5
Implementing Emission Controls on Stationary Sources

INTRODUCTION

Stationary emission sources are divided into two categories in the Clean Air Act (CAA): major stationary sources (also called point sources) and area sources (see Box 5-1). Both contribute significantly to air pollution in the United States, and the CAA has contained provisions to regulate and control emissions from many of these sources for over three decades.

In principle, stationary sources can be controlled through the imposition of a design standard or a performance standard applied to individual facilities or through the imposition of an overall cap on a specific industry or segment of sources (see Box 5-2). The CAA applies these controls through a variety of programs that generally fall into five categories:

  • Permits and standards for new sources or major modifications of existing sources (for example, the new-source review [NSR], New Source Performance Standards [NSPS], and prevention of significant deterioration [PSD]).

  • Technology-based standards for emissions reduction in a class of existing facilities (for example, the reasonably available control technology [RACT] requirements for nitrogen oxides [NOx], the acid rain NOx provisions, and the maximum achievable control technology [MACT] for hazardous air pollutants).

  • Cap-and-trade provisions (for example, the acid rain sulfur dioxide [SO2] program of the CAA Amendments of 1990).



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 174
Air Quality Management in the United States 5 Implementing Emission Controls on Stationary Sources INTRODUCTION Stationary emission sources are divided into two categories in the Clean Air Act (CAA): major stationary sources (also called point sources) and area sources (see Box 5-1). Both contribute significantly to air pollution in the United States, and the CAA has contained provisions to regulate and control emissions from many of these sources for over three decades. In principle, stationary sources can be controlled through the imposition of a design standard or a performance standard applied to individual facilities or through the imposition of an overall cap on a specific industry or segment of sources (see Box 5-2). The CAA applies these controls through a variety of programs that generally fall into five categories: Permits and standards for new sources or major modifications of existing sources (for example, the new-source review [NSR], New Source Performance Standards [NSPS], and prevention of significant deterioration [PSD]). Technology-based standards for emissions reduction in a class of existing facilities (for example, the reasonably available control technology [RACT] requirements for nitrogen oxides [NOx], the acid rain NOx provisions, and the maximum achievable control technology [MACT] for hazardous air pollutants). Cap-and-trade provisions (for example, the acid rain sulfur dioxide [SO2] program of the CAA Amendments of 1990).

OCR for page 174
Air Quality Management in the United States BOX 5-1 Major and Area Sources of Emissions The regulations and controls on stationary sources in the United States have generally been designed to focus on two types of stationary sources: Major stationary sources whose emissions exceed a nominal threshold defined in the CAA or by a regulatory agency. Area sources whose emissions fall below the threshold. Major stationary sources (such as factories and electricity-generating facilities) are defined as single sources with emissions exceeding a threshold level that depends on the pollutant. This threshold is generally 100 tons/yr for criteria pollutants and their precursors; however, the threshold for Pb is 5 tons/yr. For the most severe O3 nonattainment areas, the threshold for VOC can be as low as 10 tons/ yr. For serious CO nonattainment areas, the CO threshold is 50 tons/yr. The reporting requirements for major stationary sources are more detailed than those for area sources (65 Fed. Reg. 33268 [2000]). One important distinction between major stationary and area sources is the method used to estimate and control their emissions. Major stationary sources are inventoried individually and their emissions are generally controlled through a permitting process that requires specific regulatory review of each individual facility. Area sources, which are generally widely dispersed sources arising from relatively small industrial and business facilities (for example, agricultural fields and small copying and printing shops) or from application and use of consumer products (for example, architectural coatings), are inventoried and regulated collectively. Other trading and voluntary mechanisms (for example, pollution prevention programs and the U.S. Environmental Protection Agency’s [EPA’s] Project XL). Regulations on area sources (for example, consumer product specifications). In addition to those programs, which either provide permits for, or mandate changes in, new and existing facilities and products, the CAA Amendments of 1990 also established the Title V operating permit program, which requires comprehensive operating permits for large stationary sources to record all operating requirements for a facility as a basis for tracking compliance. Each of these programs is discussed below. This discussion is then followed by a summary of the strengths and weaknesses of the various programs and the lessons to be gleaned for future approaches to air quality management (AQM).

OCR for page 174
Air Quality Management in the United States BOX 5-2 Design Versus Performance Versus Cap and Trade Traditionally, stationary sources have been regulated through the imposition of emission standards or limitations. The CAA defines such a standard or limitation “as a requirement established by the State or the Administrator which limits the quantity, rate, or concentration of emissions of a source to assure continuous emission reduction, and any design, equipment, work practice or operational standard promulgated under this Act.” Although many specific programs regulate stationary sources in the CAA, the basic approaches that have been adopted to achieve emission reductions fall into three broad categories: technology specification standards (or design standards), performance standards, and the newer use of cap-and-trade requirements. A design standard mandates that a set of design or technological options (for example, installation of particle traps in smoke stacks) must be adopted by the managers of the regulated facilities to meet emission targets. Although this approach has the potential to achieve substantial reductions in air emissions, it has been criticized as being overly inflexible and cost-ineffective. For example, even though there is often a substantial disparity in the marginal costs of emission reductions among facilities affected by the same technology standard, technology-specification standards do not allow market forces to use this disparity to minimize the overall costs of the desired level of emission reductions (Hahn and Stavins 1992; Stavins 2002). Moreover, because firms must use the technologies specified in the standard, the approach does not encourage individual firms to pursue ways to reduce emissions through potentially more effective alternative technologies and front-end process adjustments. In contrast to a design standard, a performance standard simply specifies a maximum allowable rate of emission from a given type of source or facility, and the managers of the facility are free to choose any combination of technologies and operational practices to meet the standard. In principle, this approach provides an individual facility with greater flexibility to discover the most cost-effective way to meet the emission standard. Although the flexibility exists in theory, in practice the performance standard is normally set at the level that can only most readily be achieved by a known technology. Thus, unless readily available alternatives can meet the standards to the satisfaction of the regulators, there is likely to be a tendency for facilities to default to the known technology, thus also limiting the achieved by selection of parameters within that technology (for example, size of options for the affected industries. In that case, the necessary level of control is the control technology and flow rates of reactants). The performance standard can set different degrees of control for different sources—low-emission sources may have a lower control requirement than high-emission sources. However, regulators faced with setting a performance standard often compromise, setting a standard at a lower level than the one that can be achieved by many facilities so that the facility with the largest uncontrolled emissions will not face an impossible task of control. As a result, marginal costs of emission reductions often continue to vary substantially among facilities. Further, once a performance standard is achieved at a facility, there is little incentive to discover more efficient ways of achieving the same or greater emission reduction,

