Background

Analysis Request

On July 30, 2003, Senator James M. Inhofe requested “the Energy Information Administration to undertake analyses of S.843, The Clean Air Planning Act of 2003, introduced by Senator Thomas Carper, and S.485, Clear Skies Act of 2003.”¹ Senator Inhofe specifically asked the Energy Information Administration (EIA) to address the impact on sulfur dioxide, nitrogen oxide, mercury, and carbon dioxide emissions nationally and regionally, the marginal costs of reducing each emission, the amount of emissions control equipment needed, and the costs and electricity price impacts of each bill. Senator Inhofe also asked EIA to analyze S. 485 with and without the mercury provisions and S. 843 with and without the mercury and carbon dioxide provisions.

Bill Summary

Both bills require reductions in the emissions of sulfur dioxide (SO2), nitrogen oxides (NOx), and mercury (Hg) from electricity generating plants.² In addition, S. 843 (hereafter referred to as the Carper bill) also calls for reductions in power sector emissions of carbon dioxide (CO2). With respect to SO2, NOx, and Hg, the emissions caps and reduction timetables differ, but both bills generally call for multi-phase, cap-and-trade emission reduction programs (Table 1 and Figures 1 through 4) covering electricity-generating facilities larger than 25 megawatts. For SO2 and Hg the emissions caps take effect earlier and end up more stringent in the Carper bill than they do under Clear Skies. For NOx the final caps are the same, but the timetable of the emissions caps differs. Relative to 2000 emission levels, Clear Skies calls for reducing SO2 emissions by 73 percent while the Carper bill calls for an 80-percent reduction. For NOx both the Clear Skies and Carper bills call for a 67-percent reduction from the 2000 emission level. For mercury, Clear Skies calls for a 70-percent reduction from the 2000 level while the Carper bill calls for an 80-percent reduction. The Carper bill calls for reducing CO2 emissions from electricity generating plants in 2009 to the level projected in EIA’s Reference case for 2006 and further reducing them to the actual 2001 level by 2013. Relative to EIA’s projected CO2 emissions from electricity generators in the Reference case, the 2013 target in the Carper bill would require a 24-percent reduction in 2020 and a 30-percent reduction in 2025. However, the Carper bill allows generators to comply with the CO2 target using allowances from other domestic or international greenhouse gas trading programs or by investing in projects that reduce greenhouse gas emissions or increase sequestration.

Both bills rely primarily on emissions cap-and-trade programs to meet their specified emission targets. Under such programs, each power plant must annually submit an allowance for each unit (i.e., tons, metric tons, pounds, or ounces) of emissions. Under such programs, market forces will determine allowance prices, and each covered entity is free to determine its optimal compliance strategy. They can choose to reduce their emissions or purchase allowances from others who have reduced their emissions below the level of allowances they hold. They can also choose to overcomply in an earlier year and to use those allowances in a future period, i.e., bank allowances.

Besides differences in the timing and stringency of the emissions caps, there are several other important differences between the two bills. These include:

• Allowance price safety valves and excess emissions penalties. Clear Skies sets excess emissions penalties to the most recent auction price for each emission each year. It also sets a safety valve for each emission. Facilities can purchase allowances from the government at these safety valve prices if they are not available in the market at lower prices. The safety valve is $4,000 per ton for SO2 and NOx and $2,187.50 per ounce ($35,000 per pound) for mercury. The safety valve puts a limit on the respective allowance prices and, if utilized, will cause the
  emission targets to be exceeded.³ The Carper bill does not set safety valves, but imposes excess emissions penalties: $2,000 (1990 dollars) per ton for SO2, $5,000 per ton for NOx, $10,000 per pound for mercury, and $100 per ton for CO2 (penalty fees are to be adjusted for inflation). In addition, excess emissions must be made up in the following year or a period of time prescribed by the Administrator of the Environmental Protection Agency (EPA).
• Facility-specific mercury limits. The Carper bill requires that all coal facilities either remove a minimum percentage (50 percent between 2009 and 2012, and 70 percent in 2013 and later) of the mercury in the coal burned or meet an outputbased rate to be set by the EPA Administrator.⁴ The efforts taken to comply with the requirement to remove a certain percentage of the mercury in the coal reduce the additional efforts needed to meet the overall emissions cap and will lead to lower allowance prices but higher industry cost than would occur with only a capand-trade program.
• Output-based standards for older plants. Beginning in 2020, the Carper bill requires that plants that began construction before August 17, 1971, must emit no more than 4.5 pounds per megawatthour of SO2 and 2.5 pounds per megawatthour of NOx. This provision is not explicitly modeled in this analysis, but because of the relatively stringent limits on national NOx and SO2 emissions that will be in place by 2020 in the Carper bill, most plants are expected to comply with these limits.

