mid-1980s, as a result of relatively stable projected energy prices and increased use of electricitybased energy services. In the AEO98 projections, the demand for energy services increases markedly over current levels; however, energy consumption per capita will remain essentially stable throughout the forecast, due to efficiency improvements in many end-use services. The average home in 2010 is expected to be 3 percent larger and to rely more heavily on electricitybased technologies, but energy demand in the residential and commercial sectors is projected to grow at about the same rate as population due to efficiency improvements. In the industrial sector, efficiency gains coupled with a shift toward less energy-intensive industries cause energy demand to grow more slowly than GDP. Annual highway travel and air travel per capita in 2010 are expected to be 13 and 65 percent higher, respectively, than their 1996 levels. Demand for energy in the transportation sector grows most rapidly, driven by estimates of increased per capita travel and slow fuel efficiency gains. The average efficiency of the light-duty fleet increases from 20.2 miles per gallon in 1996 to 20.3 miles per gallon in 2010. With the assumption that light-duty vehicle efficiency standards remain at current levels, the projected low fuel prices and higher disposable personal income increase the demand for more powerful and larger vehicles. Between 1996 and 2010, the average horsepower increases from 168 to 234 for new cars and from 188 to 263 for new light trucks. Low fuel prices also limit efficiency gains for freight trucks and aircraft. On a fuel basis, natural gas has the most rapid increase in consumption, increasing by 31 percent in 2010. Demand for natural gas increases in all sectors, but the most rapid growth is for electricity generation, where consumption more than doubles by 2010, relative to 1996. Although the share of coal generation declines, it accounts for about half of the electricity generation throughout the forecast period. Total coal consumption increases by 15 percent between 1996 and 2010, primarily due to its use for generation. Petroleum continues to have the largest share of energy consumption and is projected to increase by 23 percent between 1996 and 2010. In 2010, 70 percent of petroleum use is in the transportation sector, up from 66 percent in 1996. Increases in light-duty vehicle miles traveled more than offset the increases in vehicle efficiency throughout the projection period. Continued economic growth also increases petroleum use for air and freight travel and shipping over the forecast horizon. Renewable fuel use increases 7 percent between 1996 and 2010. Electricity generation accounts for about 60 percent of the renewable fuel use and the rest is for dispersed heating and cooling, industrial uses, and blending in vehicle fuels. The growth of renewables is limited by the relatively slow increases in the prices of petroleum and natural gas and the decreasing price of coal and also by the assumptions concerning a more competitive electricity industry that weigh against more capital-intensive projects, such as coal and baseload renewables. Nuclear power for electricity generation declines significantly in the projections. Of the 101 gigawatts of nuclear capacity available in 1996, 20 gigawatts, or 28 plants, are assumed to be retired by 2010, with no new nuclear plants constructed throughout the forecast horizon. In AEO98, carbon emissions from energy combustion are expected to reach 1,803 million metric tons in 2010 (Figure 2), 34 percent above the 1990 level of 1,346 million metric tons. Total emissions increase at an average annual rate of 1.5 percent through 2010, and per capita emissions also increase at an average rate of 0.6 percent, as continued economic growth and moderate prices encourage growth in energy services and energy consumption while electricity generation from nuclear power plants declines, particularly after 2010. Electricity generation accounts for 37 percent of all carbon emissions in 2010, increasing from 35 percent in 1996. The growth in carbon emissions from generation is mitigated somewhat by the increasing share of natural gas generation, but nuclear generation, which is carbon free, declines significantly. This Figure 2 U.S. Carbon Emissions by Sector, loss of nuclear generation, as well as the additional generation needed to meet demand growth, is in general met by new gas- or coal-fired generation. Because of increased travel and slow growth in fuel efficiency, energy consumption and emissions for transportation grow the fastest of all end-use sectors, accounting for 35 percent of emissions by 2010. In the projections, per capita vehicle-miles traveled in light-duty vehicles continues to increase. Increased travel is not offset by improved fleet efficiency because the average efficiency of light-duty vehicles grows slowly. At the same time, light-duty trucks, vans, and sport utility vehicles, which are less efficient, increase their market share. Growth in air travel also contributes to the increase in transportation energy use and emissions. The residential and commercial sectors contribute 19 and 15 percent, respectively, of the carbon emissions in 2010, including the emissions from the generation of electricity used in those sectors, dampened somewhat by efficiency improvements. In both sectors, continuing growth in electricity use for new appliances and energy services contributes to increasing consumption and emissions. Although the industrial sector contributes 31 percent of the emissions in 2010, ! industrial emissions grow the slowest of all the end-use sectors, as shifts to less energy-intensive industries and efficiency gains moderate the growth in energy use. Factors That Affect Carbon Emissions Levels There are a number of uncertainties that could alter the baseline projections for energy markets in the future. If total energy demand or the fuel mix were different in 2010 than projected in AEO98, then the level of carbon emissions would also be different. AEO98 explored a number of alternative energy futures by varying key assumptions in the baseline projections. Several of these alternative cases included revised assumptions for end-use technologies, economic growth, renewable technologies, renewable portfolio standards, and nuclear retirements, all of which directly impacted the projections of carbon emissions. Impact of End-Use Technology Improvements. In keeping with the general practice of considering only current policy and regulations, the AEO98 reference case assumes no new efficiency standards or improvements in standards beyond those already approved. Thus, AEO98 includes the new refrigerator and room air conditioner standards for 2001 approved in 1997 but no additional standards. Energy intensity-energy consumption per dollar of GDP—is projected to decline by an average of 0.9 percent annually between 1996 and 2010. Frozen standards, the modest increases in the prices of petroleum and natural gas coupled with declining prices for electricity and coal in the reference case, and growing demand for energy services, such as appliances, office equipment, and travel, slow further declines in energy intensity over the projection period. The projected decline in energy intensity is significant but considerably less than the decline in the 1970s and early 1980s, which averaged 2.3 percent a year between 1970 and 1986. Approximately half the decline in energy intensity in that period resulted from structural shifts in the economy, such as shifts to service industries and other less energyintensive industries; however, the other half of the decline was due to the use of more energyefficient technologies during periods of rapid increases in energy prices. As the growth in energy prices moderated in the later part of the 1980s, growth in some energy-intensive industries resumed, which moderated the decline in energy intensity. The AEO98 reference case includes continued improvement in technologies for both energy consumption and production-for example, improvements in building shell efficiencies for both new and existing buildings; efficiency improvements for new appliances; productivity improvements for coal production; and improvements in the exploration and development costs, finding rates, and success rates for oil and gas production. Technology improvements could reduce energy consumption, and therefore energy-related carbon emissions, below that in the reference case. Conversely, slower improvement than that assumed in the reference case could increase both energy consumption and carbon emissions. AEO98 presents a range of alternative cases that vary key assumptions concerning technology improvement and penetration. A high end-use technology case is defined by assuming that the |