OCR for page 174
Air Quality Management in the United States and, most important, there is no mechanism for a company to profit from innovations that achieve emission reductions beyond the standard. The choice between a design and a performance standard is often made as part of a rule-making process. In some cases, such as fugitive emissions, it is simpler to set a design standard because of the complexity of measuring the actual emissions. However, design standards do not provide the same limit on emissions that performance standards do; they simply provide a reduction over a set of baseline emissions. In either case, the standards are related to the output of the facility. Performance standards are usually expressed in such terms as pounds of pollutant emitted per million British thermal units of fuel heat input. In these cases, the actual amount of emissions is permitted to increase as the amount of fuel is increased. In response to the limitations of both design and performance standards, a new approach (a market-based approach) based on cap and trade has emerged in the last decade. In this approach, each source category (or every source) in a given geographic area has its total emissions of a particular pollutant capped at a level below its current level, and each individual source is assigned an emissions allotment consistent in the aggregate with the overall emissions cap. The novel aspects of this total-emissions-based performance standard are (1) that it does not presume any particular technology or emissions standard for the sources, and (2) that it allows market forces to minimize costs and reward innovation. Each facility is allowed to achieve the required reductions in a variety of ways, including conventional pollution control, process change, and product substitution, as well as purchase of reductions at a more economical rate from other facilities that have exceeded their reduction target. Even with a cap-and-trade standard, an emission limit must be set that is based on feasible control technology or process operations. However, the ability to trade removes one of the problems faced by regulators when dealing with a range of existing sources. A greater control requirement can be set, and companies that cannot easily meet the requirement can trade emission-reduction credits to comply with the cap-and-trade requirement. There are challenges in applying this emission-control mechanism in every situation, as discussed later in this chapter. The mechanism does, at least in theory, offer the possibility of achieving substantial reductions while allowing individual sources to minimize costs and optimize efficiency. PERMITS AND STANDARDS FOR NEW OR MODIFIED MAJOR STATIONARY SOURCES The CAA mandates that the states implement and EPA oversee permit programs to control and regulate pollutant emissions from major stationary sources in National Ambient Air Quality Standards (NAAQS) attainment and nonattainment areas. Under these programs, each new major stationary source of air pollutants must apply for a permit before beginning construction and, within the permit application, demonstrate that the new facility will meet appropriate emission-control standards. In recognition of the substantial costs of retrofitting, existing stationary sources are required

OCR for page 174
Air Quality Management in the United States by federal law to undergo the permitting process only when nonroutine modifications are planned that will result in a significant increase in pollutant emissions from that source.1 The program for nonattainment areas is described in Part C of Title 1 of the CAA and that for attainment areas in Part D of Title 1. In the CAA, the program is the NSR in nonattainment areas and the PSD in attainment areas. Although NSR is often used generically for both types of programs, there are some differences, so the discussion below uses the CAA’s terms. Background Since 1970, the CAA has required EPA to promulgate NSPS for major and minor sources on a category-by-category basis. NSPS are national emission standards that are progressively tightened over time to achieve a steady rate of air quality improvement without unreasonable economic disruption. In recognition that the 1970 goals for attainment of air quality standards would not be met and that some attaining areas required measures to prevent conditions from worsening, NSR procedures became applicable to major stationary sources in nonattainment areas and in PSD attainment areas with the 1977 CAA Amendments. The NSR provisions required that major new sources in nonattainment areas be constructed only if they created no net increase in emissions. That requirement was intended to ensure that major new or modified stationary sources of air pollution within nonattainment areas did not inadvertently undermine the state implementation plans (SIPs) that had been developed for those areas. To accomplish that, NSR requires that new and modified facilities use control equipment that has the lowest achievable emission rate (LAER) and that those facilities obtain an “emission offset” to offset the increase in emissions anticipated from the proposed construction or modification. As discussed in more detail below, the offset is actually required to be greater than the emissions increase from the proposed project to ensure progress toward attainment. The PSD program for attainment areas was designed to do the following: Maintain public health protection in areas that already meet the NAAQS. Reduce the total amount of emissions entering the atmosphere and thereby provide general protection from public welfare damage and the impacts of pollutant transport. 1   Throughout the remainder of this discussion, “modification” will be used as it is defined in the CAA to denote “any physical change in, or change in the method of operation of, a stationary source which increases the amount of any air pollutant emitted by such source or which results in the emission of any air pollutant not previously emitted.”

