4.3.1.1 Coal and peat

Coal is the world’s most abundant fossil fuel and continues to be a vital resource in many countries ( IEA, 2003e [NPR] ). In 2005, coal accounted for around 25% of total world energy consumption primarily in the electricity and industrial sectors ( BP, 2005 [NPR] ; US EIA, 2005 [NPR, MoS] ; Enerdata, 2004 ). Global proven recoverable reserves of coal are about 22,000 EJ ( BP, 2004 [NPR] ; WEC, 2004b [NPR] ) with another 11,000 EJ of probable reserves and an estimated additional possible resource of 100,000 EJ for all types. Although coal deposits are widely distributed, over half of the world’s recoverable reserves are located in the US (27%), Russia (17%) and China (13%). India, Australia, South Africa, Ukraine, Kazakhstan and the former Yugoslavia account for an additional 33% ( US DOE, 2005 [NPR] ). Two thirds of the proven reserves are hard coal (anthracite and bituminous) and the remainder are sub-bituminous and lignite. Together these resources represent stores of over 12,800 GtCO₂. Consumption was around 120 EJ/yr in 2005, which introduced approximately 9.2 GtCO₂/yr into the atmosphere.

Peat (partially decayed plant matter together with minerals) has been used as a fuel for thousands of years, particularly in Northern Europe. In Finland, it provides 7% of electricity and 19% of district heating.

Technologies

The demand for coal is expected to more than double by 2030 and the IEA has estimated that more than 4500 GW of new power plants (half in developing countries) will be required in this period ( IEA, 2004a [NPR] ). The implementation of modern high-efficiency and clean utilization coal technologies is key to the development of economies if effects on society and environment are to be minimized ( Section 4.5.4 ).

Most installed coal-fired electricity-generating plants are of a conventional subcritical pulverized fuel design, with typical efficiencies of about 35% for the more modern units. Supercritical steam plants are in commercial use in many developed countries and are being installed in greater numbers in developing countries such as China ( Philibert and Podkanski, 2005 [NPR] ). Current supercritical technologies employ steam temperatures of up to 600ºC and pressures of 280 bar delivering fuel to electricity-cycle efficiencies of about 42% ( Moore, 2005 [Ambiguous] ). Conversion efficiencies of almost 50% are possible in the best supercritical plants, but are more costly (Equitech, 2005; IPCC, 2001 [NPR] ; Danish Energy Authority, 2005 ). Improved efficiencies have reduced the amount of waste heat and CO₂ that would otherwise have been emitted per unit of electricity generation.

Technologies have changed little since the TAR. Supercritical plants are now built to an international standard, however, and a CSIRO 2005 [NPR] ) project is under way to investigate the production of ultra-clean coal that reduces ash below 0.25%, sulphur to low levels and, with combined-cycle direct-fired turbines, can reduce GHG emissions by 24% per kWh, compared with conventional coal power stations.

Gasifying coal prior to conversion to heat reduces the emissions of sulphur, nitrogen oxides, and mercury, resulting in a much cleaner fuel while reducing the cost of capturing CO₂ emissions from the flue gas where that is conducted. Continued development of conventional combustion integrated gasification combined cycle (IGCC) systems is expected to further reduce emissions.

Coal-to-liquids (CTL) is well understood and regaining interest, but will increase GHG emissions significantly without CCS ( Section 4.3.6 ). Liquefaction can be performed by direct solvent extraction and hydrogenation of the resulting liquid at up to 67% efficiency ( DTI, 1999 [NPR] ) or indirectly by gasification then producing liquids by Fischer-Tropsch catalytic synthesis as in the three SASOL plants in South Africa. These produce 0.15 Mbbl/day of synthetic diesel fuel (80%) plus naphtha (20%) at 37–50% thermal efficiency. Lower-quality coals would reduce the thermal efficiency whereas co-production with electricity and heat (at a 1:8 ratio) could increase it and reduce the liquid fuel costs by around 10%.

Production costs of CTL appear competitive when crude oil is around 35–45 US$/bbl, assuming a coal price of 1 US$/GJ. Converting lignite at 0.50 US$/GJ close to the mine could compete with production costs of about 30 US$/bbl. The CTL process is less sensitive to feedstock prices than the gas-to-liquids (GTL) process, but the capital costs are much higher ( IEA, 2005e [NPR, MoS] ). An 80,000 barrel per day CTL installation would cost about 5 billion US$ and would need at least 2–4 Gt of coal reserves available to be viable.

4.3.1.2 Gaseous fuels

Conventional natural gas

Natural gas production has been increasing in the Middle East and Asia–Oceania regions since the 1980 s. Globally, from 1994 – 2004, it showed an annual growth rate of 2.3%. During 2005, 11% of natural gas was produced in the Middle East, while Europe and Eurasia produced 38%, and North America 27% ( BP, 2006 [NPR] ). Natural gas presently accounts for 21% of global consumption of modern energy at around 100 EJ/yr, contributing around 5.5 GtCO₂ annually to the atmosphere.

Proven global reserves of natural gas are estimated to be 6500 EJ ( BP, 2006 [NPR] ; WEC, 2004c [NPR] ; USGS, 2004b [NPR] ). Almost three quarters are located in the Middle East, and the transitional economies of the FSU and Eastern Europe. Russia, Iran and Qatar together account for about 56% of gas reserves, whereas the remaining reserves are more evenly distributed on a regional basis including North Africa ( BP, 2006 [NPR] ). Probable reserves and possible undiscovered resources that expect to be added over the next 25 years account for 2500 EJ and 4500 EJ respectively ( USGS, 2004a [NPR] ), although other estimates are less optimistic.

Natural gas-fired power generation has grown rapidly since the 1980 s because it is relatively superior to other fossil-fuel technologies in terms of investment costs, fuel efficiency, operating flexibility, rapid deployment and environmental benefits, especially when fuel costs were relatively low. Combined cycle, gas turbine (CCGT) plants produce less CO₂ per unit energy output than coal or oil technologies because of the higher hydrogen-carbon ratio of methane and the relatively high thermal efficiency of the technology. A large number of CCGT plants currently being planned, built, or operating are in the 100–500 MW_e size range. Advanced gas turbines currently under development, such as so-called ‘H’ designs, may have efficiencies approaching 60% using high combustion temperatures, steam-cooled turbine blades and more complex steam cycles.

Despite rising prices, natural gas is forecast to continue to be the fastest-growing primary fossil fuel energy source worldwide ( IEA, 2006b [NPR] ), maintaining average growth of 2.0% annually and rising to 161 EJ consumption in 2025 . The industrial sector is projected to account for nearly 23% of global natural gas demand in 2030, with a similar amount used to supply new and replacement electric power generation. The share of natural gas used to generate electricity worldwide is projected to increase from 25% of primary energy in 2004 to 31% in 2030 ( IEA, 2006b [NPR] ).

LNG

Meeting future increases in global natural gas demand for direct use by the industrial and commercial sectors as well as for power generation will require development and scale-up of liquefied natural gas (LNG) as an energy carrier. LNG transportation already accounts for 26% of total international natural gas trade in 2002, or about 6% of world natural gas consumption and is expected to increase substantially.

The Pacific Basin is the largest LNG-producing region in the world, supplying around 50% of all global exports in 2002 ( US EIA, 2005 [NPR, MoS] ). The share of total US natural gas consumption met by net imports of LNG is expected to grow from about 1% in 2002 to 15% (4.5 EJ) in 2015 and to over 20% (6.8 EJ) in 2025 . Losses during the LNG liquefaction process are estimated to be 7 to 13% of the energy content of the withdrawn natural gas being larger than the typical loss of pipeline transportation over 2000 km.

LPG

Liquefied petroleum gas (LPG) is a mixture of propane, butane, and other hydrocarbons produced as a by-product of natural gas processing and crude oil refining. Total global consumption of LPG amounted to over 10 EJ in 2004 (MCH/WLPGA, 2005 ), equivalent to 10% of global natural gas consumption ( (Venn, 2005 ) ). Growth is likely to be modest with current share maintained.

Unconventional natural gas

Methane stored in a variety of geologically complex, unconventional reservoirs, such as tight gas sands, fractured shales, coal beds and hydrates, is more abundant than conventional gas ( Table 4.2 ). Development and distribution of these unconventional gas resources remain limited worldwide, but there is growing interest in selected tight gas sands and coal-bed methane (CBM). Probable CBM resources in the US alone are estimated to be almost 800 EJ but less than 110 EJ is believed to be economically recoverable ( USGS, 2004b [NPR] ) unless gas prices rise significantly. Worldwide resources may be larger than 8000 EJ, but a scarcity of basic information on the gas content of coal resources makes this number highly speculative.

Large quantities of tight gas are known to exist in geologically complex formations with low permeability, particularly in the US, where most exploration and production has been undertaken. However, only a small percentage is economically viable with existing technology and current US annual production has stabilized between 2.7 and 3.8 EJ.

Methane gas hydrates occur naturally in abundance worldwide and are stable as deep marine sediments on the ocean floor at depths greater than 300m and in polar permafrost regions at shallower depths. The amount of carbon bound in hydrates is not well understood, but is estimated to be twice as large as in all other known fossil fuels ( USGS, 2004a [NPR] ). Hydrates may provide an enormous resource with estimates varying from 60,000 EJ ( USGS, 2004a [NPR] ) to 800,000 EJ (Encyclopedia of Energy,) to 800,000 EJ (Encyclopedia of Energy,). Recovering the methane is difficult, however, and represents a significant environmental problem if unintentionally released to the atmosphere during extraction. Safe and economic extraction technologies are yet to be developed ( USGS, 2004a [NPR] ). Hydrates also contain high levels of CO₂ that may have to be captured to produce pipeline-quality gas (Encyclopedia of Energy, that may have to be captured to produce pipeline-quality gas (Encyclopedia of Energy,).

The GTL process is gaining renewed interest due to higher oil prices, particularly for developing uneconomic natural gas reserves such as those associated with oil extraction at isolated gas fields which lie far from markets. As for CTL, the natural gas is turned into synthesis gas, which is converted by the Fischer-Tropsch process to synthetic fuels. At present, at least nine commercial GTL projects are progressing through various development stages in gas-rich countries such as Qatar, Iran, Russia, Nigeria, Australia, Malaysia and Algeria with worldwide production estimated at 0.58 Mbbl/day (FACTS, 2005 ). GTL conversion technologies are around 55% efficient and can help bring some of the estimated 6000 EJ of stranded gas resources to market. Production costs vary depending on gas prices, but where stranded gas is available at 0.5 US$/GJ production costs are around 30 US$ a barrel ( IEA 2006a [NPR] ). Higher CO₂ emissions per unit consumed compared with conventional oil products.

4.3.1.3 Petroleum fuels

Conventional oil products extracted from crude oil-well bores and processed by primary, secondary or tertiary methods represent about 37% of total world energy consumption ( Figure 4.4 and Table 4.2 ) with major resources concentrated in relatively few countries. Two thirds of proven crude oil reserves are located in the Middle East and North Africa ( IEA, 2005a [NPR] ).

Known or proven reserves are those extractable at today’s prices and technologies. Additional probable and possible resources are based on historical experience in geological basins. While new discoveries have lagged behind production for more than 20 years, reserve additions from all sources including discoveries, extensions, revisions and improvements in oil recovery continue to outpace production ( IEA, 2005b [NPR] ).

Various studies and models have been used to forecast future oil production ( US EIA, 2004 [NPR] ; Bentley, 2005 [NPR] ). Geological models take into consideration the volume and quality of hydrocarbons but do not include economic effects on price, which in turn has a direct effect on supply and the overall rate of recovery. Mathematical models generally use the historical as well as the observed patterns of production to estimate a peak (or several peaks) reached when half the reserves are consumed.

Assessments of the amount of oil consumed, the amount remaining for extraction, and whether the peak oil tipping point is close or not, have been very controversial ( Hirsch et al., 2005 [NPR] ). Estimates of the ultimate extractable resource (proven + probable + possible reserves) with which the world was endowed have varied from less than 5730 EJ to 34,000 EJ ( 1000 to 6000 Gbbl), though the more recent predictions have all ranged between 11,500–17,000 EJ ( 2000 –3000 Gbbl) ( Figure 4.8 ). Over time, the prediction trend showed increasing resource estimates in the 1940 s and 1950 s as more fields were discovered. However, the very optimistic estimates of the 1970 s were later discredited and a relatively constant estimate has since been observed.

Figure 4.8: Estimates of the global ultimate extractable conventional oil resource by year of publications.

Source: Based on Bentley, 2002a [MoS] ; Andrews and Udal, 2003 [NPR] . 

Specific analyses include ( Bentley 2002b ) ), who concluded that 4870 EJ had been consumed by 1998 and that 6300 EJ will have been extracted by 2008 . The US Geological Survey ( USGS, 2000 [NPR] ) the World Petroleum Congress and the IFP agreed that approximately 4580 EJ (800 Gbbl) have been consumed in the past 150 years and 5730 EJ ( 1000 Gbbl) of proven reserves remain. Other detailed analyses (e.g. USGS, 2000 [NPR] ) also estimated there are 4150 EJ of probable and possible resources still available for extraction. Thus, the total available potential proven reserves plus resources of around 10,000 EJ ( BP, 2004 [NPR] ; WEC, 2004b [NPR] ) should be sufficient for about 70 years’ supply at present rates of consumption. Since consumption rates will continue to rise, however, 30 to 40 years’ supply is a more reasonable estimate ( Hallock et al., 2004 [MoS] ). Burning this amount of petroleum resources would release approximately 700 GtCO₂ (200 GtC) into the atmosphere, about two thirds the amount released to date from all fossil-fuel consumption. Opportunities for energy-efficiency improvements in oil refineries and associated chemical plants are covered in Chapter 7 .

4.3.1.4 Unconventional oil

As conventional oil supplies become scarce and extraction costs increase, unconventional liquid fuels, in addition to CTL and GTL, will become more economically attractive, but offset by greater environmental costs ( Williams et al., 2006 [NPR] ). Oil that requires extra processing such as from shales, heavy oils and oil (tar) sands is classified as unconventional. Resource estimates are uncertain, but together contributed around 3% of world oil production in 2005 (2.8 EJ) and could reach 4.6 EJ by 2020 ( USGS, 2000 [NPR] ) and up to 6 EJ by 2030 ( IEA, 2005a [NPR] ). The oil industry has the potential to diversify the product mix, thereby adding to fuel-supply security, but higher environmental impacts may result and investment in new infrastructure would be needed.

Heavy oil reserves are greater than 6870 EJ ( 1200 Gbbl) of oil equivalent with around 1550 EJ technically recoverable. The Orinoco Delta, Venezuela has a total resource of 1500 EJ with current production of 1.2 EJ/yr ( WEC, 2004c [NPR] ). Plans for 2009 are to apply deep-conversion, delayed coking technology to produce 0.6 Mbbl/day of high-value transport fuels.

Oil shales (kerogen that has not completed the full geological conversion to oil due to insufficient heat and pressure) represent a potential resource of 20,000 EJ with a current production of just 0.024 EJ/yr, mostly in the US, Brazil, China and Estonia. Around 80% of the total resource lies in the western US with 500 Gbbl of medium-quality reserves from rocks yielding 95 L of oil per tonne but with 1000 Gbbl potential if utilizing lower-quality rock. Mining and upgrading of oil shale to syncrude fuel costs around 11 US$/bbl. As with oil sands (below), the availability of abundant water is an issue.