Allowance programs

• Clear Skies generally allocates NOx, SO2, and Hg allowances to existing units based on historical heat input. This is often referred to as“grandfathering” since the allocation is based on historical fuel use. Over time the allocation gradually shifts to an auction with the auction revenue going to the government.⁵ Allowances are not allocated to new units.
• For SO2, the Carper bill also allocates allowances using a grandfathering approach, while for NOx, Hg, and CO2, allowances are allocated on an output basis (i.e., pounds per megawatthour of electricity produced) that is continually updated based on the most recent three years of each facility’s generation. Essentially this is a rolling three-year generation performance standard (GPS) for NOx, Hg, and CO2. Under the Carper bill, allowances are also allocated to new units until they have operated for three years and become part of the regular GPS program.⁶ The GPS programs in the Carper bill will impact the cost and price impacts of meeting the emission targets. In general, a dynamic GPS, which is updated continuously as each facility’s generation changes, provides an incentive to facilities to increase their output so that they receive more allowances in the future. This “output subsidy” lowers the electricity price impacts of reducing emissions, but increases the cost impacts.7 As one expert said, “output based rebating sacrifices some of the efficiencies of market-based environmental policies. Allocating by market share essentially provides a subsidy to output, which creates a bias away from output substitution and toward emissions rate reduction. The result is a higher marginal cost of control, a lower equilibrium output price, and a greater cost of achieving any given level of emissions reduction, compared to an efficient policy. The size of the welfare loss from this distortion depends on how much emissions reduction would normally be performed by output substitution.”8 In layman’s terms this means if facilities are given allowances based on their output (generation) they will tend to produce more than they otherwise would have.
• The output subsidy associated with a GPS derives from its impact on covered generators’ operating costs. For example, a typical coal plant produces approximately 0.25 metric tons of carbon per megawatthour. As a result, a $100 carbon fee would raise its operating cost by $25 per megawatthour. However, under a GPS, the plant will be allocated some allowances for each megawatthour it generates. If it is assumed that the GPS is 0.15 metric tons of carbon per megawatthour, calculated by dividing the CO2 emissions cap by the generation of all covered plants, the impact on the coal plant’s operating costs of a $100 carbon fee is only $10 per megawatthour ((0.25 – 0.15) X $100). If this plant were setting the market-clearing price of electricity, consumers would face a smaller price increase under the GPS, $10 per megawatthour rather than $25 per megawatthour, and have less incentive to reduce their use of electricity. This would lead to greater generation (output) from the power sector under a GPS allocation program, than under a grandfathering allowance program.
• The Carper bill establishes an independent review board to certify projects outside of the U.S. power sector as eligible for additional CO2 allowances. It also allows the use of allowances from recognized international CO2 trading programs. Electricity facilities are able to use these allowances from certified projects as well as allowances from other U.S. or recognized international CO2 trading programs (all referred to as offsets in this report) to meet their CO2 targets rather than directly reducing their own emissions. In addition to existing fossil generators, new fossil fuel and renewable units receive CO2 allowances.⁹
• To analyze the availability and cost of greenhouse gas offsets, this analysis incorporates a set of curves representing the potential for other greenhouse reductions and sequestration. These curves, referred to as marginal abatement curves (MACs), were obtained from EPA’s Office of Air and Radiation. Essentially, MACs are simplified, reduced-form representations of emissions compliance potential as a function of a single variable, the allowance price. Because there is great uncertainty in developing these MACs, a range of results is provided based on alternative assumptions.¹⁰

