energy source, the
point at which as much energy is used to deliver a unit of energy as
that unit yields. The EROI for crude oil has declined over time, and
may continue to do so as the resource base is depleted [7]. Smaller,
deeper, and more remote fields require more energy to develop.
Alternatives to crude oil such as ethanol from corn and coal
liquefaction deliver a lower EROI because a significant amount of
energy is needed to process the feedstock itself (corn or coal) [8].
Economic growth and rising standards of living may be more
difficult to maintain than they were 50 years ago when wealth was
produced by the massive energy surplus associated with the
discovery of the Earth’s great oil fields in the first half of the
twentieth century.
EROI is a tool of net energy analysis, a methodology that seeks to
compare the amount of energy delivered to society by a technology
to the total energy required to find, extract, process, deliver, and
otherwise upgrade that energy to a socially useful form. Net energy
analysis was developed in response to the emergence of energy as
an important economic, technological, and geopolitical force
following the energy price increases of 1973–74 and 1980–81.
Interest in net energy analysis was rekindled in recent years
following another round of energy price increases, growing
concern about energy’s role in climate change, and the debate
surrounding the remaining lifetime of conventional fossil fuels,
especially crude oil. It typically is assessed along with material
flows in life cycle analysis (LCA) of energy systems (e.g., [9]).
3. Methodological issues
3.1. System boundary
The choice about system boundaries is perhaps the most
important decision made in net energy analysis, and, for that
Fig. 2. Energy outputs and energy costs of a power generation facility.
Fig. 1. Global annual installed wind power capacity (Source: Global Wind Energy Council, http://www.gwec.net/, retrieved 9 September 2008.)
I. Kubiszewski et al. / Renewable Energy 35 (2010) 218–225 219

matter, in other analytical approaches a well. One of the most
critical differences among the diverse studies is the number of
stages in the life cycle of an energy system that are assessed and
compared against the cumulative lifetime energy output of the
system. These stages include the manufacture of components,
transportation of components to the construction site, the
construction of the facility itself, operation and maintenance over
the lifetime of the facility, overhead, possible grid connection costs,
decommissioning, and recycling of component materials. Energy
systems have external costs as well, most notably environmental
and human health costs, although these are difficult to assess in
monetary and energy terms. External costs are excluded from our
analysis.
3.2. Methodology
Two individual types of net energy analysis techniques are used
to calculate the net energy derived from wind power: process
analysis and input–output analysis. A third type, called hybrid
analysis, is a combination of the two. Process analysis assesses the
energy used directly in each successive step of the production of
a good or service. The energy input–output approach is more
comprehensive than process analysis and is analogous to and
derived from the input–output matrix used in standard economic
analyses. The assumptions, strengths, and weaknesses of the two
approaches have been discussed elsewhere in detail (e.g., [10,11]).
3.3. Operating characteristics
Many analyses must make important assumptions regarding the
operating characteristics of wind turbines. These include power
rating, assumed lifetime, and capacity factors. Changes in the
assumptions made about these factors, or deviations in actual
operating conditions from assumed conditions can have a signifi-
cant impact on results.
3.4. Conceptual versus empirical studies
Some studies use the theoretical or ideal operating character-
istics of a wind turbine that are derived from simulated or assumed
costs and operating conditions, e.g., a wind turbine of a given power
rating, costing a certain dollar amount, in a location with an
assumed wind power density, withan assumed capacity factor, and
so on. Of course, actual operating conditions always deviate from
assumed conditions. Empirical analyses rely on actual costs, oper-
ating conditions, and energy outputs, and thus provide a better
metric of an energy system’s contribution to a nation’s energy
supply. This article focuses primarily on empirical studies based on
actual operational data.
4. Results
Table 1 provides the detailed technical results of the wind
studies. The data include year and location of the study, key tech-
nical assumptions such as load factor, power rating and lifetime,
system boundaries, the type of net energy method used, certain
environmental variables, and EROI. The table also distinguishes
between studies based on actual performance of a wind system and
conceptual studies based on theory or simulations.