OCR for page 174
Air Quality Management in the United States Protect visibility. Counteract the unintended incentive given to industries from the NSR program to relocate to less developed states and thereby avoid NSR permitting requirements. PSD, like NSR, requires new and modified facilities to meet emission-control standards, although PSD standards need not be as restrictive as those for NSR. Moreover, PSD does not require emission offsets. NSR and PSD Requirements Applicability NSR and PSD apply to major new stationary sources and major modifications of existing stationary sources. However, the definitions of these terms differ somewhat for the two programs: NSR applies only to criteria pollutant emissions and their precursors from major or modified sources. In most nonattainment areas, a major source for the purposes of NSR is a source that has the potential to emit 100 tons or more per year of any criteria pollutant. For ozone (O3) nonattainment areas, the definition of “major stationary source” includes smaller sources in areas of more severe nonattainment and sources with the potential to emit as little as 10 tons per year of NOx and VOCs in extreme nonattainment areas. PSD generally defines a major source as one that produces 250 tons or more per year of any pollutant. However, 28 specific source categories are identified in the CAA for which the PSD definition of a major source is broadened to include sources that produce 100 tons per year or more of any pollutant.2 A major modification is one that produces a “significant increase” in emissions as defined by PSD regulations (40 CFR 52.21(b)23). 2   The 28 source categories are coal cleaning plants (with thermal dryers), coke oven batteries, Kraft pulp mills, sulfur recovery plants, Portland cement plants, carbon black plants (furnace process), primary zinc smelters, primary lead smelters, iron and steel mills, fuel conversion plants, primary aluminum ore reduction plants, sintering plants, primary copper smelters, secondary metal production plants, municipal incinerators capable of combusting more than 250 tons of refuse per day, chemical process plants, hydrofluoric acid plants, fossil-fuel boilers (or combination thereof) totaling more than 250 million British thermal units (Btu) per hour heat input, sulfuric acid plants, petroleum storage and transfer units with a total storage capacity exceeding 300,000 barrels, nitric acid plants, taconite ore processing plants, petroleum refineries, glass fiber processing plants, lime plants, charcoal production plants, phosphate rock processing plants, and fossil-fuel-fired steam electric plants of more than 250 million Btu per hour heat input.

OCR for page 174
Air Quality Management in the United States The CAA indicates that existing facilities undergoing modifications that significantly increase pollutant emissions shall be subject to NSR or PSD review through the permitting process. The CAA also identifies the net emission increases for each of a series of pollutants that triggers such review. To implement this requirement, EPA developed rules to determine whether a proposed modification would result in a net increase in emissions. That determination involves prescriptions for estimating the present-day baseline emissions and the projected future emissions. In addition, EPA’s rules exempt routine maintenance and repair activities from the definition of a modification that can trigger NSR or PSD review. EPA also exempts modifications that cause de minimis, or insignificant, emission increases. For a modification to trigger NSR or PSD review, an emission increase must exceed a “significant level” specified by EPA. That level varies by pollutant and attainment status of the area. As discussed later in this section, EPA’s rules and regulations have been the subject of much debate and litigation and have been changed from time to time by EPA to reflect changes in policy priorities within the Executive Branch. To implement the provisions, EPA usually delegates permitting authority to states or local districts. However, before such a delegation is made, the state or district must demonstrate that its permitting process is substantially the same as that used by EPA. Operation The operation of NSR in nonattainment areas and PSD in attainment areas is conceptually similar. Major new or modified sources undergo a preconstruction review to qualify for a permit that allows the construction or modification to proceed. The new construction or modification must use stipulated control technology. However, the control technology requirements are different in attainment and nonattainment areas. Sources subject to NSR in nonattainment areas are required to use control equipment that provides the LAER and those subject to PSD in attainment areas are required to use best available control technology (BACT). Section 171(3) of the CAA defines LAER as the most stringent emission limitation based on either (1) the most stringent limitation in any SIP for that class or category of source or (2) the most stringent limitation achieved in practice for a certain class or category of source. Section 169(3) of the CAA defines BACT as an emission limitation based on the maximum degree of reduction of each pollutant subject to regulation under the CAA emitted from any major emitting facility, which the permitting authority, on a case-by-case basis, determines is achievable for such facility. The authority is required to take into account energy, environmental and economic impacts, and other costs. Thus, the emission

OCR for page 174
Air Quality Management in the United States limitations required under BACT can be less stringent than those under LAER. In determining the specifics of a BACT requirement for a facility, EPA and the states must also ensure that the facility does not exceed the NSPS. In practice, NSPS, which are determined once and are changed only if the regulation is fully revised, serve as the minimum level for BACT and LAER determinations. In nonattainment NSR, offsets are required to bring about a net decrease in emissions in the area. The size of the offset relative to the anticipated increase from the new or modified source varies from 1.1:1 for non-O3 nonattainment areas and marginal-O3 nonattainment areas to as much as 1.5:1 for extreme-O3 nonattainment areas (see Box 3-3 in Chapter 3). The offsets must be reductions that would not otherwise occur (for example, as a result of other SIP activities) and can be obtained from new emission controls on, or shutdowns of, other facilities and sources. They must be real, permanent, and enforceable. In principle, offsets provide an incentive to try innovative control technologies. In practice, offsets are usually obtained by shutting down other facilities. Offsets are not required in PSD permits. However, any emission increase after the use of BACT is limited to a fixed amount from all new facilities in the area. The amount of the increase allowed under PSD regulations, called the increment, is defined in terms of ambient air concentration and depends on the pollutant and the class of the area affected. National parks, wilderness areas, and similar sensitive areas are identified as Class I areas. Most areas are Class II. The regulations also define a Class III area, for which emission limits are more lenient than those for Class I and II areas; however, no areas have been designated Class III. The increments for various pollutants are shown in Table 5-1 for each area class. PSD permit applications are also required to analyze the effects of the source on visibility, soils, and vegetation; the economic growth likely to result from the new source; and the effects of this new growth on the new emissions. When a Class I area is affected by the source, the federal land manager for that area must play a role in the permit process. According to PSD regulations, the manager can provide an analysis showing that there is an “adverse air quality impact” on the Class I area, even if all other permit requirements are met. In such cases, the manager can recommend that the permit not be issued, and the permit may be denied. Issues with the Application of NSR and PSD NSR and PSD have been a positive force in AQM in the United States. NSR has provided a mechanism for economic development and new con-