Around 80% of the known global tar sand resource of 15,000 EJ is in Alberta, Canada, which has a current production of 1.6 EJ/yr, representing around 15% of national oil demand. Around 310 Gbbl is recoverable ( CAPP, 2006 [NPR] ). Production of around 2 Mbbl/day by 2010 could provide more than half of Canada’s projected total oil production with 4 Mbbl/day possible by 2020 . Total resources represent at least 400 Gt of stored carbon and will probably be added to as more are discovered, assuming that natural gas and water (steam) to extract the hydrocarbons are available at a reasonable cost.

Technologies for recovering tar sands include open cast (surface) mining where the deposits are shallow enough (which accounts for 10% of the resource but 80% of current extraction), or injection of steam into wells in situ to reduce the viscosity of the oil prior to extraction. Mining requires over 100m³ of natural gas per barrel of bitumen extracted and in situ around 25m³. In both cases cleaning and upgrading to a level suitable for refining consumes a further 25–50m³ per barrel of oil feedstock. The mining process uses about four litres of water to produce one litre of oil but produces a refinable product. The in situ process uses about two litres of water to one of oil, but the very heavy product needs cleaning and diluting (usually with naptha) at the refinery or sent to an upgrader to yield syncrude at an energy efficiency of around 75% ( NEB, 2006 [NPR] ). The energy efficiency of oil sand upgrading is around 75%. Mining, producing and upgrading oil sands presently costs about 15 US$/bbl ( IEA, 2006a [NPR] ) but new greenfield projects would cost around 30–35 US$/bbl due to project-cost inflation in recent years ( NEB, 2006 [NPR] ). If CCS is integrated, then an additional 5 US$ per barrel at least should be added. Comparable costs for conventional oil are 4–6 US$/bbl for exploration and production and 1–2 US$/bbl for refining.

Mining of oil sands leaves behind large quantities of pollutants and areas of disturbed land.

The total CO₂ emitted per unit of energy during production of liquid unconventional oils is greater than for a unit of conventional oil products due to higher energy inputs for extraction and processing. Net emissions amount to 15–34 kgCO₂ (4–9 kgC) per GJ of transport fuel compared with around 5-10 kgCO₂ (1.3-2.7 kgC) per GJ for conventional oil ( IEA, 2005d [NPR, MoS] , Woyllinowicz et al., 2005 [NPR] ). Oil sands currently produce around 3–4 times the pre-combustion emissions (CO₂/GJ liquid fuel) compared with conventional oil extraction and refining, whereas large-scale production of oil-shale processing would be about 5 times, GTL 3–4 times, and CTL around 7–8 times when using sub-bituminous coal. The Athabascan oil-sands project has refining energy expenditures of 1 GJ energy input per 6 GJ bitumen processed, producing emissions of 11 kgCO₂ (3 kgC) per GJ from refining alone, but with a voluntary reduction goal of 50% by 2010 (Shell, 2006 ).

4.3.2.1 Risks and environmental impacts

Regulations demand that public and occupational radiation doses from the operation of nuclear facilities be kept as low as reasonably achievable and below statutory limits. Mining, milling, power-plant operation and reprocessing of spent fuel dominate the collective radiation doses ( OECD, 2000 [NPR] ). Protective actions for mill-tailing piles and ponds have been demonstrated to be effective when applied to prevent or reduce long-term impacts from radon emanation. In the framework of the IAEA’s Nuclear Safety Convention ( IAEA, 1994 [NPR] ), the IAEA member countries have agreed to maintain high safety culture to continuously improve the safety of nuclear facilities. However, risks of radiation leakage resulting from accidents at a power plant or during the transport of spent fuel remain controversial.

Operators of nuclear power plants are usually liable for any damage to third parties caused by an incident at their installation regardless of fault ( UIC, 2005 [NPR] ), as defined by both international conventions and national legislation. In 2004, the contracting parties to the OECD Paris and Brussels Conventions signed Amending Protocols setting the minimum liability limit at 700 million € with additional compensation up to 800 € through public funds. Many non-OECD countries have similar arrangements through the IAEA’s Vienna Convention. In the US, the national Price-Anderson Act provides compensation up to 300 million US$ covered by an insurance paid by each reactor and also by a reactor-operator pool from the 104 reactors, which provides 10.4 billion US$.

4.3.2.2 Nuclear-waste management, disposal and proliferation aspects

The main safety objective of nuclear waste management ( IAEA, 1997 [NPR] ; IAEA, 2005b [NPR] ) is that human health and the environment need to be protected now and in the future without imposing undue burdens on future generations. Repositories are in operation for the disposal of low- and medium-level radioactive wastes in several countries but none yet exist for high-level waste (HLW) such as spent light-water reactor (LWR) fuel. Deep geological repositories are the most extensively studied option but resolution of both technical and political/societal issues is still needed.

In 2001, the Finnish Parliament agreed to site a spent fuel repository near the Olkiluoto nuclear power plant. After detailed rock-characterization studies, construction is scheduled to start soon after 2010 with commissioning planned for around 2020 . In Sweden, a repository-siting process is concentrating on the comparison of several site alternatives close to the Oskarshamn and Forsmark nuclear power plants. In the US, the Yucca Mountain area has been chosen, amidst much controversy, as the preferred site for a HLW repository and extensive site-characterization and design studies are underway, although not without significant opposition. It is not expected to begin accepting HLW before 2015 . France is also progressing on deep geological disposal as the reference solution for long-lived radioactive HLW and sets 2015 as the target date for licensing a repository and 2025 for opening it ( DGEMP, 2006 [NPR] ). Spent-fuel reprocessing and recycling of separate actinides would significantly reduce the volume and radionuclide inventory of HLW.

The enrichment of uranium (U-235), reprocessing of spent fuel and plutonium separation are critical steps for nuclear-weapons proliferation. The Treaty on Non-Proliferation of Nuclear Weapons (NPT) has been ratified by nearly 190 countries. Compliance with the terms of the NPT is verified and monitored by the IAEA. Improving proliferation resistance is a key objective in the development of next-generation nuclear reactors and associated advanced fuel-cycle technologies. For once-through uranium systems, stocks of plutonium are continuously built up in the spent fuel, but only become accessible if reprocessed. Recycling through fast-spectrum reactors on the other hand allows most of this material to be burned up in the reactor to generate more power, although there are vulnerabilities in the reprocessing step and hence still the need for careful safeguards. Advanced reprocessing and partitioning and transmutation technologies could minimize the volumes and toxicity of wastes for geological disposal, yet uncertainties about proliferation-risk and cost remain.

4.3.2.3 Development of future nuclear-power systems

Present designs of reactors are classed as Generations I through III ( Figure 4.9 ). Generation III+ advanced reactors are now being planned and could first become operational during the period 2010 – 2020 ( GIF, 2002 [NPR] ) and state-of-the-art thereafter to meet anticipated growth in demand. These evolutionary reactor designs claim to have improved economics, simpler safety systems with the impacts of severe accidents limited to the close vicinity of the reactor site. Examples include the European design of a pressurized water reactor (EPR) scheduled to be operating in Finland around 2010 and the Flamanville 3 reactor planned in France.

Figure 4.9: Evolution of nuclear power systems from Generation I commercial reactors in the 1950 s up to the future Generation IV systems which could be operational after about 2030 .

Source: GIF, 2002 [NPR] .

Notes: LWR = light-water reactor; PWR = pressurized water reactor; BWR = boiling-water reactor; ABWR = advanced boiling-water reactor; CANDU = Canada Deuterium Uranium.

Generation IV nuclear-energy technologies that may become operational after about 2030 employ advanced closed-fuel cycle systems with more efficient use of uranium and thorium resources. Advanced designs are being pursued mainly by the Generation-IV International Forum (GIF, a group of ten nations plus the EU and coordinated by the US Department of Energy) as well as the International Project on Innovative Nuclear Reactors and Fuel Cycles (INPRO) coordinated by the IAEA. The Global Nuclear Energy Partnership ( US DOE, 2006 [NPR] ), proposed by the US, has similar objectives. These initiatives focus on the development of reactors and fuel cycles that provide economically competitive, safe and environmentally sound energy services based on technology designs that exclude severe accidents, involve proliferation-resistant fuel cycles decoupled from any fuel-resource constraints, and minimize HLW. Much additional technology development would be needed to meet these long-term goals so strategic public RD&D funding is required, since there is limited industrial/commercial interest at this early stage.

GIF has developed a framework to plan and conduct international cooperative research on advanced (breeder or burner) nuclear-energy systems ( GIF, 2002 [NPR] ) including three designs of fast-neutron reactor, (sodium-cooled, gas-cooled and lead-cooled) as well as high-temperature reactors. Reactor concepts capable of producing high-temperature nuclear heat are intended to be employed also for hydrogen generation, either by electrolysis or directly by special thermo-chemical water-splitting processes or steam reforming. There is also an ongoing development project by the South African utility ESKOM for an innovative high-temperature, pebble-bed modular reactor. Specific features include its smaller unit size, modularity, improved safety by use of passive features, lower power production costs and the direct gas-cycle design utilizing the Brayton cycle ( Koster et al., 2003 [MoS] ; NER, 2004 [NPR] ). The supercritical light-water reactor is also one of the GIF concepts intended to be operated under supercritical water pressure and temperature conditions. Conceivably, some of these concepts may come into practical use and offer better prospects for future use of nuclear power.

Experience of the past three decades has shown that nuclear power can be beneficial if employed carefully, but can cause great problems if not. It has the potential for an expanded role as a cost-effective mitigation option, but the problems of potential reactor accidents, nuclear waste management and disposal and nuclear weapon proliferation will still be constraining factors.

4.3.2.4 Uranium exploration, extraction and refining

In the long term, the potential of nuclear power is dependent upon the uranium resources available. Reserve estimates of the uranium resource vary with assumptions for its use ( Figure 4.10 ). Used in typical light-water reactors (LWR) the identified resources of 4.7 Mt uranium, at prices up to 130 US$/kg, correspond to about 2400 EJ of primary energy and should be sufficient for about 100 years’ supply ( OECD, 2006b [NPR] ) at the 2004 level of consumption. The total conventional proven (identified) and probable (yet undiscovered) uranium resources are about 14.8 Mt (7400 EJ). There are also unconventional uranium resources such as those contained in phosphate minerals, which are recoverable for between 60 and 100 US$/kg ( OECD, 2004a [NPR] ).

If used in present reactor designs with a ‘once-through’ fuel cycle, only a small percentage of the energy content is utilized from the fissile isotope U-235 (0.7% in natural uranium). Uranium reserves would last only a few hundred years at current rate of consumption ( Figure 4.10 ). With fast-spectrum reactors operated in a ‘closed’ fuel cycle by reprocessing the spent fuel and extracting the unused uranium and plutonium produced, the reserves of natural uranium may be extended to several thousand years at current consumption levels. In the recycle option, fast-spectrum reactors utilize depleted uranium and only plutonium is recycled so that the uranium-resource efficiency is increased by a factor of 30 ( Figure 4.10 OECD, 2001 [NPR] ). Thereby the estimated enhanced resource availability of total conventional uranium resources corresponds to about 220,000 EJ primary energy ( Table 4.2 ). Even if the nuclear industry expands significantly, sufficient fuel is available for centuries. If advanced breeder reactors could be designed in the future to efficiently utilize recycled or depleted uranium and all actinides, then the resource utilization efficiency would be further improved by an additional factor of eight ( OECD, 2006c [NPR] ).

Figure 4.10: Estimated years of uranium-resource availability for various nuclear technologies at 2004 nuclear-power utilization levels.

Source: OECD, 2006b [NPR] ; OECD, 2006c [NPR] . 

Nuclear fuels could also be based on thorium with proven and probable resources being about 4.5 Mt ( OECD, 2004a [NPR] ). Thorium-based fast-spectrum reactors appear capable of at least doubling the effective resource base, but the technology remains to be developed to ascertain its commercial feasibility ( IAEA, 2005a [NPR] ). There are not yet sufficient commercial incentives for thorium-based reactors except perhaps in India. The thorium fuel cycle is claimed to be more proliferation-resistant than other fuel cycles since it produces fissionable U-233 instead of fissionable plutonium, and, as a by-product, U-232 that has a daughter nuclide emitting high-energy photons.

4.3.2.5 Nuclear fusion

Energy from the fusion of heavy hydrogen fuel (deuterium, tritium) is actively being pursued as a long-term almost inexhaustible supply of energy with helium as the by-product. The scientific feasibility of fusion energy has been proven, but technical feasibility remains to be demonstrated in experimental facilities. A major international effort, the proposed international thermonuclear experimental reactor ( ITER, 2006 [NPR] ), aims to demonstrate magnetic containment of sustained, self-heated plasma under fusion temperatures. This 10 billion US$ pilot plant to be built in France is planned to operate for 20 years and will resolve many scientific and engineering challenges. Commercialization of fusion-power production is thought to become viable by about 2050, assuming initial demonstration is successful (Smith et al., 2006a [NotFound] ; Cook et al., 2005 [NPR] ).

4.3.3.1 Hydroelectricity

Large (>10 MW) hydroelectricity systems accounted for over 2800 TWh of consumer energy in 2004 ( BP, 2006 [NPR] ) and provided 16% of global electricity (90% of renewable electricity). Hydro projects under construction could increase the share of electricity by about 4.5% on completion ( WEC, 2004d [NPR] ) and new projects could be deployed to provide a further 6000 TWh/yr or more of electricity economically ( BP, 2004 [NPR] ; IEA, 2006a [NPR] ), mainly in developing countries. Repowering existing plants with more powerful and efficient turbine designs can be cost effective whatever the plant scale. Where hydro expansion is occurring, particularly in China and India, major social disruptions, ecological impacts on existing river ecosystems and fisheries and related evaporative water losses are stimulating public opposition. These and environmental concerns may mean that obtaining resource permits is a constraint.

Small (<10 MW) and micro (<1 MW) hydropower systems, usually run-of-river schemes, have provided electricity to many rural communities in developing countries such as Nepal. Their present generation output is uncertain with predictions ranging from 4 TWh/yr ( WEC, 2004d [NPR] ) to 9% of total hydropower output at 250 TWh/yr ( Martinot et al., 2006 [NPR, SRC] ). The global technical potential of small and micro hydro is around 150–200 GW with many unexploited resource sites available. About 75% of water reservoirs in the world were built for irrigation, flood control and urban water-supply schemes and many could have small hydropower generation retrofits added. Generating costs range from 20 to 90 US$/MWh but with additional costs needed for power connection and distribution. These costs can be prohibitive in remote areas, even for mini-grids, and some form of financial assistance from aid programmes or governments is often necessary.

The high level of flexibility of hydro plants enables peak loads in electricity demand to be followed. Some schemes, such as the 12.6 GW Itaipu plant in Brazil/Paraguay, are run as baseload generators with an average capacity factor of >80%, whereas others (as in the 24 GW of pumped storage plant in Japan) are used mainly as fast-response peaking plants, giving a factor closer to 40% capacity. Evaluations of hybrid hydro/wind systems, hydro/hydrogen systems and low-head run-of-river systems are under review ( IEA, 2006d [NPR] ).