The average EROI for all studies (operational and conceptual) is
25.2 (n¼114; std. dev ¼22.3). The average EROI for just the oper-
ational studies is 19.8 (n¼60; std. dev ¼13.7).
5. Discussion
5.1. EROI and power rating
One of the striking features of the studies is that the average
EROI generally increases with the power rating of the turbine
(Fig. 3). Fig. 3 solely looks at operational wind turbines with
a power rating below 1 MW. Turbines above 1 MW were not
included due to lack of reliable data.
The results found in Fig. 3 may be due to a combination of
factors, including larger wind turbines creating economies of scale,
more efficient technology, greater rotor diameters increasing the
load factor, and higher hub heights providing access to greater wind
speeds. These factors are derived from those found to be important
in the creation of experience curves which show a decrease of wind
turbine price over time [12].
First, smaller wind turbines represent older, less efficient tech-
nologies. The new turbines nearing the megawatt (MW) range
embody many important technical advances that improve the
overall effectiveness of energy conversion. Such developments
include improved aerodynamic profiles, increasing the peak effi-
ciency approximately 8% between the early 1980 and early 1990
[13]. Although larger turbines require greater initial energy
investments in materials, the increase in power output due to
improvements more than compensates for this over the lifetime of
the turbine.
Second, larger turbines have a greater rotor diameter, which
determines swept area, probably the most important design
element that affects generating power potential. High annual
energy output will be difficult to obtain if the rotor diameter limits
the ability for the turbine to capture the wind power at lower wind
speeds, even if turbine power rating is respectable. Again, larger
rotors require greater initial energy investments in materials, but
the increase in power output more than compensates for this.
Fig. 4 demonstrates how an increase in rotor diameter produces
an increase in EROI. These conclusions are consistent with the
finding that commercial wind farms have moved towards larger
turbines that are less expensive on a levelized basis with regard to
installation, operation, and maintenance. The greater cost efficiency
of larger turbines is largely attributed to economies of scale and
learning by doing. Accordingly, under a similar assumption, larger
turbines have a greater EROI.
Another reason that larger turbines have a larger EROI is the
well known ‘‘cube rule’’ of wind power, i.e., the power available
from the wind varies as the cube of the wind speed. Thus, if the
wind speed doubles, the power of the wind increases eight times.
New turbines are taller than in earlier technologies, and thus
extract energy from the higher winds that exist at greater heights.
Fig. 5 shows the affect of wind speed on EROI. Surface roughness –
determined mainly by the height and type of vegetation and
buildings – reduces wind velocity near the surface. Over flat, open
terrain in particular, the wind speed increases relatively quickly
with height. EROI at location with high wind speeds is also often
affected due to limited accessibility to those areas. The installation
of wind turbines on mountaintops or far off shore, areas with
greatest wind speeds, significantly increases the input energy
required in transportation, construction, and connection to the
gird.
5.2. Comparison with other power systems
The EROI for wind turbines compares favorably with other
power generation systems (Fig. 6). Coal accounts for about 40% of
global electricity generation [2] and has an EROI of about 8.0. It is
a mature technology where technical improvements are not likely
I. Kubiszewski et al. / Renewable Energy 35 (2010) 218–225220

Table 1
Metadata analysis of wind power systems.
