Analysis

Reference and RPS Cases

Generation

The imposition of the RPS is projected to have impacts on all aspects of the electricity business, including the fuels and technologies used to generate electricity, the types of capacity built, the various fuels consumed and their prices, power plant emissions, electricity prices, and resource costs. In the AEO 2002 reference case, plants using fossil fuels are projected to meet most of the growth in demand expected over the next 20 years (Table 3). Increased generation from natural gas is expected to be especially important. For example, between 1999 and 2020 the generation from natural gas is projected to increase from 561 billion kilowatt-hours to 1,733 billion kilowatt-hours. The share of total generation coming from natural gas is projected to increase from 15 percent to 32 percent over the same time period. New natural gas-fired combustion turbine and combined cycle facilities are expected to be the most economical option for meeting the growing demand for electricity in most cases over the next 20-years. These technologies are generally less expensive and more efficient than other combustion options.

The generation from nonhydroelectric renewable fuels is projected to grow from 79 billion kilowatt-hours in 1999 to 159 billion kilowatt-hours in 2020 in the AEO 2002 Reference case. Much of this growth in generation from nonhydroelectric renewable fuels is expected to be encouraged by various State programs, with only a small amount coming from new merchant power plants. However, even with this doubling of generation, the share of generation coming from these fuels is only projected to increase from 2.1 percent in 1999 to 2.9 percent in 2020.

Even with the increase in renewable generation projected in the RPS 10 case the mix of fuels used to produce electricity is not expected to change dramatically (Figure 1). For example, while generation from natural gas is projected to account for 32 percent of total generation in 2020 in the Reference case, it is projected to account for 30 percent in the RPS 10 case. Similarly, generation from coal is projected to account for 46 percent of total generation in 2020 in the Reference case and 43 of total generation in the RPS 10 case. Because the RPS in S. 1766 is defined as a percentage of sales (excluding small utilities) minus renewable generation, when converted into the percentage of generation required to come from all nonhydroelectric renewables in 2020, it amounts to approximately 8.7 percent.

The lower coal and gas generation projected in the RPS 10 case is offset by the higher renewable generation stimulated by the RPS. In the Reference case, the generation from qualifying renewable generators (as defined from S. 1766) is projected to reach 1.7 percent of electricity sales in 2020. In the RPS 10 case, the 2020 share for qualifying renewables is projected to reach 8.4 percent. The generation from qualifying renewables is not projected to reach the share called for in S. 1766 in 2020 (Figure 2). This is projected to occur because of the 3-cent per kilowatt-hour credit price cap and the 2020 sunset of the RPS. In the later years of the projections, as 2020 gets closer, the number of years during which new renewable power plants will receive credits declines and, as a result, the value of the credit over the remaining years must increase to make them competitive with other generation options. In 2018 and beyond, in the RPS 10 case, the credit price needed to make new renewable plants competitive is projected to exceed 3-cents per kilowatt-hour. This causes retail electricity suppliers to pay the penalty rather than build new renewables or purchase additional credits.

Wind, biomass, and to a much lesser extent geothermal, are projected to be the most important renewable fuels stimulated by the RPS. The increased wind and geothermal generation is projected to come from new power plants while the increased biomass generation is projected to come primarily from the increased use of biomass in coal plants – what is referred to as cofiring.

Capacity

As with generation, the addition of renewable capacity to comply with the RPS is not projected to lead to a dramatic shift in the mix of generating capacity (Figures 3 and 3a).

Only wind capacity is projected to make a significant change between the Reference and RPS cases. As is the case with generation by fuel, coal and gas capacity are lower in the RPS 10 case than in the Reference case. However the combined reduction in coal and gas capacity is much less than the increase in renewable capacity. Total capacity is higher in the RPS 10 case than in the Reference case because of the intermittent nature of wind resources. In addition, there is a shift in the type of natural gas capacity added when the RPS is imposed. Over the 2000 to 2020 period, relative to the Reference case, 16 gigawatts fewer natural gas combined cycle plants are projected to be added while 7 gigawatts more natural gas combustion turbines are added in the RPS 10 case. Because generation from wind plants is only available when the wind is blowing, more backup capacity – generally natural gas turbines - is needed to ensure that consumers’ demands can be met at all times.

