International Energy Outlook 1999 - Environmental Issues and World Energy Use

In the coming decades, global environmental issues could significantly affect patterns of energy use around the world. This chapter examines the factors that govern national levels of energy-related carbon emissions.

In recent years, the principal international energy issues have shifted from supply interruptions and their implications for energy security and price stability to the impact of energy production and consumption on regional and global environments. Frequently, regional and global environmental goals are in conflict. For example, nuclear or hydropower energy projects may be opposed within a given country, while on a global scale they lessen emissions of carbon dioxide—the principal greenhouse gas. Although the focus of this analysis is on global environmental issues such as climate change, it should be understood that local environmental concerns and political decisions based on them may affect the ability of the world community to meet global environmental goals. In the coming decades, global environmental issues and their policy implications could significantly affect patterns of energy use.

The challenges of energy use and environmental quality facing the industrialized countries differ from those for the developing world. The industrialized countries have predicated their economic development on the availability of relatively low-cost fossil fuels. Accordingly, the infrastructure has been built to accommodate private vehicle travel and single-family dwelling units with relatively large amounts of space per person, especially in North America. Given the amount of capital investment in place, policies that modify underlying sources of energy inputs and end-use patterns will require time for turnover of existing capital stock if potentially large economic displacements are to be avoided. The principal challenge then, to the industrialized countries, is to implement policies that protect the global environment while allowing for flexible adjustment of their energy systems.

The developing world, while seeking to grow economically, is confronted with the environmental lessons learned in the process of the economic growth achieved by the industrialized countries. Within developing countries, much of the infrastructure that would support an industrialized economy is not yet in place. This presents an advantage in terms of identifying development paths that will allow greater scope for alternative energy sources and patterns of end-use consumption as new capital stock is put in place. On the other hand, developing economies are not likely to adopt policies that encourage alternative patterns of energy production and consumption if there is the perception that such policies are more costly and undermine near-term growth objectives.

This chapter examines the link between energy use and emissions of carbon dioxide in the context of the International Energy Outlook 1999 (IEO99) baseline projections and in light of possible greenhouse gas initiatives and reduction scenarios such as those proposed under the Kyoto Protocol. Implications are examined for the industrialized Annex I countries in terms of what combinations of energy intensity and carbon intensity reductions would be required to reduce carbon emissions.

The focus of international debate in recent years has been the Framework Convention on Climate Change developed in Rio de Janeiro, Brazil, in 1992; the resulting Protocol developed in Kyoto, Japan; and the Conference of the Parties (COP-4) held in Buenos Aires, Argentina, in the first 2 weeks of November 1998—all under the auspices of the United Nations Intergovernmental Panel on Climate Change [1]. The Kyoto Protocol agreement, if ratified, calls for quantifiable goals for carbon emissions from the Annex I countries (Table 20). Although it is difficult to predict the outcome of such broad-reaching international agreements, it is helpful to study the underlying factors that contribute to energy-related emissions of carbon dioxide.

The Kyoto Protocol, while calling for substantial cuts in carbon emissions by Annex I countries, will not mean stabilization of carbon emissions on a global basis (Figure 94). Without the Kyoto agreement, emissions are projected to increase by 38 percent between 1990 and 2010; if the agreement goes into effect, they are projected to increase by 31 percent. In the long term, carbon stabilization will require participation by the world’s developing nations.

Figure 94. Carbon Emissions by Annex I and Non-Annex I Nations in Two Cases, 1990, 2010, and 2020

Sources: 1990: Energy Information Administration (EIA), International Energy Annual 1996, DOE/EIA-0219(96) (Washington, DC, February 1998). Projections: EIA, World Energy Projection System (1999).

The so-called “clean development mechanism” (CDM) may offer developing nations participation under the Kyoto Protocol. Under the CDM, projects undertaken by Annex I nations to reduce emissions in developing countries could be counted toward Annex I country goals. Thus, the CDM may provide a way to fund sustainable energy projects in developing countries.

