Most scenarios suggest that during the next decades the growth in carbon emissions will increasingly take place in the developing countries. According to the data summarized in Chapter 12 [THERE IS NO CHAPTER 12. PLEASE CHANGE] of this report, the mean world annual growth rate of CO2 emissions for over twenty scenarios is 1.56%, the corresponding mean rate for China is 2.83%, for Eastern Europe and ex-USSR 0.76%, and for Africa 3.85%. Consequently, according to one of the scenarios in that chapter, ECS 92 (dynamics as usual), the share of developing countries in global CO2-emissions is projected to reach 46% by 2020 (as compared to 34% for the OECD and 20% for countries in transition). However, according to the various World Energy Council (WEC) scenarios, the developing countries' share in 2020 would be over 60%. In any case, it seems most likely that the developing countries as a group will start to become the major CO2 emitters within a few decades. This picture is reinforced if the emissions of CH, from wetland rice cultivation and from enteric fermentation are also taken into account.

At the same time it is clear that, although the scope for effectively applying policy options in the developing countries seems to be significant (for a recent evaluation of various technical options at the country level, see UNEP, 1994), so are the obstacles to be encountered. Indeed, the availability of technical options for higher energy efficiency, to give just one example, does not guarantee their adoption on a large scale. There may need to be a significant stimulus to achieve widespread efficiency improvements, particularly in markets characterized by high implicit discount rates. But a combination of education, financial incentives, and minimum efficiency standards coupled with freedom from distortionary policies can effectively transform energy use markets so that large energy savings and emission reductions are achieved along with net economic savings (Geller and Nadel, 1994).

The literature on the adoption and diffusion of technology clearly indicates that while profitability is probably the most straightforward determinant of the adoption of a new idea, a new technology, or new equipment, various other factors may also be important. A review of recent research into the diffusion of energy technologies in developing countries shows that there are many financial, institutional, and other factors that influence the successful adoption of these technologies (Barnett, 1990; Ghai, 1994). In Africa, for example, social resistance has impeded the diffusion of drought-tolerant crop varieties, and such resistance could also inhibit the adoption of new energy technologies. Often the initial awareness of benefits and new opportunities may be contingent upon such factors as, say, winning ever women to introduce the support of women for more energyefficient cooking stoves.

Moreover, one necessary ingredient for the adoption of new technology, namely a pool of local skills to draw upon, may be lacking or inadequate in many cases, so that even proven technologies may spread rather slowly in these countries. For all these reasons an adequate and timely process of energy efficiency institution building seems imperative, especially in developing countries. There is evidence that the existence of such separate institutions has, for instance, helped Indonesia, South Korea, and Thailand make greater headway in the scope and coverage of their energy efficiency policies and programmes (Byrne et al., 1991).

23-558 96-25

1. The "engineering efficiency" approach determines the financial costs and benefits of various options to an individual agency or other entity in terms of CO2 emission reduction/absorption; in the "welfare economics" approach the broadly defined costs and benefits of options to society are determined. These two approaches will be further discussed in Section 7.3.

2. For an example of energy conservation, see, for example, Rubin (1992). Here 25% of employerprovided parking places and tax remaining places-are eliminated and the remainder taxed to reduce solo commuting by 15-20% in the U.S. Net costs are estimated to be -$22/tC (ie-savingsa negative the same as a saving)

3. It may seem that, although conceptually the above categories are conceptually distinct, in real life they are not strictly mutually exclusive; that is, measures are conceivable that can be classified in more than one category. An example would be the plantation of forests or biomass that are used for energy purposes. These measures seem to fall both in category 3 (renewable energy) and in category 6 (enhancing carbon sinks). However, this is not the case. The measures are an example of how easily markedly different processes that underlie the measures can be confused.

