from the 1993 CCAP analysis for each major greenhouse gas and source category.

Energy-Related Actions

The projected decrease in natural gas prices and increased electricity sales compared to the 1993 CCAP have increased the projected market share for new natural gasfired electric-generating capacity. Although the reduction in projected natural gas prices is beneficial from a climate change perspective because natural gas is a less carbon-intensive fuel per unit of energy than other fossil fuels and because natural gas technologies tend to be more efficient, it reduces the efficacy of climate change policies designed to reduce electricity use.

Many of the 1993 CCAP actions reduce carbon emissions through their impact on electricity demand. Changes in the marginal fuel used for generation has important implications for translating electricity savings into carbon reductions. Marginal fuel is the fuel consumed to produce the last "unit" of electricity generated. In this instance, the unit is defined as the kilowatt-hour savings from electricity-related actions.

In the original 1993 CCAP, the marginal fuel mix for electricity production was 80 percent coal and 20 percent oil and natural gas in the year 2000. This resulted in carbon emissions decreasing by 0.28 MMTCE for every decrease in 1 billion kilowatt-hours of electricity (0.28 MMTCE/bkWh). In the cur

rent estimates, due to an increased market share of natural gas-fired generation, coal accounts for 32 percent and natural gas and oil for the remaining 68 percent of marginal electricity production in the year 2000. As natural gas garners a greater market share of electricity production, the marginal carbon savings per unit of electricity reduced becomes smaller. In the 1997 CAR, the coefficient of reduction decreases from 0.20 MMTCE/bkWh in 2000 to 0.13 MMTCE bkWh by 2020. This change decreases carbon emission reductions by electricity-saving actions by 10 MMTCE in 2000.

Overall, due to funding shortfalls and other factors, carbon emission reductions from energy-related actions have decreased by 34 MMTCE in 2000 compared to the 1993 CCAP. However, after 2000, 1997 CAR-projected reductions are larger than those envisioned in the 1993 CCAP for 2000.

Methane

Overall, carbon-equivalent emission reductions of 16 MMTCE from methanerelated actions in 2000 are about the same as the 1993 CCAP.

HFCs and PFCs

Action 40 (Narrowing the use of High GWP Chemicals) is being expanded to form partnerships with newly identified sources described in the beginning of this chapter. Increases in the global warming potentials for HFCs and PFCs have lead to increased effectiveness in Action 42 (Voluntary Aluminum Industrial Partnership Program).

Overall, HFC and PFC reductions are about the same as the 1993 CCAP in 2000.

The addition of policies to reduce newly identified gases, such as the Environmental Stewardship Initiative, results in reductions in other gases exceeding the amount claimed in the 1993 CCAP.

Nitrous Oxide

The revised global warming potential for nitrous oxide increases the carbon-equivalent measure of actions taken to reduce nitrous oxide by about 0.8 MMTCE in 2000.

Forest Sinks

Decreases in other federal tree-planting and technical assistance programs for forest landowners increases the potential for participation in this program.

Key Uncertainties Affecting Projected Emissions

Any projection of future emissions, even for a period as short as four years, is subject to considerable uncertainty. Key factors that can increase emissions include more rapid growth in electricity demand, flat rather than slightly rising real energy prices, more rapid economic growth, and further cuts in 1993 CCAP funding or effectiveness. Key factors that can reduce emissions include slower growth, increased CCAP program efficacy, greater penetration of baseline energy-efficiency measures, higher energy prices, increased program funding levels, and relatively mild weather in 2000. A qualitative analysis of key uncertainties suggests that net greenhouse emissions in 2000 could exceed their 1990 level by 150-230 MMTCE.

CCAP Program Funding Levels (+ or -)

The point estimate assumes that CCAP funding through 2000 reflects an extrapolation of fiscal year 1996 funding. Increases or decreases in 1993 CCAP program funding relative to the "current funding" level in fiscal years after 1996 would result in higher or lower levels of projected emis

sions in 2000.

Required Legislative Authority (-)

Included in the estimates of emission reductions are the assumed adoption of policies that require no additional funding, but require some congressional or regulatory action, such as tire-labeling and energyefficiency standards. Many of the actions in this category are still assumed to occur, but their deployment has been adjusted to account for delay in their implementation. If legislative authority to initiate these programs is not received, emissions will be higher than projected.

Energy Prices (+ or -)

The relationship between energy prices and emissions is complex. Lower energy prices generally reduce the incentive for energy conservation. However, reductions in the price of natural gas relative to other fuels also encourages fuel switching that can reduce carbon emissions.

