The Proposed WEPCo Rule: Making the Problem Fit the Solution

Richard E. Ayres and Richard W. Parker

Editors' Summary: EPA's final decision on its proposed WEPCo rule, which addresses how new Clean Air Act provisions apply to electric utilities, is expected soon. This Article provides a glimpse into the regulatory machinery needed to deal with implementing just one aspect of the Clean Air Act Amendments of 1990: whether an electric utility's proposed renovations at one of its facilities constitutes a "modification" triggering new source performance standards and new source review programs. Although Congress passes legislation and the President signs bills into law, executive branch agencies are usually tasked with the regulatory implementation of those laws. Thus, what Congress writes and the President signs into law is often not what interested parties expect. The Article first explores the origins of the controversy, the parties involved, and reaction to EPA's first ruling on the issue. The authors then analyze EPA's response, the proposed WEPCo rule, and evaluate its potential ramifications if made final. Finally, the authors conclude that the proposed rule softens the modification rule as applied to electric utilities in pursuit of utility cost savings, but possibly at the expense of local impacts and technological innovation.

Richard E. Ayres is a partner with O'Melveny & Myers in Washington, D.C., specializing in environmental and related energy regulatory matters. Prior to joining the firm, he was Senior Attorney at the Natural Resources Defense Council, which he cofounded in 1970. From 1975 to 1991, he was also cofounder and chairman of the National Clean Air Coalition, where he led successful efforts to reshape the federal Clean Air Act in 1977 and 1990. He was appointed by the President to the National Commission on Air Quality, serving from 1979-81. Over the past 20 years, he has been involved in most of the major rulemakings under the Clean Air Act, and has handled numerous air pollution related cases in the federal courts. He earned his law degree and an M.A. in political science from Yale University in 1968.

Richard W. Parker is an associate with O'Melveny & Myers specializing in the practice of environmental law. He earned his law degree from Yale University in 1985.

[22 ELR 10201]

In 1987, the Wisconsin Electric Power Company (WEPCo) decided to overhaul five units at its 50-year-old Port Washington power plant, rather than scrap the plant and build a new one as previously planned. Under the new source performance standards (NSPS) and the new source review (NSR) programs promulgated under Title I of the Clean Air Act,¹ any project involving "new" construction or a "modification" of an existing plant must incorporate advanced pollution control technologies and undergo air quality impact review.² WEPCo wanted to know whether its proposed overhaul of the Port Washington facility would constitute a "modification" triggering the NSPS and NSR requirements.

In 1988, the U.S. Environmental Protection Agency (EPA) notified Region V officials in a letter³ that the NSPS and the NSR applied to WEPCo's proposed renovations, which triggered widespread protest, a court challenge by WEPCo,⁴ and strong political pressure for a corrective ruling by EPA.⁵

In response, EPA issued its "WEPCo fix" as a proposed new rule on June 14, 1991.⁶ The proposed rule goes beyond the confines of EPA's original WEPCo decision⁷ and constitutes a victory for utilities on all fronts. As proposed, the rule would ease the test for determining whether a particular utility plant change is a "modification" and would grant a broader "safe harbor" from modification status for utility pollution control projects. In addition, it would lighten the burden on a utility making modifications by holding that relatively inexpensive "low-NO_x" burners will satisfy the "best available" technology requirements for nitrogen oxide (NO_x) emissions.

EPA's proposed WEPCo rule continues to arouse controversy for several reasons. For example, allegations of improper influence by the administration in EPA's formulation of the proposed rule have been made.⁸ In addition, the proposed rule may have a significant impact on utility [22 ELR 10202] costs and emissions levels. Further, it could set a precedent for the treatment of manufacturing firms generally.

Significance of the WEPCo Issue

The most puzzling aspect of the WEPCo controversy may be that it exists and is so intense. Why does it matter so much whether particular changes are "modifications" for purposes of applying the NSPS and the NSR? The reasons relate to the size of the electric utility pollution problem and the tangled history of attempts to reduce it.