OCR for page 174
Air Quality Management in the United States TABLE 5-1 Allowable Concentration Increments (micrograms per cubic meter) for Prevention of Significant Deterioration (PSD) Pollutanta Measurement Increment in Areas That Are Class I Class II Class III Particulate matter (PM10) Annual arithmetic mean 4 17 34   24-hr maximum 8 30 60 Sulfur dioxide (SO2) Annual arithmetic mean 2 20 40   24-hr maximum 5 91 182   3-hr maximum 25 512 700 Nitrogen dioxide (NO2) Annual arithmetic mean 2.5 25 50 aPM10 is particulate matter with an equivalent aerodynamic diameter of 10 micrometers or less. SOURCE: Clean Air Act, Section 163. struction to proceed in nonattainment areas while maintaining progress toward NAAQS compliance. Moreover, both NSR and PSD have mandated the use of modern, clean technologies and practices in new facilities and in modified existing facilities throughout the nation. The application of BACT in conjunction with the requirements for NSPS has encouraged, indeed required, the continual development and application of new technologies that are more cost-effective, cleaner, or both. However, NSR and PSD have some limitations as well. Some of the more prominent aspects are discussed below. Complexity and Inefficiency The NSR- and PSD-permitting processes have become complex and time consuming, especially if there are disagreements between the permit seeker and the permitting agency. The ever-growing nature of the process is illustrated by EPA’s documentation describing NSR and PSD regulations, manuals, and guidance. About 30 pages in length in the early stages of the program, the documentation now exceeds 1,000 pages and is contained in numerous documents with which permit applicants and writers must be familiar (EPA 2002k). Representatives of industry complain that the process fosters inefficiencies and unduly discourages economic growth and innovation (NAM 2002). The process as currently organized can lead to conflicts between the goal of implementing improved emission-control technology as quickly as possible on the basis of BACT or LAER requirements and the need for firms to know what control technology will be required for new construction or modifications to existing facilities. Because of the lengthy time required to complete a permitting process and the rate at

OCR for page 174
Air Quality Management in the United States which control technologies are developing, the BACT or LAER requirements for a given project could change between the time that a project is proposed and the time that it is permitted.3 Older, Dirtier Facilities Remain in Operation When the NSR and PSD programs were enacted, Congress did not require that emission controls be placed on any existing facilities, in effect “grandfathering” these facilities, although the provisions left open the possibility that controls could be required on these older facilities as part of a SIP filed in a nonattainment area.4 Such facilities were expected to reach the end of their operating lives about 30 years after their initial construction. Experience over the past 25 years, however, shows that older high-polluting facilities throughout the nation have continued to operate with minimal modernization or have undergone so-called lifetime-extension projects. These projects have been able to maintain the economic viability of the facilities well beyond their initial design lifetime without triggering NSR or PSD review and the addition of modern emission-control technology (see Box 5-3). In addition, placing controls on such facilities as part of a SIP has proved politically difficult. Facility owners and operators can demonstrate that the cost per ton of controls on such facilities, assuming a short remaining lifetime, is much greater than the equivalent cost for new facilities. Some maintain that the onerous nature of the NSR and PSD permitting process contributes to this state of affairs by raising bureaucratic hurdles against new construction. Others argue that industry keeps old facilities operating, sometimes unlawfully, to avoid the NSR and PSD requirements to use clean technologies, as discussed further below. 3   Such requirement changes occurred in the South Coast Air Quality Management District during the 1980s. At that time, one of the BACT and LAER requirements for gas turbines was for the combustion devices to be designed and operated to produce low amounts of NOx. Several companies were proposing to permit new gas turbines with low NOx combustors, as required. However, the district reviewed its BACT requirements and determined that selective catalytic reduction (SCR), not low NOx combustors, should be used as the BACT. That decision caused some of the projects with low NOx combustors to be dropped. 4   One reason that new and modified emission sources were subjected to more stringent regulations than existing sources is that it was considered easier to develop an appropriate design standard for a facility that is being constructed or modified. Promulgating a design standard for existing facilities can represent a challenge to regulators because of the spectrum of operating conditions and preexisting emissions that are present in the field. Individual sources within a particular group, such as glass furnaces, might have significantly different operating temperatures and, consequently, different NOx emissions. If the sources were required to install the same technology or to obtain the same percentage of emission reduction, some sources would have higher emissions than others. Conversely, if a regulation were passed that required all sources to achieve the same level of emissions, sources that had low emissions would be able to comply with much less effort than other sources.