The following discussion focuses on the industrialized Annex I countries. They are grouped into three categories: North America, which consists of the United States and Canada (Mexico is not an Annex I country); Western Europe; and industrialized Asia, which consists of Japan, Australia, and New Zealand. The transitional Annex I countries in the EE/FSU region are projected to generate net carbon credits in relation to Kyoto Protocol goals. Current estimates indicate that credits for the equivalent of 374 million metric tons of carbon emissions would be available from the EE/FSU countries, nearly double last year’s estimate of 196 million metric tons. The Kyoto targets for Bulgaria, Hungary, Poland, and Romania—which currently account for some 66 percent of all emissions from Eastern European countries—were recalculated in this year’s IEO to reflect Article 3.5 of the Protocol, which allows the four countries to use base years other than 1990. Bulgaria and Romania are using 1989 as a base year; Poland is using 1988; and Hungary is using the average emissions for the years 1985 to 1987. As a result, the Kyoto target for total carbon emissions for Eastern Europe in 2010 is 320 million tons in IEO99, up from 277 million metric tons in the International Energy Outlook 1998 (IEO98), freeing the difference as possible emissions credits. The rest of the increase in the estimate of available credits is accounted for by changes in the economic outlook and expected mix of energy fuel use in the FSU region.

The reduction in expected energy consumption for the FSU region in this year’s projections could substantially change the amount of effort required by the Annex I countries as a whole to meet their Kyoto Protocol targets. In IEO98, energy demand in the FSU was expected to recover to its 1990 level by the end of the projection period. Carbon emissions were also expected to rise, but they remained below the 1990 level because the recovery featured increases in the use of less carbon-intensive natural gas rather than more carbon-intensive coal. As a result, IEO98 projected that in 2010 credits available from the FSU would contribute 199 million metric tons to the total of 822 million metric tons that the industrialized Annex I countries would need to eliminate from the baseline projection to meet the Annex I targets. In this year’s projection, however, the potential contribution in 2010 from the FSU has increased by 62 percent, to 324 million metric tons.

In IEO99, the credits projected to be generated from the transitional economies represent 45 percent of the total emissions reductions that would be required to meet the Kyoto targets if they are ratified—up from the 21 percent estimated in IEO98. In addition to the reasons described above, the increase is the result of adjustments that changed the 1990 baseline emissions for the EE/FSU region, the 2010 projections, and the resulting Kyoto targets for the EE/FSU nations.

The economic collapse in Russia has meant reductions from the IEO98 projections of FSU fossil fuel use in 2010: 21 percent for oil, 10 percent for natural gas, and 22 percent for coal. IEO99 projects that, by 2010, the resulting emissions from this lowered outlook for fossil fuel use in the FSU reach only 666 million metric tons, nearly 16 percent less than projected in IEO98. Because the transitional Annex I countries currently account for about 86 percent of the EE/FSU region’s total emissions, much of the projected emissions reduction could be used as tradable emissions units with the industrialized Annex I countries as they attempt to meet Kyoto Protocol emissions targets. Accordingly, to meet their Kyoto Protocol targets, Annex I countries would need to reduce emissions by 10 percent from the reference case projection rather than by 16 percent as reported in IEO98. Indeed, emissions are expected to grow by 7 percent between 1990 and 2010 in the industrialized Annex I countries and the EE/FSU combined, because the 27-percent decrease in emissions expected for the EE/FSU offsets the 23-percent increase projected for the industrialized Annex I countries.

Two factors, in combination with the level of economic activity, determine the energy-related carbon dioxide emissions of a given country at a given point in time. Differences in these factors from one country or region to the next determine the amount of emissions, their sources, and the relative baseline position from which a country could move to carbon reduction and stabilization.

A more formal perspective on the components of carbon emissions levels is provided by the following simple equation, where total energy-related carbon (C) is equal to the ratio of carbon to energy (C/E) times the ratio of energy to gross domestic product (E/GDP) times GDP [3]:

Clearly, government policy is not formed with the intention of reducing a country’s GDP. The primary goal of carbon reduction policy, therefore, is one of constraining carbon emissions while minimizing adverse affects on GDP growth.

