The recent literature on long-term energy and emission scenarios increasingly emphasizes that both resource availability and technology are interrelated and inherently dynamic (see, e.g., IIASA-WEC, 1995; Watson et al. 1996; Nakicenovic et al., 1998a). The state of the art of theories and models of technological change is reviewed in Section 3.4.5. The literature suggests that models of endogenous technological change are still in their infancy, and that no methodologies are established that reduce the substantial uncertainties with respect to direction and rates of change of future technology developments. Differences in opinions as to the likelihood and dynamics of change in future technologies will therefore persist. Such future uncertainties are best captured by adopting a scenario approach. The following discussion on changes in energy supply and end-use technologies therefore reviews the literature with emphasis on empirically observed historical and conjectured future changes. The principal message is that while the future is uncertain, the certainty is that future technologies will be different from those used today. Hence the most unlikely scenario of future development is that of stagnation, or absence of change.
Fossil-fueled power stations traditionally have been designed around steam turbines to convert heat into electricity. Conversion efficiencies of new power stations can exceed 40% (on a lower heating value basis - when the latent heat of steam from water in the fuel or the steam arising from the hydrogen content of the fuel has been excluded). New designs, such as supercritical designs that involve new materials to allow higher steam temperatures and pressures, enable efficiencies of close to 50%. In the long run, further improvements might be expected. However, the past decade or so has seen the dramatic breakthrough of combined cycle gas turbines (CCGTs). The technology involves expanding very hot combustion gases through a gas turbine with the waste heat in the exhaust gases used to generate steam for a steam turbine. The gas turbine can withstand much higher inlet temperatures than a steam turbine, which produces considerable increases in overall efficiency. The latest designs currently under construction can achieve efficiencies of over 60%, a figure that has been rising by over 1% per year for a decade. The low capital costs and high availability of CCGTs also make them highly desirable to power station operators. Gregory and Rogner (1998) estimate that maximum efficiencies of 71 to 73% are achievable within a reasonable period (on a lower heating basis; around 65 to 68% on a higher heating basis).
CCGTs can also be used with more difficult fuels, such as coal and biomass, by adding a gasifier to the front end to form an integrated gasification combined cycle (IGCC) power station. The gases need to be hot cleaned prior to combustion to avoid energy losses and this is one of the key areas of development. The added benefit is that coal flue gas desulfurization (FGD) becomes unnecessary as sulfur is removed before the combustion stage. In addition to FGD and IGCC, fluidized bed combustion (FBC) technology facilitates sulfur abatement (adding limestone during combustion to retain sulfur) and allows the utilization of low quality fossil fuels because of the high sulfur retention capacity. However, FGD reduces overall conversion efficiencies (i.e., increases CO2 emissions). Also, both FGD and FBC sulfur abatement technologies use calcium carbonate to reduce emissions, which increases CO2 emissions because of CaO liberation from CaCO3 capture. The use of high sulfur coals in FBC requires high limestone consumption, which results in increased CO2 emissions. Only IGCC reduces SO2 and CO2 emissions, because of the higher power generation efficiency.
Biomass is particularly suited to gasification. Stoll and Todd (1996) and Willerboer (1997) estimate that current designs for coal are around 9% less efficient than a standard CCGT burning gas. Developments to reduce heat losses through better heat recovery and by hot gas cleaning could potentially increase efficiency significantly in the next 10 to 15 years, and the technology can yield efficiencies of 51% now and perhaps 65% in the longer run.
The major potential competitor to CCGT technology is the fuel cell, which may be able to offer similar efficiencies at much lower plant sizes and so may be an ideal candidate for distributed combined heat and power generation. Another promising fuel cell application is vehicle propulsion. In contrast to other fossil-sourced electricity generation, fuel cells (similar to batteries) convert the chemical energy of fuels electrochemically (i.e., without combustion) into electricity and heat, and thus offer considerably higher conversion efficiencies than internal combustion engines. In the past, high costs and durability problems have restricted their use to some highly specialized applications, such as electricity generation in space. Recent advances in fuel cell technology, however, have led to their commercial production and application in niche markets for distributed combined heat and power production (Penner et al., 1995).
Although operating internally on hydrogen, fuel cells can be fueled with a hydrocarbon fuel, such as natural gas, methanol, gasoline, or even coal. Before entering the fuel cells, these fuels would be converted on-site or on-board into hydrogen via steam reforming, partial oxidation, or gasification and hydrogen separation. In the longer run and to make fuel cells truly zero-emission devices, non-fossil derived pure hydrogen, supplied and stored as compressed gas, cryogenic liquid, metal hydrate, or other storage form, could replace hydrocarbon fuels. Current fuel cell conversion efficiencies (45 to 50%) have yet to approach their potentials; some designs report efficiencies as high as 60% electrical efficiency in simple systems and 74% total efficiency in hybrid fuel cell and gas combined cycle systems (FETC, 1997). Hydrogen production efficiencies range from 65 to 85% for fossil-based systems, 55 to 73% for biomass-based systems, and 80% to close to 90% for electrolysis (Ouellette et al., 1995; Williams, 1998).
Nuclear power is a proven technology that provides 17% of global electricity supply. There is currently no consensus concerning the future role of nuclear power. While it stagnates in Europe and North America, it continues as a strong option in a number of Asian countries and countries undergoing economic reform. Economics and security of supply are considerations in the choice of nuclear power, along with its environmental trade-offs - on a full energy chain basis, from mining to waste disposal and decommissioning, nuclear power emits little GHG emissions.