OCR for page 174
Air Quality Management in the United States BOX 5-3 The Grandfathering of Facilities When the CAA Amendments of 1970 were written, it was assumed that electric power plants would be decommissioned when they completed their expected functional lifetime. Instead, many older power plants have received extensive maintenance to extend their operation and remain in use. Although no specific exemption was created for these facilities, their extended operation amounts to a de facto grandfathering, and that term is commonly used to describe facilities with little or no emissions control that continue in operation well beyond their original expected lives. Table 5-2 provides a breakdown by vintage of NOx emission rates for coal-fired power plants operating in 1999 in pounds per megawatt hours (the amount of NOx emitted per unit of electricity generated) and the percent of total NOx from coal-fired boilers emitted by each vintage. The table indicates that some facilities have remained operational long after their expected 30-year lifetime and that these older facilities emit NOx at a much higher rate than newer facilities. They also emit a disproportionately large fraction of NOx relative to the power they produce, as indicated in the far right column of the table. Substantial reductions in emissions from coal-fired power plants could be obtained by retrofitting plants or by replacing them with newer coal-fired or gas power plants. Although there are reasons for focusing regulation primarily on new facilities or major modifications—most notably the relatively lower cost to incorporate control technology in the planned construction—the net effect of de facto grandfathering of facilities can be a substantial source of emissions. The use of a total cap on emissions with emissions trading can provide a monetary incentive for older facilities to reduce emissions. (See discussion on advantages and challenges of such programs later in this chapter.) TABLE 5-2 NOx Emissions from Coal-Fired Boilers in 1999 by Vintage Power Plant Vintage Avg. NOx Emission Factors (lb/MW-hr) % of Coal-Fired Electricity Capacity % of Coal-Fired Electricity Generation % of Total NOx Emitted % of NOx Emitted per % of Electricity Generated Pre-1950 7.44 1 0.4 0.6 1.50 1950–1959 5.97 15 12.9 14.9 1.16 1960–1969 5.95 20.6 18.8 21.6 1.15 1970–1979 5.37 37.6 38.6 40.1 1.04 1980–1989 4.09 23.5 26.6 21.1 0.79 1990–1999 3.55 2.4 2.6 1.8 0.69 Abbreviation: lb/MW-hr, pound per megawatt-hour. SOURCE: Data from Burtraw and Evans 2003.

OCR for page 174
Air Quality Management in the United States Evaluation of Cap-and-Trade Approaches to Air Regulations The advent of cap-and-trade systems over the past decade has provided an opportunity to achieve substantial reductions in emissions at a cost that appears to be substantially below that required in a traditional emission-control program. At the same time, these programs have met with varying success. Even in the case of the acid rain SO2 emissions-trading program, which has been the most successful application, there are issues to be addressed and improvements that could be made. The following identifies the major issues and opportunities for ensuring that cap-and-trade programs can be an effective and growing part of the AQM system. Spatial Redistribution of Emissions One of the major reservations expressed about cap-and-trade programs is that they may produce spatially and temporally heterogeneous patterns in emission reductions with undesirable environmental consequences. Although more conventional prescriptive approaches can address regional and seasonal issues by defining technology or performance standards that are more restrictive in areas where or, at times, when environmental problems are more critical, the least costly trading programs allow trading across regions and banking of emission allowances without regard to the possible environmental consequences. Such trading might worsen ecological hot spots or increase the number of persons exposed to pollution. That possibility is especially of concern for toxic air pollutants. Spatially heterogeneous emission reductions are also of concern for a pollutant like NOx, whose emissions can have little or no effect on O3 pollution in one region of the nation and a dominant effect in another. On the other hand, heterogeneous emission reductions are not of concern for nonreactive, long-lived gases, such as CO2 and methane (CH4). Some analysts of cap and trade point out that there is little possibility that any given area will have negative impacts from the program, provided the cap is set low enough to reduce emissions by a large percentage (Burtraw and Mansur 1999). They also point out that emissions trading cannot be used to avoid meeting NAAQS (or other health and safety regulations) and that states retain the authority to set tighter emission limits to meet NAAQS. Both Massachusetts and Connecticut have adopted more stringent SO2 emission levels for electricity generators for that reason, and New York is in the process of developing them. These arguments appear to be largely borne out by the results of the acid rain SO2 emissions trading program, which set a stringent national emissions cap on a criteria pollutant. As illustrated in Figure 5-2, the acid rain emissions trading program has achieved SO2 emission reductions in each of five major multistate regions of the contiguous

OCR for page 174
Air Quality Management in the United States FIGURE 5-2 Regional SO2 emission from electric utilities. Northeast Region: CT, ME, MA, NJ, NY, NH, RI, and VT; Mid-Atlantic: DE, MD, PA, VA, WV, and DC; Southeast: AL, FL, GA, KY, MS, NC, SC, and TN; North Central: IL, IN, MI, MN, OH, and WI; and Western: AZ, CA, LA, NM, OK, TX, IA, KS, MO, NE, NV, CO, MT, ND, SD, UT, and WY. SOURCE: EPA 2001d. United States. Even more significant, regions with the highest emissions, such as the north-central region, have had the largest reductions.5 On the other hand, some emission increases have occurred on smaller geographic scales (for example, some states have had emission increases) (NAPAP 1998). To guard against even the possibility of regional disbenefits, a modified cap-and-trade program with “zones” can be adopted. For example, the RECLAIM program in California allows trading between two zones but only in one geographic direction. This approach is guaranteed to prevent an increase in emissions in the zone that has the greatest potential to affect 5   The pattern of trading also appears to have added slightly to the aggregate benefits to ecosystem health, but by far the most significant benefits have been realized as a result of the dramatic overall reduction in emissions (Burtraw and Mansur 1999; Swift 2000).

OCR for page 174
Air Quality Management in the United States ambient air quality. However, the existence of multiple zones serves to reduce the potential number of trades and to raise the cost of environmental regulations. Indeed, the original design for the acid rain SO2 emissions-trading program would have divided the nation into two trading zones to prevent trades that might worsen ecological problems in northeastern states. However, policy-makers recognized that a more constrained market would increase the cost, and they opted for the uniform national market approach currently contained in the statute (Hausker 1992). As discussed above, that decision does not appear to have caused substantial problems. Banking Emission Allowances for the Future Another major issue concerning the environmental performance of market-based programs is the role of banking emission allowances. Though not as common, prescriptive approaches can explicitly account for temporal issues, for example, by placing restrictions on open burning or implementing seasonal O3 controls. The opportunity to bank emission allowances from one time period to the next can contribute substantially to cost savings, because it can facilitate a rational investment plan across facilities. Some facilities may be retrofitted early if they are utilized heavily. The emission allowances that are not used at that facility may be used to keep another with a shorter remaining operating life running until it is retired, thus avoiding expensive retrofits. Banking can provide an important element of economic stability to a permit market (for example, the absence of banking for the RECLAIM program was one of the factors contributing to its instability). However, the accrual of banked emission allowances can distort the perception of environmental performance under a trading program. In the early years, emissions appear to be below their allowable level, suggesting unexpected environmental success. In later years, when the banked allowances are drawn down, emissions appear to be above the allowable level, suggesting that the program is unsuccessful. Banking within a single emissions season can also have beneficial or adverse effects. For example, seasonal limits on NOx emissions that allow emissions to vary in a manner that might enhance O3 concentrations over the summer O3 season can be temporally mismatched with peak O3 concentrations that result from the reaction of NOx and VOCs during specific episodic meteorological events (for example, warm stagnant conditions). More recent cap-and-trade programs, such as the Ozone Transport Commission’s NOx budget, contain flow-control provisions that attempt to limit increases in emissions during meteorological conditions conducive to high O3 concentrations by discounting the value of allowances when a high number of allowances are banked.