As a result of the petroleum supply disruptions and price spikes of the 1970s, energy intensity has dropped consistently among most of the industrialized nations. The continuing decrease projected in the IEO99 reference case is based on an expected shift away from heavy industry toward information-based service economies and the adoption of inherently more efficient technologies, even in the face of stable energy prices.

Although there has been incentive in the past to reduce energy intensity, there has been little incentive to reduce carbon intensity. Until recently, carbon dioxide has been viewed as a benign presence in the atmosphere, and there have been no costs associated with carbon emissions. What changes have taken place have been the result of changes in technology, such as the introduction of nuclear power, or market forces, such as the emergence of natural gas as a competitive fuel for electricity generation. Increasing the use of low-carbon or carbon-free energy sources may require costly long-term changes to the energy infrastructure, but there is little chance of stabilizing carbon emissions without them.

The remainder of this chapter examines energy-related carbon dioxide emissions from the industrialized Annex I countries. Possible reduction scenarios are examined in the context of total energy demand and fuel mix and their implications for energy and carbon intensity.

Energy intensity differs from one region to the next in the industrialized world (Figure 95). The differences result from many factors, such as population density, weather, existing infrastructure, availability of energy supply, taxation, and cultural tastes and preferences.

Note: North America does not include Mexico.
Sources: History: Energy Information Administration (EIA), Office of Energy Markets and End Use, International Statistics Database and International Energy Annual 1996, DOE/ EIA-0219(96) (Washington, DC, February 1998). Projections: EIA, World Energy Projection System (1999).

North America’s energy intensity—the highest in the industrialized world—is more than twice that of industrialized Asia. Energy intensity in North America dropped sharply in the 1980s following the price increases of the late 1970s and early 1980s, and it is projected to continue dropping as the economy grows more rapidly than energy demand. The energy intensity of the industrialized Annex I countries is projected to decline by about 24 percent between 1996 and 2020 in the IEO99 reference case. Without such a decline, projected energy requirements would rise by about 135 quadrillion British thermal units (Btu), as compared with the projected rise of 55 quadrillion Btu, between 1996 and 2020.

Carbon intensity also differs across regions and over time (Figure 96); however, the differences tend to be smaller than the differences in energy intensity, inasmuch as fossil fuels are a ubiquitous energy source throughout the world. Before 1990, the carbon intensity for North America was the lowest of the industrialized regions. From 1990 to 1995, however, coal became a less important energy source in Western Europe with the shutting down of lignite production in Germany and hard coal production in the United Kingdom, even as productivity gains in the U.S. coal industry made coal a relatively inexpensive fuel for U.S. electricity generation [4]. As a result, North America had the highest carbon intensity among the industrialized regions, a circumstance that is expected to continue through 2020. A further problem with regard to carbon intensity in the United States is the expected loss of nuclear generation capacity, which at this point is likely to be replaced by fossil fuel generation.

The carbon intensity for Western Europe, which was just slightly more than that for North America in 1990, has dropped in recent years and is projected to stay below 14 million metric tons per quadrillion Btu of energy produced, as significant amounts of coal use are replaced by natural gas and by nuclear power, particularly in France. On the other hand, a reversal in the current trend of declining energy intensity in Western Europe is possible, as Sweden and Germany are leaning toward early nuclear retirements [5]. If a significant amount of Europe’s nuclear capacity is retired early, reductions in carbon intensity will become more difficult.

Industrialized Asia has shown, and is projected to show, a carbon intensity similar to that for Western Europe, stabilizing at about 14 million metric tons per quadrillion Btu. Japan is somewhat limited in its ability to expand natural gas use due to the relatively high cost of liquefied natural gas.¹⁹ The main source of non-carbon- emitting energy in Japan is nuclear power, which is projected to increase.

In all major regions of the industrialized world, nuclear power plays an important role in keeping carbon intensities below what they would be otherwise. However, of the industrialized Annex I countries, political and economic pressures have prevented any new nuclear capacity from being built outside France and Japan. It remains to be seen how climate change policy will affect the prognosis for the nuclear industry worldwide.