Magnitude of Utility Emissions

Electric utilities are the largest single source category of sulfur dioxide (SO₂) and NO_x emissions, the two principal precursors of acid rain. While approximately 23 million tons of SO₂ and 19 million tons of NO_x are emitted annually from all sources in the United States, about 16 million tons of SO₂ and 7 million tons of NO_x are emitted by electric utilities.⁹ Thus, any rule affecting the environmental obligations of electric utilities is of major environmental importance. The proposed WEPCo rule raises fundamental issues about the nation's past and future efforts to clean up utility air emissions.

History of Efforts to Curtail Utility Pollution

Shortfalls of localized measures. Traditionally, air pollution control has been conceptualized primarily in local terms. Title I of the Clean Air Act, as enacted in 1970, requires states and localities to adopt programs to curtail emissions sufficient to achieve health-based ambient air quality standards¹⁰ in "nonattainment areas," which currently fail to meet such standards. In 1977, Congress added the prevention of significant deterioration (PSD) program, which seeks to prevent significant deterioration of air quality in regions that meet the national ambient standards.¹¹

Sources subject to the NSPS must install the "best demonstrated" pollution control technology.¹² Projects subject to the NSR must install pollution control technology that is at least as sophisticated as the applicable NSPS.¹³ In addition, sources subject to the NSP and the NSR must undergo public and agency air quality review.¹⁴ In this review, a source located in a "clean air" area must demonstrate that its emissions will not consume more than the available "increment" of pollution allowed in that area. A source in a so-called nonattainment area must obtain greater than one-for-one offsets for any increased emissions attributable to that source. These requirements apply to new plants and to plants that undergo major modifications. Existing plants that do not undergo a modification are generally exempted from both NSR and NSPS regulations, except in nonattainment areas where existing sources occasionally may be required to retrofit control technologies. Implicit in this focus on new and modified plants is the expectation that local air quality goals are sufficiently rigorous and that they will become easier to attain as new, or modified, clean plants replace old, more polluting, plants.

Subsequent developments have forced a reexamination of this premise. In the early 1970s, William Ruckelshaus, EPA's first administrator, decided to permit utilities to achieve local air quality goals by using tall smokestacks to disperse pollution, rather than reducing emissions.¹⁵ Tall stacks allowed the low-cost improvement of local air quality, but did so at the expense of broadcasting pollutants into areas previously untouched.

The second development was gradual recognition of the Methuselah-like persistence of large, old power plants. By 1985, the typical large coal-fired plant was 30-50 years old. Only a third of the nation's coal-fired generating capacity was operating under new source controls.¹⁶ 1985 SO₂ emissions from coal-fired electric utilities measured 15 million tons per year, the same as in 1970, while NO_x emissions increased from 3.5 million tons in 1970 to 6.5 million tons in 1985.¹⁷ By the late 1980s it was clear that old power plants were not being phased out at a rate sufficient to achieve reductions in overall utility emissions.

Indeed, utilities concluded in the 1980s that it was much [22 ELR 10203] cheaper to "life-extend" an existing power plant than to scrap it and build anew. As a result, the U.S. Department of Energy (DOE) predicted in 1990 that over 70 percent of the nation's 1989 fossil-fuel generating capacity will be life-extended by 2010.¹⁸ Applying the NSR would increase the cost and administrative burden of these projects. But not imposing the NSR on life-extension projects could indefinitely postpone the cleanup of a substantial number of these large, dirty power plants.

The third development that forced a revision of thinking was the scientific discovery of the environmental¹⁹ and public health²⁰ effects of acid rain precursors carried by prevailing winds beyond the immediate vicinity of the source. This led to widespread recognition, beginning in the 1980s, of the need for a federal program to control atmospheric loading across the nation. In a 1983 study,²¹ the National Academy of Sciences took the position that for purposes of developing a policy to reduce acid deposition, an area at least as large as the eastern half of the United States could be considered a single "airshed." In gross terms, emission reductions occurring anywhere within this airshed would help reduce damages from acid deposition across the airshed.

