5. Electricity Supply

Background

Historically, the electricity supply sector has used a diverse mix of fuels to meet consumers’ electricity needs (Figure 5.1). The fuels used include coal, oil, natural gas, nuclear, hydroelectric, wood, waste, geothermal, solar, and wind. In the early 20th century the industry began with small hydroelectric facilities built to provide electricity for city lights. As the uses of and demand for electricity grew, the industry increasingly turned to fossil fuels—coal, oil, and natural gas. By 1950 fossil fuels accounted for nearly 70 percent of total U.S. electricity generation, and their share continued to grow, exceeding 82 percent by 1970. Through the 1970s and 1980s the growth of nuclear and hydroelectric power, together with the declining use of oil, reduced the role of fossil fuels in electricity production. By 1990, the share of electricity accounted for by fossil fuels was just under 70 percent. Since 1990, however, almost all the new capacity added has been fueled by natural gas. Even with the increasing output from existing nuclear plants and the growth in some renewable technologies—particularly wind—the share of electricity accounted for by fossil fuels has again started to grow.

Because of its strong dependence on fossil fuels, the electricity supply sector accounted for 39 percent of total U.S. carbon dioxide emissions in 2000.¹⁴⁵ As a result, the imposition of a greenhouse gas emissions limit will affect all aspects of the electricity supply sector. It will affect the choice of fuels used to produce electricity, the types of plants built to meet growing consumer electricity needs, the future price of electricity that consumers will face and their responses to them, and the level of other emissions— sulfur dioxide, nitrogen oxide, and mercury—often associated with electricity production from fossil fuels. As might be expected, numerous uncertainties exist. Key uncertainties for the electricity sector include the role that new technologies might or might not play, and how emission allowances might be treated in electricity pricing in various regions of the country.

Generation by Fuel

Over the next 20 years, without a greenhouse gas emissions cap, the power sector is projected to remain heavily dependent on fossil fuels, particularly coal, and, to a growing degree, natural gas. In the reference case, fossil fuels are projected to account for 76 percent of total generation in 2020 and 78 percent in 2025 (Figure 5.2). The vast majority of new power plants built over the next 20 years are expected to be fueled by natural gas. Relative to other technologies, new natural gas combustion turbine and combined cycle plants are less expensive to build, and their improving efficiencies help to offset the higher cost of natural gas relative to other fuels, such as coal. As the price of natural gas rises over time, new coal plants are projected to become increasingly economical later in the projections. Without a greenhouse gas emissions limit, new plants using non-carbon-based fuels such as renewables and nuclear are not expected to be widely competitive when new generating capacity is needed. New renewable plants, particularly new wind plants, are projected to play a role in some areas, but not enough to increase their share of total generation. Total renewable generation is projected to accounted for 9.8 percent of total generation in 2010, 8.9 percent in 2020 and 8.4 percent in 2025. Between 2000 and 2025 wind capacity, stimulated in part by State and Federal programs, is projected to more than quadruple; however, generation from wind plants still is expected to account for just over 0.5 percent of total generation in 2025.

The role of non-carbon-based fuels in the generation of electricity is projected to change dramatically if a greenhouse gas emissions cap is imposed. In addition, generation from fossil technologies equipped with carbon capture and sequestration equipment is also expected to grow (Table 5.1). As discussed in previous chapters, the electric power sector is projected to account for a large portion of the greenhouse gas emission reductions needed to meet the cap on the covered sectors. To do so, the electric power sector will have to increasingly turn to low- or zero-carbon technologies.

To comply with the greenhouse gas limit the electric power sector is projected to turn away from coal generation towards renewable, natural gas and nuclear generation (Figures 5.3 and 5.4). By 2025, coal generation is projected to be 2,243 billion kilowatthours (80 percent) lower in the S.139 case than in the reference case. In contrast, renewable generation is projected to be 699 billion kilowatthours (143 percent) higher in the S.139 case than in the reference case in 2025. Natural gas (695 billion kilowatthours and 42 percent) and nuclear (393 billion kilowatthours and 50 percent) generation are also projected to be higher in the S.139 case in 2025, while overall electricity demand is lower by 593 billion kilowatthours (11 percent) as consumers reduce their use of electricity. The imposition of a greenhouse gas emission cap simply makes it uneconomical to continue using coal in existing plants without carbon capture and sequestration equipment. For example, in 2025 in the S.139 case, the delivered price of coal to the power sector is projected to be $0.90 per million Btu, $0.21 less than the $1.11 price projected in the reference case. However, the effective cost of coal to the power sector in the S.139 case is projected to be $6.53 per million Btu - $0.90 per million Btu for the coal plus the $5.62 per million Btu for the allowances needed to use it. In effect, coal costs in the S.139 case in 2025 are 488 percent higher than the $1.11 per million Btu projected in the reference case.

