Energy Policy 31 (2003) 13151326

Carbon emission and mitigation cost comparisons between fossil fuel,

nuclear and renewable energy resources for electricity generation
Ralph E.H. Simsa,*, Hans-Holger Rognerb, Ken Gregoryc
b

a
Centre for Energy Research, Massey University, Palmerston North, New Zealand
Department of Nuclear Energy, International Atomic Energy Agency (IAEA), P.O. Box 100, Wagramerstr. 5, A-1400 Vienna, Austria
c
Centre for Business and Environment, 80 Cecil Park, Pinner, Middlesex HA5 5HH, UK

Abstract
A study was conducted to compare the electricity generation costs of a number of current commercial technologies with
technologies expected to become commercially available within the coming decade or so. The amount of greenhouse gas emissions
resulting per kWh of electricity generated were evaluated. A range of fossil fuel alternatives (with and without physical carbon
sequestration), were compared with the baseline case of a pulverised coal, steam cycle power plant. Nuclear, hydro, wind, bioenergy
and solar generating plants were also evaluated. The objectives were to assess the comparative costs of mitigation per tonne of
carbon emissions avoided, and to estimate the total amount of carbon mitigation that could result from the global electricity sector
by 2010 and 2020 as a result of fuel switching, carbon dioxide sequestration and the greater uptake of renewable energy. Most
technologies showed potential to reduce both generating costs and carbon emission avoidance by 2020 with the exception of solar
power and carbon dioxide sequestration. The global electricity industry has potential to reduce its carbon emissions by over 15% by
2020 together with cost saving benets compared with existing generation.
r 2002 Elsevier Science Ltd. All rights reserved.
Keywords: Electricity generation; Carbon emissions; Mitigation potential

1. Introduction
This paper reviews and compares the major technological advances and carbon dioxide mitigation options
for the global electricity supply industry. It is a summary
of a review study undertaken by the authors for the
Third Assessment Report of the Intergovernmental
Panel on Climate Change (IPCC, 2001). The IPCCs
policy specically excludes policy prescriptions or
commentary. Whether or not liberalised electricity
markets are likely to adopt the cost-saving, emissionreducing technologies included in the paper, or whether
there are particular institutional obstacles to their
adoption is not discussed. Sufce to say that if an
investment opportunity exists to increase prot for a
privately owned power generating company (or even for
a state-owned enterprise), serious consideration will be
given. If this same investment would also serve to offset
*Corresponding author. Tel.: +64-6-3505288; fax: +64-6-3505604.
E-mail addresses: [email protected] (R.E.H. Sims), [email protected] (H.-H. Rogner), [email protected]
(K. Gregory).

the threat of a future carbon emissions charge, then it

would make good commercial sense to invest.
The paper concentrates on carbon emissions and
largely ignores the impacts of other emissions from
power generation plants (other than the desulpurisation
of ue gases). Other more minor greenhouse gas (GHG)
emissions such as nitrogen oxides NOx are detailed in
the Third Assessment Report (IPCC, 2001). These,
together with other possible emissions from some power
plants such as dioxins, whilst certainly needing to be
considered in any environmental impact assessment
necessary for a resource consent when developing a new
plant, were not considered since they have relatively low
importance.
The global electricity supply sector accounts for the
release to the atmosphere of over 7700 million tonnes of
carbon dioxide annually 2100 Mt C=yr; being 37.5%
of total CO2 emissions. Under business as usual
conditions, the annual carbon emissions associated with
electricity generation, including from combined heat
and power cogeneration, is projected to surpass the
4000 Mt C level by 2020 (IEA, 1998). GHG emissions
are easier to monitor and control from a limited number

0301-4215/03/$ - see front matter r 2002 Elsevier Science Ltd. All rights reserved.
PII: S 0 3 0 1 - 4 2 1 5 ( 0 2 ) 0 0 1 9 2 - 1

1316

R.E.H. Sims et al. / Energy Policy 31 (2003) 13151326

of centralised, large power stations than from millions of

vehicles, small boilers and even ruminant animals.
Therefore the electricity sector is likely to become a
prime target in any future world where GHG emission
controls are implemented and GHG mitigation is
valued.
Several broad methods for mitigation of carbon
dioxide emissions exist:
*

More efficient conversion of fossil fuels: Technological

development has the potential to increase the present
world average power station efciency from 30% to
more than 60% in the longer term. Also, the use of
cogeneration plants replacing the separate generation
of power and heat offers a signicant rise in the
utilisation effectiveness of fuel.
Switching to low-carbon fossil fuels and suppressing
emissions: A switch to gas from coal for example
allows the use of high efciency, low capital cost,
combined cycle gas turbine (CCGT) technology,
resulting in lower carbon emissions per kWh of
electricity generated.
Decarbonisation of fuels and flue gases, and carbon
dioxide sequestration: Decarbonisation of fossil fuel
feedstocks before combustion for electricity generation can be used to make hydrogen-rich secondary
fuels which in the longer term can be used in fuel
cells. Decarbonisation of ue gases after combustion
can become an effective GHG abatement option. In
both cases the carbon dioxide can then be stored over
geological time frames, for example, in depleted gas
elds.
Increasing the use of nuclear power: Nuclear energy
could replace baseload fossil fuel electricity generation in many parts of the world if acceptable
responses can be found to concerns over reactor
safety, radioactive waste transport, waste disposal
and proliferation.
Increasing the use of renewable sources of energy:
Technological advances offer new opportunities and
declining costs for renewable energy technologies
which, in the longer term, could meet a greater share
of the rapidly growing world energy demand.

Since the onset of the industrial revolution in the mid

19th century, some 290 Gt C have been oxidised from
fossil fuels and released to the atmosphere. The known
fossil fuel resource base represents a further carbon
volume of some 5000 Gt C (excluding methane clathrates) indicating there are good reserves of coal, gas and
oil (as well as uranium), although there is still some
uncertainty. The technical potential of renewable energy
sources is far higher though it currently meets only
around 20% of the global energy demand, mainly as
traditional biomass and hydro power. Modern renewable energy systems have the technical potential to

provide all global energy services in sustainable ways

and with low or virtually zero GHG emissions.
In relation to world electricity production, coal
continues to have the largest share at 38% followed
by renewables (principally hydro power) at 20%,
nuclear at 17%, natural gas at 16% and oil at 9%.
Electricity production is expected to almost double by
2020 (Table 1). However average carbon emissions per
unit generated will decline over time mainly due to using
new technologies with improved conversion efciencies.
Coal-red power generation is projected to have a
major increased share by 2020 due to strong growth in
countries such as India and China reecting its
importance there, together with steady growth in the
USA, but a decline in Western Europe. Gas-red plant
is projected to continue to grow strongly in many world
regions reecting the increasing availability of natural
gas whilst oil will lose market share. Nuclear power is
projected to decline slightly after 2010 with capacity
additions in developing countries and economies in
transition roughly balancing plants being retired in
OECD countries. Hydropower is projected to grow by
around 60%, mainly in China and other Asian
countries. New renewables have expanded substantially
throughout the 1990s in absolute terms including wind
by 21% per year and solar photovoltaics (PV) by 30%
per year. Biomass and geothermal projects are also
experiencing good growth. So overall, renewables are
projected to continue to grow till 2020, but without
signicant government intervention, they will still only
supply less than 2% of the electricity market share.