GHG emissions vary with reservoir location, power density (W capacity per m² flooded), flow rate, and whether dam or run-or-river plant. Recently, the GHG footprint of hydropower reservoirs has been questioned ( Fearnside, 2004 [JoC] ; UNESCO, 2006 [NPR] ). Some reservoirs have been shown to absorb CO₂ at their surface, but most emit small amounts as water conveys carbon in the natural carbon cycle ( Tremblay, 2005 [NPR] ). High emissions of CH₄ have been recorded at shallow, plateau-type tropical reservoirs where the natural carbon cycle is most productive ( Delmas, 2005 [NPR] ). Deep water reservoirs at similar low latitudes tend to exhibit lower emissions. Methane from natural floodplains and wetlands may be suppressed if they are inundated by a new reservoir since the methane is oxidized as it rises through the covering water column ( Huttunen, 2005 [Ambiguous] ; dos Santos, 2005 [Ambiguous] ). Methane formation in freshwater produces by-product carbon compounds (phenolic and humic acids) that effectively sequester the carbon involved ( Sikar, 2005 [NPR] ). For shallow tropical reservoirs, further research is needed to establish the extent to which these may increase methane emissions.

Several Brazilian hydro-reservoirs were compared using life-cycle analyses with combined-cycle natural gas turbine (CCGT) plants of 50% efficiency ( dos Santos et al., 2004 [NPR] ). Emissions from flooded reservoirs tended to be less per kWh generated than those produced from the CCGT power plants. Large hydropower complexes with greater power density had the best environmental performance, whereas those with lower power density produced similar GHG emissions to the CCGT plants. For most hydro projects, life-cycle assessments have shown low overall net GHG emissions ( WEC, 2004a [NPR] ; UNESCO, 2006 [NPR] ). Since measuring the incremental anthropogenic-related emissions from freshwater reservoirs remains uncertain, the Executive Board of the UN Framework Convention on Climate Change (UNFCCC) has excluded large hydro projects with significant water storage from the CDM. The IPCC Guidelines for National GHG Inventories ( 2006 ) recommended using estimates for induced changes in the carbon stocks.

Whether or not large hydro systems bring benefits to the poorest has also been questioned ( Collier, 2006 [NPR] ; though this argument is not exclusive to hydro). The multiple benefits of hydro-electricity, including irrigation and water-supply resource creation, rapid response to grid-demand fluctuations due to peaks or intermittent renewables, recreational lakes and flood control, need to be taken into account for any given development. Several sustainability guidelines and an assessment protocol have been produced by the industry ( IHA, 2006 [NPR] ; Hydro Tasmania, 2005 [NPR] ; WCD, 2000 [NPR] ).

4.3.3.2 Wind

Wind provided around 0.5% of the total 17,408 TWh global electricity production in 2004 ( IEA, 2006b [NPR] ) but its technical potential greatly exceeds this ( WEC, 2004d [NPR] ; GWEC, 2006 [NPR] ). Installed capacity increased from 2.3 GW in 1991 to 59.3 GW at the end of 2005 when it generated 119 TWh at an average capacity factor of around 23%. New wind installation capacity has grown at an average of 28% per year since 2000, with a record 40% increase in 2005 ( BTM, 2006 [NPR, MoS] ) due to lower costs, greater government support through feed-in tariff and renewable energy certificate policies ( Section 4.5 ), and improved technology development. Total offshore wind capacity reached 679 MW at the end of 2005 ( BTM, 2006 [NPR, MoS] ), with the expectation that it will grow rapidly due to higher mean annual wind-speed conditions offsetting the higher costs and public resistance being less. Various best-practices guidelines have been produced and issues such as noise, electromagnetic (EMF) interference, airline flight paths, land-use, protection of areas with high landscape value, and bird and bat strike, are better understood but remain constraints. Most bird species exhibit an avoidance reaction to wind turbines, which reduces the probability of collision ( NERI, 2004 [NPR] ).

The average size of wind turbines has increased in the last 25 years from less than 50 kW in the early 1980 s to the largest commercially available in 2006 at around 5MW and having a rotor diameter of over 120 m. The average turbine size being sold in 2006 was around 1.6–2 MW but there is also a market for smaller turbines <100 kW. In Denmark, wind energy accounted for 18.5% of electricity generation in 2004, and 25% in West Denmark where 2.4 GW is installed, giving the highest generation per capita in the world.

Capital costs for land-based wind turbines can be below 900 US$/kW with 25% for the tower and 75% for the rotor and nacelle, although price increases have occurred due to supply shortages and increases in steel prices. Total costs of an onshore wind farm range from 1000 – 1400 US$/kW, depending on location, road access, proximity to load, etc. Operation and maintenance costs vary from 1% of investment costs in year one, rising to 4.5% after 15 years. This means that on good sites with low surface roughness and capacity factors exceeding 35%, power can be generated for around 30–50 US$/MWh ( IEA, 2006c [NPR] ; Morthorst, 2004 [NPR, MoS] ; Figure 4.12 ).

Figure 4.12: Development of wind-generation costs based on Danish experience since 1985 with variations shown due to land surface and terrain variations (as indicated by roughness indicator classes which equal 0 for open water and up to 3 for rugged terrain).

Source: Morthorst, 2004 [NPR, MoS] .

A global study of 7500 surface stations showed mean annual wind speeds at 80 m above ground exceeded 6.9 m/s with most potential found in Northern Europe along the North Sea, the southern tip of South America, Tasmania, the Great Lakes region, and the northeastern and western coasts of Canada and the US. A technical potential of 72 TW installed global capacity at 20% average capacity factor would generate 126,000 TWh/yr ( Archer and Jacobsen, 2005 [JoC] ). This is five times the assumed global production of electricity in 2030 ( IEA, 2006b [NPR] ) and double the 600 EJ potential capacity estimated by Johansson et al. 2004 [NPR] ) ( Table 4.2 ).

The main wind-energy investments have been in Europe, Japan, China, USA and India (Wind Force 12, 2005 ). The Global Wind Energy Council assumed this will change and has estimated more widespread installed capacity of 1250 GW by 2020 to supply 12% of the world’s electricity. The European Wind Energy Association set a target of 75 GW (168 TWh) for EU-15 countries in 2010 and 180 GW (425 TWh) in 2020 ( EWEA, 2004 [NPR] ). Several Australian and USA states have similar ambitious targets, mainly to meet the increasing demand for power rather than to displace nuclear or fossil-fuel plants. Rapid growth in several developing countries including China, Mexico, Brazil and India is expected since private investment interest is increasing ( Martinot et al., 2005 [NPR, SRC] ).

The fluctuating nature of the wind constrains the contribution to total electricity demand in order to maintain system reliability. To supply over 20% would require more accurate forecasting ( Giebel, 2005 [NPR, MoS] ), regulations that ensure wind has priority access to the grid, demand-side response measures, increases in the use of operational reserves in the power system ( Gul and Stenzel, 2005 [NPR] ) or development of energy storage systems ( EWEA, 2005 [NPR] ; Mazza and Hammerschlag, 2003 [NPR, MoS] ). The additional cost burden in Denmark to provide reliability was claimed to be between 1–1.5 billion € ( Bendtsen, 2003 [NPR] ) and 2–2.5 billion € per annum ( Krogsgaard, 2001 [NPR] ). However, the costs for back-up power decrease drastically with larger grid area, larger area containing distributed wind turbines and greater share of flexible hydro and natural-gas-fired power plants ( Morthorst, 2004 [NPR, MoS] ).

A trend to replace older and smaller wind turbines with larger, more efficient, quieter and more reliable designs gives higher power outputs from the same site often at a lower density of turbines per hectare. Costs vary widely with location ( Table 4.7 ). Sites with wind speeds of less than 7–8 m/s are not currently economically viable without some form of government support if conventional power-generation costs are above 50 US$/Wh (Oxera, 2005 ). A number of technologies are under development in order to maximize energy capture for lower wind-speed sites. These include: optimized turbine designs; larger turbines; taller towers; the use of carbon-fibre technology to replace glass-reinforced polymer in longer wind-turbine blades; maintenance strategies for offshore turbines to overcome difficulties with access during bad weather/rough seas; more accurate aero-elastic models and more advanced control strategies to keep the wind loads within the turbine design limits.

4.3.3.3 Biomass and bioenergy

Biomass continues to be the world’s major source of food, stock fodder and fibre as well as a renewable resource of hydrocarbons for use as a source of heat, electricity, liquid fuels and chemicals. Woody biomass and straw can be used as materials, which can be recycled for energy at the end of their life. Biomass sources include forest, agricultural and livestock residues, short-rotation forest plantations, dedicated herbaceous energy crops, the organic component of municipal solid waste (MSW), and other organic waste streams. These are used as feedstocks to produce energy carriers in the form of solid fuels (chips, pellets, briquettes, logs), liquid fuels (methanol, ethanol, butanol, biodiesel), gaseous fuels (synthesis gas, biogas, hydrogen), electricity and heat. Biomass resources and bioenergy use are discussed in several other chapters (Fig. 4.13) as outlined in Chapter 11 . This chapter 4 concentrates on the conversion technologies of biomass resources to provide bioenergy in the form of heat and electricity to the energy market.

Figure 4.13: Biomass supplies originate from a wide range of sources and, after conversion in many designs of plants from domestic to industrial scales, are converted to useful forms of bioenergy. 

Bioenergy carriers range from a simple firewood log to a highly refined gaseous fuel or liquid biofuel. Different biomass products suit different situations and specific objectives for using biomass are affected by the quantity, quality and cost of feedstock available, location of the consumers, type and value of energy services required, and the specific co-products or benefits IEA Bioenergy, 2005 [NPR] ). Prior to conversion, biomass feedstocks tend to have lower energy density per volume or mass compared with equivalent fossil fuels. This makes collection, transport, storage and handling more costly per unit of energy ( Sims, 2002 [NPR, SRC] ). These costs can be minimized if the biomass can be sourced from a location where it is already concentrated, such as wood-processing residues or sugar plant.

Globally, biomass currently provides around 46 EJ of bioenergy in the form of combustible biomass and wastes, liquid biofuels, renewable MSW, solid biomass/charcoal, and gaseous fuels. This share is estimated to be over 10% of global primary energy, but with over two thirds consumed in developing countries as traditional biomass for household use ( IEA, 2006b [NPR] ). Around 8.6 EJ/yr of modern biomass is used for heat and power generation ( Figure 4.14 ). Conversion is based on inefficient combustion, often combined with significant local and indoor air pollution and unsustainable use of biomass resources such as native vegetation ( Venkataraman et al., 2004 [JoC] ).

Figure 4.14: World biomass energy flows (EJ/yr) in 2004 and their thermochemical and biochemical conversion routes to produce heat, electricity and biofuels for use by the major sectors.

Note: much of the data is very uncertain, although a useful indication of biomass resource flows and bioenergy outputs still results.

Residues from industrialized farming, plantation forests and food- and fibre-processing operations that are currently collected worldwide and used in modern bioenergy conversion plants are difficult to quantify but probably supply approximately 6 EJ/yr. They can be classified as primary, secondary and tertiary ( Figure 4.15 ). Current combustion of over 130 Mt of MSW provides more than 1 EJ/yr though this includes plastics, etc. ( Chapter 10 ). Landfill gas also contributes to biomass supply at over 0.2 EJ/yr ( Chapter 10 ).

Figure 4.15: Biomass sources from land used for primary production can be processed for energy with residues available from primary, secondary and tertiary activities.

Source: van den Broek, 2000 [NotFound] .

A wide range of conversion technologies is under continuous development to produce bioenergy carriers for both small- and large-scale applications. Organic residues and wastes are often cost-effective feedstocks for bioenergy conversion plants, resulting in niche markets for forest, food processing and other industries. Industrial use of biomass in OECD countries was 5.6 EJ in 2002 ( IEA, 2004a [NPR] ), mainly in the form of black liquor in pulp mills, biogas in food processing plants, and bark, sawdust, rice husks etc. in process heat boilers.

The use of biomass, particularly sugarcane bagasse, for cogeneration (CHP) and industrial, domestic and district heating continues to expand ( Martinot et al., 2005 [NPR, SRC] ). Combustion for heat and steam generation remains state of the art, but advancing technologies include second-generation biofuels ( Chapter 5 ), biomass integrated-gasification combined-cycle (BIGCC), co-firing (with coal or gas), and pyrolysis. Many are close to commercial maturity but awaiting further technical breakthroughs and demonstrations to increase efficiency and further bring down costs.

Biochemical conversion using enzymes to convert ligno-cellulose to sugars that, in turn can be converted to bioethanol, biodiesel, di-methyl ester, hydrogen and chemical intermediates in biorefineries is not yet commercial. Biochemical- and Fischer-Tropsch-based thermochemical synthesis processes can be integrated in a single biorefinery such that the biomass carbohydrate fraction is converted to ethanol and the lignin-rich residue gasified and used to produce heat for process energy, electricity and/or fuels, thus greatly increasing the overall system efficiency to 70–80% ( OECD, 2004b [NPR] ; Sims, 2004 [NPR, MoS, SRC] ).

Combustion and co-firing

Biomass can be combined with fossil-fuel technologies by co-firing solid biomass particles with coal; mixing synthesis gas, landfill gas or biogas with natural gas prior to combustion. There has been rapid progress since the TAR in the development of the co-utilisation of biomass materials in coal-fired boiler plants. Worldwide more than 150 coal-fired power plants in the 50–700 MW_e range have operational experience of co-firing with woody biomass or wastes, at least on a trial basis ( IEA, 2004c [NPR] ). Commercially significant lignites, bituminous and sub-bituminous coals, anthracites and petroleum coke have all been co-fired up to 15% by energy content with a very wide range of biomass material, including herbaceous and woody materials, wet and dry agricultural residues and energy crops. This experience has shown how the technical risks associated with co-firing in different types of coal-fired power plants can be reduced to an acceptable level through proper selection of biomass type and co-firing technology. It is a relatively low-cost, low-risk method of adding biomass capacity, particularly in countries where coal-fired plants are prevalent.

Gaseous fuels

Gasification of biomass (or coal, Section 4.3.1.1 ) to synthesis (producer) gas, mainly CO and H₂, has a relatively high conversion efficiency (40–45%) when used to generate electricity through a gas engine or gas turbine. The gas produced can also be used as feedstock for a range of liquid biofuels. Development of efficient BIGCC systems is nearing commercial realization, but the challenges of gas clean-up remain. Several pilot and demonstration projects have been evaluated with varying degrees of success ( IEA, 2006d [NPR] ).

Recovery of methane from anaerobic digestion plants has increased since the TAR. More than 4500 installations (including landfill-gas recovery plants) in Europe, corresponding to 3.3 Mt methane or 92 PJ/yr, were operating in 2002 with a total market potential estimated to be 770 PJ (assuming 28 Mt methane will be produced) in 2020 (Jönsson, 2004 ). Biogas can be used to produce electricity and/or heat. It can also be fed into natural gas grids or distributed to filling stations for use in dedicated or dual gas-fuelled vehicles, although this requires biogas upgrading ( Section 10.4 ).