In the S.139 case, renewable power sources, especially biomass, wind, and geothermal energy resources, are projected to play an increasing role in meeting the growing demand for electricity while also helping to reduce greenhouse gas emissions. Renewable energy resources are generally considered to be net zero emitters of carbon. Electric generation from most renewable resources, such as wind, solar, hydroelectric, or geothermal, involves no direct emissions of carbon dioxide or other gaseous carbon compounds. On the other hand, the use of biomass and other organic materials does produce direct carbon emissions when electricity is produced. However, biomass fuel sources, such as agricultural wastes, urban wastes, or dedicated energy crops, all fix atmospheric carbon on a short enough time scale (years or at most a few decades) to be effectively considered net zero emitters of carbon in the mid- to long term. In other words, while carbon is released when biomass is burned, almost the same amount of carbon is “captured” when the biomass products—agricultural wastes, urban wastes, or dedicated energy crops—are grown, resulting in near zero net emissions over time.

Among the renewables, the largest response to the greenhouse gas emissions cap is projected to come from biomass, wind, and, to a lesser extent, geothermal (Table 5.2 and Figure 5.5). For example, in 2025, biomass, wind and geothermal generation in the S.139 case are projected to be 406, 248, and 39 billion kilowatthours, respectively, above the reference case level. In terms of generation shares, biomass, wind, and geothermal account for 1.3, 0.6, and 0.7 percent of total generation, respectively, in 2025 in the reference case, and are projected to account for 9.1, 5.3 and 1.5 percent of total generation, respectively, in the S.139 case. The imposition of a greenhouse gas emissions cap is expected to make new dedicated biomass¹⁴⁶ and wind plants relatively attractive. Except for a small amount of carbon released through the use of fossil fuels in the cultivation and transportation of biomass material, new biomass plants will be part of a closed loop system, sequestering carbon during the growing of the crops and releasing it when the crops are gasified and burned. Total net carbon emissions from biomass are expected to be near zero. Wind and geothermal produce no greenhouse emissions. New dedicated biomass plants are especially attractive because they are fully dispatchable,¹⁴⁷ whereas new wind plants produce power intermittently, only when the wind is blowing. Biomass can also be used in conjunction with other fuels, particularly coal (often referred to as biomass co-firing).

Capacity Additions and Retirements by Plant Type

The change in capacity additions generally parallels the change in generation by fuel. As mentioned, in the reference case, over the next 20 years or so the vast majority of new power plants built are expected to be fueled by natural gas. In fact, of the 440 gigawatts of new capacity projected to be added in the power sector in the reference case, 347 gigawatts (79 percent) is expected to be natural gas combustion turbine, combined-cycle, fuel cell, or distributed generation plants (Figure 5.6). In the later years of the projections, as natural gas prices rise, new coal plants become increasingly economical. A relatively small amount of new renewable capacity is also projected. No new nuclear plants are expected in the reference case.

The introduction of a greenhouse gas emissions cap is projected to lead to a dramatic change in the mix of new capacity built and capacity retired. To meet the greenhouse gas emissions limit, electricity suppliers are expected to turn to new renewable, nuclear, and fossil (with carbon capture and sequestration) plants. These technologies are generally uneconomical in the reference case. For example, construction costs for new natural gas plants with carbon capture and sequestration equipment are about 75 percent more than the costs of similar plants without such equipment (Table 5.3). In addition, the efficiency of such a plant is approximately 25 percent less than that of a plant without carbon capture and sequestration equipment.

In all cases, as the commercialization of a technology progresses, capital costs are assumed to decline as a result of “learning-by-doing” effects, which indicate that costs fall as experience increases. This is represented by assuming a specified cost reduction for each doubling of capacity. The greatest amount of learning is assumed to occur during the initial stages of development. As a technology matures, the cost declines due to learning slow down.