2. New technological options

2.1. Fossil fuels
The efciencies of modern thermal power stations
using the steam cycle can exceed 40% based on lower
heating value (LHV), although the average efciency of
the installed stock world wide is closer to 30%. The cost
of a modern coal-red power station with SOx and NOx
controls is typically $US1 1300=kWe but less efcient
designs with fewer environmental controls are cheaper
and therefore often built in developing countries. Costs
vary considerably depending on location. Recently,
efciencies of 48.5% have been reported and, with
further development, by 2020 they could reach 55%
at costs only slightly higher than current technology
(DTI, 1999).
Some new technological options such as CCGTs have
the ability to penetrate the market place whereas others,
such as new and emerging renewables, need government
support by improving market efciency; by nding new
1

All costs quoted are in 2000$US.

R.E.H. Sims et al. / Energy Policy 31 (2003) 13151326

Table 1
Past and projected global production from the electricity generating
sector (TWh/yr) and average C emissions per kWh due to fuel
switching and efciency gains
1971

1995

2000

2010

2020

Coal
Natural gas
Oil
Nuclear
Hydro
Other renewables

2100
691
1100
111
1209
36

4949
1932
1315
2332
2498
177

5758
2664
1422
2408
2781
215

7795
5063
1663
2568
3445
319

10,296
8243
1941
2317
4096
433

Total
Average GHG
emissions (g C/kWh)

5247

13,203

15,248

20,853

27,326

200

158

157

151

147

Source: Adapted from IEA (1998).

ways to internalise external costs; by accelerating

RD&D; and by providing temporary incentives for
early market development as such new technologies
approach commercial readiness. The efciency of the
best available natural gas-red CCGTs currently being
installed is now around 60% (LHV) and cost around
$450500=kWe ; including selective catalytic reduction
for NOx : Project costs can be substantially higher in
some regions especially where new infrastructure is
required. However in general the capital costs have been
falling and efciencies improving. So, together with
good availability and short construction times, CCGTs
are the highly favoured option where gas is available at
reasonable prices. Developments in the liqueed natural
gas markets could further expand the use of CCGTs and
efciencies could ultimately rise to 70% (Gregory and
Rogner, 1998).
Integrated gasication combined cycle (IGCC) systems utilise the efciency and low capital cost advantages of a CCGT by rst gasifying the solid or liquid
fuel. Gasiers are usually uidised bed designs, oxygen
blown and still at an early commercial stage. Coal and
difcult fuels such as bitumens, tar and pyrolysis derived
bio-oil can be used as feedstocks. Biomass fuels contain
oxygen and little sulphur so can be easier and cheaper to
gasify which may lead to improved thermal efciency.
Gas clean up prior to combustion in the gas turbine
which is sensitive to contaminants, is one of the key
areas needing further development. Based on the latest
CCGTs with 60% efciency, the potential fuel to
electricity conversion efciency of an IGCC system is
around 51%, with efciencies over 60% feasible by
2020. In addition, IGCC offers one of the more
promising routes to CO2 capture and disposal by
converting the gas from the gasier into separate
streams of H2 and CO2 via a shift reaction then
removing the CO2 before the gas enters the gas turbine
or fuel cell.

1317

2.2. Nuclear power

Nuclear power is a mature technology with 434
nuclear reactors totalling around 349GWe operating in
32 countries in 1999 (IAEA, 2000). Worldwide, the
majority of current nuclear power plants are competitive
on a marginal generating cost basis in a deregulated
market environment because of low operating costs and
the fact that many are already fully depreciated. The
life-cycle GHG emissions per unit of electricity from
nuclear power plants are at least two orders of
magnitude lower than those from fossil fueled electricity
generation and comparable to most renewables at near
zero. Hence nuclear power generation is an effective
GHG mitigation option, especially by way of investments to extend the lifetime of existing plants. Whether
or not building more nuclear power plants will be
accepted depends on new designs becoming economically competitive and on the industrys ability to restore
public condence in its safe use.
The future of nuclear power will therefore depend on
whether it can meet several objectives simultaneously
economics, operating safety, proliferation safeguards
and effective solutions to waste disposal. While present
new nuclear power plants already incorporate unprecedented levels of safety based on in-depth designs, their
economics need further improvement to be competitive
in most markets. New nuclear power plants at $1700
$3100=kWe cannot compete against natural gas CCGT
technology at current and expected gas prices where gas
supply infrastructures are already in place (OECD,
1998). However, new nuclear power at generating costs
between 3.9 and 8:0 c=kWh can be competitive with coal
and natural gas where coal has to be transported over
long distances or natural gas pipelines and infrastructures are not in place. Safe waste disposal for
approximately 1 million years is claimed to be technically feasible and to add only 0:02 c=kWh to generating
costs. Proliferation is primarily a political issue, but can
also be addressed by technology.
Technological approaches for safe and long-term
disposal of high-level radioactive waste have been
extensively studied, one possible solution involving deep
geological repositories. In the longer run, fundamentally
new reactor congurations may need to be developed
based on innovative designs that integrate inherent
operating safety features and waste disposal using
previously generated radioactive waste as fuel. By way
of transmutation, this would also convert nuclear waste
or plutonium to less hazardous and short-lived isotopic
substances.
Major nuclear reactor vendors have now developed
modied reactors that offer both improved safety and
lower costs. One of the innovative designs is the pebble
bed modular reactor (PBMR) developed by the South
African utility ESKOM. The fundamental concept of