The concept of a regional or national airshed, within which reductions anywhere carried benefits for the region or nation as a whole, naturally led to the notion of a regional or national system of emission reduction trading,²² and ultimately to an explicit pollution rights "market." A trading system (or market) would allow sources that could reduce emissions cheaply to transfer (or sell) pollution rights to sources facing higher costs of compliance, thereby yielding greater efficiency systemwide. The result of this thinking was the "market-based" acid rain program proposed by President Bush and enacted in 1990 as Title IV of the Clean Air Act.²³

The Nationwide Acid Rain Program. Under Title IV of the Clean Air Act, aggregate SO₂ emissions from utility power plants are to be cut by approximately half from 1980 levels and permanently capped by the year 2000.²⁴ Sources will be allowed to trade SO₂ pollution rights ("allowances," defined as the right to emit one ton of SO₂), within the overall cap. NO_x emissions from power plants will not be capped per se, but many existing power plants will have to install low-NO_x burners or their cost equivalent.²⁵ This promises to provide significant reductions in aggregate NO_x emissions. Title IV provides no additional controls for particulate emissions (soot is not an acid rain precursor), but particulates from power plants are a diminishing aspect of the nation's air pollution problem.²⁶

Title IV added a new issue to the WEPCo controversy. By mandating for the first time major SO₂ and NO_x emissions reductions from old power plants,²⁷ Title IV created the prospect that pollution control projects undertaken to comply with Title IV might be considered "modifications," which would trigger the NSR requirements of Title I. A dilemma was presented. A policy of treating all pollution control projects as "modifications" would greatly amplify the cost of such projects.²⁸ Yet a policy of categorically exempting these projects would be unacceptable on environmental grounds, because installing a "pollution control" device can result in increased emissions of co-pollutants. In fact, a pollution control project might even result in increased emissions of the controlled pollutant, if plant utilization subsequently increased by more than an offsetting amount.

Enactment of Title IV also precipitated a policy debate over the relative emphasis to be given to the preexisting new source requirements of Title I. If reducing total national or regional [22 ELR 10204] SO₂ emissions is the policy goal, a flexible, market-based acid rain program is a more economically "efficient" means of achieving that goal than NSR and NSPS requirements of Title I. There is wide speculation that this reasoning led the White House staff to push for a WEPCo ruling that would deemphasize the new source requirements of Title I. Such is the effect of the proposed WEPCo rule.²⁹

The problem with this reasoning is that Title IV, with its emphasis on controlling nationwide aggregate SO₂ emissions, does not purport to serve the localized environmental and public health values protected by Title I. Title IV does not cap loadings of pollutants other than SO₂. It does not require the best available NO_x controls. Although it offers some incentives for advanced SO₂ controls (market forces will ensure that some, but not all, power plants install them), Title IV does little to further the "technology-forcing" objectives of Title I.

Most important, Title IV does not prevent unacceptable local concentrations of pollution. Despite the widespread use of tall stacks, utility emissions are not exclusively a national problem. Most of the time, the majority of a power plant's emissions still come down to earth within range of the power plant. Because Title IV places no limit on local impacts, it offers no protection to residents living near power plants that decide to buy allowances, rather than clean up their emissions. The growth of a nationwide market for allowances will virtually guarantee that there will be a substantial number of such power plants. Title IV is thus the indispensable means for regulating local impacts. It is not surprising, then, that the Clean Air Act expressly provides that the requirements of Title I shall be unaffected by passage of Title IV.³⁰

The Potential Precedential Effect of the WEPCo Rule

The third factor fueling the WEPCo controversy is recognition that the outcome may have precedential force for regulating other industries. EPA is bound by an existing court-approved consent agreement with the Chemical Manufacturers Association to propose, in early 1992, a new modification rule for nonutility sources.³¹ Utilities traditionally have been treated simply as a category of industrial polluter, which suggests to some the extension of the WEPCo rule to manufacturing firms at large. By prior agreement, EPA's proposal will include both a continuation of past practice and a test even more favorable to industry than the one contained in the proposed WEPCo rule. The WEPCo rule would thus occupy the middle ground between more extreme alternatives and might well win acceptance as a "compromise" candidate. Utilities have paved the way for this result by publicizing their view that the proposed rule represents not only good policy but the only correct interpretation of existing law.³²

Analysis of the Proposed WEPCo Rule

The Clean Air Act grandfathers (exempts) old plants from the NSR unless they undergo a "modification." The Act defines "modification," for NSPS and NSR purposes, as a physical or operational change that results in increased emissions from the facility.³³ But what is a physical or operational change? How does one measure "increased emissions" before the fact, or determine whether a change will "result in" increased emissions? EPA's proposed WEPCo rule would make four principal changes in the rules for applying the modification test to electric utility power plants.