New technologies are projected to become more competitive in a case that assumes more rapid technological improvement (as shown in the high technology assumptions in Table 5.3). The newer technologies tend to experience a greater relative reduction in cost and greater performance improvement over time than do the existing commercial technologies. Thus, the competitive gap closes with more rapid technology improvement.

Among the renewables a large increase in biomass, wind, and to a smaller extent, geothermal capacity is projected when a greenhouse gas emission cap is imposed. Solar and conventional hydroelectric capacity is not expected to be significantly affected. Although they are available throughout the United States, biomass resources generally are better in some portions of the country than others. For electric generation, biomass fuel is utilized both to “co-fire” with coal in existing coal-steam plants or as the primary fuel in dedicated biomass power plants. The first option, co-firing, is characterized by low capital costs and short lead times. In both the reference case and the S.139 case, co-firing grows significantly between 2005 and 2010. However, in the S.139 case, generation from co-firing declines precipitously from its 2009 peak and ceases to provide significant generation by 2012 as host coal capacity declines and biomass feedstocks find higher value use in dedicated biomass facilities. Growth in dedicated facilities in the early portion of the forecast is limited by long construction lead times for the plants, high capital costs relative to co-firing, and a small existing capacity base, with resulting low levels of the knowledge and infrastructure necessary to sustain quick and large levels of market growth. Further limiting early-year growth is the limited supply of urban and agricultural wastes used for low-cost supply. Establishment of dedicated energy crop feedstocks is not expected to occur before 2010, but this source of supply provides a significant source of fuel for the fast growing biomass sector in the later years of the forecast. Through most of the forecast period, the expansion rate of dedicated biomass capacity in the S.139 case is limited primarily by the high costs imposed through production bottlenecks encountered with any fast growing technology. Despite these growth limits, biomass becomes the leading renewable fuel by 2025 in the S.139 case, with more generation than all other renewable fuels combined, although only half as much as either natural gas or nuclear capacity.

High-quality, economically exploitable wind resources are somewhat more widespread than some other renewables, such as geothermal, but they still are not available throughout the United States. Projected to be nearly cost competitive in the base case, wind generation becomes highly competitive early in the S.139 case forecast despite the assumed expiration of the production tax credit in 2003.¹⁴⁸ With relatively short project lead times and the higher costs of fossil alternatives when a greenhouse gas emissions cap is imposed, wind power reaches 83 gigawatts of capacity in the S.139 case, compared with just 11 gigawatts in the reference case. At this level, wind contributes 5.8 percent of total U.S. generation sold to the grid. In the S.139 case, most of this growth occurs in the years from 2005 to 2015. By 2015, despite the tightening greenhouse gas emissions cap, further development of wind is increasingly curtailed by the rising costs of integrating this intermittent generation source into regional grid operations, the need to upgrade transmission networks to access ever more remote wind resources, and the increasingly competitive costs of other low-carbon or carbon-free resources such as biomass and nuclear fuels. In the best wind regions, wind power produces as much as 20 percent of electric generation. At these levels, wind may require extensive “backup” from firm generating capacity resources (such as combustion turbines or hydropower) to ensure grid reliability, and some wind output may have to be curtailed to avoid operational difficulties with the few remaining coal and increasingly important nuclear plants in the regions. Furthermore, at these levels it is likely that most of the prime, low-cost sites with easy access to load centers will already be developed. Additional development will require utilizing sites that are more expensive to develop and require costly upgrade or expansion of transmission systems to bring the power where it is needed. These expenses begin to manifest themselves in the face of declining costs and increasing availability of competing non-carbon resources such as biomass and nuclear fuels.

Although the growth of geothermal capacity in the S.139 case is significant relative to the reference case, with over 80 percent more installed capacity by 2025, the overall contribution of the resource to national electricity supply is still somewhat limited. While relatively inexpensive to exploit, high-quality geothermal resources are limited. Taking advantage of naturally occurring formations of underground steam or high-temperature/high-pressure water, the known supply of geothermal resources in the United States is limited to 51 sites in a few western States and Hawaii. Technology to exploit the “hot dry rock” formations that underlie the entire continent has not yet been developed and is not projected to be available in the forecast horizon.