Ref Year of
study
Location Operational/
conceptual
EROI CO
2
Intensity
(gCO
2
/kWh)
Power rating
(kW)
Lifetime
(yr)
Capacity
factor (%)
Energy payback
time (yr)
Analysis
type
Scope as
stated
Turbine
information
On/off
shore
Rotor
diameter (m)
Hub height
(m)
Wind speed
(m/s)
[4] 1977 USA c 43.5 1500 30 50.4 I/O BCEMT 2 blades 60 50 10.5
[4] 1980 UK c 12.5 1000 25 18.3 I/O CM on 46 18.4
[4] 1980 UK c 6.1 1000 25 18.3 I/O CM 46 18.4
[4] 1981 USA o 1.0 3 20 26.8 I/O CMO 4.3 20 10.1
[4] 1983 Germany o 2.3 2 15 45.7 I/O CM
[4] 1983 Germany o 3.4 6 15 45.7 I/O CM
[4] 1983 Germany o 5.0 12.5 15 45.7 I/O CM
[4] 1983 Germany o 8.3 32.5 15 45.7 I/O CM
[4] 1983 Germany o 1.3 3000 20 45.7 I/O CM 2 blades 100 100
[4] 1990 Denmark o 71.4 95 20 25.2 PA M!3 blades on 19 22.6
[4] 1990 Denmark o 47.6 8.81 150 25 30.1 PA M
[4] 1990 Germany o 32.3 300 20 28.9 PA CMT 3 blades 32 34 11.5
[4] 1991 Germany o 18.9 45 20 33.5 PA M 12.5
[4] 1991 Germany o 32.3 225 20 39.9 PA M 27
[4] 1991 Germany c 27.0 300 20 39.9 PA M 32
[4] 1991 Germany c 22.2 3000 20 34.2 PA M 80
[4] 1991 Germany o 11.8 30 20 14.4 PA CGMOT 2 blades 12.5 14.8 13
[4] 1991 Germany o 20.4 33 20 29.4 PA M 2 blades 14.8 22 11
[4] 1991 Germany o 14.7 95 20 20.5 PA CGMT 3 blades on 19 22.6
[4] 1991 Germany o 19.6 95 20 20.5 PA M 3 blades 19 22.6
[4] 1991 Germany o 16.7 100 20 20.9 PA M 2 blades 34 24.2 8
[4] 1991 Germany o 20.4 150 20 25.6 PA M 3 blades 23 30 13
[4] 1991 Germany o 27.0 165 20 23.2 PA M 3 blades 25 32 13.5
[4] 1991 Germany o 18.9 200 20 21 PA M 3 blades 26 30 13
[4] 1991 Germany o 15.6 265 20 19 PA M 2 blades 52 30.5 8.5
[4] 1991 Germany o 20.8 450 20 20 PA GM 3 blades 35 36 18
[4] 1991 Germany o 15.4 3000 20 30.4 PA GM 2 blades 100 100 12
[4] 1991 Japan o 4.0 71.7e 100 20 31.5 I/O CMT
[4] 1992 Germany o 11.2 0.3 20 38.8 PA CDMOT 3 blades 1.5 11.6 9
[4] 1992 Germany c 37.0 300 20 41.9 PA CDGMOT 3 blades 32 34
[4] 1992 Japan o 2.9 95.6e 100 20 31.5 I/O CMOT
[4] 1992 Japan o 30.3 33.7 100 30 28 I/O CMOT 30 13
[4] 1992 Japan o 18.5 100 30 40 I/O CMOT 1983 30 10
[4] 1993 Germany o 21.7 11e 300 20 22.8 PA CDMOT
[4] 1994 Germany o 18.2e 500 20 27.4 I/O CM
[4] 1994 Germany o 45.5 300 20 22.8 PA MO(D)
[4] 1994 Germany o 14.7 8.1 500 20 36.5 PA M 2/3 blades 39 41
[4] 1995 UK o 23.8 9.1 350 20 30 PA M 3 blades 30 30 15
[4] 1996 Switzerland o 3.1 52 30 20 7.9 PA CDGMOT 2 blades 12.5 22 11.4
[4] 1996 Switzerland o 5.0 28 150 20 7.6 PA CDGMOT 3 blades 23.8 30
[4] 1996 Germany o 14e 1000 20 18.5 PA CMO 3 blades 54 55
[4] 1996 Germany o 22e 1000 20 18.5 I/O CMO 3 blades 54 55
[4] 1996 UK o 25 6600 20 29 I/O CDMO
[4] 1996 Japan o 2.3 123.6e 100 30 20 I/O CMO
[4] 1996 Japan o 2.2 123.7e 100 20 18 I/O CMO 1984 30
[4] 1996 Japan o 5.8 47.4e 170 20 22.5 I/O CMO 27
[4] 1996 Japan o 8.5 34.9e 300 20 18 I/O CMO 28
[4] 1996 Japan o 11.4 24.1e 400 20 18 I/O CMO 31