A New Pollution Control Project Exclusion From the NSR

The preexisting rule. NSPS regulations traditionally have exempted pollution control projects by excluding the addition of pollution control equipment from the category of "physical or operational change." They also have excluded the removal or replacement of pollution control equipment unless EPA determines that the change is not "environmentally beneficial."³⁴ This means that pollution control projects generally have not been subject to an NSPS emissions test.

EPA's NSR regulations contain no comparable exemption. Although EPA once took the view that the NSPS exclusion is automatically incorporated into the NSR program by virtue of the Clean Air Act's single definition of "modification," the Agency changed its view in 1986.³⁵ Since then, EPA has maintained that the "special orientation" of the NSR program towards protecting ambient air quality precludes automatic incorporation of the NSPS exclusion [22 ELR 10205] into NSR rules.³⁶ Thus under the existing regime, the source bears the burden of demonstrating the compatibility of the plant, as altered by the pollution control project, with the local air quality objectives.

Agency Disagreement

EPA opposed incorporating the NSPS exclusion into NSR rules because of its concern that installing a control device for one pollutant may result in increased emissions of a co-pollutant, or even of the same pollutant if utilization increases by more than an offsetting amount. For example, installing a scrubber will reduce SO₂ emissions, but not emissions of NO_x or particulates. If the utility then uses the scrubbed power plant more intensively, the result could be increased NO_x and/or particulate emissions, even if a low-NO_x burner were installed. Passage of the acid rain program enhanced this concern by creating a market-based system that gives utilities monetary incentives to maximize their utilization of "scrubbed" or "fuel-switched" facilities.

EPA considered that the statutory purpose of the modification rule is to focus the NSR at a "logical and efficient" point, namely, the occasion of a major capital expenditure or long-term decision regarding pollution levels.³⁷ Installing a stack-gas scrubber, or physically altering the plant to accommodate a lower-sulfur coal, would appear to be just such an occasion and should not bar the NSR if the project seems likely to result in increased emissions from the plant.

Accordingly, in the interagency deliberations leading up to the proposed rule, EPA favored, at most, a limited NSR pollution control exclusion. The limited exclusion would employ a rebuttable presumption that adding a pollution control device would not increase emissions from a source. The presumption would be rebutted whenever "there is a clear likelihood that the addition of the control system or device would increase the utilization [and hence the emissions] of the source."³⁸ The practical effect of this rule would have been to incorporate a de facto emissions test into the determination of whether a pollution control project is a "physical change" that triggers an emissions test. The DOE viewed this as a sort of "non-exemption" and pushed for something more.

The DOE sought a broad exclusion from the NSR equivalent to the existing NSPS exemption.³⁹ The DOE argued that the single statutory definition of "modification" in both the NSPS and the NSR programs compels a unitary regulatory definition as well. Also, many utilities will be undertaking pollution control projects in coming years to comply with new acid rain requirements. The DOE felt they deserved a broad NSR exclusion for these projects.⁴⁰ Finally, the DOE argued that the current definition of "physical or operational change" (the first prong of the "modification" test) already excludes changes involving only "increased hours of operation and production rates," even if the net result is increased emissions from the unit.⁴¹ Any plant physically and legally capable of increasing its hours of operation apart from the pollution control project ought to be able to do so, the DOE argued, without triggering the NSR.