Although central-station solar electric capacity remains too expensive for adoption in the S.139 case, enduser installed photovoltaics are projected to show substantial growth relative to the reference case. Central-station solar technologies, including solar thermal and photovoltaic systems, are hampered by high capital costs and low utilization rates and are unable to compete in wholesale power markets, even with substantial price support that might come from greenhouse gas allowance trading. However, distributed photovoltaic applications, such as panels installed on commercial buildings or residential rooftops, do not require investment in transmission or distribution facilities, and with higher retail electricity rates they are competitive.¹⁴⁹

With respect to conventional hydroelectric power, the prime, low-cost, high-output hydroelectric sites in the United States are already largely developed. Remaining sites face numerous obstacles to significant future development, including small capacity potential at most sites, legal constraints on developing “wild and scenic rivers,” and other environmental sensitivities, even if no legal prohibition exists at a site. The incentives from greenhouse gas allowance trading may serve to make development of remaining sites more attractive, but the possible increase in hydroelectric capacity would expected to be small. While not addressed in this report, the expansion or development of some sites is possible, but it is not expected to amount to a large amount of capacity.

While the last nuclear plant order in the United States occurred 30 years ago, the imposition of a greenhouse gas emissions cap is projected to make new advanced nuclear technologies economical in the future. In the S.139 case 17 gigawatts of new nuclear capacity is projected to be built by 2020 and 49 gigawatts by 2025. The first new nuclear plants are expected to be quite expensive, costing about $2,100 per kilowatt (in 2001 dollars). However, as with most new technologies, their costs are expected to decline after the initial units are brought on, so they become increasingly competitive in the later years of the projections as the greenhouse gas emissions cap tightens, natural gas prices rise, and the most attractive sites for new renewable plants are exploited. By 2020 in the S.139 case, the cost of new nuclear plants is projected to have fallen to $1,660 per kilowatt. While there is uncertainty about the costs of these new plants, they are also likely to face more difficulty in siting and permitting than other technologies. The results of a sensitivity case discussed later will illustrate the impact of an assumption that neither this technology nor fossil plants with carbon capture and sequestration equipment will be available.

The same behavior is expected for new fossil plants—both natural gas and coal—with carbon capture and sequestration equipment. Initially they are expected to be quite expensive, but after the initial projects penetrate the market their costs are expected to fall. For example, new coal plants with carbon capture and sequestration equipment are projected to cost $2,070 per kilowatt initially, but by 2020 in the S.139 case their costs are projected to have fallen to approximately $1,660 per kilowatt. Similarly, new gas plants with carbon capture and sequestration are initially projected to cost $1,068 per kilowatt, but in the S.139 case their costs drop to approximately $780 per kilowatt. In the S.139 case 45 gigawatts of natural gas combined-cycle capacity with sequestration is projected to be built by 2020, and 102 gigawatts is projected by 2025. For coal with sequestration, the amount is 12 gigawatts by 2020 and 38 gigawatts by 2025. As with new nuclear plants, there is uncertainty about the costs and performance of these new plants. It is possible that some insurmountable engineering obstacle will arise that causes costs to remain relatively high. Again, the results of a sensitivity case discussed later will illustrate the impact of this technology not being available.

The rapid expansion of the markets for new nuclear capacity and fossil capacity with carbon capture and sequestration equipment could also face significant market hurdles. As mentioned earlier, it has been 30 years since the last order for a new nuclear plant was made in the United States. The infrastructure needed to plan, site, build and maintain the amount of new advanced nuclear capacity projected in the S.139 case could take considerable time to develop. The same is true for fossil plants with carbon capture and sequestration. Industry—and the public that will have to accept them—currently has little or no experience with these technologies. If expanded very rapidly, their costs could be higher than expected.

In addition to changing the mix of new capacity added, the imposition of a greenhouse gas emissions cap also has an impact on projected capacity retirements (Figure 5.7). In the reference case, almost all existing coal capacity is projected to continue operating. On the other hand, a significant amount of the existing oil and gas steam capacity is expected to retire. Typically older oil and gas steam plants are relatively inefficient and as natural gas prices rise and new more efficient gas capacity is added, it will no longer be economical to continue operating these plants.