Overall wind capacity in 2020 is projected to be more than 5 times the Reference case level in the RPS 10 case. Though not broadly competitive in the Reference case, a small number of unsubsidized new wind plants are expected to be built in the later years of the projections when natural gas prices rise. Over the last 10 to 20 years, the cost and performance of new wind plants has improved and they are expected to continue to improve as new plants are built. In the Reference case, the basic cost¹¹ of new wind plants is expected to decline from just under $918 per kilowatt in 1999 ($982 with contingencies) to approximately $773 per kilowatt-hour ($826 with contingencies) in 2020. When the RPS is imposed, the revenue from credit sales is expected to make more new wind plants competitive and lead to more wind capacity being built. As more wind plants are built their costs are expected to decline further as manufacturers and project developers learn more about their construction and operation. For example, in the RPS 10 case the cost of new wind plants is projected to decline to $725 ($776 with contingencies) per kilowatt by 2020. However, at the same time, to reach the quantity of new wind capacity called for in the RPS 10 case – from just 2 gigawatts in 1999 to 52 gigawatts of wind capacity by 2020 – developers are projected to have to build on less attractive sites, such as those requiring upgrades to existing transmission lines, those with more expensive land, and those having more difficult terrain. After adjusting the $725 per kilowatt to reflect these factors the cost of new wind plants in the RPS 10 case in 2020 is expected to be $916 per kilowatt, very close to the current value.¹² As might be expected, the costs of all new power plants are sometimes influenced by these factors.

All new plants must incur some site-specific development and transmission interconnection costs and these costs are incorporated in this analysis. However, while wind plants have no choice but to locate where high quality wind resources are available, new natural gas plants are more flexible in their location and their developers will attempt to avoid sites that require above average development expenditures. Little change in dedicated biomass capacity is projected even though biomass generation is projected to increase significantly. The increased biomass generation comes from increased use of biomass in existing coal plants rather than in dedicated biomass facilities. In this analysis, it is assumed that coal plants can use biomass for up to 5 percent of their total fuel use if sufficient biomass supplies are available within the region the plant is located. Studies have shown that coal plants can use this level of biomass without major plant modifications or changes in other operating costs. Without the RPS, few coal plants are expected to find it economical to displace relatively low cost coal with higher cost biomass fuels. It is possible that with the RPS incentive it might be economical for some coal plants to make modifications to allow them to use even larger shares – 10 percent or more - of biomass fuels. If this occurred these plants could satisfy a large percentage of the RPS requirement. For example, if 10 percent of the projected coal generation in 2020 in the Reference case – 247 billion kilowatt-hours – were to come from using biomass rather than coal, that could satisfy approximately 60 percent of the RPS generation requirement in S. 1766. However, in today’s market, coal plant operators are focused on how future environmental regulations, particularly any efforts to reduce U.S. carbon emissions, might impact them and they are wary about making investments in their plants. If the power sector were required to significantly reduce its carbon emissions, the opportunities for increased biomass cofiring to comply with the RPS would be much lower because many coal plants would probably retire and those that continued to operate would be running much less intensively. In addition, many coal plants would probably not have sufficient low-cost biomass available to reach a 10 percent share.

Besides wind, only geothermal and municipal solid waste (landfill gas facilities) are projected to appreciably increase capacity in response to the RPS. Geothermal is projected to play a role in the west where economically accessible geothermal resources are located. However, even with the RPS credit many of the potential sites are expected to remain uneconomical. New landfill gas facilities are limited by the amount of waste that is expected to be put into relatively large landfills where gas collection facilities are economical.