The Proposed Rule

The result of this interagency clash is a proposed rule that significantly shifts the burden of protecting air quality. Whereas the previous rule required sources to demonstrate compliance with air quality requirements through the NSR process, the proposed rule places the burden of protecting air quality on pollution control officials. The proposal would exclude from NSPS and NSR review the addition, replacement, or use of a "pollution control project" except when the reviewing authority (1) determines that the proposed change is "not environmentally beneficial," (2) has reason to believe that the project would result in a significant net increase in the source's projected emissions over the level used for that source in the most recent air quality impact analysis, or (3) determines that the increase will cause or contribute to a violation of any national ambient air quality standard (NAAQS), PSD increment, or visibility limitation.⁴² "Environmentally beneficial" would not be defined explicitly. Its interpretation would be left to the discretion of the reviewing authority, which would reserve the right to require additional air quality modeling, on an ad hoc basis, whenever it determines that the last two conditions may not be met.⁴³

An exclusion for fuel switching. Because much of the nation's low-sulfur coal supply is lower in energy content [22 ELR 10206] and higher in ash than higher sulfur coals, switching to low-sulfur coal can increase particulate emissions unless particulate controls are upgraded at the same time. Switching to low-sulfur coal accompanied by increased plant utilization can increase yearly emissions of NO_x and SO₂ as well. The existing EPA rule allows fuel switching without the NSR only if the unit was physically capable of accommodating the switch as of a specified date.⁴⁴ The proposed rule goes further by defining "pollution control project" to include changes made to accommodate switching to a less polluting fuel.⁴⁵ This means that decisions to switch to less polluting low-sulfur fuels would receive the same treatment accorded to all other pollution control projects.

The determination of whether life extensions and other physical and operational changes result in "increased emissions." Once a project is determined to be a "physical or operational change," the second test of a "modification" is to determine whether it will increase emissions from the source.⁴⁶ EPA regulations apply different tests for "increased emissions" under the NSPS and the NSR programs, reflecting "the fundamental distinctions between the technology-based provisions of NSPS and the air quality-based provisions of NSR."⁴⁷ Here lies the core issue of the WEPCo controversy: how strictly to scrutinize major capital expenditures at existing plants to determine whether they will result in "increased emissions" from the source.

NSPS Emissions Test

To determine whether a change in a unit will increase emissions, thereby triggering the NSPS, EPA traditionally has compared maximum hourly emissions rates, expressed in kilograms per hour, immediately before and after the change. In making the comparison, the permitting authority focused on the unit's actual capacity, which may be considerably smaller than design capacity if, as in the WEPCo case, the plant was wholly or partially idle at the time of the change.⁴⁸

EPA's proposed WEPCo rule would relax this test. It would allow the utility to compare the post-change emissions rate with the "maximum hourly emissions rate achievable at that unit during the five years prior to the change."⁴⁹ The proposed rule does not clarify whether the baseline would be maximum achievable rate at any time, or the maximum achievable rate that is representative of the plant's actual capacity over the period; the preamble seems to support both interpretations.⁵⁰ Most commentators have accepted the former interpretation as the intended one.⁵¹ If this is indeed the test, the proposed new rule would offer plant operators wide latitude to pick and choose a high baseline emissions rate for determining whether emissions rates would increase after a change.

NSR Emissions Test

Whereas the NSPS test focuses on maximum hourly rates, the NSR test focuses on total yearly emissions.⁵² Emissions rates can be estimated with some accuracy on the basis of physical tests or parameters known at the time of the change. However, comparing past and future yearly emissions requires choosing a representative baseline and an estimate of future capacity utilization.

The Pre-WEPCo Rule

Prior to WEPCo, EPA applied an approach that restricted utility options in choosing baseline emissions levels and provided a strong incentive for utilities to agree to enforceable limits on post-modification emissions. Utilities were required to use the immediately preceding two-year period as the baseline unless they could persuade EPA that a prior period was more representative.⁵³ Whenever a plant had not yet "begun operation" (i.e., prior to recommencing operations after a major change), EPA would compare (1) actual average yearly emissions during the baseline period with (2) the potential to emit after the change, after subtracting any contemporaneous reductions achieved at the source.⁵⁴

To define the post-change "potential to emit," EPA offered two options. If the operators were prepared to enter into a federally enforceable emission limit, then that would define the potential to emit. If not, EPA would assume emissions that reflected operating 24 hours per day, 365 days per year.⁵⁵ A requirement of federal enforceability ensured that the utility would offer a forecast of actual capacity utilization realistic enough for the utility to abide by. After the projection, enforceability ensured that utilization [22 ELR 10207] did not increase beyond the projected levels and that emissions did not rise.