In the S.139 case, a large proportion of existing coal capacity is projected to be retired. It is simply not economical to continue operating these plants. For example, in 2020 with a greenhouse gas allowance costing $178 per metric ton carbon equivalent, the effective cost of coal is projected to be $5.53 per million Btu (the $0.99 per million Btu delivered price of coal plus the $4.54 per million Btu cost of the allowances needed to use it for a plant without carbon capture and sequestration equipment), 394 percent above the cost projected in the reference case. Even with lower electricity demand, because of the large number of projected retirements, the total amount of new capacity added in the S.139 case exceeds that added in the reference case by more than 185 gigawatts.

It is impossible to say which of the relatively low carbon technologies discussed—new nuclear, biomass, geothermal, wind, gas with sequestration, or coal with sequestration—might prove the most attractive over the next 20 years or so. Any one of them might hit cost or performance hurdles that cannot be overcome, or, vice versa, one or more of them might prove extremely economical and capture a very large portion of the market for low-carbon generating technologies. The mix of technologies chosen is also sensitive to the assumed cost of capital. Some of these technologies are very capital intensive; for others, operating costs are more important. As a result, changes in the assumed cost of capital can lead to a different mix of technologies being built.

Electricity Prices and Consumer Demand

The National Energy Modeling System explicitly reflects the status of electric industry regulation by region.¹⁵⁰ The handling of the electric power sector’s opportunity costs¹⁵¹ of allowances varies depending on the status of regulation in the region. For this analysis, the opportunity cost of allowances is included in the generation price in competitive regions, when a fossil-fired unit is on the margin and sets the market-clearing price in the region.

The opportunity costs of allocated allowances are handled differently in cost-of-service regulated regions. In cost-of-service regulated regions, Public Utility Commissions will determine how they are treated for ratemaking purposes. They might only allow the recovery of allowance costs in electricity prices if they were purchased by the utility, requiring that any revenue associated with allowance sales be returned to ratepayers. However, if regulators followed this strategy completely they would significantly reduce the incentive for utilities to reduce emissions. For this analysis, it is assumed that regulators will apportion the benefits of freely allocated allowances, with ratepayers receiving 90 percent of the benefits and company shareholders receiving 10 percent of the benefits. This will act to encourage cost-of-service regulated utilities to make optimal environmental compliance decisions while distributing most of the benefit to ratepayers.¹⁵² This distribution is based on analysis of the regulatory treatment of freely allocated SO2 allowances under the SO2 allowance-trading program created in the Clean Air Act Amendments of 1990.^{153, 154}

The imposition of a greenhouse emission cap on the electric, transportation, and industrial sectors is projected to lead to significant increases in electricity prices (Figure 5.8). The higher prices result from the increased reliance on higher cost generating technologies and the need to hold allowances for all generation from fossil fuel plants without carbon capture and sequestration equipment.¹⁵⁵ In the early years of the greenhouse gas reduction efforts the relatively low cost of greenhouse gas allowances, $79 per metric ton carbon equivalent in 2010, is projected to lead to an increase in electricity prices of 9 percent above the reference case level. However, the impact on electricity prices grows in the later years. Relative to the reference case, the price of electricity is projected to be 33 percent higher in 2020, and 46 percent higher in 2025, in the S.139 case. As mentioned earlier, the effective cost of using fossil fuels— where the effective cost of the fuel is its delivered price plus the cost of allowances needed when it is used—in plants without carbon capture and sequestration equipment is much higher in cases with a greenhouse gas emission cap.

For example, in 2020 in the S.139 case, the greenhouse gas allowance price is $178 per metric ton carbon equivalent. Translating this into its impact on the delivered price of fossil fuels results in $4.54 per million Btu for coal and $2.57 per million Btu for natural gas. The higher value for coal results from its greater carbon content. Translating this into the cost of running the power plants would mean a 4.5-cent per kilowatthour increase for a coal plant and a 1.9-cent per kilowatthour increase for a natural gas combined cycle plant.¹⁵⁶ The larger impact on the operating cost of the coal plant is driven by both the higher carbon content of coal and the fact that the coal steam plant is less efficient—consuming more Btu per kilowatthour generated—than a natural gas combined-cycle plant.