1318

R.E.H. Sims et al. / Energy Policy 31 (2003) 13151326

the design was to achieve a plant that has no physical

process which could cause a radiation induced hazard
outside the site boundary. The prospect of a core melt
scenario is therefore zero. The capital cost of a
production of 10 modules of 1000 MWe was estimated
by ESKOM to be around $10001200=kWe ; being
cheaper than other designs and giving attractive
generating costs with unprecedented safety aspects. As
a base load station with a depreciation period of 20
years and at a 10% discount rate, the expected cost of
power would be approximately 2 c=kWh including the
full fuel cycle and decommissioning (Nicholls, 1998).
These costs are far less than typical present nuclear
technologies and in part perhaps the result of engineering optimism. Other assessments quote double this
generating cost (WEA, 2000).
Increased performance and lifetime extension of the
currently existing nuclear reactors often present zero
cost GHG mitigation options. Current nuclear power
plants avoid carbon emissions of 600 Mt=year compared with using coal-red plant to generate the same
quantity of power. However, given current market
conditions, new nuclear power plants remain a leastcost alternative only in countries usually characterised
by limited indigenous fossil fuel resources or with large
distances between energy resource location and demand
centres.
2.3. Renewable energy systems
Natural energy ows vary from location to location
and make the techno-economic performance of renewable energy conversion highly site specic. Intermittent
sources such as wind, solar, tidal and wave energy,
require back-up if not grid connected, while high
penetration into grids may eventually require storage
and/or back-up to guarantee reliable power supply
(Ackermann et al., 1999). Therefore, it is difcult to
generalise costs and potentials.
Hydro-electricity remains the most developed renewable resource worldwide but is now constrained due to
societal and environmental barriers. The market potential resulting from detailed geological, environmental,
social and technical evaluations is difcult to establish
because societal preferences are inherently uncertain and
difcult to predict. Other limitations to further development include the remote locations of many potential
hydro sites resulting in high transmission costs, and the
high, up front, capital investment cost, which recently
privatised power companies are unlikely to accept due to
the relatively low rates of return compared with gas or
coal. An overall assessment of global economic hydropower potential is between 7000 and 9000 TWh per
year. In addition numerous small o10 MWe ; mini
o1 MWe and micro o100 kWe scale hydro schemes
with low environmental impacts continue to be devel-

oped globally for generation costs around 612 c=kWh

particularly in rural communities currently without
electricity. Under certain conditions ooding of land
upstream of a hydro dam can cause loss of biological
carbon stocks as well as produce methane emissions as
the vegetation decays over time. These losses and
emissions are very site specic and do not occur with
run-of-river schemes so were not considered in this
analysis.
Wind power accounts for around 0.3% of the global
installed generation capacity due to its relatively recent
emergence but, due to its intermittent nature, it supplies
around only 0.1% of total global electricity. Capacity
reached 13; 000 MWe by 2000, with estimates of
increases to over 30; 000 MWe operating by 2005
(EWEA, 1999). Denmark aims to provide 4050% of
its national electricity generation from wind power by
2030 and remains the main exporter of wind turbine
technology. To meet future demand, many wind
turbines will be sited off-shore, exceed 3 MWe maximum rated output, have lower operating and maintenance costs, be more reliable, and have a greater
content of local manufacture. The need for resource
planning consents and limited areas of shallow seas
suitable for off-shore wind farms may be constraints.
The cost of wind turbines continues to fall as more
new capacity is installed. In high-wind areas, wind
power is competitive with other forms of electricity
generation at between 3 and 5 c=kWh: The global
average price is expected to drop to 2.73 c=kWh by
around 2020 due to economies of scale from mass
production and improved turbine designs. The installed
costs will fall from $1000 to $635=kWe and operating
costs will fall from 0:01 c=kWh to 0:005 c=kWh (EPRI/
DOE, 1997). However, on poorer sites of around 5 m=s
mean annual wind speed, the generating costs would
remain high at around 1012 c=kWh (8% discount
rate).
Biomass resources include agricultural and forestry
residues, landll gas, municipal solid wastes and energy
crops. Since biomass is widely distributed, it has good
potential to provide rural areas with a renewable source
of energy. The challenge is to provide the sustainable
management, conversion and delivery of bioenergy to
the market place in the form of modern and competitive
energy services. Agricultural and forest residues such as
bagasse, rice husks, bark and sawdust often have a
disposal cost. Therefore, waste-to-energy conversion for
heat and power generation can have good economic and
market potential, particularly, in rural industry and
community applications where it is widely used. In
Denmark, about 40% of electricity generated is from
biomass cogeneration plants using wood waste, straw
and animal wastes for biogas, and in Finland it is 10%
using sawdust, forest residues and pulp liquors. In other
countries biomass cogeneration is utilised to a lesser

R.E.H. Sims et al. / Energy Policy 31 (2003) 13151326

degree due to unfavourable regulatory practices and

limiting structures within the electricity industry. Energy
crop feedstocks have less potential in the short term due
to their higher delivered costs in terms of $/GJ of
available energy as well as competition for land use.
Biomass fuels are generally easier to gasify than coal
and the development of efcient biomass IGCC systems
is nearing commercial realisation. Capital investment
for a high-pressure, direct gasication combined-cycle
plant is estimated to fall from over $2000=kWe at
present to around $1100=kWe by 2020, with operating
costs, including fuel supply, declining from 3.98 to
3:12 c=kWh (EPRI/DOE, 1997). By way of comparison,
the higher current operating costs of 5:50 c=kWh for
traditional combustion boiler/steam turbine technology,
(reecting poorer fuel conversion efciency compared
with gasication), were predicted to lower to
3:87 c=kWh: Capital costs were also predicted to fall
from the present $1965=kWe to $1100=kWe in the same
period.
Solar radiation as intercepted at the Earths surface
may be reasonably high in many regions but the market
potential for its capture is low due to the current
relatively high costs of solar collectors. In addition there
are:
*

time variations from daily and seasonal uctuations

and hence the need for energy storage;
geographical variations, areas near the equator
receiving approximately twice the annual solar
radiation than at 601 latitudes; and
diffuse characteristics with low power such that largescale generation from direct solar energy requires
signicant amounts of equipment and land, even with
solar concentrating techniques.