The WEPCo Case

The Wisconsin Electric Power Company, however, persuaded the Seventh Circuit Court of Appeals to take a less benign view of this test. Rather than see the actual-potential test as a means of extracting realistic projections of likely actual emissions, the court dismissed the test as simply an "apples-oranges" comparison stacked against the utility.⁵⁶ The court opined that plants undergoing "like-kind replacement" of existing equipment should be deemed to have "begun operations" and therefore should not trigger the actual-potential test.⁵⁷ The court remanded the case to EPA with instructions to apply an "actual-projected actual" test to the WEPCo facility, but left EPA the option of returning to its original approach after notice and comment rulemaking.

Agency Disagreement

After WEPCo, there was no question of going back to the pre-WEPCo status quo. The court decision became the starting point for an interagency debate between the DOE and EPA over how, not whether, the NSR emission test should be reformulated.

The DOE argued for a potential-potential test that would compare the maximum potential yearly emissions of the plant before and after the change in much the same way as the NSPS test currently compares maximum hourly emissions rates.⁵⁸ This would conform NSPS and NSR regulatory definitions of "modification." But it would exclude all physical changes from the NSR unless they significantly increased the operating capacity — and emissions rate — of the plant. The DOE approach would ignore both capacity utilization and the overall level of yearly emissions from the plant.

EPA opposed the DOE approach because it believed that economic and physical constraints had long kept many power plants operating below their design capacity. As EPA pointed out to the White House with some alarm, "[because of the low rates of actual capacity utilization at many utility power plants] the 'potential-to-potential' test of the DOE approach would allow real world emissions to increase by a factor of two or three without consideration of air quality impacts."⁵⁹

The Proposed New Rule

The proposed new rule for NSR emissions testing once again represents a "compromise" between DOE and EPA positions. It would provide an actual-to-actual test for all electric utility steam generating units (other than new units and replacements of existing units) but without a federally enforceable limit on future actual emissions.⁶⁰ Therule also would give utilities added flexibility in choosing the components of the test. Under the proposed new rule any two consecutive years within the prior five years would be presumed representative baseline years.⁶¹

The other side of the comparison, future emissions, would be adjusted by the creation of a new concept: representative actual annual emissions. This figure would be derived by multiplying (1) the hourly emissions rate based on the unit's post-change operational capability, or a smaller rate provided it is guaranteed by a federally enforceable limitation, by (2) projected capacity utilization based on the unit's historical utilization and all available information, but without any requirement for a federally enforceable limitation on capacity utilization.⁶²

From this number, the utility would be allowed to subtract the portion of increased emissions arising from factors unrelated to the particular change, including any increased "utilization due to the rate of electricity demand growth for the utility system … as a whole."⁶³ The only restriction on the right to subtract projected future demand growth in arriving at "representative actual annual emissions" would be that the plant must have been physically able and legally permitted to accommodate the demand growth during the representative baseline period chosen by the utility.

Utility BACT presumption for NO_x. The final feature of the proposed WEPCo fix bears no direct relation to the WEPCo decision. It proposes that sources subject to NSR requirements (because they are new or are found to have undergone a "modification") may meet the "best available control technology" (BACT) requirements of the PSD program by using the NO_x control technology required under § 407 of the Clean Air Act (the acid rain program). Section 407 calls for the use of "combustion modification and/or low-NO_x burners or similar, cost-effective technologies."⁶⁴ However, traditional NSR standards contain no such limitation and would permit a requirement of technologies considerably more advanced than low-NO_x burners.

[22 ELR 10208]

The Proposed WEPCo Rule: An Evaluation

The proposed WEPCo fix raises a number of significant legal and policy issues arising from the proposed pollution control exemption, the emissions tests for determining whether life-extension projects are modifications, and the treatment of low-NO_x burners.