If one looks at the generation component of electricity prices, excluding the costs of transmitting and distributing the power, the projected changes in electricity prices are even larger (Figure 5.9). In all cases the price of electricity distribution and transmission services is assumed to continue to be based on costof-service regulation. Because no greenhouse gas emissions occur in the transmission or distribution of electricity, these sectors of the market are not directly impacted by the imposition of a greenhouse gas emission limit. Focusing solely on generation prices illustrates the impacts of the greenhouse emission cap on the price of producing electricity. In 2020, generation prices in the S.139 case are projected to be 48 percent above those in the reference case, and by 2025 the difference is 68 percent.

The allowance cost regulatory treatment assumed in this report leads to different electricity price impacts in the cases where the share of allowances going to the Climate Change Credit Corporation (hereafter referred to as the Corporation) is changed from the main S.139 case (see Figure 5.8 and Table 5.1).¹⁵⁷

When allowances are allocated to the Corporation and then purchased by a cost-of-service regulated entity, their costs will be fully passed onto consumers. Thus, in the case where it is assumed that 80 percent of the allowances are allocated to the Corporation (the corp80 case), electricity prices are projected to show a larger price impact than in the case where only 20 percent of the allowances are assumed to be allocated to the Corporation (the corp20 case). In 2025, electricity prices in the corp80 case are projected to be 9.8 cents per kilowatthour, 3.11 cents per kilowatthour (46 percent) higher than in the reference case. Conversely, in the corp20 case electricity prices in 2025 are projected to be 9.05 cents per kilowatthour, 2.34 cents per kilowatthour (35 percent) higher than the reference case. Note that the projected electricity prices in the S.139 case are close to those in the corp20 case in the early years and close to those in the corp80 case in the later years. This occurs because the share going to the Corporation in the S.139 case is assumed to start at 20 percent in 2010 and gradually increase to 80 percent by 2025.

The electricity price differences across these cases are even more pronounced at the individual cost of service regional level (Figure 5.10). For example, electricity prices in the Southeastern Electric Reliability Council (SERC) region (southeastern states) are assumed to continue to be set using cost of service

procedures. As a result, full allowance costs will only be reflected in prices when they are purchased. Thus, in the case where it is assumed that 80 percent of the allowances are allocated to the Corporation, the electricity price impacts in SERC will be relatively large because most of the allowances needed in the region will have to be purchased. In 2025, generation prices in SERC for the corp80 case are projected to be 7.3 cents per kilowatthour, 3.5 cents per kilowatthour (90 percent) higher than in the reference case. In the corp20 case electricity prices in SERC in 2025 are projected to be 6.2 cents per kilowatthour, 2.4 cents per kilowatthour (63 percent) higher than the reference case.

Because the opportunity costs of holding allowances are always assumed to be passed on in regions where electricity prices are set competitively and fossil fuel plants set the marginal electricity price, those regions do not see significant prices differences in the corp20 and corp80 cases. Their prices always reflect the full costs of holding allowances. The different regional impacts may lead regulators and legislators to look for ways to allocate allowances to reduce the divergent regional price impacts. To the extent that this is achieved by reducing or eliminating the full passthrough of the allowance value into electricity prices in competitive areas, electricity demand reductions will be less than projected. In such a scenario, allowance prices would have to rise above the levels projected in this study to achieve the emissions targets in S.139.

Consumers are projected to respond to the higher electricity prices by reducing their use of electricity. For example, in 2010 in the S.139 case electricity sales are projected to be 54 billion kilowatthours (1.3 percent) below the reference case level. This difference is projected to increase to 381 billion kilowatthours (7.9 percent) in 2020 and 593 billion kilowatthours (11.3 percent) in 2025. (See Chapter 4 for more information on consumers’ responses to fuel price changes.)

Emissions

In addition to carbon dioxide emissions reductions, efforts to comply with the greenhouse gas emissions cap are also projected to lead to large reductions in power sector emissions of NOx, SO2, and mercury (Figures 5.11, 5.12, 5.13 and 5.14). The reference case incorporates the existing national SO2 cap and trade program and the 19-State NOx cap and trade program. Pending regulations, such as those that may be required to reduce mercury or fine particulate emissions, and proposed legislation such as the President’s Clear Skies proposal are not represented. In the S.139 case, power sector carbon dioxide emissions are expected to be fall well below their reference case level. For example, in 2020, power sector carbon dioxide emissions are projected to be 802 million metric tons in the reference case and 352 million metric tons in the S.139 case. By 2025, the difference grows even larger, 868 million metric tons in the reference case and 205 million metric tons in the S.139 case. To put this change in perspective it should be noted that the 1990 greenhouse gas emissions in the power sector were close to 500 million metric tons. Thus, the level expected in 2025 in the S.139 case is almost 60 percent below the 1990 level.