Other nonhydroelectric technologies such as solar thermal, solar photovoltaic and ocean technologies are not projected to respond to an RPS. The relatively high capital costs of solar technologies make them uneconomical when compared to other renewable options such as wind and biomass. The various ocean technologies, either kinetic (including ocean wave, tidal, or ocean current) or thermal (taking advantage of temperature differences between surface and deep water) technologies, are in a very early stage of development and they are not expected to contribute to meeting the RPS called for in S. 1766. Ocean thermal efforts in Hawaii over the past 20 years have not lead to commercial development. No commercial ocean wave projects are currently operating in the United States, although a 500-kilowatt project in Britain has been completed and plans for a 1-megawatt ocean wave demonstration plant some miles off the Washington State coast are ongoing. Current costs appear to be well over $2,000 per kilowatt, making them more expensive then other renewables, such as wind or biomass¹³

Emissions

While the RPS is projected to have little impact on sulfur dioxide (SO2) or nitrogen oxide (NOx) emission levels, it is projected to have a significant impact on the SO2 allowance market. The 9-million ton emission cap established in the Clean Air Amendments of 1990 governs the level of power plant SO2 emissions and it is projected to be met with or without an RPS. However, because the RPS is projected to induce biomass co-firing in coal plants thereby reducing coal generation, the incremental costs of complying with this cap are expected to be lower when an RPS is imposed. As a result, in 2020, the cost of SO2 allowances is projected to be 31 percent lower in the RPS 10 case than in the Reference case, while SO2 emissions remain at the CAAA cap. However, the increase in co-firing does not have the same impact on NOx emissions, because NOx emissions are mainly determined by a plants’ boiler type and emissions control equipment, rather than the fuel it is using. The RPS is projected to lead to lower carbon dioxide emissions because fossil fuel generation is displaced by carbon free renewable generation (Figure 4). In 2010, power sector carbon dioxide emissions in the RPS 10 case are projected to be 3 percent below the level projected in the Reference case, while in 2020 they are 7 percent lower. However, even with this reduction they will remain 55 percent above the 1990 level for the power sector in 2020.

Electricity Price and Costs

The impact of the RPS requirement on retail electricity prices is projected to be small. This occurs because of the relatively low renewable share required – about 5 percentage points higher than is forecast without an RPS - and the impact on natural gas prices of displacing some gas capacity with higher cost renewables when the RPS is imposed. As mentioned, S. 1766 nominally calls for a 10 percent RPS in 2020, but because of the definition of qualifying renewables used and that credits are only required to cover nonrenewable generation, the actual non-hydroelectric renewable share of generation needed to meet the target is 8.7 percent.

In simple terms, an RPS is a way of subsidizing qualifying facilities (renewables) through a fee on non-qualifying facilities (coal, gas, nuclear, and oil facilities). Without the credit revenue from the non-qualifying facilities, the renewable facilities would require higher electricity prices to be economically viable. The overall cost and price impacts of an RPS program are driven by the combination of the higher costs spent on renewables minus any change in costs for other technologies that occurs because of the RPS. In this analysis, the RPS is projected to lead to a fall in natural gas prices that just about offsets the higher costs of the new renewables. The retail price of electricity in the RPS 10 case is only projected to be appreciably above the Reference case in the last few years of the projections when the renewable credit price is expected to reach 3 cents per kilowatt-hour (Figure 5). In 2020, the nation’s electricity bill is projected to be $3.1 billion higher in the RPS 10 case than in the Reference case. The 3-cent penalty is reached in 2018 and beyond because, with only a few years left when the credit will be available (it sunsets in 2020), it would have to be much higher than 3 cents per kilowatt-hour to make additional renewables economic.