In the case of forests, broadly three types of measures are conceivable to fix carbon: (1) to afforest new lands to let the forest simply mature; (2) to plant forest and sequester the timber derived from it; and (3) to use the wood for energy purposes on a sustainable basis, and thereby avoiding the alternative use of fossil fuels. In the following, (1) and (2) are discussed in Section 4.6 (forestry options), whereas (3) belongs to the renewable/biomass category. With respect to (3), it should be borne in mind that sometimes a significant amount of additional energy may be required to turn the biomass into energy. This is, for instance, the case for the production of ethanol from corn, where additional energy requirements are of the order of the energy content of the produced ethanol itself (Swisher et al., 1993).

Another example would be to classify an Integrated Gasification Combined Cycle (IGCC) or the hydrocarb process both in category 1 (energy saving and efficiency) and in category 3 (clean fossil technologies). In the present paper both are considered primarily clean technologies and changingboth change the energy conversion process to the extent that it violatesof violating the definition for the energy saving and efficiency category. However, ultimately no clear distinction can be made as modifications in the energy conversion process become minor (due to further technological progress).

4. For instance, a set of three studies for Poland, Hungary, and the former USSR indicate that a combination of energy efficiency improvements, fuel substitution, and structural change (Chandler, 1990), could reduce carbon emissions by 40-60% from base case projected levels by 2030: in the case of Poland, from 260 MMT in 2030 to 117 MMT (Sitnicki et al. (1990) suggest that baseline emissions of 260 Mt [megatonnes - confirm] could be reduced to 117 Mt by 2030; andfor the former USSR, -40% (Makarov and Bashmakov (1990) suggest that a reduction of 40% would be feasible,

6. The WEC distinguishes four scenarios for the energy mix in 2020: scenario A assumes high annual world economic growth (especially in developing countries), high annual energy intensity reduction, and very high total energy demand; scenario B1 assumes moderate annual world economic growth rates, moderate annual energy intensity reduction, and high possible total energy demand; scenario B, the reference scenario, assumes high annual energy intensity reduction; scenario C assumes moderate annual economic growth, very high energy intensity reductions, and relatively low total energy demand in 2020.

7. Here a set of definitions is used oftaken from Rogner et al. (1993) is used to distinguish between different levels of geological certainty and economical and technical feasibility. The resource base is defined to consist of (proven) reserves and resources. Reserves are those occurrences that are identified, measured, and known to be economically and technically recoverable at current prices and using current technologies. Resources comprise the remainder of occurrences with less certain geological and economic characteristics. Additional quantities with unknown certainty of occurrence or with unknown or no economic significance at present are referred to simply as occurrences.

8. Total global energy consumption amounted to 10 TWyr in 1990, whereas identified fossil energy reserves are estimated at 1,280 TWyr (Rogner et al. 1993, p. 463). Obviously, the use of an aggregate figure for fossil fuels (which is mostly coal reserves) should not obscure the fact that the corresponding time span for the individual fossil fuels differs widely. The ratio of proven reserves to annual production (R/P) is estimated at about 55 years for natural gas, at about 45 years for oil, and at about-235 years for coal.

9. IPCC carbon emission rates are 13.8, 18.7, and 26.6 kg of carbon per GJ for natural gas, crude oil, and coal respectively (Inaba, 1990, pp. 40-45.).

10. This is because type of transport and combustion technologies are roughly the same for natural gas and hydrogen (H2).

11. Adopting Assuming a 100-year time horizon. For the various ways the GWP measure for methane could be calculated, see, for example, Reilly and Richards (1993), pp. 41-61.

12. This would imply that 3-41% (for distribution) and 1-63% (for production) of the carbon reduction effrom a 100% coal-to-natural-gas fuel switch iswould be offset by the detrimental effects of leakage.