The cost of PV, at around $5000=kWe installed, is

slowly falling due to manufacturing scale-up and mass
production techniques as more capacity is installed.
Generating costs are relatively high at 2040 c=kWh:
However, PV is often deployed at the point of electricity
use such as buildings, and this can offset the high costs
by giving a competitive advantage over power transmitted long distances from central power stations with
high losses and distribution costs. Advances in inverters
(including incorporation into the modules to give AC
output), together with net metering systems and
government sponsorship, have encouraged the market
uptake of PV panels for grid-connected building
integration projects, either in large-scale installations
(up to 1 MWe ) or on residential buildings (up to 5 kWe ).
Other growing markets for PV power generation
systems include off-grid applications for rural locations
particularly in developing countries where two billion
people still have no access to electricity.
The size of the world market for PV rose to
300 MWe =yr in 2000 with anticipated growth to over

1319

1000 MWe =yr by 2005. Conversion efciencies of silicon

cells continue to improve in the laboratory, though
commercial monocrystalline based modules are still
obtaining only 1317% efciency and multicrystalline
1214%. Recent studies showed a $660 M investment
in a single factory capable of producing 400 MWe
(5 million panels) a year would reduce manufacturing
costs by 75% as a result of economies of scale (KPMG,
1999). Possible further cost reductions from project
learning experience could be offset by predicted higher
cost of silicon wafers in future. Neij (1997) calculated a
$100 billion investment in manufacturing capacity would
be needed in order to reach an acceptable generating level
of 5 c=kWh (excluding back-up supply or storage costs).
Several types of high temperature, solar thermal,
independent and utility-owned grid connected power
stations have also been demonstrated. Dish systems
giving concentration ratios up to 2000, and therefore
performing at temperatures up to 15001C; can supply
steam directly to a standard steam turbine generator.
Capital costs are projected to fall from around
$4000=kWe now to $2500=kWe by 2030 with other
estimates even lower (AGO, 1998).
Wave power, ocean currents and tidal power are other
renewable energy technologies but these are unlikely to
nd signicant markets by 2020 and were not considered
in this analysis. Geothermal generation was also not
included.
2.4. Carbon sequestration
Carbon sequestration technologies have become
much better understood during the past few years, so
they can now be seriously considered as GHG mitigation options alongside the better established options.
They include the use of biological sinks such as
plantation forests but only physical sequestration is
discussed here. Once proven, the potential for generation of electricity from fossil fuels linked with the
capture and storage of CO2 is unlikely to be constrained
by either the availability of resources of fossil fuels or by
the capacity for storage. The technology is mature and
readily available for both piping CO2 over large
distances, and for providing underground storage.
Carbon dioxide exists naturally in underground
reservoirs in various parts of the world and some is
often released during oil, gas and geothermal extraction.
Hence potential sites for the storage of captured CO2 are
underground reservoirs such as depleted oil and gas
elds or deep saline reservoirs. Also CO2 injected into
coal measures may be preferentially absorbed, displacing methane from the coal (which would be captured
for commercial use) such that sequestration would be
achieved providing the coal is never mined. Another
possible storage location for captured CO2 is in the deep
ocean, but this option is at an earlier stage of

1320

R.E.H. Sims et al. / Energy Policy 31 (2003) 13151326

development than underground reservoirs. So far only

small-scale experiments for preliminary investigation
have been carried out.
Physical carbon sequestration is best suited to dealing
with the emissions of large point sources of CO2 such as
power plants. It can be captured either from the fuel gas
before combustion in, for example, an IGCC or
steam methane reforming process or from the ue gas
stream after combustion using regenerable amine
solvents or other systems. The concentration of CO2
in power station ue gas is between about 4% (for gas
turbines) and 14% (for a pulverised coal-red plant).
These low-concentrations mean that large volumes
of gas have to be handled and powerful solvents
therefore have to be used, resulting in high-energy
consumption for solvent regeneration. However,
8090% of the CO2 in a ue gas stream can be captured
by such techniques. Other CO2 capture techniques
available or under development, include cryogenics,
membranes and adsorption. After capturing the CO2 ; it
is pressurised, typically up to 100 bar; before transportation to storage areas using high-pressure pipelines or
even by ship.
Carbon dioxide capture and compression imposes a
penalty on the thermal efciency of a power generation
plant estimated to be 813%. The cost of CO2 capture in
power stations is estimated to be approximately $30
50=t CO2 of emissions ($110180=t C), being equivalent
to an increase of about 50% in the cost of electricity.
The cost of CO2 transport depends greatly on the
transport distance and the capacity of the pipeline
but is approximately $13=t CO2 per 100 km: The
cost of underground storage, excluding compression
and transport, would be approximately $12=t CO2
stored. The overall cost of transport and storage for
a transport distance of 300 km would therefore be
about $8=t CO2 stored, equivalent to about $37=t
of carbon emissions avoided. Hence the overall cost
of CO2 capture and storage would be around $150
220=t C emissions avoided. As with most new technologies, there is scope to reduce these costs in
future through technical developments and wider
application.
If CO2 storage is to be used as a basis for emissions
trading, or to meet national commitments under the
UN Framework Convention for Climate Change, it will
be necessary to establish the quantities of CO2 stored
in a veriable manner and to ensure a retention
time sufcient to avoid any adverse effect on the
climate. Most verication requirements for geologically
stored CO2 can be achieved with technology available
today but the permanency of storage is yet to be
determined. Validation of CO2 storage in the ocean
would be more difcult but it should be possible to
verify quantities of CO2 stored in concentrated deposits
on the seabed.

3. Technological and economic potential of power

generation systems
Several electricity generation technologies were analysed and compared for both their costs and carbon
mitigation potential. Previous studies attempted to
compare power generation technologies on cost alone
(US DOE/EIA, 2000; WEA, 2000; OECD, 1998). The
OECD data resulted from a survey of power stations
due for completion between 2000 and 2005 in a wide
cross-section of countries. It showed that costs can vary
considerably between projects due to national and
regional differences including the need for additional
infrastructure, the trade-off between capital costs and
efciency, the ability to run on baseload and the cost
and availability of various fuels. The costs of reducing
GHG emissions will also vary because of local
variability in the costs of the baseline generation
technology chosen (pulverised coal or CCGT) and the
alternative technologies available. This large variation in
local circumstances prevents specic generating costs
from being used even within the boundaries of one
country. In addition, it was not always possible to
accurately account for the varying assumptions made
for each of the variables used in each study when
comparing the technologies in this analysis. Costs and
mitigation potentials are highly location-dependent and
consequently ranges had to be used.
This paper presents typical CO2 emissions and costs/
kWh from conventional pulverised coal-red power
generation plants and compares them with alternative
types of generation and also with GHG mitigation
technologies expected to be rmly in place by 2010.
These include natural gas-red CCGT, CO2 capture and
storage, nuclear power, hydro, wind, biomass, PV and
solar thermal. Two principle sources of data were used:
OECD, 1998 (Tables 25) and US DOE/EIA, 2000
(Table 6) for a single country in an attempt to reduce
some of the variability in costs seen in multi-country
studies. The OECD (1998) survey of power stations
provided data on actual projects due to come on stream
during 20002005 in 19 countries including Brazil,
China, India and Russia, together with a few projects
for 20062010 based on more advanced technologies.
Additional data from other sources were used where
necessary and are identied in the text.
Pulverised coal, steam turbine technology was used as
the baseline for comparative purposes (Tables 2 and 3)
but for regions where natural gas is the rst choice
of generation, a second comparison was made with
the baseline technology assumed to be CCGT (Tables 4
and 5). New generation plant investment costs differ
between Annex I (developed) countries (Tables 2 and 4)
and non-Annex 1 countries (Tables 3 and 5) as dened
in the UN Framework Convention on Climate Change.
Costs and carbon emissions of the coal baseline were