[4] 1996 Germany o 8.3 17 100 20 31.4 PA CMO 3 blades 20 30
[4] 1996 Germany c 28.6 10 1000 20 36.2 PA CMO 3 blades 60 50
[4] 1997 Denmark o 8.3 15 20 20.5 I/O CMO 1980 10 18
[4] 1997 Denmark o 8.1 22 20 19.9 I/O CMO 1980 10.5 18
[4] 1997 Denmark o 10.0 30 20 19 I/O CMO 1980 11 19
[4] 1997 Denmark o 15.2 55 20 20.6 I/O CMO 1980 16 20
[4] 1997 Denmark o 27.0 600 20 26.5 I/O BCDEGMOT 3 blades 47 50 15
(continued on next page)
I. Kubiszewski et al. / Renewable Energy 35 (2010) 218–225 221

Table 1 (continued)
Ref Year of
study
Location Operational/
conceptual
EROI CO
2
Intensity
(gCO
2
/kWh)
Power rating
(kW)
Lifetime
(yr)
Capacity
factor (%)
Energy payback
time (yr)
Analysis
type
Scope as
stated
Turbine
information
On/off
shore
Rotor
diameter (m)
Hub height
(m)
Wind speed
(m/s)
[4] 1997 Denmark c 33.3 1500 20 38.4 I/O CMO 3 blades off 64 55 17
[4] 1997 Denmark o 50.0 15.9 400 20 22.8 PA M(O)
[4] 1998 Argentina c 5.9 42 2.5 20 22 PA CMT(O)
[4] 1998 Argentina c 8.3 29 30 20 22 PA CMT(O)
[4] 1998 Argentina c 12.5 18 225 20 22 PA CMT(O)
[4] 1998 Germany o 23.8 500 20 29.6 PA CGMOT 3 blades 40.3 44
[4] 1998 Germany o 15.4 500 20 29.6 I/O CGMOT 3 blades 40.3 44
[4] 1998 Germany o 21.7 1500 20 31 PA CGMOT 3 blades 66 67
[4] 1998 Germany o 14.1 1500 20 31 I/O CGMOT 3 blades 66 67
[4] 1999 Germany c 26.3 1500 20 31 PA CDGMOT 66 67
[4] 1999 India c 31.3 1500 20 45.9 PA CDGMOT E-66 66 67
[21] 1999 USA o 23.0 14.4 342.5 30 24 I/O (B)CDMOT Kenetech KVS-
33
on 32.9 36.6
[21] 1999 USA o 17.0 20.2 600 20 31 I/O (B)CDMOT Tacke 600e on 46.0 60.0 6.1
[21] 1999 USA o 39.0 8.9 750 25 35 I/O (B)CDMOT Zond Z-46 on 46.0 48.5
[4] 2000 Denmark o 51.3 16.5 500 20 40 0.39 MTCGOD 3-blades off 39 40.5 16
[4] 2000 Denmark o 76.9 9.7 500 20 40 0.26 MTCGOD on 41.5
[22] 2000 Italy o 7.7 36.15 2500 I/O MCO
[4] 2000 Belgium o 30.3 9.2e 600 20 34.2 PA DM(O)
[4] 2000 Belgium o 27.8 7.9e 600 20 34.2 I/O DM(O)
[4] 2001 Japan o 6.3 39.4 100 25 34.8 I/O CMT 30 30
[4] 2001 Brazil o 14.5 500 20 29.6 I/O CGMOT 3 blades; E-40 40.3 44
[23] 2002 USA c 80.0 8.16e TCO
[24] 2003 Canada c 123.5 10 500 20 PA MCTOD
[24] 2003 Canada c 125.8 7.1 500 20 PA MCTOD
[24] 2003 Canada c 109.6 3.7 500 20 PA MCTOD
[25] 2004 Germany c 8.4 45 500 PA-I/O MTCOD Enercon E-40 on 40.3 44 7.5
[25] 2004 Germany c 7.8 48 500 PA-I/O MTCOD Enercon E-40 on 40.3 55 7.5
[25] 2004 Germany c 6.2 61 500 PA-I/O MTCOD Enercon E-40 on 40.3 55 7.5
[25] 2004 Germany c 4.7 81 500 PA-I/O MTCOD Enercon E-40 on 40.3 55 7.5
[25] 2004 Germany c 4.9 77 500 PA-I/O MTCOD Enercon E-40 on 40.3 65 7.5
[25] 2004 Germany and
Brazil
c 22.5 15 500 PA-I/O MTCOD Enercon E-40 on 40.3 44 7.5
[25] 2004 Germany and
Brazil
c 21.2 16 500 PA-I/O MTCOD Enercon E-40 on 40.3 55 7.5
[25] 2004 Germany and
Brazil
c 16.4 20 500 PA-I/O MTCOD Enercon E-40 on 40.3 55 7.5
[25] 2004 Germany and
Brazil
c 12.0 27 500 PA-I/O MTCOD Enercon E-40 on 40.3 55 7.5