The actions taken to lower carbon dioxide emissions—reduced use of coal and the increased use of nuclear, renewables, and natural gas in electricity generation—in the S.139 case also lead to reductions in power sector NOx, SO2, and mercury emissions. By 2020 power sector NOx emissions are projected to be 1.5 million tons in the S.139 case, 63 percent below the reference case level and 68 percent below the level seen in 2001. By 2025, the power sector NOx emissions level is projected to fall further to 0.7 million tons. In fact, power sector NOx emissions in the S.139 case are projected to fall below the lowest level seen in the last 30 years or so—the 4.75 million tons emitted in 2001. The story is similar for power sector SO2 emissions. In the S.139 case, power sector SO2 emissions are projected to be 5.9 million tons in 2020 and 1.9 million tons in 2025. This compares to emissions of 17.3 million tons in 1970 and 10.6 million tons in 2001. The results are similar for mercury, with projected emissions in the S.139 case falling to 19.1 tons in 2020 and 7.2 tons 2025, as compared to 54.1 and 54.8 tons in the reference case.

Uncertainties and Sensitivity Cases

As with any mid- to long-term forecast there is considerable uncertainty surrounding the projections. In the power sector, the cost and performance of new generating technologies, especially those that are relatively low carbon emitters, is an important area of uncertainty. While the cost and performance improvement that is typically seen as new technologies enter the market is represented in the reference and S.139 cases, it is possible that the changes could be better or worse than projected, or that technologies that do not penetrate the market under normal circumstances might play a bigger role when a greenhouse gas emission cap is imposed. To assess the impact of more rapid improvements in the cost and performance of new technologies—in the residential, commercial, industrial, transportation, and electricity sectors—high technology assumptions have been incorporated into both the reference and S.139 cases (Table 5.4). The results of these cases should not be seen as predicting which of the emerging technologies might prove most successful in the marketplace but, rather, as indicative of the impacts of a general improvement in all these technologies.

In the electricity sector, the greenhouse gas emissions cap in S.139 is projected to result in lower electricity demand, higher electricity prices, and reliance on a mix of new technologies—coal and gas plants with sequestration equipment, new nuclear plants, and new renewable facilities. The improved technology cost and performance assumptions reduce the impact of the greenhouse gas emissions cap, but lower electricity demand and higher electricity prices still are projected. Compared to the high technology reference case, there is also projected to be a greater reliance on new technologies in the S.139 high technology case—coal and gas plants with sequestration equipment, new nuclear plants, and new renewable facilities. The improved technology assumptions in the residential, commercial, and industrial sectors contribute to lower electricity demand in both of the high technology cases (Figure 5.15). Consumers also reduce their demand further in response to higher electricity prices (Figure 5.16) when the greenhouse emission cap is imposed in the S.139 high technology case.

Compared to the high technology reference case, the electricity price in 2020 is about 25 percent higher in the S.139 high technology case and the greenhouse gas allowance price is projected to be $133 per metric ton carbon equivalent. In contrast, the corresponding greenhouse gas allowance price in the S.139 case is projected to be $178 per metric ton carbon equivalent in 2020. By 2025, the demand for electricity is projected to be 4,481 billion kilowatthours in the S.139 high technology case, compared to 4,997 billion kilowatthours in the high technology reference case and 4,653 billion kilowatthours in the S.139 case. Note that the end-use efficiency improvements in the S.139 high technology case lead to lower electricity demand than in the S.139 case, even though electricity prices are not projected to be as high. For example, in 2025 electricity prices are projected to be 6.7 cents per kilowatthour in the reference case, 6.3 cents per kilowatthour in the high technology reference case, 9.8 cents per kilowatthour in the S.139 case, and 8.6 cents per kilowatthour in the S.139 high technology case. Thus, relative to the reference case, the high technology assumptions reduce the projected increase in electricity prices. However, when compared to the high technology case, electricity prices in 2025 in the S.139 high technology case are 37 percent higher. Relative to the S.139 case, the lower greenhouse gas allowance prices in the S.139 high technology case allow power companies to continue operating more of their existing coal plants (those without carbon capture and sequestration equipment), leading to higher power sector emissions in the later projection years.