Analysis

EIA analyzed the bills using the National Energy Modeling System (NEMS). The Reference case for the analysis was based on EIA’s Annual Energy Outlook 2003 (AEO2003).¹¹ It was updated in June 2003 to reflect changes in electric generating capacity since the AEO2003 was completed; to incorporate revised expectations about near-term trends in natural gas prices; to incorporate revised mercury emissions factors; and to reflect recent changes in corporate average fuel economy (CAFE) standards. In addition potential CO2 offsets from reductions in other greenhouse gases and sequestration projects were reviewed.

It should be noted that the projections in the cases in this report are not statements of what will happen but of what might happen, given the assumptions and methodologies used. The Reference case projections are business-as-usual trend forecasts, given known technology, technological and demographic trends, and current laws and regulations. Thus, they provide a policy-neutral Reference case that can be used to analyze policy initiatives. EIA does not propose, advocate, or speculate on future legislative and regulatory changes. All laws are assumed to remain as currently enacted; however, the impacts of planned regulatory changes, when defined, are reflected. In addition to the uncertainties inherent in the Reference case projection itself, there are several important uncertainties in evaluating the bills. Of particular concern in this analysis are the cost and performance of technologies to remove mercury and the availability and cost of greenhouse gas offsets.

In order to respond to the requests from Senator Inhofe, the following cases were analyzed for the Clear Skies and Carper bills.

Reference

• Clear Skies Bill Cases

• Clear Skies 3-P - NOx, SO2, and Hg caps with the safety valves

Clear Skies bill sensitivity cases requested by Senator Inhofe

• Clear Skies 2-P - NOx and SO2 caps only

• Clear Skies 3-P No Mercury Safety Valve - NOx, SO2, and Hg caps without the mercury safety valve.

• Carper Bill Cases

• Carper 4-P High Offset - NOx, SO2, Hg, and CO2 caps with high greenhouse gas offsets. Assumes that the independent review board allows electricity generators to use certified allowances from 1) Annex 1 countries, 2) projects that reduce the emissions of other greenhouse gases in the United States, and 3) U.S. and international sequestration projects.¹²

• Carper 4-P Mid Offset - NOx, SO2, Hg, and CO2 caps with mid greenhouse offsets. Assumes that the independent review board allows electricity generators to use certified projects that reduce the emissions of other greenhouse gases in the United States.

• Carper 4-P No Offset - NOx, SO2, Hg, and CO2 caps without any CO2 offsets. Assumes that electricity generators must directly reduce their emissions to the targets in the Carper bill.

Carper bill sensitivity cases requested by Senator Inhofe

• Carper 2-P - NOx and SO2 caps only

• Carper 3-P - NOx, SO2, and Hg caps only

The assumptions about offsets in the three Carper 4-P cases are not meant to be predictions about how the independent review board established in the Carper bill might act. Rather they provide a range of results regarding the availability and cost of offsets, which are highly uncertain. They should be seen as representing the uncertainty about the potential availability and cost of offsets.

Generation and Fuel Use

Coal

Power sector efforts to reduce NOx, SO2, and Hg emissions are projected to lead to lower coal generation and increased generation from natural gas. For example, if the NOx and SO2 provisions of Clear Skies were imposed, coal generation in 2010 is projected to be 2.4 percent lower than it otherwise would have been (Figure 5). By 2020, this difference is projected to be 5.6 percent. When the Hg cap with the safety valve is also imposed, the change in coal generation is projected to be slightly larger, falling to 6.4 percent below Reference case projections in 2020 and 7.4 below the Reference case level when the Hg safety valve is removed. However, even with these changes, coal generation in 2020 and 2025 is projected to be well above current levels with or without Clear Skies.