[25] 2004 Germany and
Brazil
c 12.4 26 500 PA-I/O MTCOD Enercon E-40 on 40.3 65 7.5
[25] 2004 Germany and
Brazil
c 27.7 8 500 PA-I/O MTCOD Enercon E-40 on 40.3 44 7.5
[25] 2004 Germany and
Brazil
c 25.7 8 500 PA-I/O MTCOD Enercon E-40 on 40.3 55 7.5
[25] 2004 Germany and
Brazil
c 20.0 10 500 PA-I/O MTCOD Enercon E-40 on 40.3 55 7.5
[25] 2004 Germany and
Brazil
c 15.6 13 500 PA-I/O MTCOD Enercon E-40 on 40.3 55 7.5
[25] 2004 Germany and
Brazil
c 16.4 12 500 PA-I/O MTCOD Enercon E-40 on 40.3 65 7.5
[25] 2004 Brazil c 32.7 3 500 PA-I/O MTCOD Enercon E-40 on 40.3 44 7.5
[25] 2004 Brazil c 30.0 3 500 PA-I/O MTCOD Enercon E-40 on 40.3 55 7.5
[25] 2004 Brazil c 24.0 3 500 PA-I/O MTCOD Enercon E-40 on 40.3 55 7.5
[25] 2004 Brazil c 18.9 4 500 PA-I/O MTCOD Enercon E-40 on 40.3 55 7.5
[25] 2004 Brazil c 18.9 4 500 PA-I/O MTCOD Enercon E-40 on 40.3 65 7.5
[25] 2004 Brazil c 40.0 2 500 PA-I/O MTCOD Enercon E-40 on 40.3 44 7.5
I. Kubiszewski et al. / Renewable Energy 35 (2010) 218–225222

to significantly improve generation efficiency, and thus the EROI
will remain fairly stable. Adding carbon sequestration technology
to coal combustion will increase the energy cost of power genera-
tion. Hydropower has a relatively high EROI (about 12), but on
a global scale it has a modest potential for expansion. The average
EROI for hydropower is based on a literature review of published
life cycle energy assessments (N¼7). Similar literature reviews
were done for coal (N¼12), solar thermal (N¼9) and geothermal
(N¼11) power generation systems.
The comparison with nuclear power is complicated by a number
of factors. The system boundary looms large for nuclear power
because the fuel cycle has many steps, and because many of the
important stages are upstream (mining, milling, enrichment) or
downstream (decommissioning, waste disposal) from the genera-
tion stage. The data presented in Fig. 3 are from Lenzen’s [14]
comprehensive survey of the life cycle energy and greenhouse gas
emissions of nuclear energy based on 52 unique analyses. The
complete sample yields an average EROI ¼15.8, but with a very
large standard deviation (28.0). The 52 studies exhibit a wide range
in the number of stages that are assessed, which explains some of
the huge variation in EROI. Most of the studies with EROI in the
upper range shown in Fig. 3 exclude multiple stages of the fuel
cycle, and thus generate unrealistically high EROI. Excluding those
outliers produces an average EROI ¼9.1, but still with a large
standard deviation (8.0). Readers should also note that two-thirds
of the analyses in Lenzen’s nuclear review date from 1980 or earlier,
and thus do not represent nuclear power plants currently being
built, or any plants that will be built in the future. Suffice it to say
that there remains significant uncertainty regarding the energy
costs associated with nuclear power.