Costs and reduction opportunities

Costs vary widely for biomass fuel sources giving electricity costs commonly between 0.05 and 0.12 US$/kWh ( Martinot et al., 2005 [NPR, SRC] ) or even lower where the disposal cost of the biomass is avoided. Cost reductions can occur due to technical learning and capital/labour substitution. For example, capital investment costs for a high-pressure, direct-gasification combined-cycle plant up to 50 MW are estimated to fall from over 2000 US$/kW to around 1100 US$/kW by 2030, with operating costs, including delivered fuel supply, also declining to give possible generation costs down to 0.03 US$/kWh ( Martinot et al., 2005 [NPR, SRC] ; Specker, 2006 [NPR] ; EIA/DOE, 2006 ). Commercial small-scale options using steam turbines, Stirling engines, organic Rankin-cycle systems etc. can generate power for up to 0.12 US$/kWh, but with the opportunity to further reduce the capital costs by mass production and experience.

4.3.3.4 Geothermal

Geothermal resources from low-enthalpy fields located in sedimentary basins of geologically stable platforms have long been used for direct heat extraction for building and district heating, industrial processing, domestic water and space heating, leisure and balneotherapy applications. High-quality high-enthalpy fields (located in geodynamically active regions with high-temperature natural steam reached by drilling at depths less than 2 km) where temperatures are above 250ºC allow for direct electricity production using binary power plants (with low boiling-point transfer fluids and heat exchangers), organic Rankin-cycle systems or steam turbines. Plant capacity factors range from 40 to 95%, with some therefore suitable for base load ( WEC, 2004b [NPR] ). Useful heat and power produced globally is around 2 EJ/yr ( Table 4.2 ).

Fields of natural steam are rare. Most are a mixture of steam and hot water requiring single- or double-flash systems to separate out the hot water, which can then be used in binary plants or for direct use of the heat ( Martinot et al., 2005 [NPR, SRC] ). Binary systems have become state-of-the-art technologies but often with additional cost. Re-injection of the fluids maintains a constant pressure in the reservoir and hence increases the life of the field, as well as overcoming any concerns at environmental impacts. Sustainability concerns relating to land subsidence, heat-extraction rates exceeding natural replenishment ( Bromley and Currie, 2003 [NPR] ), chemical pollution of waterways (e.g. with arsenic), and associated CO₂ emissions have resulted in some geothermal power-plant permits being declined. This could be partly overcome by re-injection techniques. Deeper drilling up to 8 km to reach molten rock magma resources may become cost effective in future. Deeper drilling technology could also help to develop widely abundant hot dry rocks where water is injected into artificially fractured rocks and heat extracted as steam. Pilot schemes exist but tend not to be cost effective at this stage. In addition, the growth of ground-to-air heat pumps for heating buildings ( Chapter 6 ) is expected to increase.

Capital costs have declined by around 50% from the 3000–5000 US$/kW in the 1980 s for all plant types (with binary cycle plants being the more costly). Power-generation costs vary with high- and low-enthalpy fields, shallow or deep resource, size of field, resource-permit conditions, temperature of resource and the applications for any excess heat ( IEA, 2006d [NPR] ; Table 4.7 ). Operating costs increase if CO₂ emissions released either entail a carbon charge or require CCS.

Several advanced energy-conversion technologies are becoming available to enhance the use of geothermal heat, including combined-cycle for steam resources, trilateral cycles for binary total-flow resources, remote detection of hot zones during exploration, absorption/regeneration cycles (e.g., heat pumps) and improved power-generation technologies ( WEC, 2004c [NPR] ). Improvements in characterizing underground reservoirs, low-cost drilling techniques, more efficient conversion systems and utilization of deeper reservoirs are expected to improve the uptake of geothermal resources as will a decline in the market value for extractable co-products such as silica, zinc, manganese and lithium ( IEA, 2006d [NPR] ).

4.3.3.5 Solar thermal electric

The proportion of solar radiation that reaches the Earth’s surface is more than 10,000 times the current annual global energy consumption. Annual surface insolation varies with latitude, ranging between averages of 1000 W/m² in temperate regions and 1200 W/m² in low-latitude dry desert areas.

Concentrating solar power (CSP) plants are categorized according to whether the solar flux is concentrated by parabolic trough-shaped mirror reflectors (30–100 suns concentration), central tower receivers requiring numerous heliostats (500– 1000 suns), or parabolic dish-shaped reflectors ( 1000 –10,000 suns). The receivers transfer the solar heat to a working fluid, which, in turn, transfers it to a thermal power-conversion system based on Rankine, Brayton, combined or Stirling cycles. To give a secure and reliable supply with capacity factors at around 50% rising to 70% by 2020 ( US DOE, 2005 [NPR] ), solar intermittency problems can be overcome by using supplementary energy from associated natural gas, coal or bioenergy systems ( IEA, 2006 [Ambiguous] g) as well as by storing surplus heat.

Solar thermal power-generating plants are best sited at lower latitudes in areas receiving high levels of direct insolation. In these areas, 1 km² of land is enough to generate around 125 GWh/yr from a 50 MW plant at 10% conversion of solar energy to electricity ( Philibert, 2004 [NPR] ). Thus about 1% of the world’s desert areas (240,000 km²), if linked to demand centres by high-voltage DC cables, could, in theory, be sufficient to meet total global electricity demand as forecast out to 2030 ( Philibert, 2006 [NPR] ; IEA, 2006b [NPR] ). CSP could also be linked with desalination in these regions or used to produce hydrogen fuel or metals.

The most mature CSP technology is solar troughs with a maximum peak efficiency of 21% in terms of conversion of direct solar radiation into grid electricity. Tower technology has been successfully demonstrated by two 10 MW systems in the USA with commercial development giving long-term levelized energy costs similar to trough technology. Advanced technologies include troughs with direct steam generation, Fresnel collectors, which can reduce costs by 20%, energy storage including molten salt, integrated combined-cycle systems and advanced Stirling dishes. The latter are arousing renewed interest and could provide opportunities for further cost reductions ( WEC, 2004d [NPR] ; IEA 2004b [NPR] ).

Technical potential estimates for global CSP vary widely from 630 GW_e installed by 2040 ( Aringhoff et al., 2003 [NPR] ) to 4700 GW_e by 2030 ( IEA, 2003 [Ambiguous] h; Table 4.2 ). Installed capacity is 354 MW_e from nine plants in California ranging from 14 to 80 MW_e with over 2 million m² of parabolic troughs. Connected to the grid during 1984 – 1991, these generate around 400 GWh/yr at 100–126 US$/MWh ( WEC, 2004d [NPR] ). New projects totalling over 1400 MW are being constructed or planned in 11 countries including Spain (500 MW supported by a new feed-in tariff) ( ESTIA, 2004 [NPR] ; Martinot et al., 2005 [NPR, SRC] ) and Israel for the first of several 100 MW plants ( Sagie, 2005 [NPR] ). The African Development Bank has financed a 50 MW combined-cycle plant in Morocco that will generate 55 GWh/yr, and two new Stirling dish projects totalling 800 MW_e planned for the Mojave Desert, USA ( ISES, 2005 [NPR] ) are estimated to generate at below 90 US$/MWh Stirling,) are estimated to generate at below 90 US$/MWh Stirling,). Installed capacity of 21.5 GW_e, if reached by 2020, would produce 54.6 TWh/yr with a further possible increase leading towards 5% coverage of world electricity demand by 2040 .

4.3.3.6 Solar photovoltaic (PV)

Electricity generated directly by utilizing solar photons to create free electrons in a PV cell is estimated to have a technical potential of at least 450,000 TWh/yr (Renewables, 2004; WEC, 2004d [NPR] ). However, realizing this potential will be severely limited by land, energy-storage and investment constraints. Estimates of current global installed peak capacity vary widely, including 2400 MW (Greenpeace, 2004 ); 3100 MW ( (Maycock, 2003 ) ); >4000MW generating more than 21 TWh ( Martinot et al., 2005 [NPR, SRC] ) and 5000 MW (Greenpeace, 2006 ). Half the potential may be grid-connected, primarily in Germany, Japan and California, and grow at annual rates of 50–60% in contrast to more modest rates of 15–20% for off-grid PV. Expansion is taking place at around 30% per year in developing countries where around 20% of all new global PV capacity was installed in 2004, mainly in rural areas where grid electricity is either not available or unreliable ( WEC, 2004c [NPR] ). Decentralized generation by solar PV is already economically feasible for villages with long distances to a distribution grid and where providing basic lighting and radio is socially desirable. Annual PV module production grew from 740 MW in 2003 to 1700 MW in 2005, with new manufacturing plant capacity built to meet growing demand ( Martinot et al., 2005 [NPR, SRC] ). Japan is the world market leader, producing over half the present annual production ( IEA, 2003 [Ambiguous] f). However, solar generation remains at only 0.004% of total world power.

Most commercially available solar PV modules are based on crystalline silicon cells with monocrystalline at up to 18% efficiency, having 33.2% of the market share. Polycrystalline cells at up to 15% efficiency are cheaper per W_p (peak Watt) and have 56.3% market share. Modules costing 3–4 US$/W_p can be installed for around 6–7 US$/W_p from which electricity can be generated for around 250 US$/MWh in high sunshine regions (US Climate Change Technology Program, 2005 ). Cost reductions are expected to continue ( UNDP, 2000 [NPR] ; Figure 4.11 ), partly depending on the future world price for silicon; solar-cell efficiency improvements as a result of R&D investment; mass production of solar panels and learning through project experience. Costs in new buildings can be reduced where PV systems are designed to be an integral part of the roof, walls or even windows.

Thinner cell materials have prospects for cost reduction, including thin-film silicon cells (8.8% of market share in 2003 ), thin-film copper indium diselenide cells (0.7% of market share), photochemical cells and polymer cells. Commercial thin-film cells have efficiencies up to 8%, but 10–12% should be feasible within the next few years. Experimental multilayer cells have reached higher efficiencies but their cost remains high. Work to reduce the cost of manufacturing, using low-cost polymer materials, and developing new materials such as quantum dots and nano-structures, could allow the solar resource to be more fully exploited. Combining solar thermal and PV power-generation systems into one unit has good potential as using the heat produced from cooling the PV cells would make it more efficient ( (Bakker et al., 2005 ) ).

4.3.3.7 Solar heating and cooling

Solar heating and cooling of buildings can reduce conventional fuel consumption and reduce peak electricity loads. Buildings can be designed to use efficient solar collection for passive space heating and cooling ( Chapter 6 ), active heating of water and space using glazed and circulating fluid collectors, and active cooling using absorption chillers or desiccant regeneration (US Climate Change Technology Program, 2003 ). There is a risk of lower performance due to shading of windows or solar collectors by new building construction or nearby trees. Local ‘shading’ regulations can prevent such conflicts by identifying a protected ‘solar envelope’ ( Duncan, 2005 [NPR] ). A wide range of design measures, technologies and opportunities are covered by the IEA Solar Heating and Cooling implementing agreement (www.iea-shc.org).

Active systems of capturing solar energy for direct heat are used mainly in small-scale, low-temperature, domestic hot water installations; heating of building space; swimming pools; crop drying; cook stoves; industrial processes; desalination plants and solar-assisted district heating. The estimated annual global solar thermal-collector yield of domestic hot water systems alone is around 80 TWh (0.3 EJ) with the installations growing by 20% per year. Annual solar thermal energy use depends on the area of collectors in operation, the solar radiation levels available and the technologies used including both unglazed and glazed systems. Unglazed collectors, mainly used to heat swimming pools in the USA and Europe, represented about 28 million m² in 2003 .

More than 130 million m² of glazed collector area was installed worldwide by the end of 2003 to provide around 0.5 EJ of heat from around 91 GW_th capacity ( Weiss et al., 2005 [NPR] ). In 2005, around 125 million m² (88 GW_th) of active solar hot-water collectors existed, excluding swimming pool heating ( Martinot et al, 2005 [NPR, SRC] ). China is the world’s largest market for glazed domestic solar hot-water systems with 80% of annual global installations and existing capacity of 79 million m² (55 GW_th) at the end of 2005 . Most new installations in China are now evacuated-tube in contrast with Europe (the second-largest market), where most collectors are flat-plate ( (Zhang et al., 2005 ) ). Domestic solar hot-water systems are also expanding rapidly in other developing countries. Estimated annual energy yields for glazed flat-plate collectors range between 400 kWh/m² in Germany and 1000 kWh/m² in Israel ( IEA, 2004d [NPR] ). In Austria, annual solar yields were estimated to be 300 kWh/m² for unglazed, 350 kWh/m² for flat-plate, and 550 kWh/m² for evacuated tube collectors ( Weiss et al., 2005 [NPR] ). The retail price for a solar water heater unit for a family home differs with location and any government support schemes. Installed costs range from around 700 US$ in Greece for a thermo-siphon system with a 2.4 m² collector and 150 L tank, to 2300 US$ in Germany for a pumped system with antifreeze device. Systems manufactured in China are typically 200–300 US$ each.

Nearly 100 commercial solar cooling technologies exist in Europe, representing 24,000 m² with a cooling power of 9 MW_th. High potential energy savings compared with conventional electric vapour-pressure air-conditioning systems do not offset the higher costs ( Philibert and Podkanski, 2005 [NPR] ).

4.3.3.8 Ocean energy

The potential marine-energy resource of wind-driven waves, gravitational tidal ranges, thermal gradients between warm surface water and colder water at depths of > 1000 m, salinity gradients, and marine currents is huge (Renewables, 2004 ), but what is exploitable as the economic potential is low. All the related technologies (with the exception of three tidal-range barrages amounting to 260 MW, including la Rance that has generated 600 GWh/yr since 1967 ) are at an early stage of development with the only two commercial wave-power projects totalling 750 kW. To combat the harsh environment, installed costs are usually high. The marine-energy industry is now in a similar stage of development to the wind industry in the 1980 s (Carbon Trust, 2005 [NPR] ). Since oceans are used by a range of stakeholders, siting devices will involve considerable consultation.

The best wave-energy climates ( Figure 4.16 ) have deep-water power densities of 60–70 kW/m but fall to about 20 kW/m at the foreshore. Around 2% of the world’s 800,000 km of coastline exceeds 30 kW/m, giving a technical potential of around 500 GW assuming offshore wave-energy devices have 40% efficiency. The total economic potential is estimated to be well below this ( WEC, 2004d [NPR] ) with generating cost estimates around 80–110 US$/MWh highly uncertain, since no truly commercial scale plant exists ( IEA, 2006d [NPR] ).

Figure 4.16: Annual average wave-power density flux (kW/m at deep water)

Extracting electrical energy from marine currents could yield in excess of 10 TWh/yr (0.4 EJ/yr) if major estuaries with large tidal fluctuations could be tapped, but cost estimates range from 450– 1350 US$/MWh ( IEA, 2006a [NPR] ). A 1 km-stretch of permanent turbines built in the Agulhas current off the coast of South Africa, for example, could give 100 MW of power ( Nel, 2003 [NPR] ). However, environmental effects on tidal mud flats, wading birds, invertebrates etc. would need careful analysis. In order for these new technologies to enter the market, sustained government and public support is needed.

Ocean thermal and saline gradient energy-conversion systems remain in the research stage and it is still too early to estimate their technical potential. Initial applications have been for building air conditioning (www.makai.com/p-pipelines/) for desalination in open- and hybrid-cycle plants using surface condensers and in future could benefit tropical island nations where power is presently provided by expensive diesel generators.