As shown in Table 5.1, the S.139 case results indicate that technologies which under reference case conditions are projected to play a fairly small role in the power sector over the next 20 years—i.e., coal and natural gas generators with carbon capture and sequestration, advanced nuclear, wind and biomass— are expected to be important options for reducing greenhouse gas emissions. However, the future availability, cost, and performance of these technologies cannot be known with certainty. Also other factors, including their environmental impacts and/or their lack of public acceptance, might limit their market penetration. An alternative case that assumes that new nuclear plants and fossil plants with carbon capture and sequestration equipment are not available was prepared for this analysis (Table 5.5). The results in this case should not be seen as predicting that these technologies might not be available or economical but rather as illustrating the impact on the results if their development or deployment were not successful.

Without new nuclear plants or fossil plants with carbon capture and sequestration equipment, meeting the greenhouse gas emission cap will be more difficult, requiring a higher greenhouse gas allowance fee. For example, in 2025 the greenhouse gas allowance fee is projected to be $221 per metric ton carbon equivalent in the S.139 case, but $297 in the no nuclear, no geologic sequestration case. The higher allowance cost leads to higher electricity prices and lower electricity demand. Relative to the S.139 case, electricity prices in 2025 are projected to be 9.3 percent higher, 10.68 cents per kilowatthour versus 9.79 cents per kilowatthour (Figure 5.17). Electricity sales in 2025 are lower at 4,573 billion kilowatthours in the no nuclear, no geological sequestration case, compared with 4,653 billion kilowatthours in the S.139 case and 5,246 billion kilowatthours in the reference case (Figure 5.18). Without nuclear or geologic sequestration technologies, the power sector is projected to turn to even more renewables than are expected in the S.139 case. When compared to the S.139 case, nearly 60 gigawatts of additional renewable generating capacity is expected to be added in the no nuclear, geologic sequestration case. Most of this additional increase in renewable capacity is projected come from newly dedicated biomass and, to a lesser extent, new wind plants. The dedicated biomass plants are attractive because they can be used to replace retiring baseload coal plants, whereas wind plants are only available intermittently.

Another critical factor that could impact efforts to reduce greenhouse gas emissions in the power sector is the extent to which companies can use offsets to meet their allowance requirements. As discussed in Chapter 2, S.139 places explicit limits—no more 15 percent in Phase I and 10 percent in Phase II—on the share of a company’s allowance requirements that can be satisfied with offsets. To test the sensitivity of the analysis to these limits, a case was prepared in which the maximum offset share was increased to 50 percent. In the power sector this change has a significant impact on greenhouse gas emissions, electricity prices, and the technologies chosen to meet future electricity demand. When the share limit is raised to 50 percent, power companies are projected to purchase additional offsets instead of reducing their own emissions as much as they did in the S.139 case (Figure 5.19). For example, in 2020, the power sector is projected to emit 132 million metric tons less carbon in the offset 50 case than in the S.139 case, and the difference continues to grow over time, reaching 195 million metric tons in 2025.

Because the ability to use more offsets eases the need for power companies to reduce their own emissions, the increase in electricity prices in the offset 50 case is smaller than in the S.139 case (Figure 5.20). In 2020, electricity prices in the S.139 case are projected to be 8.8 cents per kilowatthour; in the offset 50 case they are 8.3 cents per kilowatthour, or 6 percent lower. Electricity prices in 2025 are projected to be 9.8 cents per kilowatthour in the S.139 case and 9.1 cents in the offset 50 case.

With less pressure to reduce their own emissions in the offset 50 case, relative to the S.139 case, power generators are projected to reduce their dependence on low- or zero-carbon technologies, particularly new coal and gas plants with carbon capture and sequestration equipment (Table 5.6). New renewable and nuclear technologies are projected to continue to play an important role in the offset 50 case, but the penetration of fossil plants with carbon capture and sequestration is much lower than projected in the S.139 case.

5. Electricity Supply - Tables

Notes