The EROI for wind is demonstrably higher then the current EROI
for photovoltaic (PV) power generation. A literature (N¼62) review
of LCAs and net energy assessments for PV systems from 1997
through 2007 produced an average EROI of about 6.5 (s.d. ¼4.7).
The vast majority of these studies were simulations that assumed
specific lifetimes, locations, module efficiencies, solar intensities,
and other operating characteristics. Like the wind and nuclear
analyses, the PV studies exhibit a wide range in terms of scope, with
decommissioning and recycling stages often excluded. Ceteris par-
ibus, the ongoing improvements in PV module efficiency will tend
to improve the EROI over time.
5.3. Challenges facing wind energy
Does the high EROI for wind power presented here guarantee
that wind will assume a major role in the world’s power generation
system? There are a number of issues surrounding wind energy
that require resolution before that happens. These issues have been
discussed in detail elsewhere, and are summarized here:
%The dramatic cost reductions in the manufacture of new wind
turbines that has characterized the past two decades may be
slowing [15] due to a variety of economic, financial, and tech-
nical reasons. Recently this is particularly true in light of the
rising energy and commodity prices, which are slowly esca-
lating turbine costs. The rising global demand for turbines is
also driving prices upward.
%The uncontrolled, intermittent nature of wind poses unique
challenges to grid management relative to operator-controlled
(baseload) resources such as coal, gas, or nuclear generation
[16].
%Much of the wind resource base is located in remote locations,
so costs exist in getting wind-generated electricity from the
local point-of-generation to a potentially distant load center.
[25] 2004 Brazil c 40.0 2 500 PA-I/O MTCOD Enercon E-40 on 40.3 55 7.5
[25] 2004 Brazil c 32.7 2 500 PA-I/O MTCOD Enercon E-40 on 40.3 55 7.5
[25] 2004 Brazil c 25.7 3 500 PA-I/O MTCOD Enercon E-40 on 40.3 55 7.5
[25] 2004 Brazil c 25.7 3 500 PA-I/O MTCOD Enercon E-40 on 40.3 65 7.5
[26] 2004 Germany c 14.8 5000 20 50 0.33 PA MTCOD Repower
Systems AG
off 126.5 95 9.2
[27] 2004 Germany c 70.0 500 20 0.29 PA MCTO Enercon E-40 on 40.3 44
[27] 2004 Germany c 53.0 500 20 0.38 PA MCTO Enercon E-40 on 40.3 55
[27] 2004 Germany c 38.0 500 20 0.53 PA MCTO Enercon E-40 on 40.3 65
[27] 2004 Germany c 64.0 1500 20 0.32 PA MCTO Enercon E-66 on 66 67
[27] 2004 Germany c 50.0 1500 20 0.4 PA MCTO Enercon E-66 on 66 67
[27] 2004 Germany c 39.0 1500 20 0.52 PA MCTO Enercon E-66 on 66 67
[28] 2005 Japan c 29.5 300 30 20 PA-I/O CMO
[28] 2005 Japan c 20.3 400 30 20 PA-I/O CMO
[29] 2006 Italy o 19.2 14.8e 7260 20 I/O MTCOD on 50 55
[30] 2006 Germany c 30.0 10.2 1500 MTCOD
[30] 2006 Germany c 32.7 8.9 2500 MTCOD
[30] 2006 Germany c 29.4 10.2 1500 MCOTD on
[30] 2006 Germany c 32.3 8.9 2500 MCOTD off
Notes: I/O ¼Input–output-based analysis, PA ¼Process analysis, c ¼conceptual, o ¼operating, B ¼Business management, M ¼Manufacture, T ¼Transport, C ¼Construction, G ¼Grid connection, O ¼Operation and maintence,
D¼Decommissioning, e ¼CO
2
equivalents including CH
4
and N
2
O, () ¼partly covered.
I. Kubiszewski et al. / Renewable Energy 35 (2010) 218–225 223

%The remoteness of the wind resource base also generates
increased costs of developing land with difficult terrain or that
which is increasingly removed from development infrastruc-
ture (such as major roads, rivers, or rails capable of trans-
porting the bulky and heavy construction equipment). Little is
known about the extent of these costs.