4.3.4.1 Electricity

Electricity is the highest-value energy carrier because it is clean at the point of use and has so many end-use applications to enhance personal and economic productivity. It is effective as a source of motive power (motors), lighting, heating and cooling and as the prerequisite for electronics and computer systems. Electricity is growing faster as a share of energy end-uses ( Figure 4.18 ) than other direct-combustion uses of fuels with the result that electricity intensity (Electricity/GDP) has remained relatively constant even though the overall global-energy intensity (Energy/GDP) continues to decrease. If electricity intensity continues to decrease due to efficiency increases, future electricity demand could be lower than otherwise forecast (Sections 4.4.4 and 11.3.1 ).

Figure 4.18: Ratio of electricity to total primary energy in the US since 1900 .

Life-cycle GHG-emission analyses of power-generation plants ( WEC 2004a [NPR] ; Vattenfall, 2005; ( Dones et al., 2005; ) van de Vate, 2002 [NotFound] ; Spadaro, 2000 [NPR] ; Uchiyama and Yamamoto, 1995 [NPR, SRC] ;( Hondo, 2005 ) ) show the relatively high CO₂ emissions from fossil-fuel combustion are 10–20 times higher than the indirect emissions associated with the total energy requirements for plant construction and operation during the plant’s life ( Figure 4.19 ). Substitution by nuclear or renewable energy decreases carbon emissions per kWh by the difference between the full-energy-chain emission coefficients and allowing for varying plant-capacity factors ( WEC 2004a [NPR] ; Sims et al., 2003a [SRC] ). The average thermal efficiency for electricity-generation plants has improved from 30% in 1990 to 36% in 2002, thereby reducing GHG emissions.

Figure 4.19: GHG emissions for alternative electricity-generation systems.

Notes: 1 tCO₂ –eq/GWh = 0.27 tC –eq/GWh. The high estimate for hydro includes possible GHG emissions from reservoirs ( Section 4.3.3.1 )

Electricity generated from traditional coal-fired, steam-power plants is expected to be displaced over time with more advanced technologies such as CCGT or advanced coal to reduce the production of GHG and increase the overall efficiency of energy use. Previous IPCC 2001 [NPR] and WEC 2001 [NPR] ) scenarios suggested that nuclear, CCGT and CCS could become dominant electricity-sector technologies early this century ( Section 4.4 ). Although CCS can play a role, its potential may be limited and hence some consider it as a transitional bridging technology.

4.3.4.2 Heat and heat pumps

Heat, whether from fossil fuels or renewable energy, is a critical energy source for all economies. Its efficient use could play an important role in the development of transition and developing economies ( UN, 2004 [NPR] ; IEA, 2004e [NPR] ). It is used in industrial processes for food processing, petroleum refining, timber drying, pulp production, etc. ( Chapter 7 ), as well as in commercial and residential buildings for space heating, hot water and cooking ( Chapter 6 ). Many industries cogenerate both heat and electricity as an integral part of their production process ( Section 4.3.5 ; Chapter 7 ), in most cases being used on-site, but at times sold for other uses off-site such as district heating schemes.

Heating and cooling using renewable energy ( Section 4.3.3.7 ) can compete with fossil fuels ( IEA, 2006 [Ambiguous] f). In some instances, the best use of modern biomass will be co-firing with coal at blends up to 5–10% biomass or with natural gas.

Heat pumps can be used for simple air-to-air space heating, air-to-water heating, and for utilizing waste heat in domestic, commercial and industrial applications ( Chapter 6 ). Thermo-dynamically reverse Carnot-cycle heat pumps are more demand-side technologies but also linked with sustainable energy supply by concentrating low-grade solar heat in air and water. Their efficiency is evaluated by the coefficient of performance (COP), with COPs of 3 to 4 available commercially and over 6 using advanced turbo-refrigeration (www.mhi.co.jp/aircon/). A combination of CCGT with advanced heat-pump technology could reduce carbon emissions from supplying heat more than using a conventional gas-fired CHP plant of similar capacity.

4.3.4.3 Liquid and gaseous fuels

Coal, natural gas, petroleum and biomass can all be used to produce a variety of liquid fuels for transport, industrial processes, power generation and, in some regions of the world, domestic heating. These include petroleum products from crude oil or coal; methanol from coal or natural gas; ethanol and fatty acid esters (biodiesel) from biomass; liquefied natural gas; and synthetic diesel fuel and di-methyl ether from coal or biomass. Of these, crude oil is the most energy-efficient fuel to transport over long distances from source to refinery and then to distribute to product demand points. After petrol, diesel oil and other light and medium distillates are extracted at the refinery, the residues are used to produce bitumen and heavy fuel oil used as an energy source for industrial processes, oil-fired power plants and shipping.

Gaseous fuels provide a great deal of the heating requirements in the developed world and increased use can lead to lower GHG and air-pollution emissions.

Hydrogen

Realizing hydrogen as an energy carrier depends on low-cost, high-efficiency methods for production, transport and storage. Most commercial hydrogen production today is based on steam reforming of methane, but electrolysis of water (especially using carbon-free electricity from renewable or nuclear energy) or splitting water thermo-chemically may be viable approaches in the future. Electrolysis may be favoured by development of fuel cells that require a low level of impurities. Current costs of electrolysers are high but declining. Producing hydrogen from fossil fuels on a large scale will need integration of CCS if GHG emissions are to be avoided. A number of routes to produce hydrogen from solar energy are also technically feasible ( Figure 4.20 ).

Figure 4.20: Routes to hydrogen-energy carriers from solar-energy sources.

Source: EPRI, 2003 [NPR]

Hydrogen has potential as an energy-storage medium for electricity production or transport fuel when needed. The prospects for a future hydrogen economy will depend on developing competitively priced fuel cells for stationary applications or vehicles, but fuel cells are unlikely to become fully commercial for one or two decades. International cooperative programmes, such as the IEA Hydrogen Implementing Agreement ( IEA, 2005 [Ambiguous] f), and more recently the International Partnership for the Hydrogen Economy ( www.iphe.net ) aim to advance RD&D on hydrogen and fuel cells across the application spectrum ( IEA, 2003 [Ambiguous] g; EERE, 2005 [NPR] ).

Hydrogen fuel cells may eventually become commercially viable electricity generators, but because of current costs, complexity and state of development, they may only begin to penetrate the market later this century ( IEA, 2005 [Ambiguous] g). Ultimately, hydrogen fuel could be produced in association with CCS leading to low-emission transport fuels. Multi-fuel integrated-energy systems or ‘energyplexes’ ( Yamashita and Barreto, 2005 [Ambiguous] ) could co-produce electricity, hydrogen and liquid fuels with overall high-conversion efficiencies, low emissions and also facilitating CCS. FutureGen is a US initiative to build the world’s first integrated CCS and hydrogen-production research power plant ( US DOE, 2004 [NPR, MoS] ).

4.3.6.1 Costs

Cost estimates of the components of a CCS system vary widely depending on the base case and the wide range of source, transport and storage options ( Table 4.5 ). In most systems, the cost of capture (including compression) is the largest component, but this could be reduced by 20–30% over the next few decades using technologies still in the research phase as well as by upscaling and learning from experience ( IPCC, 2005 [NPR] ). The extra energy required is a further cost consideration. CO₂ storage is economically feasible under conditions specific to enhanced oil recovery (EOR), and in saline formations, avoiding carbon tax charges for offshore gas fields in Norway. Pipeline transport of CO₂ operates as a mature market technology ( IPCC, 2005 [NPR] ), costing 1–5 US$/tCO₂ per 100 km (high end for very large volumes) ( IEA, 2006a [NPR] ). Several thousand kilometres of pipelines already transport 40 Mt/yr of CO₂ to EOR projects. The costs of transport and storage of CO₂ could decrease slowly as technology matures further and the plant scale increases.

Table 4.5: Current cost ranges for the components of a CCS system applied to a given type of power plant or industrial source

4.3.7.1 Energy storage

Energy storage allows the energy-supply system to operate more or less independently from the energy-demand system. It addresses four major needs: utilizing energy supplies when short-term demand does not exist; responding to short-term fluctuations in demand (stationary or mobile); recovering wasted energy (e.g. braking in mobile applications), and meeting stationary transmission expansion requirements ( Testor et al., 2005 [NPR, SRC] ). Storage is of critical importance if variable low-carbon energy options such as wind and solar are to be better utilized, and if existing thermal or nuclear systems are to be optimized for peak performance in terms of efficiencies and thus emissions. Advanced energy-storage systems include mechanical (flywheels, pneumatic), electrochemical (advanced batteries, reversible fuel cells, hydrogen), purely electric or magnetic (super- and ultra-capacitors, superconducting magnetic storage), pumped-water (hydro) storage, thermal (heat) and compressed air. Adding any of these storage systems necessarily decreases the energy efficiency of the entire system ( WEC, 2004d [NPR] ). Overall, cycle efficiencies today range from 60% for pumped hydro to over 90% for flywheels and super-capacitors ( Testor et al., 2005 [NPR, SRC] ). Electric charge carriers such as vanadium redox batteries and capacitors are under evaluation but have low energy density and high cost. Cost and durability (cycle life) of the high-technology systems remains the big challenge, possibly to be met by more advanced materials and fabrication. Energy storage has a key role for small local systems where reliability is an important feature.

4.4.3.1 Plant efficiency and fuel switching

Reductions in CO₂ emissions can be gained by improving the efficiency of existing power generation plants by employing more advanced technologies using the same amount of fuel. For example, a 27% reduction in emissions (gCO₂/kWh) is possible by replacing a 35% efficient coal-fired steam turbine with a 48% efficient plant using advanced steam, pulverized-coal technology ( Table 4.9 ). Replacing a natural gas single-cycle turbine with a combined cycle (CCGT) of similar output capacity would help reduce CO₂ emissions per unit of output by around 36%.

Table 4.9: Reduction in CO₂ emission coefficient by fuel substitution and energy conversion efficiency in electricity generation.

Switching from coal to gas increases the efficiency of the power plant because of higher operating temperatures, and when used together with the more efficient combined-cycle results in even higher efficiencies ( IEA, 2006a [NPR] ). Emission savings (gCO₂-eq/kWh) were calculated before and after each substitution option (based on IPCC 1996 [Ambiguous] emission factors). The baseline scenario ( IEA, 2004a [NPR] ) assumed a 5% CO₂ reduction from fossil-fuel mix changes (coal to gas, oil to gas etc.) and a further 7% reduction in the Alternative Policy scenario from fuel switching in end uses (see Chapters 6 and 7 ). By 2030, natural gas CCGT plants displacing coal, new advanced steam coal plants displacing less-efficient designs, and the introduction of new coal IGCC plants to replace traditional steam plants could provide a potential between 0.5 and 1.4 GtCO₂ depending on the timing and sequence of economics and policy measures ( IEA, 2006a [NPR] ). IEA analysis also showed that up to 50 GW of stationary gas-fired fuel cells could be operating by 2030, growing to around 3% of all power generation capacity by 2050 and giving about 0.5 Gt CO₂ emissions reduction ( IEA, 2006 [Ambiguous] j). This potential is uncertain, however, as it relies on appropriate fuel-cell development and is not included here.

By 2030, a proportion of old heat and power plants will have been replaced with more modern plants having higher energy efficiencies. New plants will also have been built to meet the growing world demand. It is assumed that after 2010 only the most efficient plant designs available will be built, though this is unlikely and will therefore increase future CO₂ emissions above the potential reductions. The coal that could be displaced by gas and the additional gas power generation required is assessed by region ( Table 4.10 ). A plant life time of 50 years; a 2%-per-year replacement rate in all regions starting in 2010; 20% of existing coal plants replaced by 2030 and 50% of all new-build thermal plants fuelled by gas, are among the most relevant assumptions. The cost of fuel switching partly depends on the difference between coal and gas prices. For example if mitigation costs below 20 US$/tCO₂-eq avoided, this would imply a relatively small price gap between coal and gas, although since fuel switching to a significant degree would affect natural gas prices, actual future costs are difficult to estimate with accuracy. Generation costs are assumed to be 40–55 US$/MWh for coal-fired and 40–60 US$/MWh for gas-fired power plants.

Table 4.10: Potential GHG emission reductions by 2030 from coal-to-gas fuel switching and improved efficiency of existing plant.

4.4.3.2 Nuclear

Proposed and existing fossil fuel power plants could be partly replaced by nuclear power plants to provide electricity and heat. Since the nuclear plant and fuel system consumes only small quantities of fossil fuels in the fuel cycle, net CO₂ emissions could be lowered significantly. Assessments of future potential for nuclear power are uncertain and controversial. The 2006 WEO Alternative scenario ( IEA, 2006b [NPR] ) anticipated a 50% increase in nuclear energy (to 4106 TWh/yr) by 2030 . The ETP report ( IEA, 2006a [NPR] ) assumed a mitigation potential of 0.4–1.3 GtCO₂ by 2030 from the construction of Generation II, III, III+ and IV nuclear plants ( Section 4.3.2 ). From a review of the literature and the various scenario projections described above (for example Figure 4.25 ), it is assumed that by 2030 18% of total global power-generation capacity could come from existing nuclear power plants as well as new plants displacing proposed new coal, gas and oil plants in proportion to their current share of the baseline ( Table 4.11 ). The rate of build required is possible (given the nuclear industry’s track record for building reactors in the 1970 s) and generating costs of 25–75 US$/MWh are assumed ( Section 4.4.2 ). However, there is still some controversy regarding the relatively low costs shown by comparative life-cycle analysis assessments reported in the literature ( Section 4.4.2 ) and used here.

Table 4.11: Potential GHG emission reduction and cost ranges in 2030 from nuclear-fission displacing fossil-fuel power plants.

4.4.3.3 Renewable energy

Fossil fuels can be partly replaced by renewable energy sources to provide heat (from biomass, geothermal or solar) or electricity (from wind, solar, hydro, geothermal and bioenergy generation) or by CHP plants. Ocean energy is immature and assumed unlikely to make a significant contribution to overall power needs by 2030 . Net GHG emissions avoided are used in the analysis since most renewable energy systems emit small amounts of GHG from the fossil fuels used for manufacturing, transport, installation and from any cement or steel used in their construction. Overall, net GHG emissions are generally low for renewable energy systems ( Figure 4.19 ) with the possible exception of some biofuels for transport, where fossil fuels are used to grow the crop and process the biofuel.

Hydro

The ETP ( IEA, 2006a [NPR] ) stated the technical potential of hydropower to be 14,000 TWh/yr, of which around 6000 TWh/yr (56 EJ) could be realistic to develop ( IHA, 2006 [NPR] ). The WEO Alternative scenario ( IEA, 2006b [NPR] ) assumed an increased share for hydro generation above baseline, reaching 4903 TWh/yr by 2030 ^[2] IEA 2006a [NPR] ) suggested hydropower (both small and large) could offset fossil-fuel power plants to give a mitigation potential between 0.3–1.0 GtCO₂/yr by 2030 . Here it is assumed that enough existing and new sites will be available to contribute around 5500 TWh/yr (17% of total electricity generation) by 2030 as a result of displacing coal, gas and oil plants based on their current share of the base load ( Table 4.12 ). Future costs range from 30–70 US$/MWh for good sites with high hydrostatic heads, close proximity to load demand, and with good all-year-round flow rates. Smaller plants and those installed in less-favourable terrain at a distance from load could cost more. GHG emissions from construction of hydro dams and possible release of methane from resulting reservoirs ( Section 4.3.3.1 ) are uncertain and not included here.

Table 4.12: Potential GHG emission reduction and cost ranges in 2030 as a result of hydro power displacing fossil-fuel thermal power plants.