R.E.H. Sims et al. / Energy Policy 31 (2003) 13151326

1321

Table 2
Cost estimates of alternative mitigation technologies in the power generation sector compared to baseline coal-red power stations and potential
reductions in C emissions to 2010 and 2020 for Annex I countries
Technology

PF fgd; IGCC and CCGT

NOx ; etc. supercritical

PF fgd CCGT Nuclear Hydro

CO2
CO2
capture capture

Wind
Biomass IGCC PV and
turbines
solar
thermal

Energy source
Generating costs (c/kWh)
Emissions (g C/kWh)
Cost of C reduction ($/t C avoided)
Reduction potential to 2010 (Mt C/yr)
Reduction potential to 2020 (Mt C/yr)

Coal
4.90
229
Baseline
Baseline
Baseline

Coal
7.9
40
159
210
550

Wind
3.08
0
82135
51
128

Coal
3.66.0
190198
1040
13
55

Gas
4.96.9
103122
0156
18
103

Gas
6.48.4
17
71165

Uranium
3.98.0
0
38135
30
191

Water
4.27.8
0
31127
6
37

Biofuel
2.87.6
0
92117
9
77

Solar
8.740.0
0
1751400
2
20

PF, pulverised fuel; fgd, ue gas desulphurisation; IGCC, integrated gasication combined cycle.

Table 3
Cost estimates of alternative mitigation technologies in the power generation sector compared to baseline coal-red power stations and potential
reductions in C emissions to 2010 and 2020 for non-Annex I countries
Technology

PF fgd IGCC and CCGT

NOx ; etc. supercritical

PF fgd CCGT Nuclear Hydro

CO2
CO2
capture capture

Wind
Biomass IGCC PV and
turbines
solar
thermal

Energy source
Generating costs (c/kWh)
Emissions (g C/kWh)
Cost of C reduction ($/t C avoided)
Reduction potential to 2010 (Mt C/yr)
Reduction potential to 2020 (Mt C/yr)

Coal
4.45
260
Baseline
Baseline
Baseline

Coal
7.45
40
136
0
550

Wind
3.08
0
56137
12
45

Coal
3.66.0
190198
10200
36
85

Gas
4.456.9
103122
017
20
137

Gas
5.958.4
17
62163

Uranium
3.98.0
0
2077
36
220

Water
4.27.8
0
10129
20
55

Biofuel
2.87.6
0
63121
5
13

Solar
8.740.0
0
1641370
0.5
8

Table 4
Cost estimates of alternative mitigation technologies in the power generation sector compared to gas-red CCGT power stations and the potential
reductions in C emissions to 2010 and 2020 for Annex I countries
Technology

CCGT

PF fgd
CO2
capture

CCGT
CO2
capture

Nuclear

Hydro

Wind
turbines

Biomass IGCC

PV and
solar
thermal

Energy source
Generation costs (c/kWh)
Emissions (g C/kWh)
Cost of C reduction ($/t C avoided)
Reduction potential to 2010 (Mt C/yr)
Reduction potential to 2020 (Mt C/yr)

Gas
3.45
108
Baseline
Baseline
Baseline

Coal
7.610.6
40
6101050

Gas
4.95
17
165
210
550

Uranium
3.98.0
0
46421
62
181

Water
4.27.8
0
66400
3
18

Wind
3.08
0
4392
23
61

Biofuel
2.87.6
0
60224
4
36

Solar
8.740.0
0
5003800
0.8
9

Table 5
Cost estimates of alternative mitigation technologies in the power generation sector compared to gas-red CCGT power stations and the potential
reductions in C emissions to 2010 and 2020non-Annex I countries
Technology

CCGT

PF fgd
CO2
capture

CCGT
CO2
capture

Nuclear

Hydro

Wind
turbines

Biomass IGCC

PV and
solar
thermal

Energy source
Generation costs (c/kWh)
Emissions (g C/kWh)
Cost of C reduction ($/t C avoided)
Reduction potential to 2010 (Mt C/yr)
Reduction potential to 2020 (Mt C/yr)

Gas
3.45
108
Baseline
Baseline
Baseline

Coal
6.98.7
40
507772

Gas
4.95
17
165
0
550

Uranium
3.98.0
0
46421
10
70

Water
4.27.8
0
66400
9
26

Wind
3.08
0
4392
5
21

Biofuel
2.87.6
0
60224
1
6

Solar
8.740.0
0
5003800
0.2
4

taken from the average of several coal-red projects

under construction as surveyed in the OECD database.
Flue gas desulphurisation (fgd) was assumed to be

included in all Annex I plants and in around 20% of the

non-Annex I cases. Due to this additional cost, and also
reecting the lower efciencies of older power station

1322

R.E.H. Sims et al. / Energy Policy 31 (2003) 13151326

Table 6
Estimated costs of alternative baseline and mitigation technologies in the USA power generation sector, in idealised conditions
Technology

PF fgd; IGCC
NOx ; etc.