The reduction in coal use in the Carper 2-P and 3-P cases is projected to be slightly smaller than in the comparable Clear Skies cases (Figure 6). Even though the SO2 and Hg emissions caps in Carper are more stringent, the GPS allowance allocation scheme and the minimum facility-level removal requirements for mercury dampen the impact on coal that would otherwise be seen. As discussed in the background section, the GPS provides an output subsidy that leads facilities to rely more on emissions control technologies rather than reducing their output to comply with the emissions limits. In the Carper 3-P case, coal generation is actually projected to be slightly higher than in the Carper 2-P case because the allocation of mercury allowances to new coal plants, which are assumed to remove 90 percent of the mercury in the coal they use, make them more economically attractive. Overall, coal generation in 2020 is projected to be 4.9 percent below the Reference case in the Carper 2-P case and 4.5 percent below in the Carper 3-P case.

The projected change in coal generation could be much larger in the Carper 4-P cases, but it is very sensitive to the availability and cost of CO2 offsets. There is significant uncertainty about the potential price of CO2 offsets. There is also uncertainty about the requirements the independent review board created in the Carper bill might establish before a project can be certified to receive additional CO2 allowances and what might be required regarding the use of international programs. Across the three Carper 4-P cases, coal generation in 2020 is projected to be between 12 percent and 32 percent below the Reference case level. The low impact occurs if CO2 offsets are readily available with their price growing from $4 per metric ton carbon equivalent (2001 dollars) in 2010 to $26 per metric ton in 2025. Conversely, if the U.S. power sector can not rely on offsets, the impact would be much larger, with the CO2 allowance price growing from $66 per metric ton carbon equivalent in 2010 to $135 per metric ton in 2025.

In aggregate, the changes in coal production are expected to parallel the changes in coal generation in the Clear Skies cases. However, regional coal production is expected to react differently. In the Clear Skies 2-P case, western coal production is projected to be 4.5 percent higher in 2020 than in the Reference case because of the tighter SO2 cap, which makes low-sulfur subbituminous western coal more attractive (Figure 7). In the Clear Skies 3-P cases, particularly the one without the mercury safety valve, the pattern reverses with western coal production falling below Reference case levels. The imposition of a mercury cap makes western coal less attractive because it is more difficult to remove mercury from the lower rank (subbituminous and lignite) coals. In the Clear Skies 3-P case, western coal production is projected to be 7.4 percent below the Reference case level in 2020. This change widens to 16.6 percent in the Clear Skies 3-P case without the mercury safety valve. However, even with these changes, western coal production is projected to increase from current levels in all Clear Skies cases.

The projected changes in eastern coal production under Clear Skies are nearly the mirror opposite of those for western coal. In the Clear Skies 2-P case, the production of eastern bituminous coal is projected to be 15.8 percent below the Reference case level in 2020 as power plants switch to low-sulfur western coal to comply with the tightening SO2 emissions cap (Figure 8). However, in the Clear Skies 3-P cases, particularly when the mercury safety valve is removed, the production of eastern bituminous coal is projected to be above the level seen in the Clear Skies 2-P case. Mercury is generally easier to remove from bituminous coal, so the imposition of a mercury cap makes such coal more
economic. In 2020, eastern coal production is projected to be only 4.3 percent below the Reference case level in the Clear Skies 3-P case. In the Clear Skies 3-P case without the mercury safety valve, it is projected to be 2.9 percent above the Reference case level in 2020.

As was the case with coal generation, the largest changes in projected coal production are seen in the Carper 4-P cases, particularly those with less offsets and higher CO2 allowance prices (Figure 9). Because of its high carbon content relative to other fuels, coal generation and production are projected to be very sensitive to the CO2 allowance price. For example, on a Btu basis, natural gas contains less than 60 percent as much carbon as coal does. In the Carper 4-P High Offset case, where power generators are assumed to be able to buy offsets from 1) Annex 1 countries, 2) projects that reduce the emissions of other greenhouse gases in the United States, and 3) U.S. and international sequestration projects (up to the limits of the Marrakech accords)¹³, the CO2 allowance price is projected to remain fairly low, reaching $26 per metric ton carbon equivalent in 2025, and coal production is projected to be 12 percent below the Reference case level in 2020 and 15 percent below the Reference case level in 2025. However, the impact on coal is much larger if CO2 allowance prices are higher. In the Carper 4-P Mid Offset and Carper 4-P No Offset cases, coal production in 2020 ranges between 16 percent and 30 percent below the Reference case level.