OCR for page 174
Air Quality Management in the United States Fairness in Allocating Emission Allowances For a market-based trading program to be effective, it must be perceived as being fair. Key to a fairness perception is the method used to assign emission allowances. In most previous programs, including the acid rain SO2 emissions-trading program, fairly simple principles served to guide the regulatory process. The general principle was to allocate emission allowances on the basis of historical generation (heat input at the plant). This grandfathering approach provides cost-free allowances to larger emitting sources that will be most affected by the program. That can result in giving larger allowances to those facilities that have not taken action to reduce—in effect penalizing those utilities that have taken earlier actions. A second method is an allowance allocation program based on output. Fixed numbers of allowances are allocated on the basis of the percentage of electricity (or other product) generated by each source. Typically, the allocations are updated as the generation mix changes. Many stakeholders argue that this method encourages greater efficiency than the grandfathering method (EPA 1999g). A third method is to hold an auction where all affected sources must go to purchase allowances to cover their emissions. An analysis of allocating CO2 allowances under the three methods found that the auction is the least costly to society (Burtraw et al. 2001b). A series of important general equilibrium studies over the last decade also focused on the virtues of an auction for allocating emission permits (Goulder et al. 1997, 1999). The social cost of an allowance auction is expected to be dramatically less than allocation at zero cost. The reason is that the regulatory program (regardless of how allowances are allocated) raises costs similar to a new tax and thereby serves to exacerbate distortions from preexisting taxes. However, an auction provides a source of revenue that can be used to reduce preexisting distortionary taxes. The main criticisms of auctions are that they increase the cost of a program from the viewpoint of companies, which have to pay for emission allowances, as they pay for other inputs to productive activity, such as fuel and labor. Setting and Revising the Emissions Cap The concept of placing an overall limit on total emissions was an important innovation in the cap-and-trade approach in environmental policy. A shortcoming of the approach as currently practiced is the absence of a mechanism for changing the cap in response to new scientific and economic information. During the 1990s, for example, the cost of controlling SO2 emissions from coal-fired power plants fell to less than half of the amount predicted at the time of the 1990 CAA Amendments (Ellerman et

OCR for page 174
Air Quality Management in the United States al. 2000). Moreover, as will be discussed in Chapter 6, there is growing scientific evidence that NOx emissions contribute more to the deleterious ecological effects of acid rain than had been recognized at the time of enactment of the 1990 Amendments, an understanding that came after the relatively modest target set for reducing NOx emissions in the acid rain rule. Despite these new understandings, regulators have not been able to adjust the SO2 emission cap and NOx emission targets assigned by Congress in Title IV (Zuckerman and Weiner 1998), and any further adjustment now awaits the passage of the multipollutant legislation discussed above or some other legislation. An alternative to a firm cap would be one that was adjusted in response to new information. One example is the cap on specific O3 depleting substances. The cap was dramatically revised downward when new scientific evidence was discovered, because the initial cap definition did not prevent change in the cap.6 Others have suggested similar trigger mechanisms on emission caps to provide economic relief if costs are greater than expected (Pizer 2002), but such an approach might better be coupled with a mechanism that provides environmental improvement when costs are less than expected. One possibility is a cap that is set to decline over time in anticipation of new technologies and less expensive ways to achieve emission reductions. However, if the cost of reductions exceeded a ceiling, the cap would be frozen until costs again fell below the ceiling. The value of a tradable emissions allowance could serve as an index for such a mechanism. Implicit Emission Increases Following Transition to a Trading Program Another concern about market-based programs, but one that can be addressed in their design, is that an important source of cost savings has to do with implicit emission increases at some facilities under market-based trading compared with those under prescriptive regulation (Oates et al. 1989). Under a prescriptive regulation, each facility must meet or exceed an emission-reduction target. Because of the nature of emission-control investments and the liabilities incurred from noncompliance, a facility will generally exceed the target reduction, meaning that emissions will be less than allowable. An emissions-trading program provides the flexibility to overcomply at one facility and to apply the unused emission allowances to increase emissions at another facility. Even if facilities have identical costs, when faced with uncertainty about production and emissions, a group of facilities can come closer to meeting allowable aggregate emissions with the 6   Another example is fisheries, which define a tradable right to be a share of the cap (not a specific number of tons). In this way the cap could change, and the allowances could be changed automatically in response (see Tietenberg 2002).