Hydropower, which has played a role in moderating carbon intensities in various countries, faces limited prospects for the future. Most of the best available sites in the industrialized countries have long since been exploited, and hydropower is no longer viewed as an environmentally benign energy source. New restrictions on facility relicensing could lead to decrements in hydroelectric generating capacity. The renewable energy sources of the future are most likely to be wind, solar (especially solar photovoltaics), and closed-loop biomass.²⁰ Although the use of these non-carbon- producing energy sources has increased only slightly in recent years, binding limits on carbon emissions such as those specified in the Kyoto Protocol could greatly enhance their economic prospects.

In addition to the reference case, IEO99 includes low and high capacity cases for nuclear power—an energy source that does not emit any greenhouse gases. The projected impacts of the different nuclear capacity cases on worldwide carbon emissions can be seen if it is assumed that world electricity demand, and the amount of energy needed to generate electricity, does not vary across the reference, high nuclear, and low nuclear cases. The resulting projections of fossil fuel consumption and carbon emissions in the three cases show the effects of the higher and lower nuclear capacity assumptions.

In the high nuclear case, projected worldwide nuclear generation of electricity is higher than the reference case projection by 5 percent in 2010 and 29 percent in 2020. In the reference case, fossil fuels provide 124 quadrillion Btu of energy for electricity generation in 2010 (65 percent of the total 191 quadrillion Btu used to produce electricity) and 156 quadrillion Btu in 2020 (68 percent of the total 227 quadrillion Btu). In the high nuclear case, fossil energy consumption for electricity generation in 2020 would be reduced by only 6.4 quadrillion Btu. Accordingly, worldwide carbon emissions from electricity generation would be reduced by only 52 million metric tons in 2010 and 206 million metric tons in 2020 in the high nuclear case from the reference case projections of 2,664 and 3,271 million metric tons, respectively (Figure 97).

Assuming that the increase in nuclear generation would displace coal, which is the most carbon-intense of the fossil fuels, a 206 million metric ton reduction would represent an upper bound for the carbon reduction in the high nuclear case. Alternatively, if nuclear generation displaced natural gas, carbon emissions would be reduced by about 115 million metric tons in the high nuclear case in 2020 relative to the reference case projection.

In the low nuclear case, projected carbon emissions are higher than in the reference case. Nuclear electricity generation worldwide would be reduced by 10 percent in 2010 and by nearly 37 percent in 2020—representing losses of 2.6 and 8.0 quadrillion Btu of nuclear power, respectively. It is technically possible that renewable energy sources could be substituted for the lost nuclear generation with no relative increase in carbon emissions; however, this would require an 88-percent increase in worldwide consumption of renewable energy sources between 1996 and 2020, as compared with the projected increase of 62 percent in the reference case. If natural gas were used instead, carbon emissions from electricity generation would be 38 million metric tons higher in 2010 and 116 million metric tons higher in 2020 than in the reference case (Figure 97).

Figure 97. Carbon Emissions From Electricity Generation in Three Cases, 1990, 2010, and 2020

Note: The high nuclear case assumes that nuclear generation replaces coal-fired generation. The low nuclear case assumes that gas-fired generation replaces nuclear generation.

Sources: 1990: Derived from Energy Information Administration (EIA), International Energy Annual 1996, DOE/ EIA-0219(96) (Washington, DC, February 1998). Projections: EIA, World Energy Projection System (1999).

Natural gas is considered the most likely substitute for the lost nuclear capacity because it is the cleanest of the fossil fuels and would minimize the impact on carbon emissions. If coal were substituted for the lost nuclear generation, an additional 204 million metric tons of carbon would be emitted in 2020—88 million metric tons more than would be emitted if gas were used.

Despite projections of decreasing energy intensities and relatively stable carbon intensities, total carbon emissions from the industrialized nations are projected to grow (Figure 98), with GDP growth rates exceeding the rates of reduction in energy intensity. Growth is expected to be particularly strong in North America, where relatively robust economic growth and flat carbon intensity more than offset reductions in energy intensity. Carbon emissions are projected to increase in North America by about 47 percent in the 30-year period between 1990 and 2020. Smaller increases are projected for Western Europe and industrialized Asia—19 percent and 32 percent, respectively. For the industrialized nations as a group, GDP growth is expected to be the most significant factor underlying the increase in carbon emissions (Figure 99).