Wind

The 2006 WEO Reference scenario baseline ( IEA, 2006b [NPR] ) assumed 1132 TWh/yr (3.3% of total global electricity) of wind generation in 2030 rising to a 4.8% share in the Alternative Policy scenario. However, wind industry ‘advanced’ scenarios are more optimistic, forecasting up to a 29.1% share for wind by 2030 with a mitigation potential of 3.1 GtCO₂/yr ( GWEC, 2006 [NPR] ). The ETP mitigation potential assessment ( IEA, 2006a [NPR] ) for on- and offshore wind power by 2030 ranged between 0.3 and 1.0 GtCO₂/yr. In this analysis on- and offshore wind power is assumed to reach a 7% share by 2030, mainly in OECD countries, and to displace new and existing fossil-fuel power plants according to the relevant shares of coal, oil and gas in the baseline for each region ( Table 4.13 ). Intermittency issues on most grids would not be limiting at these low levels given suitable control and back-up systems in place. The costs are very site specific and range from 30 US$/MWh on good sites to 80 US$/MWh on poorer sites that would also need to be developed if this 7% share of the total mix is to be met.

Table 4.13: Potential GHG emission reduction and costs in 2030 from wind power displacing fossil-fuel thermal power plants.

Bioenergy (excluding biofuels for transport)

Large global resources of biomass could exist by 2030 (Chapters 8 , 9 and 10 ), but confidence in estimating the bioenergy heat and power potential is low since there will be competition for these feedstocks for biomaterials, chemicals and biofuels. Bioenergy in its various forms (landfill gas, combined heat and power, biogas, direct combustion for heat etc.) presently contributes 2.6% to the OECD power mix, 0.4% to EIT and 1.5% to non-OECD. The WEO 2006 ( IEA, 2006b [NPR] ) assumed 805 TWh of biomass power generation in 2030 rising 22% to 983 TWh under the Alternative scenario to then give 3% of total electricity generation. The ETP gave a bioenergy potential ranging between 0.1 and 0.3 GtCO₂ /yr by 2030 . The baseline ( IEA, 2004a [NPR] ) assumed biomass and waste for heat and power generation will rise from 2% of primary fuel use (3.2 EJ) in 2002 to 4% (10.8 EJ) by 2030 .

Heat and CHP estimates are wide ranging so cannot be included in this analysis, even though the bioenergy potential could be significant. However, any heat previously utilized from displaced coal and gas CHP plants could easily be supplied from biomass, with more biomass available for use in stand-alone heat plants ( Chapter 11 ). In this analysis, a 5% share in OECD regions is assumed feasible, relying on co-firing in existing and new coal plants and with 7–8% of the total replacement capacity built being bioenergy plants. In EIT regions, the available forest biomass could be utilized to gain 5% share and in non-OECD regions, where there are less stock turnover issues than in the OECD, 10% of power could come from new bioenergy plants ( Table 4.14 ). A total potential by 2030 of 5% is assumed based on costs of 30–100 US$/MWh. The biomass feedstock required to meet these potentials, assuming thermal-conversion efficiencies of 20–30%, would be around 9–13 EJ in OECD, 1–3 EJ in EIT, and 18–27 EJ in non-OECD regions. Little additional bioenergy capacity above that already assumed in the baseline is anticipated in EIT regions where only a small contribution is expected compared with developing countries. Small inputs of fossil fuels are often used to produce, transport and convert the biomass ( IEA, 2006 [Ambiguous] h), but the same is true when using the fossil fuels it replaces. Since both are of a similar order of magnitude, and these emissions are already accounted for in the overall total for fossil fuels, bioenergy is credited with zero emissions (in compliance with IPCC guidelines).

Table 4.14: Potential emissions reduction and cost ranges in 2030 from bioenergy displacing fossil-fuel thermal power plants.

Geothermal

The installed geothermal-generation capacity of over 8.9 GW_e in 24 countries produced 56.8 TWh (0.3%) of global electricity in 2004 and is growing at around 20%/yr ( Bertani, 2005 [NPR] ) with the baseline giving 0.05% of total generation by 2030 . IEA WEO 2006 ( IEA, 2006b [NPR] ) assumed 174 TWh/yr by 2030 rising 6% to 185 TWh under the Alternative scenario. The ETP ( IEA, 2006a [NPR] ) gave a potential of 0.1–0.3 GtCO₂-eq/yr by 2030 .

In this analysis, generation costs of 30–80 US$/MWh are assumed to provide a 2% share of the total 2030 energy mix. Direct heat applications are not included. Although CO₂ emissions are assumed to be zero, as for other renewables, this may not always be the case depending on underground CO₂ released during the heat extraction.

Solar

Concentrating solar power (CSP) and photovoltaics (PV) can theoretically gain a maximum 1–2% share of the global electricity mix by 2030 even at high costs. The 2006 WEO Reference scenario ( IEA, 2006b [NPR] ) estimated 142 TWh/yr of PV generation in 2030 rising to 237 TWh in the Alternative scenario but still at <1% of total generation. EPRI 2003 [NPR] ) assessed total PV capacity to be 205 GW by 2020 generating 282 TWh/yr or about 1% of global electricity demand. Other analyses range from over 20% of global electricity generation by 2040 ( Jäger-Waldau, 2003 [NPR] ) to 0.008% by 2030 with mitigation potential for both PV and CSP likely to be <0.1 GtCO₂ in 2030 ( IEA, 2006a [NPR] ) The calculated minimum costs for even the best sites resulted in relatively high costs per tonne CO₂ avoided ( Table 4.16 ). The baseline ( IEA, 2004a [NPR] ) gave the total solar potential as 466 TWh or 1.4% of total generation in 2030 .

In this analysis, generating costs from CSP plants could fall sufficiently to compete at around 50–180 US$/MWh by 2030 ( Trieb, 2005 [NPR] ; IEA, 2006a [NPR] ). PV installed costs could decline to around 60–250 US$/MWh, the wide range being due to the various technologies being installed on buildings at numerous sites, some with lower solar irradiation levels. Penetration into OECD and EIT markets is assumed to remain small with more support for developing country electrification.

Table 4.15: Potential emissions reduction and cost ranges in 2030 from geothermal displacing fossil-fuel thermal power plants.

Table 4.16: Potential emission reduction and cost ranges in 2030 from solar PV and CSP displacing fossil-fuel thermal power plants.

4.4.3.4 Carbon dioxide capture and storage

In the absence of explicit policies, CCS is unlikely to be deployed on a large scale by 2030 ( IPCC, 2005 [NPR] ). The total CO₂ storage potential for each region ( Hendriks et al. 2004 [NPR] ; Table 4.17 ) appears to be sufficient for storage over the next few decades, although capacity assessments are still under debate ( IPCC, 2005 [NPR] ). The proximity of a CCS plant to a storage site affects the cost, but this level of analysis was not considered here. CCS does not appear in the baseline ( IEA, 2004a [NPR] ). Penetration by 2030 is uncertain as it depends both on the carbon price and the rate of technological advances in costs and performance.

Coal CCS

ABARE ( Fisher, 2006 [NPR, ARC] ) suggested that worldwide by 2030, 1811 TWh/yr would be generated from coal with CCS (17 EJ); 7871 TWh (73 EJ) from coal without; 1492 TWh (14 EJ) from gas with; and 6315 TWh (59 EJ) from gas without. CCS would thus result in around 4.4 GtCO₂ of GHG emissions avoided in 2030 giving a 17% reduction from the reference base case level ( Figure 4.25 ). In contrast, the ETP mitigation assessments for CCS with coal plants ranged between only 0.3 and 1.0 GtCO₂ in 2030 ( IEA, 2006a [NPR] ), given that commercial-scale CCS demonstration will be needed before widespread deployment.

In this analysis, CCS is assumed to begin only after 2015 in OECD countries and after 2020 elsewhere, linked mainly with advanced steam coal plants installed with flue gas separation, although these IGCC plants and oxyfuel systems are only just entering the market Dow( Jones, 2006 ) ). Assuming a 50-year life of coal plants ( IEA, 2006a [NPR] ) and that 30% of new coal plants built in OECD and 20% elsewhere will be equipped with CCS, then the replacement rate of old plants by new designs with CCS incorporated is 0.6% per year in OECD and 0.4% elsewhere. Then 9% of total new and existing coal-fired plants will have CCS by 2030 in the OECD region and 4% elsewhere. Assuming 90% of the CO₂ can be captured and a reduced fuel-to-electricity conversion efficiency of 30% (leading to less power available for sale – IPCC, 2005 [NPR] ), then the additional overall costs range between 20 and 30 US$/MWh depending on the ease of CO₂ transport and storage specific to each plant ( Table 4.17 ).

Table 4.17: Potential emissions reduction and cost ranges in 2030 from CCS used with coal-fired power plants.

Gas CCS

The assumed life of a CCGT plant is 40 years, and with 20% of new gas-fired plants utilizing CCS starting in 2015 in OECD countries and 2020 elsewhere, then the replacement rate of old plants by new designs integrating CCS is 0.5% per year. By 2030 7% of all OECD gas plants will have CCS and 5% elsewhere. Assuming 90% of the CO₂ is captured, a reduction of gas-fired power plant conversion efficiency of 15% ( IPCC, 2005 [NPR] ), and an additional overall cost ranging between 20 and 30 US$/MWh generated, then the costs and potentials by 2030 (compared with the IEA 2004a [NPR] ) baseline of no CCS) are assessed ( Table 4.18 ). The costs for both coal and gas CCS compare well with the IPCC 2005 [NPR] ) range of 15–75 US$/tCO₂ ( Table 4.5 ).

Table 4.18: Potential emissions reduction and cost ranges in 2030 from CCS used with gas-fired power plants.

4.4.3.5 Summary

The cost ranges (US$/tCO₂-eq avoided) for each of the technologies analysed in Section 4.4.3 are compared ( Table 4.19 ). The percentage share of the total potential is shown spread across the defined cost class ranges for each region and technology. This assumes that a linear relationship exists between the lowest and highest costs as presented in Section 4.4.3 for each technology and region.

Since each technology is assumed to be promoted individually and crowding-out by other technologies under real-world constraints is ignored, the potentials in Table 4.19 are independent and cannot be added together.

4.4.4.1 Mitigation potentials of the electricity supply sector

Based on the method described above and the results from the analysis ( Table 4.20 ), the following conclusions can be drawn.

With reference to the baseline:

a) power plants existing in 2010 that remain in operation by 2030 ( Table 4.20 ), including coal, oil and gas-fired, continue to generate 12,111 TWh/yr in 2030 (38% of the total power demand) and produce 5.77 GtCO₂-eq/yr of emissions;

b) new additional plants to be built over the 20-year period from 2010 generate 11,471 TWh/yr by 2030 and new plants built to replace old plants generate 8074 TWh/yr;

c) the share of all new build plants burning coal, oil and gas produce around 10 GtCO₂-eq/yr by 2030, thereby giving total baseline emissions of 15.77 GtCO₂-eq/yr ( Table 4.8 ).

For costs < 20 US$/tCO₂-eq avoided:

a) The baseline generation from fossil fuel-fired plants in 2030 of 22,602 TWh (including 14,618 TWh from new generation) reduces by 22.5% to 17,525 TWh (including 9541 TWh of new build generation) due to the increased uptake of low- and zero-carbon technologies. This is a reduction from the 71% of total generation in the baseline to 55%.

b) Of this total, fuel switching from coal to gas results in additional gas-fired plants generating 1,495 TWh/yr by 2030, mainly in non-OECD/EIT countries, and thereby avoiding 0.67 GtCO₂-eq/yr of emissions.

c) Renewable energy generation increases from the 2030 baseline of 6126 TWh/yr to 7904 TWh/yr (6122 TWh/yr from new generation plus 2336 TWh/yr remaining in operation from 2010 ). The share of generation increases from 19.4% in 2010 to 26.7% by 2030 .

d) The nuclear power baseline of 2929 TWh/yr by 2030 (9.3% of total generation) increases to 5673 TWh/yr (17.9% of generation), of which 3882 TWh/yr is from newly built plants.

e) Overall, GHG emissions are reduced by 3.95 GtCO₂-eq giving 25.0% lower emissions than in the baseline. Around half of this potential occurs in non-OECD/EIT countries with OECD countries providing most of the remainder.

f) Should just 70% of the individual power-generation shares assumed above for all the mitigation technologies be achieved by 2030, the mitigation potential would reduce to 1.69 GtCO₂-eq.

g) This range is in reasonable agreement with the TAR analysis potential of 1.3 to 2.5 GtCO₂-eq/yr for less than 27 US$/tCO₂-eq avoided ( IPCC, 2001 [NPR] ), because this potential was only out to 2020, the baseline has since been adjusted, and since the TAR was published there has been increased acceptance for improved designs of nuclear power plants, an increase in development and deployment of renewable energy technologies and a better understanding of CCS technologies.

For costs <50 US$/tCO₂-eq avoided:

a) Fossil-fuel generation reduces further to 13,308 TWh/yr (of which 5324 TWh/yr is from new build plants) and accounts for 42% of total generation.

b) Renewable-energy generation increases to 10,673 TWh/yr by 2030 giving a 33.7% share of total generation. Solar PV and CSP are more costly ( Table 4.19 ) so they can only offer substitution for fossil fuels above 50 US$/tCO₂-eq avoided.

c) Nuclear power share of total generation remains similar since other technologies now compete.

d) CCS now becomes competitive and 2003 TWh/yr is generated by coal and gas-fired plants with CCS systems installed.

e) Overall GHG emissions in 2030 are now reduced by 6.44 GtCO₂-eq/yr below the baseline, although if only 70% of the assumed shares of total power generation for all the mitigation technologies are reached by 2030, the potential declines to 3.05 GtCO₂-eq. Non-OECD/EIT countries continue to provide half of the mitigation potential.

For costs <100 US$/tCO₂-eq avoided:

a) As more low- and zero-carbon technologies become competitive, fossil-fuel generation without CCS further reduces to 11,824 TWh in 2030 and is now only 37% of total generation.

b) New renewable energy generation increases to 8481 TWh/yr by 2030, which together with the plants remaining in operation from 2010, gives a 34% share of total generation.

c) Nuclear power provides 3574 TWh or 17% of total generation.

d) Coal- and gas-fired plants with CCS account for 3650 TWh/yr by 2030 or 12% of total generation.

e) The overall mitigation potential of the electricity sector is 7.22 GtCO₂-eq/yr which is a reduction of around 45% of GHGs below the baseline. If only 70% of the assumed shares of power generation by all low- and zero-emission technologies are achieved, then the potential would be around 45% lower at 3.97 GtCO₂-eq. Non-OECD/EIT countries contribute over half the total potential.

No single technological option has sufficient mitigation potential to meet the economic potential of the electricity-generation sector. To achieve these potentials by 2030, the relatively high investment costs, the difficulties in rapidly building sufficient capacity and expertise, and the threats resulting from introducing new low-carbon technologies as perceived by the incumbents in the existing markets, will all need to be addressed.

This analysis concentrates on the individual mitigation potentials for each technology at the high end of the wide range found in the literature ( Figure 4.29 b; IEA, 2006a [NPR] ; IEA, 2006b [NPR] ). This serves to illustrate that significant reductions in emissions from the energy-supply sector are technically and economically feasible using both the range of technology solutions currently available and those close to market. Reducing the individual assumed shares of the technologies in the 2030 power generation mix by 30% gives less ambitious potentials that are closer to the lower end of the ranges found in the literature ( Figure 4.29 a). Energy-efficiency savings of electricity use in the buildings ( Chapter 6 ) and industry ( Chapter 7 ) sectors will reduce these total emissions potentials ( Section 11.3.1 ).