Energy source
Generating costs (c/kWh)
Emissions (g C/kWh)
Cost of C reduction compared
to pulverised coal steam cycle ($/t C)
Cost of C reduction compared
to natural gas CCGT ($/t C)

Coal
3.33.7
247252

Coal
Gas
Coal
3.23.9 2.93.4 6.36.7
190210 102129 40

Baseline

80168 538

CCGT

PF fgd IGCC CCGT Nuclear

CO2
CO2 CO2
capture capture capture

Wind
Biomass Biomass PV and
turbines
solar
thermal

Coal
Gas
Uranium Wind
Residues Crops
5.76.4 4.44.9 5.06.0 3.35.5 4.06.7 6.47.5
37
17
0
0
0
0

141145 93148

3070

52102

1685 12138

Baseline 326613 250538 134176 124304 8245 47373

designs currently being built in non-Annex I countries,

the baseline coal technology was cheaper in these
regions (4:45 c=kWh; Table 3) than for Annex I
countries (4:9 c=kWh; Table 2). However CO2 emissions
were higher at 260 g C=kWh compared with
229 g C=kWh from the most modern, more efcient
power plant designs. Other generation and carbon
mitigation technologies were then compared to the
baseline data using costs from the OECD database and
other sources as referenced in the sections above.
In Tables 25, the rst column gives the generation
costs of the baseline technology in c/kWh and the
emissions of carbon dioxide as g C=kWh: The subsequent columns give a range of the costs and emissions
for alternative technologies that could be used to reduce
carbon emissions over the next 20 years or so by
displacing the baseline technology. The costs of reducing
carbon dioxide emissions in the mitigation options
($=t C avoided) vary because of variability in the costs
of both baseline and alternative technologies. Hence a
range is given.
Estimates of the CO2 reduction potential Mt C=yr
in 2010 and 2020 are provided for each of the alternative
mitigation options. These were derived mainly from
projections of world electricity generation from different
energy sources by IEA (1998) as presented in Table 1.
IEA data were chosen in preference to various IPCC
scenarios because of the shorter time horizon and the
higher technology resolution provided.
In Tables 2 and 3 it was assumed that a maximum of
20% of currently proposed coal baseline capacity would
be replaced by either gas, nuclear or renewable technologies during 20062010 and 50% during 20112020.
These assumptions allowed for a 5-year lead time for
decisions on the alternatives to be made and construction
to be undertaken. It was assumed the alternative
generation investment options to coal would ramp up
only slowly over several years and hence this would limit
the maximum capacity that could be displaced by 2010.
After 2010, it was assumed there will be practical reasons
why half the proposed new coal capacity could not be
displaced so these plants will be built.

Sunlight
9.025.0
0

107170 210880
233450 4342167

Based on similar assumptions as used for the coal

baseline, it was assumed in Tables 4 and 5 that a
maximum of 20% of proposed new baseline gas capacity
during 20062010 could be displaced by improved
mitigation options and 50% during 20112020.
The rate of developing and building gas or nuclear
power stations that would be required using these
assumptions was considered to be realistic. For nuclear
power, the rate of building between 2011 and 2020
would be less than that seen at the peak of construction
of new nuclear plant in the 1970s. For gas, the gas
turbines are factory-made so no supply constraints
should arise from the anticipated increase in demand.
Less equipment and construction effort would be
required than the coal capacity replaced in terms of
boilers, steam turbines and cooling towers.
For hydro, wind, biomass and PV technologies,
estimated penetration rates applied to displace new coal
(Tables 2 and 3) or gas capacities (Tables 4 and 5) were
derived from the well researched and publicised Shell
sustainable growth scenario (Shell, 1996). For wind and
PV, these penetration rates imply substantial growth
until 2020, but this is less than the current rate of
expansion of 2030% per year. For biomass, most of
the fuel was assumed to be agricultural, wood process
or forest residues though some purpose grown crops
would also be used as in the 10 MWe Yorkshire
ARBRE project which uses 2200 ha of coppiced Salix
(Pitcher, 2000).
The introduction of CO2 capture and storage
technology for coal or gas plants would require similar
construction processes as for conventional power plant.
The CO2 separation facilities would need additional
equipment but, in terms of physical construction, this
would involve no more effort than, say, the establishment of a similar scale of biomass gasication plant.
CO2 storage facilities would be constructed using
available oil/gas industry technology so this was not
seen to be a limiting factor. Storage of the carbon would
be in saline aquifers or depleted oil and gas elds. It was
assumed that pilot plants, rst developed in Annex I
countries, could be operational before 2010 and that the

R.E.H. Sims et al. / Energy Policy 31 (2003) 13151326

annual mitigation potential would be 210 Mt C each

for coal and for gas technologies. In 2020, the total
mitigation potential was therefore put at 40200 Mt C;
split equally between coal and gas and between Annex I
and non-Annex I countries. This was somewhat
arbitrary but reected the potential to move forward
with the technology if no major problems are encountered, and if more pilot schemes were to be extended. It
was assumed for simplicity that fuel switching, from coal
to gas or vice versa, would not occur in addition to CO2
capture and storage, although in reality this would be an
extra option for a power generation company.
Costs for coal, gas and nuclear were derived from
OECD (1998) data using $1.70/GJ for coal and $3.20/
GJ for gas in Annex 1 countries. (These costs were the
anticipated long-run real prices, though actual prices
may uctuate widely in the short term as for USA
natural gas prices which varied in 2000 between $2 and
$11/GJ.) The additional costs of CO2 capture and
disposal were derived from the IEA GHG R&D
programme (Audus, 2000). For coal, the costs for CO2
capture and storage were assumed to be for retrotting
to pulverised fuel (PF) power stations. Data for hydropower costs were taken from WEA (2000) and the IPCC
(1996) Second Assessment Report. Wind power data
were based on recent project costs (Walker et al., 1998)
and projected future costs in IPCC (1996). For biomass,
the technology used was IGCC (rather than combustion/steam cycle) with capital and operating costs
assumed to be similar to those for coal IGCC. Biomass
fuel costs ranged between $0 and 2.7/GJ, based on either
wood process residues or sugar-cane bagasse being
free by-products produced on-site (perhaps even with
a negative cost for disposal) or on forest arisings
or crop residues such as cereal straw needing collection.
PV costs, taken from the World Energy Assessment
(WEA, 2000), were based on present day costs of
$5000 per peak kWe capacity lowering to $1000=kWe
predicted after 2015. A 10% discount rate was used
throughout.
3.1. Carbon emissions reductions
When estimating the maximum CO2 mitigation each
technology could achieve, it was assumed that only
proposals for new coal or gas power stations or old
plant would be displaced since early retirement of
existing plant and equipment assets is only likely to
occur in practice under exceptional circumstances. In all
cases, the reduction potential in 2020 is substantially
higher than in 2010. This follows from the assumptions
used for the rate of plant uptake and demand growth,
reecting the time taken to make decisions and,
especially in the case of renewables and CO2 capture
and storage, to build up manufacturing capacity, to
learn from experience and to reduce costs.