%At about 6 or 7 MW per square kilometer of net power
potential, wind plants are necessarily spread-out over a signif-
icant land area [17]. Thus, wind plants must compete with
alternative uses of these land resources. This is especially true
when the land is a significant source of aesthetic and/or
recreational value.
%Government subsidies have spurred the development of wind
energy [18]. But subsidies are always subject to political whims,
and thus constitute a significant issue for the wind industry,
creating uncertainty for long-term planning and preventing
faster market development.
%There is also concern about the impacts of wind energy on
birds and bats [19]. Considerable additional research on oper-
ational wind facilities is required to provide a comprehensive
assessment of the potential magnitude of these risks.
None of these challenges are necessarily insurmountable.
Indeed, some of them may be relatively modest in cost terms when
fully assessed. The point here is simply that an EROI is crucial but is
not independently a sufficient condition for the continued wide-
spread expansion of wind energy.
5.4. Difficulties in calculating EROI
Our analysis illustrates the longstanding observation that EROI
is sensitive to the choice of system boundaries [10,20]. Studies
using the input–output analysis have an average EROI of 12 while
those using process analysis an average EROI of 24. Process analysis
typically involves a greater degree of subjective decisions by the
analyst in regard to system boundaries, and may be prone to the
exclusion of certain indirect costs compared to input–output
analysis [10].
Operational wind turbines offer the best opportunity to calcu-
late real EROI, as concrete data for input and output parameters can
be used. However, practical obstacles interfere with data avail-
ability. For example, data retrieval related to turbine transport or
construction material/volume becomes complicated by the
involvement of multiple contractors, inconsistencies in record
keeping, and other factors. Additionally, wind turbine developers or
owners/operators may be unwilling to provide data due to confi-
dentiality and competitive restrictions (this is especially true for
production data), or the time required to collect information. These
Fig. 3. EROI for operational wind turbines below 1 MW as a function of power rating in kilowatts.
Fig. 4. EROI for operational wind turbines below 1 MW as a function of rotor diameter
in meters.
Fig. 5. EROI for operational wind turbines below 1 MW as a function of wind speed in
meters per second.
I. Kubiszewski et al. / Renewable Energy 35 (2010) 218–225224

constraints give rise to the need for estimation, which increases the
level of uncertainty even for operational turbines.
6. Conclusions
This analysis reviews the extant literature on the net energy
return from wind energy systems, ranging in date from 1977 to
2007. Our survey shows average EROI for all studies (operational
and conceptual) of 25.2 (n¼114; std. dev ¼22.3). The average EROI
for just the operational studies is 19.8 (n¼60; std. dev ¼13.7). This
places wind in a favorable position relative to other forms of power
generation, and suggests that wind energy could yield significant
economic and social benefits relative to other power generation
systems. Ongoing technical progress in wind energy technology
will undoubtedly lead to further energy cost reductions. However
technical progress and a high EROI are not sufficient conditions for
the continued rapid expansion of wind energy. A number of social,
economic, environmental and regulatory issues need resolution.
References
[1] US, China & Spain lead world wind power market in 2007. Global Wind Energy
Council; 2007.
[2] World energy outlook. Paris, France: International Energy Agency; 2008.
[3] Cleveland CJ, Costanza R, Hall CAS, Kaufmann R. Energy and the US economy:
a biophysical perspective. Science 1984;225:890–7.
[4] Lenzen M, Munksgaard J. Energy and CO
2
life-cycle analyses of wind turbines
– review and applications. Renewable Energy 2002;26:339–62.
[5] Cottrell WF. Energy and society: the relation between energy, social change,
and economic development. New York: McGraw-Hill; 1955.
[6] Hall CA, Kaufmann RK, Cleveland CJ. Energy and resource quality: the ecology
of the economic process. New York: Wiley-Interscience; 1986.
[7] Cleveland CJ. Net energy from the extraction of oil and gas in the United States.
Energy 2005;30:769–82.
[8] Cleveland CJ. Ten fundamental principles of net energy. In: Saundry P, editor.