In the Carper 4-P Mid Offset and Carper 4-P No Offset cases, the impacts on employment in the U.S. coal industry are significant but less severe than the projected declines in production would suggest. Relative to the Reference case, the largest production cuts in these two cases are projected to occur in the western coalfields, which are considerably less labor-intensive than eastern operations. In the Carper 4-P Mid Offset and Carper 4-P No Offset cases, coal mine employment in 2025 is projected to be 10 percent and 27 percent less, respectively, than in the Reference case forecast. However, while there are coal industry job losses in these cases, increased employment in the natural gas and renewable fuels industries will at least partially compensate for the coal industry job loss.

Natural Gas

In 2-P and 3-P cases, the impacts on natural gas generation and fuel use are projected to be nearly the opposite of those for coal. Imposing limits on power sector NOx, SO2, and Hg emissions is projected to lead to an increase in natural gas use in the power sector (Figure 10). Under Clear Skies, natural gas generation in 2020 is projected to 8.1 percent above the Reference case level in the 2-P case. The increase is 9.2 percent in the Clear Skies 3-P case but grows to 10.4 percent in the Clear Skies 3-P case without the mercury safety valve. The increase in gas use is projected to lead to higher natural gas imports, both liquefied natural gas (LNG) and pipeline imports from Canada. For example, in the Reference case, natural gas imports are projected to reach 6.8 trillion cubic feet in 2020, while in the Clear Skies 3-P case, they reach 7.6 trillion cubic feet.

Natural gas is projected to see an even larger change if a CO2 cap is imposed. As with coal use, the increase in gas generation and fuel consumption that occurs in the Carper 4-P cases is sensitive to the cost and availability of CO2 offsets. In the Carper 4-P cases, the increase in natural gas generation in 2020 ranges from 18 percent to 24 percent, much larger than in any of the Clear Skies or Carper 2-P and 3-P cases (Figure 11).

Increased natural gas generation is projected to lead to higher natural gas prices, particularly in the later years in the 4-P cases. In the Clear Skies 2-P and 3-P cases, natural gas wellhead prices are projected to show very little change from Reference case levels through 2020. However, in 2025 they are projected to be between 3.4 percent and 4.7 percent higher. In the Carper 4-P cases, the change from the Reference case in 2020 ranges from 1.2 percent to 4.4 percent. By 2025 the change from the Reference case ranges from 4.9 percent to 9.2 percent in the Carper 4-P cases (Figure 12). The difference in the variability in natural gas prices in Figure 12 is due to the timing of the expansion and opening of LNG facilities and the development of the Alaskan Natural Gas Transportation System. As discussed in the next section, in the Carper 4-P Mid and No Offset cases, the increase in natural gas generation relative to the Carper 4-P High Offset case is relatively small because renewable fuels become attractive in these cases.

Renewable Fuels

Besides coal and natural gas, the use of other fuels for electricity is not expected to be significantly impacted in any of the 2-P or 3-P cases. However, in the Carper 4-P cases, particularly the case without carbon offsets, renewable fuel use is expected to be much higher than in other cases (Figure 13). New renewable fuel plants become attractive in the Carper 4-P cases because they are carbon free and they are given CO2 allowances that can be sold to others who need them. In 2020, renewable generation in the Carper 4-P Mid Offset case is projected to be 19 percent above the Reference case level. In the Carper 4-P No Offset case, the difference is even larger in 2020, with renewable generation 89 percent above the Reference case level. The renewable fuels expected to play the largest role in the higher generation are biomass and wind.

Notes