1323

As a general rule, increasing the amount of carbon

abatement to be achieved will require moving to higher
cost options, the low hanging fruit being picked rst.
This applies particularly to renewables since to achieve
additional capacity will eventually require constructing
plants on sites with less favourable conditions such as
those with lower mean annual wind speeds or slower
growth of energy forests.
It is evident that each of the mitigation technologies
presented can contribute to reducing emissions in terms
of Mt C/yr with nuclear, where considered to be sociopolitically desirable, having the greatest potential.
Replacement of coal by gas (Tables 2 and 3) can make
a substantial contribution as can CO2 capture and
storage. Each of the renewables can also make a
signicant contribution to displacing proposed or
replacement coal-red power plants, although solar
power will be limited due to the high investment costs
constraining its uptake.
The potential reductions within each table are not
additive since each of the technologies assessed will be
competing with each other to displace new coal and gas
power stations. Based on the assumption about the
maximum displacement of new coal and gas power
stations being 20% for the 20062010 period and 50%
by 2020, the maximum mitigation that could be
achieved would be around 140 Mt C in 2010 and
660 Mt C in 2020. These avoided emissions can be
compared with the projected global CO2 emissions from
power stations of around 2400 Mt C in 2000,
3150 Mt C in 2010 and 4000 Mt C in 2020 (IEA,
1998). In practice, a combination of technologies could
be used to displace coal and natural gas-red generation
and the choice will often depend on local circumstances.
Not shown in the tables, oil-red generation could also
be displaced. Based on similar assumptions, this would
give a further mitigation potential of 10 Mt C by 2010
and 40 Mt C by 2020.
Not all of the mitigation options are likely to achieve
their potential for a variety of reasons such as unforeseen technical difculties, investment cost limitations in
non-Annex 1 countries, and socio-political barriers in
Annex 1 countries. Conversely the uptake of renewable
energy projects due to the Clean Development Mechanism is hard to predict at this stage. Therefore the total
mitigation potential for the electricity supply sector
using mitigation options for all three fossil fuels and
allowing for potential constraints, was estimated to be
between 50150 Mt C by 2010 and 350700 Mt C by
2020.
3.2. Costs of carbon mitigation
Where the investment cost to avoid one tonne of
carbon emissions is negative, the project can be assumed
to be an economic option regardless of any emission

1324

R.E.H. Sims et al. / Energy Policy 31 (2003) 13151326

reductions. There are a range of win/win opportunities

which can reduce emissions at either zero cost or with
cost savings. Where gas infrastructures are in place and
CCGT is a lower cost option than coal, it was assumed
that natural gas would be the fuel of choice (Tables 2
and 3). Hence no negative cost reduction options exist
for natural gas to displace coal. Where additional
investment costs are involved compared with the baseline cases, the level is an indication of the carbon value
needed for a project to proceed if and when carbon
trading occurs. For example in Table 4 in areas of
Annex 1 countries where both hydro and gas resources
exist, building hydro power plants instead of CCGT
would be unlikely unless the carbon offset value exceeds
$66=t C:
Hydro, wind and biomass power generation projects
all have the potential on good sites to avoid carbon
emissions by displacing coal-red plants and in some
specic cases, investment in the technologies can
generate power at lower costs (Tables 2 and 3). It is
also possible for some wind and biomass projects to
protably displace natural gas CCGT plants (shown as
negative values within the $=t C avoided ranges in
Tables 4 and 5). These will be located on sites where
high mean annual wind speeds or cheap wood processing or agricultural residues are available. Hydro plants
would not compete with CCGT without a carbon offset
value being included. In fact, as shown by the ranges of
costs, for the majority of sites, all renewable energy
projects are more costly and generation would be more
expensive than using CCGT.
Under the assumptions made, nuclear power is
theoretically a zero cost mitigation option in terms of
$=t C reduced. If the projected generating costs for
PBMR plants can be realised in practice, this would
certainly imply negative mitigation costs were possible.
For new nuclear plants using state-of-the art designs, a
possible value between $30 and 40/t of carbon avoided
would put them more on a par with IGCC coal-red
electricity stations (Tables 2 and 3). However, it would
take a carbon value of over $46/t (Tables 4 and 5) for
nuclear power to break even with natural gas combined
cycle electricity.
3.3. Analysis of USA data
The ranges of the data presented in Tables 25 reect
the large variations of local circumstances for a specic
project resulting in a wide range of mitigation costs
($=t C avoided) for each technology option. Therefore,
to narrow down these variations, project costs were also
analysed for USA alone, mainly based on data used in
the USA Annual Energy Outlook (US DOE/EIA, 2000)
as presented in Table 6. A 10% discount rate was used
for all options. In this US DOE/EIA study, a narrower
range of future cost estimates was used for PV and solar

thermal (925 c=kWh) than was used in Tables 25.

The range of wind costs from 3.3 to 5:5 c=kWh were
taken from high and low cases in the same study and
the costs of biomass derived electricity were based
on a fuel cost of $2.7/GJ for purpose grown crops, and a
cost range of $0$2/GJ for agricultural and forest
residues.
The USA electricity generating costs were based on
national projections of utility prices for coal of $1/GJ
and for natural gas of $2.7/GJ. Capital costs and
generating efciencies were assumed to be dynamically
improving depending on their respective rates of market
penetration. Overall, the generation costs were at the
lower end of the ranges for Annex 1 countries (Table 2)
for all technologies. Compared with the costs for Annex
1 countries (Tables 2 and 4), carbon mitigation costs in
the USA tended to be lower for IGCC coal, CCGT gas,
PF CO2 capture, CCGT CO2 capture, and PV and
in the mid to high end of the ranges for nuclear, biomass
and wind. Once sufcient capacities have been adopted
in the market place, coal-red IGCC power stations
would have similar costs but lower emissions than the
baseline pulverised coal power station due to their
higher efciency. Gas-red CCGT power stations offer
lower cost generation than coal at current gas prices and
produce around only half the CO2 emissions. Data on
CO2 capture and storage were again taken from IEA
GHG R&D Programme studies (Audus, 2000). This
technology could reduce emissions by about 80% for
additional costs around 3 c=kWh for pulverised coal,
2:5 c=kWh for coal IGCC and 1:5 c=kWh for gas
CCGT. Nuclear power is more expensive than coalred or gas-red generation but can be a cheaper option
than CO2 capture and storage. Wind power can be
competitive with conventional coal and gas power
generation at sites with high mean annual speeds of
910 m=s: Biomass can also contribute to carbon
mitigation, especially where forest or agricultural
residues are available at very low or negative costs.
Where biomass is more costly (either because in-forest
residue material used requires collection or because
purpose-grown short rotation forest energy crops are
used), or where wind conditions are poorer, the
technologies will be less competitive for reducing
emissions. PV and solar thermal technologies remain
relatively expensive for large-scale power generation,
but will be increasingly attractive in niche markets or for
off-grid generation as capital costs fall.
Carbon dioxide emissions and mitigation costs
were compared to both the coal-red pulverised fuel
power station and the gas-red CCGT (Table 6).
Compared with the coal base-case, it was projected
that in 2010 under assumptions of improved fossil
fuel technologies, an IGCC plant would offer a small
reduction in emissions at positive or negative costs
depending on the specic circumstances. A gas-red