OCR for page 174
Air Quality Management in the United States same level of collective risk of noncompliance by managing their operations as a portfolio. The slack in a regulatory program, defined as the difference between the allowable emissions and the actual emissions, inevitably will be lessened by the move from a prescriptive to a market-based approach. This may be of environmental concern; in any case, it may provide justification for lowering allowable aggregate emissions under a market-based program. Compliance Assurance and CEM Because the market ultimately determines the locations and times at which emission reductions take place, market-based approaches, such as cap and trade, require a greater investment in monitoring than other regulatory approaches (EPA 1992). In the case of the acid rain SO2 emissions-trading program, Congress recognized this concern and mandated that all affected facilities install and operate continuous emissions monitoring (CEM) systems to document compliance. In that program, monitoring is required to collect data at 15-minute intervals and report consolidated data on an hourly basis (CAA § 412(b),(c); 42 USC § 7651k(b)(c); GAO 2001b; Swift 2001). CEM systems also played a critical role in the South Coast Air Quality Management District’s RECLAIM trading program for NOx and SO2 emissions (GAO 2001b). The importance of CEM systems raises a challenge for extending the cap-and-trade approach to other pollutants for which cost-effective CEM technology is not available. OTHER TRADING AND VOLUNTARY STATIONARY-SOURCE PROGRAMS In response to a desire to achieve emission reductions in the most cost-effective manner, a number of other approaches have been proposed, tested, and in some cases implemented. These approaches include open-market and other noncapped forms of trading (described briefly in Table 5-3) and voluntary programs, such as Project XL. These efforts are reviewed briefly below. Open-Market and Other Forms of Trading Open-market trading represents an alternative to cap-and-trade emissions trading. In this approach, there is no aggregate emissions cap, usually because there is not a meaningful inventory of emission sources or a mechanism to monitor their emissions. The program is intended to provide an incentive for companies and facilities to reduce emissions voluntarily by participating in the program. Several states, including Michigan, Texas, and New Jersey, have implemented open-market trading programs for VOCs (NAPA 2000). To participate, a company must first establish its

OCR for page 174
Air Quality Management in the United States TABLE 5-3 Open-Market and Other Noncapped Forms of Trading Advantages Disadvantages Open-Market Emissions Tradinga Helps to identify the emission inventory Ill-defined market because emissions are not capped The prospect of new entrants increases economic risk to those participating in the programb The veracity of emission baselines and reductions may be doubted Pollution Offset Tradingc Allows economic growth without Administratively complex to implement emission increases Generally unable to achieve the environmental goals at the least cost aAn emissions trading program without an aggregate emissions cap and without mandatory enrollment. Each source must estimate its baseline emissions and implement an emissions monitoring program. bThe economic risk to those currently participating in the program is that their investments in pollution reduction equipment could be rendered uneconomic due to an infusion of low-price permits. Such a concern could undermine the incentive to make investments in emission reductions. cWhen a source can increase its emissions only by obtaining tradable permits sufficient to offset its impact on ambient pollution, offset trading might facilitate the permitting of the source. individual emissions inventory baseline. The company must then establish a mechanism for monitoring its emissions on an ongoing basis. If the managers of a facility believe that they can reduce emissions at a cost that is below the marginal cost in an established trading program, they have the incentive to join the program and attempt to sell some of the emission reductions achieved. The primary argument for open-market trading is that it provides a way for expanding market-based emission-control programs to reach sources beyond the large stationary sources that are amenable to cap and trade. However, open-market trading has been widely criticized by economists and environmentalists. Economists have criticized the lack of well-defined markets in such programs because emissions are not capped. The possibility of new facilities entering open-market trading and bringing substantial emission permits at low cost is a deterrent to investments by facilities that already participate in the program. Environmentalists have criticized open-market trading because it provides an opportunity for a trading program to collect emission permits for supposed reductions in emissions that might have occurred anyway. Furthermore, the establishment of an emission baseline is problematic, and many doubt the veracity of baselines

OCR for page 174
Air Quality Management in the United States that are established under existing programs. Indeed, recent problems identified with the open-market trading in New Jersey (EPA, 2002n) suggest that greater attention must be paid to documentation of baselines and to compliance monitoring if such programs are to succeed in reducing emissions and gain the trust and support of all parties. Voluntary Programs to Improve Permitting Processes In response to concerns that the current stationary-source permitting process inhibits new and cost-effective emissions control, several innovative permitting approaches have been attempted or proposed over the past decade. These generally have been directed toward (1) improving the administrative efficiency in the permit development and review process; (2) increasing flexibility and responsiveness for the regulated sources, recognizing the need for rapid modification of production processes and products in response to technology and market developments; and (3) improving the cost efficiency and the environmental benefits expected to result from the prescribed emission controls. Most of the proposed and experimental approaches have not been widely adopted, but they demonstrate a continuous search for improved efficiency in the permitting process. Some of these approaches are embodied in the NSR reforms discussed previously. In addition, EPA launched Project XL as part of the Reinventing Government Program initiated in 1995. Project XL encouraged demonstrations of flexible, multimedia (not limited to ambient air emission control), facilitywide permitting. Although several worthwhile demonstrations were initiated under the project, it was discontinued within 5 years and had little national impact on permitting practice. Nevertheless, the Project XL experience is interesting because it demonstrates the difficulties inherent in any attempt to generalize flexible permitting approaches, frequently because the operating details of single sources and the interests of the various local stakeholders are unique to each situation. Although EPA has authority in many cases to make innovative revisions to its regulations to enhance their performance, the limited progress achieved by Project XL, as well as several other flexible permitting initiatives by EPA, states, and other stakeholder groups, was partly due to provisions in existing legislation that make such experiments difficult. In some cases, legislative guidance rather than regulatory experiments may be needed to achieve nationwide improvements in permitting efficiency. AREA-SOURCE REGULATIONS Although emissions from one area source are generally small, the total emissions from all area sources can be significant. Area sources appear to