Table 4.20: Projected power demand increase from 2010 to 2030 as met by new, more efficient additional and replacement plants that will displace 60% of existing plants at the end of their life. The potential mitigation above the baseline of GHG avoided for <20 US$/t, <50 US$/t and <100 US$/tCO_2-eq results from fuel switching from coal to gas, a portion of fossil-fuel generation being displaced by nuclear, renewable energy and bioenergy in each region and CCS.

Figure 4.29: Indicative low(a) and high(b) range estimates of the mitigation potential in the electricity sector based on substitution of existing fossil-fuel thermal power stations with nuclear and renewable energy power generation, coupled with energy-efficiency improvements in power-generation plants and transmission, including switching from coal to gas and the uptake of CCS. CHP and heat are not included, nor electricity savings from energy-efficiency measures in the building and industry sectors.

Source: Based on IEA, 2006a [NPR] ; IEA, 2006b [NPR] .

4.4.4.2 Uncertainties

The wide range of energy supply-related potentials in the literature is due to the many uncertainties and assumptions involved. This analysis of the costs and mitigation potential for energy-supply technologies through to 2030 involved the following degrees of confidence.

• There is high agreement on the energy types and amounts of current global and regional energy sources used in the baseline (with the exception of traditional biomass, for which data are uncertain) because the several sources of those estimates are in close agreement.
• There is high agreement that energy supply will grow between now and 2030 with medium confidence in projections of the total energy demand by 2030. Most assumptions about population and energy use in various scenarios do not diverge greatly until after 2030, although past experience suggests that projections, even over a 25-year period, can be erroneous.
• Estimates of specific potentials out to 2030 for electricity-supply technologies based on specific studies have only low agreement that a single value can be estimated accurately. However, there is medium confidence that the true potential of a mixture of supply technologies lies somewhere within the range estimated.
• The actual distribution of new technologies in 2030 can be estimated with medium confidence by using trend analyses, technology assessments, economic models and other techniques, but cannot take into account changing national policies and preferences, future carbon-price factors, and the unanticipated evolution of technologies or their cost. Current rates of adoption for particular technologies have been identified but there is low to medium agreement that these rates may continue until 2030.
• Despite the methodological limitations, the future costs and technical potentials identified provide a medium confidence for considering strategies and decisions over the next several decades. The analysis falls within the range of other projections for specific technologies.

4.4.4.3 Transport biofuels

Assessments for the uptake of biofuels range between 20 and 25% of global transport road fuels by 2050 and beyond ( Chapter 5 ). The 2006 WEO ( IEA, 2006b [NPR] ) Reference scenario predicted biofuels will supply 4% of road fuels by 2030 with greater potential up to 7% under the Alternative Policy scenario. To achieve double this penetration, as envisaged under the Beyond Alternative Policy scenario, would avoid around 0.5GtCO₂/ yr, but is likely to require large-scale introduction of second-generation biofuels from ligno-cellulosic conversions. Based on ETP assumptions ( IEA, 2006a [NPR] ), the mitigation potential of biofuels by 2030 is likely to be less than from vehicle efficiency improvements ( Chapter 5 ; Figure 4.30 ).

Figure 4.30: Potential increased emissions from the greater uptake of unconventional oils by 2030 could offset potential reductions from both biofuels and vehicle-efficiency improvements, but will be subject to the future availability and price of conventional oil.

Source: Based on IEA, 2006b [NPR] . 

Transport emissions of 6.7 GtCO₂ in 2002 will increase under business as usual to 11.6 GtCO₂ by 2030, but could be reduced by efficiency improvements together with the increased uptake of biofuels to emit between 7.1 and 9.4 GtCO₂ ( IEA, 2006a [NPR] ). This mitigation potential of between 2.2 and 4.5 GtCO₂, however, could be partially offset by the increased uptake of unconventional liquid fuels ( Section 4.3.1.4 ). Their potential is uncertain as, being more costly per litre to produce, they will be dependent partly on the future oil price and level of reserves. Overall then, the emissions from transport fuels up to 2030 will probably continue to rise ( Chapter 5 ).

4.4.4.4 Heating and cooling

The wide range of fuels and applications used for temperature modifications and the poor data base of existing heat and refrigeration plants makes the mitigation potential for heating and cooling difficult to assess. IEA 2006a [NPR] ) calculated the mitigation potential by 2030 for buildings ( Chapter 6 ) of up to 2.6 GtCO₂/yr, including 0.1-0.3 GtCO₂/yr for solar systems, and up to 0.6 GtCO₂/yr for industry ( Chapter 7 ). The mitigation potentials of CHP and trigeneration (heating, cooling and power generation) have not been assessed here.

4.5.1.1 Emission-reduction policies for energy supply

Subsidies, incentives and market mechanisms presently used to promote fossil fuels, nuclear power and renewables may need some redirection to achieve more rapid decarbonization of the energy supply.

Subsidies and other incentives

The effects of various policies and subsidies that support fossil-fuel use have been reviewed ( IEA, 2001 [NPR] ; OECD, 2002b [NPR] ; Saunders and Schneider, 2000 [NPR] ). Government subsidies in the global energy sector are in the order of 250–300 billion US$/yr, of which around 2–3% supports renewable energy ( (de Moor, 2001; ) UNDP 2004a [NPR] ). An OECD study showed that global CO₂ emissions could be reduced by more than 6% and real income increased by 0.1% by 2010 if support mechanisms on fossil fuels used by industry and the power-generation sector were removed ( OECD, 2002b [NPR] ). However, subsidies are difficult to remove and reforms would need to be conducted in a gradual and programmed fashion to soften any financial hardship.

For both environmental and energy-security reasons, many industrialized countries have introduced, and later increased, grant support schemes for producing electricity, heat and transport fuels based on nuclear or renewable energy resources and on installing more energy-efficient power-generation plant. For example, the US has recently introduced federal loan guarantees that could cover up to 80% of the project costs, production tax credits worth 6 billion US$, and 2 billion US$ of risk coverage for investments in new nuclear plants (Energy Policy Act, 2005 ). To comply with the 2003 renewable energy directive, all European countries have installed feed-in tariffs or tradable permit schemes for renewable electricity ( EEA, 2004 [NPR] ; EU, 2003 [NPR] ). Several developing countries including China, Brazil, India and a number of others have adopted similar policies.

Quantitative targets

Setting goals and quantitative targets for low-carbon energy at both national and regional levels increases the size of the markets and provides greater policy stability for project developers. For example, EU-15 members agreed on targets to increase their share of renewable primary energy to 12% of total energy by 2010 including electricity to 22% and biofuels to 5.75% ( (EU, 2001; ) EU 2003 [NPR] ). The Latin American and Caribbean Initiative, signed in May 2002 included a target of 10% renewable energy by 2010 ( Goldemberg, 2004 [NPR] ). The South African Government mandated an additional 10 TWh renewable energy contribution by 2013 (being 4% of final energy consumption) to the existing contribution of 115 TWh/yr mainly from fuel wood and waste ( DME, 2003 [NPR] ). Many other countries outlined similar targets at the major renewable energy conference in Bonn (Renewables, 2004 ) attended by 154 governments, but not to the extent that emissions will be reduced below business as usual.

Feed-in tariffs/Quota obligations

Quota obligations with tradable permits for renewable energy and feed-in tariffs have been used in many countries to accelerate the transition to renewable energy systems ( Martinot, 2005 [NPR, SRC] ). Both policies essentially serve different purposes, but they both help promote renewable energy ( (Lauber, 2004 ) ). Price-based, feed-in tariffs (providing long price certainty for renewable energy producers) have been compared with quantity-based instruments, including quotas, green certificates and competitive bidding ( Sawin, 2003a [NPR] ;( Menanteau et al., 2003; ) ( Lauber, 2004 ) ). The total level of support provided for preferential power tariffs in EU-15, in particular Germany, Italy and Spain, exceeded 1 billion € in 2001 ( EEA, 2004 [NPR] ).

Experience confirms that incentives to support ‘green power’ by rewarding performance are preferable to a capital investment grant, because they encourage market deployment while also promoting increases in production efficiency ( Neuhoff, 2004 [NPR] ). In terms of installed renewable energy capacity, better results have been obtained with price-based than with quantity-based approaches ( EC, 2005 [NPR] ; Ragwitz et al., 2005 [NPR] ; Fouquet et al., 2005 [NPR] ). In theory, this difference should not exist as bidding prices that are set at the same level as feed-in tariffs should logically give rise to comparable capacities being installed. The discrepancy can be explained by the higher certainty of current feed-in tariff schemes and the stronger incentive effect of guaranteed prices.

The potential advantages offered by green certificate trading systems based on fixed quotas are encouraging a number of countries and states to introduce such schemes to meet renewable energy goals in an economically efficient way. Such systems can encourage more precise control over quotas, create competition among producers and provide incentives to lower costs ( (Menanteau et al., 2003 ) ). Quota-obligation systems are only beginning to have an effect on capacity additions, in part because they are still new. However, about 75% of the wind capacity installed in the US between 1998 and 2004 occurred in states with renewable energy standards. Experience shows that if certificates are delivered under long-term agreements, effectiveness and compliance can be high ( Linden et al., 2005 [NPR] ; UCS, 2005 [NPR] ).

Tradable permit systems and CDM

In recent years, domestic and international tradable emission permit systems have received recognition as a means of lowering the costs of meeting climate-change targets. Creating carbon markets can help economies identify and realize economic ways to reduce GHG emissions and other energy-related pollutants, or to improve efficiency of energy use. The cost of achieving the Kyoto Protocol targets in OECD regions could fall from 0.2% of GDP without trading to 0.1% ( Newman et al., 2002 [NPR] ) as a result of introducing emission trading in an international regime. Emission trading, such as the European and CDM schemes, is designed to result in immediate GHG reductions, but CDM also has long-term aspects, since the projects must assist developing countries in achieving sustainable development (see Chapter 13 ). The CDM successfully registered 450 projects by the end of 2006 under the UNFCCC by the Executive Board with many more in the pipeline. Since the first project entered the pipeline in December 2003, 76% of projects belong to the energy sector. If all the 1300 projects in the pipeline at the end of 2006 are successfully registered with the UNFCCC and perform as expected, an accumulated emission reduction of more than 1400 MtCO₂-eq by end of 2012 can be expected ( UNEP, 2006 [NPR] ).

Information instruments

Education, technical training and public awareness are essential complements to GHG mitigation policies. They provide direct and continuous incentives to think, act and buy ‘green’ energy and to use energy wisely. Green power schemes, where consumers may choose to pay more for electricity generated primarily from renewable energy sources, are an example of combining information with real choice for the consumer ( Newman et al., 2002 [NPR] ). Voluntary energy and emissions savings programmes, such as Energy Star ( EPA, 2005a [NPR] ), Gas Star ( EPA, 2005b [NPR] ) and Coalbed Methane Outreach ( EPA, 2005c [NPR] ) serve to effectively disseminate relevant information and reduce knowledge barriers to the efficient and clean use of energy. These programmes include public education aspects, but are also built on industry/government partnerships. However, uncertainties on the effectiveness of information instruments for climate-change mitigation remain. More sociological research would improve the knowledge on adequacy of information instruments ( Chapter 13 ).

Technology development and deployment

The need for further investments in R&D of all low-carbon-emission technologies, tied with the efficient marketing of these products, is vital to climate policy. Programmes supporting ‘clean technology’ development and diffusion are a traditional focus of energy and environmental policies because energy innovations face barriers all along the energy-supply chain (from R&D, to demonstration projects, to widespread deployment). Direct government support is often necessary to hasten deployment of radically new technologies due to a lack of industry investment. This suggests that there is a role for the public sector in increasing investment directly and in correcting market and regulatory obstacles that inhibit investment in new technology through a variety of fiscal instruments such as tax deduction incentives (Energy Policy Act, 2005; ( Jaffe et al, 2005 ) ).

Following the two oil crises in the 1970 s, public expenditure for energy RD&D rose steeply, but then fell steadily in industrial countries from 15 billion US$ in 1980 to about 7 billion US$ in 2000 ( 2002 prices and exchange rates). Shares of IEA member-country support for energy R&D over the period 1974 – 2002 were about 8% for renewable energy, 6% for fossil fuel, 18% for energy efficiency, 47% for nuclear energy and 20% on other items ( IEA, 2004b [NPR] ). During this period, a number of national governments (e.g. US, Germany, United Kingdom, France, Spain and Italy) made major cuts in their support for energy R&D. Public spending on energy RD&D increased in Japan, Switzerland, Denmark and Finland and remained stable in other OECD countries (Goldemberg and Johannson, 2004 ).

Technology deployment is a critical activity and learning from market experience is fundamental to the complicated process of advancing a technology toward economic efficiency while encouraging the development of large-scale, private-sector infrastructure ( IEA, 2003 [Ambiguous] h). This justifies new technology deployment support by governments ( Section 4.5.6 ).

4.5.1.2 Policy implementation experiences—successes and failures

Experiences of early policy implementation in the 1990 s to reduce GHG emissions exist all over the world. This section lists and evaluates some examples. The fast penetration of wind power in Denmark was due to a regulated, favourable feed-in tariff. However, a new energy act in 1999 changed the policy to one based on the trading of green certificates. This created considerable uncertainty for investors and led to a significant reduction in annual investments in wind power plants during recent years ( (Johansson and Turkenburg, 2004 ) ).

In Germany, a comprehensive renewable energy promotion approach launched at the beginning of the 1990 s led to it becoming the world leader in terms of installed wind capacity, and second in terms of installed PV capacity. The basic elements of the German approach are a combination of policy instruments, favourable feed-in tariffs and security of support to reduce investment risks ( (Johansson and Turkenburg, 2004 ) ).

When Spain passed a feed-in law in 1994, relatively few wind turbines were in operation. By the end of 2002, the country ranked second in the world, but had less success with solar PV in spite of having high solar radiation levels and setting PV tariffs similar to those in Germany. Little PV capacity was installed initially because regulations to enable legal grid connection were not established until 2001 when national technical standards for safe grid connection were implemented. PV producers who sold electricity into the grid, including individual households, had to register as businesses in order not to pay income tax on their sales ( Sawin, 2003a [NPR] ). Significant growth in Spanish PV manufacturing in recent years is more attributable to the neighbouring German market ( (Ristau, 2003 ) ).

In 1990, the UK government launched the first of several rounds of competitive bidding for renewable energy contracts, known as the Non-Fossil Fuel Obligation (NFFO). The successive tendering procedures resulted in regular decreases in the prices for awarded contract value for wind and other renewable electricity projects. The average price for project proposals, irrespective of the technology involved, decreased from 0.067 €/kWh in 1994 to 0.042 €/kWh by 1998, being only 0.015 €/kWh above the wholesale electricity pool reference purchase price for the corresponding period ( Menanteau et al., 2001 [NPR] ). Due to only relatively small volumes of renewable electricity being realized through the tender process, the government changed to a support mechanism by placing an obligation on electricity suppliers to sell a minimum percentage of power from new renewable energy sources. The annual growth rate of electricity generation by eligible renewable energy plants has significantly increased since the introduction of the obligation in April 2002 ( OFGEM, 2005 [NPR] ).