Encyclopedia of Earth. Washington, DC: Environmental Information Coalition,
National Council for Science and the Environment; 2007.
[9] Pacca S, Sivaraman D, Keoleian GA. Parameters affecting the life cycle
performance of PV technologies and systems. Energy Policy 2007;35:3316–26.
[10] Bullard CW, Penner PS, Pilati DA. Net energy analysis – handbook for
combining process and input–output analysis. Resources and Energy
1978;1:267–313.
[11] Spreng DT. Net-energy analysis and the energy requirements of energy
systems. New York, New York: Praeger; 1988.
[12] Neij L. Use of experience curves to analyse the prospects for diffusion and
adoption of renewable energy technology. Energy Policy 1997;25:1099–107.
[13] Gipe P. Wind energy comes of age. New York: Wiley; 1995.
[14] Lensen M. Life cycle energy and greenhouse gas emission of nuclear energy:
a review. Energy Conversion and Management 2008;49(8):2178–99.
[15] Junginger M, Faaij A, Turkenburg WC. Global experience curves for wind
farms. Energy Policy 2005;33:133–50.
[16] Ostergaard PA. Ancillary services and the integration of substantial quantities
of wind power. Applied Energy 2006;83:451–63.
[17] Smil V. 21st century energy: some sobering thoughts. OECD Observer; 20 06.
[18] Mulder A. Do economic instruments matter? Wind turbine investments in the
EU(15). Energy Economics 2008;30:2980–91.
[19] Kunz TH, Arnett EB, Erickson WP, Hoar AR, Johnson GD, Larkin RP, et al.
Ecological impacts of wind energy development on bats: questions,
research needs, and hypotheses. Frontiers in Ecology and the Environment
2007;5:315–24.
[20] Mulder K, Hagens NJ. Energy return on investment: toward a consistent
framework. AMBIO: A Journal of the Human Environment 2008;37:74–9.
[21] White SW, Kulcinski GL. Net energy payback and CO
2
emissions from
wind-generated electricity in the Midwest. Madison, WI: Fusion Technology
Institute; 1998.
[22] Brown MT, Ulgiati S. Emergy evaluations and environmental loading of elec-
tricity production systems. Journal of Cleaner Production 2002;10:321–34.
[23] Gagnon L, BE
`langer C, Uchiyama Y. Life-cycle assessment of electricity
generation options: the status of research in year 2001. Energy Policy
2002;30:1267–78.
[24] Khan FI, Hawboldt K, Iqbal MT. Life cycle analysis of wind-fuel cell integrated
system. Renewable Energy 2005;30:157–77.
[25] Lenzen M, Wachsmann U. Wind turbines in Brazil and Germany: an example
of geographical variability in life-cycle assessment. Applied Energy
2004;77:119–30.
[26] Tryfonidou R, Wagner HJ. Multi-megawatt wind turbines for offshore use:
aspects of life cycle assessment. International Journal of Global Energy Issues
2004;21:255–62.
[27] Wagner H-J, Pick E. Energy yield ratio and cumulative energy demand for
wind energy converters. Energy 2004;29:2289–95.
[28] Hondo H. Life cycle GHG emission analysis of power generation systems:
Japanese case. Energy 2005;30:2042–56.
[29] Ardente F, Beccali M, Cellura M, Lo Brano V. Energy performances and life
cycle assessment of an Italian wind farm. Renewable and Sustainable Energy
Reviews 2008;12:200–17.
[30] Pehnt M. Dynamic life cycle assessment (LCA) of renewable energy technol-
ogies. Renewable Energy 2006;31:55–71.
0
5
10
15
20
25
30
35
40
45
Nuclear (1)
Nuclear (2)
Coal
Hydropower
Geothermal
Wind
Solar
Thermal
PV
EROI
Fig. 6. EROI for power generation systems. Nuclear (1) represents the average and
standard deviation for the entire sample of analyses reviewed by Lenzen [14]. Nuclear
(2) omits the extreme outliers from Lenzen’s survey, and thus represents a better
assessment of what the EROI for nuclear is likely to be. See text for description of
further sources.
I. Kubiszewski et al. / Renewable Energy 35 (2010) 218–225 225