Figure 98. Energy-Related Carbon Emissions in Industrialized Countries by Region, 1970-2020

Sources: History: Energy Information Administration (EIA), Office of Energy Markets and End Use, International Statistics Database and International Energy Annual 1996, DOE/ EIA-0219(96) (Washington, DC, February 1998). Projections: EIA, World Energy Projection System (1999).

Figure 99. Changes in Key Factors Affecting Energy-Related Carbon Emissions in Industrialized Nations, 1970-2020

For the United States, the National Energy Modeling System (NEMS) has been used to analyze the impacts of the Kyoto Protocol [2]. In order to address the uncertainties of international carbon mitigation activities, offsets from other greenhouse gases, and carbon-absorbing sinks, several scenarios with different targets for energy-related carbon emissions were represented (Table 21). In the most stringent case, where all emissions reductions would be accomplished domestically, the Kyoto target reduction would be 7 percent below the 1990 level (1990-7%), reaching 1,243 million metric tons of carbon in 2010—548 million metric tons or 31 percent below the baseline projection of 1,791 million metric tons in 2010.²¹ Total U.S. energy consumption in the 1990-7% case would be reduced by 17.5 percent from the reference case. Almost half the carbon reduction would be accounted for by changes in carbon intensity and about 40 percent by changes in energy intensity.

To achieve such reductions, a carbon price would be applied to the cost of energy produced from each energy fuel according to its carbon content. The cost of all energy would rise, but the cost of energy produced from high-carbon sources would rise the most. It is assumed that the government would competitively auction permits and recycle the revenue to citizens. The cost to the economy would not be in the permits per se but in the adjustment to the price of energy relative to other inputs such as capital and labor. This price represents the marginal cost of achieving the specified reduction. In the most stringent case the carbon price in 2010 is estimated to peak at $348 per metric ton (in 1996 dollars), causing retail energy prices to increase relative to the reference case by 53 percent for gasoline, 86 percent for electricity, and 110 percent for residential natural gas.

Of the fossil fuels, coal use and related emissions would decline the most in the 1990-7% carbon reduction case and continue to fall (Figure 100). Petroleum use would drop initially but recover somewhat over time. Natural gas use would increase and stay above the baseline, raising the emissions attributable to natural gas above the baseline projection. Overall, carbon emissions per capita would return to a level last seen in the United States in about 1950 [6] (Figure 101).

Figure 100. U.S. Energy-Related Carbon Emissions by Fuel in Three Cases, 1950-2020

Sources: History: Energy Information Administration (EIA), Emissions of Greenhouse Gases in the United States 1997, DOE/EIA-0573(97) (Washington, DC, October 1998). Projections: EIA, Impacts of the Kyoto Protocol on U.S. Energy Markets and Economic Activity, SR/OIAF/98-03 (Washington, DC, October 1998).

Other, less stringent cases in the analysis measured a range of possible reductions between the most stringent case and the reference or baseline case. Those cases were based on various assumptions about levels of international permit trading, credits for carbon sinks, and offsets from other greenhouse gases. In the case with the least stringent target (1990+24%), U.S. carbon emissions would be reduced by 123 million metric tons from the baseline. In a mid-range case (1990+9%), emissions would be reduced by 329 million metric tons. The emissions targets in these two scenarios would be 425 million metric tons and 219 million metric tons higher, respectively, than the target of 1,243 million metric tons that is achieved in the 1990-7% case.²² In these cases, the cost of energy and the impacts on the U.S. economy would be substantially less than in the most stringent carbon reduction case.

The additional reductions needed to meet the Kyoto target could be acquired in the form of carbon credits or emissions permits projected to be available for purchase from the transitional Annex I countries. If, as projected, 374 million metric tons were available on the international permit market, there would still be 155 million metric tons available for other countries after the United States had acquired the amount required to meet its goal. If, however, the United States needed to acquire permits for 415 million metric tons, an additional 41 million metric tons (above the 374 available from transitional Annex I countries) would have to be obtained by additional trading of permits and the “clean development mechanism”²³ in partnership with developing countries, as well as carbon sinks and offsets from other greenhouse gases.