Swedish renewable energy policy during the 1970 s and 1980 s focused on strong efforts in technology research and demonstration. Subsequently market development took off during the 1990 s when taxes and subsidies created favourable economic conditions for new investments and fuel switching. The use of biomass increased substantially during the 1990 s (for example forest residues for district heating increased from 13 PJ in 1990 to 65 PJ in 2001 ). Increased carbon taxes created strong incentives for fuel switching from cheaper electric and oil-fired boiler for district heating to biomass cogeneration. The increase of biomass utilization led to development of the technology for biomass extraction from forests, production of short-rotation coppice Salix and implementation of more efficient district heating conversion technologies ( Johansson, 2004 [NPR] ).

Japan launched a ‘Solar Roofs’ programme in 1994 to promote PV through low-interest loans, a comprehensive education and awareness programme and rebates for grid-connected residential systems. In 1997, the rebates were opened to owners and developers of housing complexes and Japan become the world’s largest installer of PV modules ( (Haas, 2002 ) ). Government promotion included publicity on television and in newspapers ( IEA, 2003 [Ambiguous] f). Total capacity increased at an average of more than 42% annually between 1994 and 2002 with more than 420 MW installed leading to a 75% cost reduction per Watt( (Maycock, 2003; ) IEA, 2003 [Ambiguous] f). The rebates declined gradually from 50% of installed cost in 1994 to 12% in 2002 when the programme ended. Japan is now the world’s leading manufacturer, having surpassed the US in the late 1990 s.

China’s State Development and Planning Commission launched a renewable energy Township Electrification Program in 2001 to provide electricity to remote rural areas by means of stand-alone renewable energy power systems. During 2002 – 2004, almost 700 townships received village-scale solar PV stations of approximately 30–150 kW (about 20 MW total), of which few were hybrid systems with wind power (about 800 kW total of wind). Overall, the government provided 240 million US$ to subsidize the capital costs of equipment and around one million rural dwellers were provided with electricity from PV, wind-PV hybrid, and small hydropower systems ( Martinot, 2005 [NPR, SRC] ). Given the difficulties of other rural electrification projects using PV ( ERC, 2004 [NPR] ), it is too early to assess the effectiveness of this programme.

The California expansion plan to aid the installation of a million roofs of solar power in the residential sector in the next ten years was signed into law in August 2006 (Environment California, 2006 ). The law increased the cap on net metering from 0.5% of a utility’s load to 2.5%. A solar rebate programme will be created and it will be mandatory that solar panels become a standard option for new homebuyers.

4.5.4.1 Health and environment

Energy interlinks with health in two contradictory ways. It is essential to support the provision of health services, but energy conversion and consumption can have negative health impacts ( Section 11.8 ). For example, in the UK, a lack of insufficient home heating has been identified as a principal cause of high levels of winter deaths (London Health Commission, 2003 ), but emissions from oil, gas, wood and coal combustion can add to reduced air quality and respiratory diseases.

The historical dilemma between energy supply and health can be demonstrated for various sectors, although it should be noted that recent times have seen major improvements. For instance, whereas epidemiological studies have shown that oil production in developed countries is not accompanied by significant health risks due to application of effective abatement technology, a Kazakhstan study compared the health costs between the city of Atyrau (with a high rate of pollution from oil extraction) and Astana (without). Health costs per household in Atyrau were twice as high as in Astana. The study also showed that the annual benefits of investments in abatement technologies were at least five times higher than the virtual annual abatement costs. A key barrier to investment in abatement technologies was the differentiated responsibility, as household health costs are borne by individuals, while the earnings from oil extraction accrue to the local authorities ( (Netalieva et al., 2005 ) ).

Accidental spills during oil-product transportation are damaging to the environment and health. There have been many spills at sea resulting in the destruction of fauna and flora, but the frequency of such incidents has declined sharply in recent times ( Huijer, 2005 [NPR] ). There are also spills originating from cracks in pipelines due to failure or sabotage. For example, it was estimated that the trans-Ecuadorian pipeline alone has spilt 400,000 litres of crude oil since it opened in 1972 . Spills at oil refineries are also not uncommon. Verweij 2003 [NPR] ) reported that in South Africa more than one million litres of petrol leaked from the refinery pipeline systems into the soil in 2001, thus contaminating ground water. One of the most recent oil spills occurred in Nanchital, Mexico in December 2004, where it was estimated that 5000 barrels of crude oil spilled from the pipeline with much of it going into the Coatzacoalcos River. Pemex, the company owning the pipeline, indicated a willingness to compensate the more than 250 local fishermen and the owners of the 200 hardest-hit homes. Coal mining is also hazardous with many thousands of fatalities each year. Exposure to coal dust has also been associated with accelerated loss of lung function ( (Beeckman and Wang, 2001 ) ).

4.5.4.2 Equity and shared responsibility

Economies with a high dependence on oil exports tend to have a poorer economic performance ( Leite and Weidmann, 1999 [NPR] ). The local energy needs of the host countries may be overlooked by their governments in the quest for foreign earnings from energy exports. Inadequate returns to the energy resource-rich communities have resulted in organized resistance against oil-extraction companies. Insecurities associated with oil supplies also result in high military expenditure as shown by OPEC countries ( Karl and Gary, 2004 [NPR] ).

The advent of reform in the energy sector increases inequalities. Notably electricity tariffs have generally shifted upwards after commencement of reforms ( Wamukonya, 2003 [NPR, SRC] ; Dubash, 2003 [NPR] ) making electricity even more inaccessible to the lower-income earners. There are many genuine efforts to address such issues (World Bank, 2005 [NPR] ), although much still needs to be done (Lort-Phillips and Herringshaw, 2006 ). Companies whose origin countries have stringent mandatory disclosure requirements are reported to perform best on transparency. Public private partnerships in developing countries are starting to make inroads into the issue of inequity and to harmonize practices between the developed and developing world. One such example is the Global Gas Flaring Reduction Partnership (World Bank, 2004a [NPR] ) aimed at reducing wasteful flaring and conserving the hydrocarbon resources for utilization by the host country.

4.5.4.3 Barriers to providing energy sources for sustainable development

The high investment cost required to build energy-system infrastructure is a major barrier to sustainable development. The IEA 2004a [NPR] ) estimated that 5 trillion US$ will be needed to meet electricity demand in developing countries by 2030 . To meet all the eight Millennium Development Goals will require an annual average investment of 20 billion US$ to develop energy infrastructure and deliver energy services ( UNDP, 2004b [NPR] ). Access to finance for investment in energy systems, especially in developing countries, has, nonetheless, been declining.

Available infrastructure also dictates energy types and use patterns. For instance, in a study on Peruvian household demand for clean fuels, Jack 2004 [NPR] ) found that urban dwellers were more likely to use clean fuels than rural householders, due to the availability of the necessary infrastructure. Investment costs necessary to capture natural gas and divert it into energy systems and curb flaring and venting are a barrier, even though efforts are being made to overcome this problem (World Bank, 2004a [NPR] ). It is estimated that over 110 billion m³ of natural gas are flared and vented worldwide annually, equivalent to the total annual gas consumption of France and Germany ( ESMAP, 2004 [NPR] ).

Levels of investment vary across regions, with the most needy receiving the least resources. Between 1990 and 2001, private investments to developing and transition countries for power projects were about 207 billion US$. Nearly 43% went to Latin America and the Caribbean, 33% to East Asia and the Pacific and approximately 1.5% to sub-Saharan Africa ( Kessides, 2004 [NPR] ). Accessibility and affordability of clean fuels remains a major barrier in many developing countries, exacerbated when complex supply systems are required that lead to high transaction costs.

Corruption, bureaucracy and mismanagement of energy resources have often prevented the use of proceeds emanating from extraction of energy resources from being used to provide local energy systems to meet sustainable development needs. Forms of corruption have encompassed such schemes as:

• the granting of lucrative power purchase agreements (PPAs) by politicians, who then benefit from receiving a share of guaranteed prices considerably higher than the international market price (Shorrock, 2002; Vallete and Wysham, 2002);
• suspending plant operations, thereby compromising access to electricity and persuading government agencies to pay high premiums for political risk insurance (Hall and Lobina, 2004); and
• granting of lucrative sole-supplier trading rights for gas supplies (Lovei and McKechnie, 2000).

Oil-backed loans have contributed to high foreign debts in many oil-producing countries at the expense of the poor majority ( IMF, 2001 [NPR] ; Global Witness, 2004 [NPR] ). Despite heavy debts, such countries continue to sign for such loans ( AEI, 2003 [NPR] ) and potential revenues are used as collateral to finance government external debt rather than to reduce poverty or promote sustainable development. These loans are typically provided at higher interest rates than conventional concessionary loans (World Bank, 2004b [NPR] ) and so the majority of the local population fail to benefit from high oil prices ( IRIN, 2004 [NPR] ). The problem could be overcome by legal frameworks that enable the channelling of revenue into investments that provide energy systems and promote sustainable development in communities affected by energy-resource extraction. In the meantime, the problem remains a key barrier to sustainable development and, although several countries including Peru, Nigeria and Gabon have mandated enabling mechanisms for such transfers, progress in implementing these measures has been slow ( Gary and Karl, 2003 [NPR] ).

Poor policies in the international financing sector hinder the establishment of energy systems for sustainable development. A review of the extractive industries (World Bank, 2004b [NPR] ), for example, revealed that the World Bank group and the International Finance Corporation (IFC) have been investing in oil- and gas-extractive activities that have negative impacts on poverty alleviation and sustainable development. The review, somewhat controversially, recommended that the banks should pull out of oil, gas and coal projects by 2008 .

Population growth and higher per-capita energy demand are forcing the transition of supply patterns from potentially sustainable systems to unsustainable ones. Efficient use of biomass can reduce CO₂ emissions, but can only be sustained if supplies are adequate to satisfy demand without depleting carbon stocks by deforestation ( Section 4.3.3.3 ). If supplies are inadequate, it may be necessary to shift demand to fossil fuels to prevent overharvesting. In Niger, for example, despite the concerted efforts through a long-term World Bank funded project, it is not possible to provide sufficient woody biomass on a sustainable basis. As a result, the government has launched a campaign to encourage consumers, particularly industry, to shift from wood to coal and has re-launched a 3000 t/yr production unit, distributed 300 t of coal to Niamey, and produced 3800 coal-burning stoves ( ISNA, 2004 [NPR] ). Further, in the electricity sector, PPAs that are not favourable to the establishment of generation plants that promote sustainable development are increasingly common. These include long-term PPAs with payments made in foreign currency denominations, leaving the power sector extremely vulnerable to macro-economic shocks as demonstrated by the 1998 Asian crisis ( Wamukonya, 2003 [NPR, SRC] ).

4.5.4.4 Strategies for providing energy for sustainable development

Although the provision of improved energy services is not mentioned specifically in the formal Millennium Development Goals (MDGs) framework, it is a vital factor. Electrification and other energy-supply strategies should target income generation if they are to be economically sustainable. It is important to focus on improving productive uses of energy as a way of contributing to income generation by providing services and not as an end in themselves. It has been argued that the traditional top-down approaches to reform the power sector – motivated by macroeconomic factors and not aimed at improving access for the poor – should be replaced by bottom-up ones with communities at the centre of the decision process ( GNESD, 2006 [NPR] ).

Sea-level rise, tropical cyclones and large ocean waves may hamper offshore oil and gas exploration and extraction of these fossil fuels. Higher ambient temperatures may affect the efficiency and capacity ratings of fossil-fuel-powered combustion turbines. In addition, electricity transmission losses may increase due to higher ambient temperatures. Renewable-energy systems may be adversely affected ( Sims, 2003 [SRC] ), for example if solar power generation and water heating are impacted by increased cloud cover. Lower precipitation and higher evaporation due to higher ambient temperatures may cause lower water levels in storage lakes or rivers that will affect the outputs of hydro-electric power stations. Energy crop yields could be reduced due to new pests and weather changes and more extreme storm events could damage wind turbines and ocean energy devices. The need to take measures to lessen the impacts on energy systems resulting from their intrinsic vulnerability to climate change will remain a challenge for the foreseeable future.

4.5.6.1 Public and private funding

Almost all (98%) of total OECD energy R&D investment has been by only ten IEA member countries ( Margolis and Kammen, 1999 [JoC] ; WEC, 2001 [NPR] ). The amount declined by 50% between the peak of 1980 (following the oil price shocks) and 2002 in real terms ( Figure 4.32 ). Expenditure on nuclear technologies, integrated over time, has been many times higher than investment in renewable energies. The end of the cold war and lower fossil-fuel prices decreased the level of public attention on energy planning in the 1980 s, and global energy R&D investment has yet to return to these levels despite growing concerns about energy security and climate change ( Chapter 13 ).

Figure 4.32: IEA member government budgets for total renewable energy R&D annual investments for 1974 – 2003 (left, a) and investment per capita, averaged between 1990 and 2003 (right, b).

Source: IEA, 2006d [NPR] .

Ultimately, it is only by creating a demand-pull market (rather than supply-push) that technological development, learning from experience, economies of scale in production and related cost reductions can result. As markets expand and new industries grow (the wind industry for example), more private investment in R&D results, which is often more successful than public research ( Sawin, 2003b [NPR, MoS] ).

The private sector invests a significant amount in energy RD³ to seek competitive advantage through improved technology and risk avoidance in relation to commercialization. Firms tend to focus on incremental technology improvements to gain profits in the short term. R&D spending by firms in the energy industry is particularly low with utilities investing only 1% of total sales in US, UK and the Netherlands compared with the 3% R&D-to-sales ratio for manufacturing, and up to 8% for pharmaceutical, computer and communication industries.

If government policies relating to strategic research can ensure long-term markets for new technologies, then industries can see their potential, perform their own R&D and complement public research institutions ( Luther, 2004 [NPR] ). Fixed pricing laws to encourage the uptake of new energy-supply technologies have been successful but do not usually result in novel concepts. Further innovation is encouraged once manufacturers and utilities begin to generate profits from a new technology. They then invest more in R&D to lower costs and further increase profit margins ( (Menanteau et al., 2003 ) ). Under government mandatory quota systems (as used to stimulate renewable energy projects in several countries – Section 4.5.1 ), consumers tend to benefit the most and hence producers receive insufficient profit to invest in R&D.

Recent trends in both public and private energy RD³ funding indicate that the role of ‘technology push’ in reducing GHG emissions is often overvalued and may not be fully understood. Subsidies and externalities (both social and environmental) affect energy markets and tend to support conventional sources of energy. Intervention to encourage R&D and adoption of renewable energy technologies, together with private investment and the more intelligent use of natural and social sciences is warranted ( (Hall and Lobina, 2004 ) ). Obtaining a useful balance between public and private research investment can be achieved by using partnerships between government, research institutions and firms.

Current levels of public and private energy-supply R&D investment are unlikely to be adequate to reduce global GHG emissions while providing the world with the energy needs of the developing nations ( Edmonds and Smith, 2006 [NPR, ARC] ). Success in long-term energy-supply R&D is associated with near-term investments to ensure that future energy services are delivered cost-effectively and barriers to implementation are identified and removed. Sustainable development and providing access to modern energy services for the poor have added challenges to R&D investment ( IEA, 2004a [NPR] ; IEA 2006a [NPR] ; Chapter 13 ).