OCR for page 174
Air Quality Management in the United States be especially important contributors to O3 pollution (via the emissions of VOC) and PM. In recognition of their importance, states first began working together to develop rules to control VOC emissions from some types of area sources in the late 1980s. For example, California and other states began adopting product formulation requirements for low-VOC architectural coatings. These were eventually translated into national requirements in the 1990 CAA Amendments, which contained several provisions requiring EPA to identify and regulate area sources of such products. A major impediment to making progress on area-source emissions arises from the large number of uncertainties associated with emission inventories for these sources. Specific challenges include the many sources in any given category and the wide variation in the conditions and operating practices under which the emissions can occur. To address the problem, the Emission Inventory Improvement Program (EIIP) includes the development of inventories for area sources (EPA 2001b), but it is unclear, barring a more ambitious program of measuring and testing, whether such an inventory-development process will result in a more accurate understanding of the contribution of area sources to air pollution in the United States. Section 183(e) of the 1990 CAA Amendments required EPA to conduct a study of consumer and commercial products by 1993 to identify products that needed regulation and then to proceed with developing control technology guidelines (CTGs) for each priority category. EPA completed the study in 1995, identifying four groups of consumer and commercial products, including such items as aerosol spray paints, architectural coatings, and industrial cleaning solvents. EPA proceeded with developing CTGs for the categories of products within those groups, albeit slowly. Regulations for the first group were scheduled to be in place by 1997, and regulations for all four groups were to be completed by 2003 (64 Fed. Reg. 13422 [1997]). Currently, however, only three of the six product categories within the first group have been regulated; no regulations for the other three groups (scheduled for completion 1999, 2001, and 2003) have been completed. EPA noted in its 1999 Federal Register notice (64 Fed. Reg. 13422) that it “intends to exercise discretion in scheduling its actions under section 183(e) in order to achieve an effective regulatory program.” The slow pace appears to be due to the perception that such rules are not central to efforts to attain the NAAQS for O3. However, in the absence of high-quality inventories and the contributions of these sources to pollution, it is difficult to verify whether the perception is accurate. Title III of the 1990 CAA Amendments also requires EPA to list and regulate area sources of HAPs. In some cases, MACT standards have been developed for several categories of area sources, including commercial dry cleaning, hazardous waste incineration, secondary aluminum production, and secondary lead smelting. The amendments also required EPA to address

OCR for page 174
Air Quality Management in the United States area sources in an integrated urban air toxics strategy. The 1999 strategy document identifies 32 high-priority HAPs (plus diesel exhaust) and lists 29 area-source categories that have been or will be targeted for regulation, including 13 new categories for which standards are to be issued by 2004. For area sources, the agency has the discretion to issue either MACT standards or less stringent generally available control technology (GACT) standards. EPA is now proceeding with implementing these regulations. Status of Area-Source Controls To date, the efforts to control area sources have been relatively scattered and have slipped far behind mandated implementation schedules. Recent efforts, such as the integrated air toxics strategy, have begun to bring some strategic vision to the control of such sources. However, in the absence of a high-quality inventory of such sources, it is nearly impossible to quantify their emission contributions and to set priorities. Yet, those few analyses that have been done (for example, by EPA in the HAPs inventory) suggest that area-source emissions are significant and will be even more important after the imposition of MACT has reduced emissions from the major industrial sources. A renewed focus on improving area-source inventories and controlling key sources at an accelerated pace will be important in crafting effective strategies to reduce emissions and exposure to many important pollutants. SUMMARY OF KEY EXPERIENCES AND CHALLENGES FOR STATIONARY-SOURCE CONTROL Strengths of Stationary-Source Control Programs CAA programs have achieved substantial reductions of pollutants from existing stationary sources, for example, SO2 (through the Acid Rain Program), VOCs (for example, through RACT), and HAPs (through MACT). The Ozone Transport Commission, and the upcoming NOx SIP call trading program have the potential to achieve substantial reductions in NOx emissions. NSR and PSD requirements have encouraged the continuous development and application of cleaner technologies and emission controls for major new stationary sources and have resulted in reduced emissions from those sources. The development of emission cap-and-trade programs that use market forces to limit the cost of pollution control has provided the AQM system with a mechanism that is capable of achieving substantial emission reductions at reduced costs.

OCR for page 174
Air Quality Management in the United States Limitations of Stationary-Source Control Programs7 Although the NSR and PSD programs appear to have been effective for new facilities, such programs and related economic factors (1) provide an incentive for industries to extend the life of higher-emission (grandfathered) facilities and (2) lead to litigation alleging facility modifications (primarily to extend their useful lifetime) without prior approval. The NSR and PSD programs do not affect a large fraction of the existing facilities that have not undergone major modifications but have remained in operation. With the exception of CEM, there is limited ability to quantify stationary-source emissions. Facility-specific emission standards have lowered emissions, but because they are often production-based standards, the potential remains for emissions to increase as economic activity or product demand increases. The next phase of control on HAP emissions is predicated on the conclusions of a residual risk analysis that is fraught with scientific uncertainty. Achieving the full potential of cap and trade will require applying the technique to a broader range of pollutants, implementing a less cumbersome process for revising caps and targets, developing enhanced and more cost-effective CEM and other monitoring technologies, and guarding against deleterious geographical and temporal distribution of emission reductions. Controls on area sources lack focus and are hampered by a large number of uncertainties in the magnitude of the emissions from these sources. To date, many emission-control programs for stationary sources have addressed pollutants separately, resulting in different time lines and different requirements for reductions of various pollutants at the same facility. That approach can raise the cost of emission control without adding any appreciable benefit in emission reductions. Recent legislative proposals for multipollutant reductions at electricity-generating facilities offer an opportunity to merge one of the most successful techniques—cap and trade—with a multipollutant approach. This merger would enable these facilities to develop long-term plans for capital improvements that minimize costs and reduce all relevant pollutant emissions at the same time. 7   Recommendations are provided in Chapter 7.