The potential impacts on economic growth in the various Kyoto Protocol analysis cases are an important consideration. Even in the most stringent case, 1990-7%, GDP is projected to continue growing between 1996 and 2010, but the projected increase is less than the increase expected in the reference case. Two factors were considered in evaluating the impacts of carbon reduction efforts on economic growth: (1) loss in potential GDP and (2) macroeconomic adjustment cost.

The loss in potential GDP is a measure of the opportunities that would be forgone in the process of reconfiguring the inputs to the economy (capital, labor, energy, and materials) to a new set of relative prices. The EIA analysis estimated the loss in potential U.S. GDP to be from $13 billion to $72 billion (1992 dollars) in 2010, or between 0.1 percent and 0.8 percent of the total.

Because the adjustment to the economy would not happen instantly, there would also be a transition or macroeconomic adjustment cost that would reduce actual GDP. For example, workers in industries that produce high levels of carbon emissions might be laid off, potentially either becoming unemployed or taking lower- paying jobs during the transition period. As a result, actual GDP would decrease.

The macroeconomic adjustment period and resulting costs could be greatly affected by government policy. The EIA analysis assumed the introduction of a carbon permit trading system, in the form of an auction run by the Federal Government (to focus on the most economically efficient means of reducing carbon emissions). It was assumed that revenues from the permit auction would be recycled to the economy and would, as a result, counteract the adverse short-term effects of higher energy prices. Revenue recycling in a way that encouraged greater investment—for example, through a reduction in social security taxes—would be more effective in reducing the transition costs than would a revenue recycling approach that provided less stimulus for investment, such as a reduction in personal income taxes. Greater investment would allow the necessary adjustments to be made more rapidly.

The EIA analysis estimated that, with revenues recycled through social security taxes, actual GDP losses would range from $96 billion (in the 1990+24% case) to $397 billion (in the 1990-7% case), in the context of an economy that totaled $6,928 billion (1992 dollars) in 1996 and was projected to increase to $9,429 billion in 2010 in the reference case.

If Western Europe and industrialized Asia are to achieve Kyoto target reductions without permit trading or other offsets, they will have to make similar changes in either energy intensity or carbon intensity (or a combination of both). For Western Europe to achieve all its emissions reduction (160 million metric tons) through a reduction in carbon intensity would require a decrease from 13.69 to 11.56 million metric tons per quadrillion Btu, or about 16 percent below the projected baseline. This would be a difficult goal to achieve. Using a simple calculation for achieving the necessary reduction in emissions, electricity generation from nonfossil fuels would have to increase by 12 percent above the baseline, oil demand would have to decrease by 50 percent below the baseline, coal use would have to be eliminated, and natural gas use would have to increase by 102 percent above the baseline, all within about a decade’s time.

If Western Europe achieved its reduction target solely on the basis of cuts in energy intensity, then the resulting ratio would be 5.39 rather than 6.39 thousand Btu per dollar of GDP or again about 16 percent below the projected baseline. This would mean total energy demand of about 63 quadrillion Btu in 2010 as opposed to a baseline of about 75. Such a decline would require a decrease in energy intensity of 27 percent between 2000 and 2010, which would be considerably greater than the 15 percent decline recorded between 1980 and 1990.

For industrialized Asia, if energy intensity were held constant, carbon intensity would have to decline from 14.08 to 11.49 million metric tons of carbon per quadrillion Btu to achieve the Kyoto goal (a reduction of 81 million metric tons). If carbon intensity were held constant, energy intensity would have to drop to 4.54 thousand Btu per dollar of GDP (18 percent below the baseline), reducing total demand by 25 quadrillion Btu, to approximately the 1994 level of energy demand.

Because both regions have relatively low energy intensities, it is likely that targeted reductions would be achieved through changes in carbon intensities rather than energy intensities. Industrialized Asia may have a particularly difficult time reaching the Kyoto goal. The region already has achieved below-average energy intensity, and because of logistics and economics, reducing carbon intensity through greater use of natural gas may be difficult. Emissions trading may be the key to achieving reductions with minimal adverse economic impacts [7].