13. BEPA/((MER*GWP)+A)A/[(MER:GWP)+A], with BEP = break-even point, A = (26.613.8)x3.67, MER = mass:energy ratio for methane = 22 Tg CH/EJ, GWP = global warming potential index of methane = 21. The term A is the additional mass of carbon dioxide released by coal compared to methane per GJ of energy and is composed of the difference between the carbon emission rates of coal and methane (26.6 and 13.8 kgC/GJ respectively, sec note 9) times the mass ratio of CO2:C (3.67), and assuming The calculation assumes a zero leakage rate of methane in coal production; for other figures see note 7. Similarly, a 100% oil-to-gas switch shows a break-even point at about 4% leakage.

14. See, for example, Jackson (1991) for an analysis of cost-effectiveness in the UK, explicitly incorporating CH, leakage.

16. See, for example, IPCC (1991), Johansson et al. (1993), WEC Commission (1993), or WEC (1994). With respect to the classification of renewables as commonly used, it should be noted that classifyingthe classification of geothermal as a renewable resource is technically not correct, as the Earth's core will slowly but surely cool down.

17. Solar can broadly be subdivided into solar thermal, solar architecture, solar thermal-electric, photovoltaic systems, and thermochemical and photochemical systems. Wind and hydro are relatively homogeneous energy technologies, the largest differences stemming from scale of operation. Here, a distinction is made between small/medium-scale and large-scale conversion systems. In contrast, biomass appears to be the most complex of all technologies. A wide range of

conversion technologies exist, depending on the type of feedstock used and the form of energy output required. Geothermal consists of hydrothermal, hot dry rock, geopressured, and magma resources technologies. Current ocean technologies encompass tidal, wave, biomass, and salt and thermal gradient technologies.

18. Different approaches exist to get hold ofdescribe the concept of "practicable," i.e., realizable, potential. Most The most common dimensions of distinctioncategorizations are physical, technical, and economical, thein that order, with each ensuing categoryensuing potentials each being a subset of the earlier mentioned one. The physical potential would denote the maximum potential that is constrained by geological, geophysical, and meteorological factors only. Technical potential would refer to that part of physical potential that can be exploited given the state of technology at hand. Finally, the remainder of technical potential after excluding what is not deemed feasible due to prevailing economical constraints (such as a prohibitive level of costs, institutional constraints in the energy markets, etc.) would pass for economical potential. Notice that, for the present purpose, the former of the three can be considered constant in time, whereas the others prevail only at a certain moment.

Practicable potential now would be defined as somewhere between technical and economical potential. This is because the two do not hold independently but are interlinked in time: e.g., technical potential is enlarged by investments that stimulate technological progress. Conversely, the impact of improvements of, say, silicon films in photovoltaic systems on the price of solar energy is obvious.

19. For example, wind energy costs depend heavily on wind speed, and solar energy costs on solar irradiance, features that are not equally favourable throughout countries (/locations) and seasons (/hours of the day) for all locations, seasons, or times of day

20. It is not possible to tellderive cost developments for individual subclasses of technologies from the enlisted figures, as they are aggregated into ranges of similar technologies. The same holds for disparities stemming from differences efamong sites. It should be realized that thisthese limitations significantly hampers a direct comparison. However, a greater sense of detail was avoided for the purpose of clarity.

21. Estimates are 21.3-29.6% in 2020 (WEC, 1993b), 15% (6% of which comes from hydro) in 2020 (Grübler et al., 1993), and close to 43% in 2025 (Johansson et al., 1993). According to Grübler et al. (1993) the latter estimate is most likely too high. It would imply an unprecedented rate of change of technology and infrastructure. For comparison, it took about 80 years for the market share of oil to grow to 40% of global primary energy supply ((Grübler et al. 1993, p. 580). In the past the mean durationinterval for replacing most technological systems was about 30 to 40 years.

22. For an extensive discussion of the nuclear option, see Volume 2 of this report.

23. Only Relatively minor fossil fuel inputs are used thatto support the overall functioning of the breeder reactors.

24. Recovery of carbon at power plants has the advantage of removing carbon from energy before it is distributed to highly dispersed end-users.

25. As the use of coal and natural gas is predominant in power plants, virtually no technologies are based on oil.