While retail electricity prices are not expected to be significantly impacted by the imposition of an RPS, the industry is projected to face higher total costs and there will be large wealth transfers between nonqualifying generators and qualifying renewable generators. Over the 2000 to 2020 time period, the cumulative total electricity supplier resource costs that include fuel, non-fuel operating and maintenance costs, the capital, financing, and tax costs for new plant and equipment, and any civil penalty payments, are projected to be $7 billion higher in the RPS than in the Reference case (Table 4).¹⁴ Relative to the total resource costs of the industry over the 2000 to 2020 time period, this change is small, a 1 percent increase relative to the Reference case.

The market for renewable credits that retail electricity suppliers will have to hold for generation for nonqualifying generators is expected to grow as the RPS share and credit price increases over time (Figure 6). In 2020 in the RPS 10 case, the renewable credit market together with penalty costs paid by retail electricity suppliers is projected to reach $12 billion ($10 billion in credits and $2 billion in penalty payments). For existing coal, nuclear and oil facilities who are not projected to see significantly lower fuel prices or higher electricity prices in the RPS 10 case, the costs of holding renewable credits will reduce their operating profits. On the other hand, for existing natural gas plants, the costs of holding renewable credits are projected be offset by lower natural gas costs.

The lower natural gas prices stimulated by the RPS does have impacts outside of the electricity sector – leading to lower residential, commercial and industrial sector natural gas bills. For example, in 2010 the total residential natural gas bill is projected to be $534 million (1 percent) lower in the RPS 10 case than in the Reference case. For the commercial and industrial sectors the bills in 2010 are $387 million (2 percent) and $1,403 million (4 percent) lower in the RPS 10 case than in the Reference case.

Regional Impacts

Because renewable resources are not distributed equally throughout the US, some regions of the country are expected to be impacted more than others (Figure 7). For example, most of the 52 gigawatts of wind capacity called for in 2020 in the RPS 10 case are projected to be located in the Northwest Power Pool (NWP), the Rocky Mountain, Arizona, New Mexico, Southern Nevada (RA) and the Mid-Continent Area Power Pool (MAIN) regions which each have substantial wind resources (Figure 8). For biomass, the key regions are East Central Area Reliability Coordination Agreement (ECAR) and South Eastern Electric Reliability Council (SERC) (Figure 9). These two regions have a large amount of coal capacity that is projected to find it economical to cofire with biomass when an RPS is imposed. Most are generally expected to see small price changes because of the RPS. In the later years, when the credit price reaches 3 cents per kilowatt-hour, consumers in regions which develop large amounts of renewables, such as the MAIN and NWP regions, are projected to see lower prices because the additional money that generators in these regions make from selling renewable credits is assumed to be returned to customers in these regulated regions.

RPS 10 Case Without Sunset Provision

Removing the sunset provision from the RPS proposed in S. 1766 has a significant impact on the estimated renewable credit prices and the level of qualifying renewable generation reached in the last few years of the projections. In the RPS 10 case (which incorporates the sunset provision called for in S. 1766) the price of renewable credits is projected to reach the 3 cent per kilowatt-hour penalty in 2018 through 2020 and the level of qualifying renewables developed does not reach the RPS target. When the sunset provision is removed, the renewable credit price in 2020 is projected to be 1.7 cents per kilowatt-hour rather than the 3.0 cents per kilowatt-hour value reached in the RPS 10 case with the sunset provision. As shown in Figure 10, relative to the RPS 10 case, when the sunset provision is removed, additional generation from wind, dedicated biomass and geothermal facilities is expected to be added to comply with the RPS requirement. The generation from biomass cofiring is actually lower when the sunset provision is removed. This occurs because capital-intensive renewable technologies like wind, geothermal, and dedicated biomass plants become more attractive when they can receive the revenue from selling renewable credits for a longer period of time. As in the RPS 10 case, the electricity price impacts in the RPS 10 case without sunsetting are projected to be small. However, because more renewables are built to comply with the RPS, the cumulative resource costs between 2001 and 2020 are $10 billion higher than in the Reference case, $3 billion higher than in the RPS 10 case with sunsetting.

Notes