R.E.H. Sims et al. / Energy Policy 31 (2003) 13151326

CCGT would generally have negative mitigation costs

compared with the coal-red baseline, reecting its
lower generating costs. Carbon dioxide capture and
storage options would enable signicant reductions in
emissions from coal-red generation but the cost would
be between $100 and 150=t C depending on the
technology used. Gas-red CCGT with CO2 capture
and storage appears more attractive, principally because
switching to CCGT is attractive in the rst place
wherever natural gas production and delivery infrastructures exist. Nuclear power mitigation costs are in
the range $50100=t C versus coal but $125300=t C
versus gas. It is uncertain whether there would be
sufcient capacity available for wind or biomass to
deliver as much electricity as could be produced by fossil
fuel-red plants, but certainly the installed capacities
would be limited at the low costs shown due to a scarcity
of good sites.
When compared with the gas-red CCGT baseline
under USA conditions, most of the mitigation options
were found to be considerably more expensive with
the exception of wind on good sites. Carbon dioxide
capture and storage appeared relatively unattractive for
both coal or gas-red plants, but could achieve
signicant reductions in emissions if carbon was valued
at $150=t C or above. Biomass and nuclear were
attractive options under certain circumstances whereas
PV and solar thermal were again expensive mitigation
options.

4. Conclusions
Compared with burning coal or gas in conventional
power generating plant designs, there are several
alternative technological ways to generate electricity
and reduce greenhouse gas emissions cost effectively.
They include using plant designs which offer more
efcient power generation conversion of fossil fuels,
greater use of renewable energy or nuclear power,
and the capture and disposal of CO2 : The choice, in
terms of cost savings and carbon emission reduction
benets, is very site specic and the least-cost option in
terms of $=t C avoided will differ from case to case.
Most of the technologies considered in the discussion
have a role to play since, for each, it is possible to obtain
both cost and carbon emission reductions under certain
circumstances. The exceptions are solar power and
carbon dioxide sequestration, though this new concept
gives future opportunity for costs to be reduced with
further experience. Compared with business as usual,
the global electricity sector has the potential to lower its
carbon emission reductions by between 1.54.7% by
2010 and 8.718.7% by 2020 based on the current
literature and the range of assumptions used in this
analysis.

1325

Acknowledgements
The secretariat of Working Group III of the IPCC
Third Assessment Report agreed for this section of
the report to be published in this revised and condensed
form. The other co-authors of Chapter 3 are acknowledged, several having provided useful reviews of this
analysis during the preparation of the IPCC report.

References
Ackermann, T., Garner, K., Gardiner, A., 1999. Wind power
generation in weak gridseconomic optimisation and power
quality simulation. Proceedings of the World Renewable Energy
Congress, Perth, Australia, Murdoch University, pp. 527532.
ISBN 0-86905-695-6.
AGO, Australian Greenhouse Ofce, 1998. Renewable energy showcase projects. Australian Greenhouse Ofce, Canberra.
www.greenhouse.gov.au/renewable/renew3.html.
Audus, H., 2000. Leading options for the capture of CO2 at power stations.
Proceedings of the Fifth International Conference on Greenhouse Gas
Control Technologies, Cairns, Australia, 1316 August.
DTI, Department of Trade and Industry, 1999. New and emerging
renewable energy prospects for the 21st century, Department of
Trade and Industry Report DTI/Pub 4024/3K/99/NPURN99/744
London, United Kingdom.
EPRI/DOE, 1997. Renewable energy technology characterizations,
Electric Power Research Institute and US Department of Energy
Report EPRI TR-109496, December.
EWEA, 1999. Wind energythe facts European Wind Energy
Association Report prepared for the Directorate-General XVIIEnergy, European Commission.
Gregory, K., Rogner, H.-H., 1998. Energy resources and conversion
technologies for the 21st century. Mitigation and Adaptation
Strategies for Global Change 3, 171229.
IAEA, 2000. Nuclear power reactors in the world. International
Atomic Energy Agency Reference Data Series No. 2, April Edition,
Vienna, Austria.
IEA, 1998. World Energy Outlook1998 Update. International
Energy Agency Report, IEA/OECD, Paris, France.
IPCC, 1996. Second Assessment Report. United Nations Intergovernmental Panel on Climate Change. Cambridge University Press,
England.
IPCC, 2001. Third Assessment Report. United Nations Intergovernmental Panel on Climate Change, Working Group III, Cambridge
University Press, England (Chapter 3).
KPMG, 1999. Solar energyfrom perennial promise to competitive
alternative. Project Number 562. KPMG Bureau voor Economische Argumentatie, Hoofddorp, Netherlands, p. 61.
Neij, L., 1997. Use of experience curves to analyze the prospects for
diffusion and adoption of renewable energy technology. Energy
Policy 23, 10991107.
Nicholls, D.R., 1998. Status of the pebble bed modular reactor.
ESKOM Report, South Africa.
OECD, 1998. Projected costs of generating electricity: update 1998.
Organization for Economic Cooperation and Development
(OECD). International Energy Agency (IEA) and Nuclear Energy
Agency (NEA) Joint Report, Paris, France.
Pitcher, K., 2000. Turning willow into megawattsthe ARBRE
project. Renewable Energy World 3 (6), 3445.
Shell, 1996. The evolution of the worlds energy systems. Shell
International Limited report, Shell Centre, London SE1 7NA,
UK, 7pp.

1326

R.E.H. Sims et al. / Energy Policy 31 (2003) 13151326

US DOE/EIA, 2000. International energy outlook 2000. US Department of Energy Information Administration, Report No. DOE/
EIA-0554(2000), Washington, DC.
Walker, D., Botha, P., White, G., 1998. Tararua wind farm. Proceedings
of the 36th Annual Conference, Australian and New Zealand Solar
Energy Society, Christchurch, November, pp. 637641.

WEA, 2000. World Energy Assessment: energy and the challenge of

sustainability. United Nations Development Programme (UNDP),
UN Department of Economic and Social Affairs (UNDESA) and
World Energy Council (WEC) Joint Report, New York, USA.