The energy price projections from the 1997 Annual Energy Outlook used in developing the updated emissions baseline are significantly lower than those used for the 1993 CCAP (U.S. DOEЛEIA 1996a). However, real prices for oil and gas are still projected to rise at respective average annual rates of 1.1 percent and 2.5 percent between 1995 and 2000.

The Annual Energy Outlook also provides sensitivity scenarios to changes in oil prices. In the year 2000 high oil price scenario, emissions are lower by about 4 MMTCE than projections using the reference-case scenario oil price assumptions. In the year 2000 low oil price scenario, emissions are higher by about 13 MMTCE not using the reference case scenario.

Economic Growth (+ or -)

Higher economic growth increases the demand for energy services, such as vehicle miles of travel, square feet of lighted and ventilated space, and process heat used in industrial production. However, faster growth also reduces the average age of the capital stock, increasing its average energy efficiency. The energy-service demand and energy-efficiency effects of higher growth work in offsetting directions. The effect on service demand is the stronger of the two, so that levels of primary energy use are positively correlated with the size of the economy.

In addition to the reference case used in developing the updated baseline, the Annual Energy Outlook provides high and low economic growth cases.

In the high-growth case, the per-
centage change increase in energy
use is slightly more than half the per-
centage increase in the size of the
economy. By 2000, the high-growth
economy is 3.5 percent larger than
the reference economy, but energy
consumption is only 1.8 percent
higher. In addition, carbon emissions
are 33 MMTCE larger than the refer-

ence case.

In the low-growth case, a 2.7 percent reduction in the size of the 2000 economy translates into a 1.9 percent reduction in primary energy use. In this case, carbon emissions were 28 MMTCE lower than the reference scenario in the year 2000.

Electricity Demand Growth (+)

While the annual rate of growth in elec. tricity demand from 1995 to 2000 is appreciably higher in the present analysis than in the 1993 CCAP, there is a strong possibility of even faster growth. Regulatory changes to allow competition in wholesale and retail electricity markets could significantly lower prices to electricity end users, while at the same time reduce utility investments in demand-side management and other conservation activities.

If electricity demand grows at 2.1 percent annually (as projected by the Gas Research Institute), rather than by 1.9 per. cent (as projected in this analysis), carbon emissions will be about 7 MMTCE higher in 2000. The Annual Energy Outlook also evaluates a-sensitivity that assumes electricity sales to grow at 3.3 percent annually between 1995 and 2000. In that scenario, emissions are 56 MMTCE higher in the year 2000.

Forest Carbon Sequestration (+ or -)

The estimates used here for annual carbon sequestration in U.S. forests include above-ground carbon plus harvested carbon in wood products and landfills. The tree car. bon estimates are derived from two independent measurements of forest inventories and growth, and have standard errors of plus or minus 3 percent. The reported forest esti

mates do not include sequestration in the forest floor understory complex.

Estimates of other carbon stocks (e.g., forest floor and understory) are likely to be less certain, since there are no comprehen. sive, statistically valid inventories of non-tree organic matter for large areas of the United States. USDA estimates their uncertainty at plus or minus 15 percent.

Additional unquantified sources of uncertainty should also be noted. First, deriving annual stock change estimates from standing stock estimates would increase uncertainty further. Second, estimates projected from historical data using econometric models will be less certain due to the unknown uncertainty of the assumptions made in the econometric models. Estimates for all years after 1992 are projected from 1992 data. Additionally, certain lands have not been included in these stock estimates.

Weather (+ or -)

Energy use for heating and cooling is directly responsive to weather variation. The updated baseline assumes thirty-year average values for population-weighted heating- and cooling-degree days. Figure 4-6, which compares average population-weighted heating and cooling-degree days with actual values for 1990, an unusually mild year, illustrates the importance of interannual weather variation for energy use and emissions. Under average weather conditions, primary energy consumption for heating and cooling in 1990 would have been 1.1 percent higher than its actual value, raising carbon emissions by roughly 16 MMTCE.

Unlike other sources of uncertainty, for which deviations between assumed and

actual trends may become apparent over time, the effect of weather on energy use and emissions in any particular year is revealed only in that year. For the United States, a swing in either direction of the magnitude experienced in 1990 could raise or lower emissions by plus or minus 20 MMTCE relative to a year with average weather. While small relative to total emissions, a change of this magnitude is significant relative to the aim of returning emissions to their 1990 level. Some European countries, which also experienced low levels of energy use and emissions in 1990 due to mild winter weather, have opted to compare 1990 and 2000 emissions levels on a "climate-adjusted" basis in their first national communications.