Public opinion is opposed to the use of nuclear power in many countries because of concerns regarding operating safety, final disposal of high-level radioactive waste, and proliferation of weapon-grade fissile materials, as well as uranium mining and its environmental implications. These concerns, perceived or real, cannot be ignored if nuclear power is to regain the position of an accepted technology. In addition, changed market conditions call for new reactor technologies - smaller in scale, reduced construction periods, and improved economics with no compromise in safety. The industry is striving continuously to develop advanced reactor designs of much lower cost and with inherent safety concepts (i.e., designs that make safety less dependent on specifically activated technology components and human performance). The latest reactor technology already prevents the release of fission products or health-damaging radiation to the environment even under highly unlikely severe accident conditions.
The 1998 update of the OECD cost comparison for power plants expected to be commercially available around 2005 reports nuclear generating cost in the range 2.5 to 9.0 c/kWh (US cents per kilowatt-hour), depending on location, type of reactor, plant factor, plant life time, discount rate, and underlying fuel price escalation (OECD, 1998c). Reconciliation of the assumptions of the OECD report with those used in the IPCC WGII SAR leads to an almost identical cost range of 2.9 to 5.4 c/kWh, at 5% discount rate, and 7.7 c/kWh, at 10% discount rate (Ishitani and Johansson, 1996).
Since the birth of the modern wind power industry in the mid-1970s, it has seen a continuous chain of innovations and cost reductions. Christiansson (1995) discusses a learning curve that relates cost reductions to installed capacity. She notes that the experience curve for the USA indicates a progress ratio of 0.84 (i.e., for each doubling of installed capacity, costs of new installations are reduced to 0.84 of the previous level). In the UK, developers have the opportunity to bid to supply renewable electricity to the grid under the Non-Fossil Fuel Obligation. In 1998, the average bid price for large wind power schemes was about 5 c/kWh, nearly 20% lower than in 1996 (ENDS, 1998). The bids are made on a full commercial cost basis and they indicate that wind energy is cheaper than most of its competitors for new schemes in the UK for modest increments of capacity. (The exception is gas-powered CCGT generation.) However, wind, by nature, is intermittent and back-up capacity will be required as the proportion of electricity provided by wind increases, which reduces its economic attractiveness. Large integrated electricity systems and systems that contain significant amounts of hydropower (especially with storage) are able to cope with the fluctuations in wind power best (Grubb and Meyer, 1993; Johansson et al., 1993).
Solar voltaic power is on a similar learning curve to wind, with progress ratios of 0.82 in the USA and 0.81 in Japan (Christiansson, 1995; Watanabe, 1997). Various alternative technologies are being developed and solar voltaics can be envisaged as providing electricity through large arrays of cells in central power generation, through arrays built into cladding or roofing on buildings, or through single arrays to meet specific purposes. In this last form, solar cells have already established themselves as economic and reliable power sources in the provision of light, clean water, and improved health services to isolated rural communities (WEC, 1994). Based on the principle of a learning curve, solar cell costs are expected to fall as capacity builds. For example, the US Department of Energy projects costs to fall from 38-55 c/kWh in 1995 to 3.5-5 c/kWh in 2030 (US DOE, 1994). For another system, WEC estimates costs to fall from 13-23 c/kWh in 1990 to 5-10 c/kWh in 2020 (WEC, 1994).
Biomass, particularly wood, has been the main initial source of energy as countries develop and remains a substantial energy source in many developing countries (WEC, 1994). In addition to this, byproducts of agriculture and forestry can be useful sources of energy. The extent to which biomass can contribute beyond this to provide energy crops for use in, say, power stations or for conversion into liquid fuels depends firstly on the competition for land with agriculture for food production. This, in turn, depends on improved productivity in food production, the amount and type of meat and other animal (e.g., dairy) products in the diet, and the growth in human population. The future contribution of biomass will also depend on increased productivity in biomass production coupled with limitations in energy input to grow, harvest, and use energy crops (Ishitani and Johansson, 1996; Leemans, 1996). Current costs of biomass (for a eucalyptus plantation in Uruguay) are typically put at US$1.8/GJ with projected costs of US$1.4/GJ and productivity at 360GJ/hectare per year (Shell International Ltd., 1996). Agricultural productivity is also growing at 2% per year with the prospects of a similar productivity gain for energy crops (Shell International Ltd., 1996). WEC (1994) notes that Brazil has had a program to grow sugar cane to produce transport fuel since the mid 1970s and that production has reached 62% of the country's needs; this may be one option for wider use as the availability of conventional oil starts to fall.
Finally, other forms of renewable energy, such as geothermal energy, tidal energy, wave energy, ocean thermal energy conversion, and solar thermal power plants, could make significant contributions at some stage in the future, as geothermal energy already does in specific markets. Also, technologies and processes that lead to carbon sequestration as a by-product include enhanced oil recovery using CO2 to improve the viscosity of crude oil or reforestation for reasons of soil preservation (Ishitani and Johansson, 1996).
The fuel effectiveness of all energy conversion processes that involve combustion, but also of fuel cells, can be raised substantially by combined heat and power generation. Utilities located in the vicinity of urban areas may divert the waste heat from combustion for residential or commercial heating purposes. Industrial producers of high-temperature process heat may consider the generation of electricity when process temperature requirements are lower than the temperatures supplied. Combined heat and power production can accomplish fuel utilization rates of 90% or more (Ishitani and Johansson, 1996).
Other reports in this collection