The Pollution Control Exclusion

The proposed exclusion for pollution control projects at utility power plants reflects, in part, EPA's recognition of the fact that many existing plants will be required either to switch to lower sulfur fuels or to retrofit pollution control equipment in order to comply with the acid rain program created in Title IV of the Clean Air Act. The pollution control project exclusion seeks to avoid biasing the utility towards any particular compliance strategy by assuring, within broad bounds, that neither strategy is likely to trigger the NSR.⁶⁵

However, there is a danger of adverse local air impacts resulting from the new exclusion. Title IV will create significant incentives for utilities to maximize utilization of particular facilities where SO₂ control measures have been implemented, because SO₂ allowances will soon have monetary value. The result may be increased emissions of NO_x, particulates, and sulfuric acid mist, which may adversely affect residents in the vicinity of these power plants. Affected citizens may find little comfort in the observation that preexisting regulations would have permitted the same result absent a pollution control project. The onus will be on local and state officials to prevent or to cope with the resulting local air impact.

The proposed restrictions on the exclusion — that it must be "environmentally beneficial," and must not invalidate planning assumptions about that source's emissions or violate an NAAQS standard, PSD increment, or visibility limitation — have limited preventive value. The "environmentally beneficial" test of pollution control projects has long inhabited the NSPS regulations, where it has suffered from vagueness and infrequent use. Moreover, a source could significantly increase emissions, worsen local air quality, and crowd out opportunities for new sources to enter the area without flunking any of these tests. The EPA proposal would provide no notice to the public of a proposed pollution control project and would afford no mandatory occasion for the relevant tests to be applied. Even if EPA chooses to scrutinize a project, the test for increased emissions would involve the "representative actual annual emissions" concept that, it will be seen, is highly subjective and arguably downward biased.⁶⁶

There is undeniable merit in the view that utilities should not be penalized or deterred from undertaking pollution control projects. By the same token, it seems clear that there needs to be some realistic supervision of what passes for a "pollution control project" and the likely impact of the project on the local environment. It would not appear unreasonable to ask utilities to file a forecast of future utilizations and emissions from the facility in question, with and without the demand growth exclusion factored in, on the occasion of each major, new "pollution control" project. Permitting authorities could use this forecast beforehand in ascertaining the environmental benefit of the project and afterwards in analyzing actual emissions.

EPA also may wish to reexamine its policy of allowing unlimited increases in "hours of operation" without the NSR. Utilities need flexibility to vary their use of particular plants periodically, up and down, in accordance with the vagaries of the seasons, local and national business cycles, and dispatch considerations without triggering the NSR. But when a utility makes a long-term, strategic decision to increase the utilization of a power plant from 30 percent on average to 60 percent on average over an extended period, it is hard to understand why this is not an "operational change." Such a change, if accompanied by a significant emissions increase, would appear to require the NSR under the statutory definition of "modification," whether or not the change is accompanied by the installation of a pollution control project.

NSPS and NSR Emissions Increase Tests

The proposed amendments to the emissions tests in the modification rule raise a number of issues relevant to the way baseline emissions and post-change emissions are calculated.

Calculation of the Baseline

The proposed amendments to NSPS and NSR baselines will allow utilities a broad scope to choose a favorable baseline. Under the NSPS program, they can choose the "maximum rate achievable" at any time over the past five years, and under the NSR they can choose annual emissions over any two consecutive years within the prior five years. Yet the stated purpose of the emissions test is to compare "representative" emissions before and after a physical change. The proposed rule creates a disjunction between means and ends.

If EPA wishes to have "representative" baseline emissions figures used in the emissions test, the Agency would be better served to simply require an average of the last five or ten years of emissions, possibly throwing out the high and low extremes. In this way, aberrations would tend to cancel out, yielding a more "representative" figure.

Likewise, the proposed changes in methods for calculating post-change emissions raise significant issues of reliability and bias. Projections of utilization and demand growth are notoriously subjective and difficult. Moreover, under the proposed rule the utility's projections are not binding. An investment that extends the life of a pollution source by 15-20 years can be sheltered from the NSR on the basis of a two-year projection of emissions and demand growth. After two years, the utility is not bound and may increase utilization and emissions from the plant. Indeed, EPA does not propose to require the utility to file its projections with air regulators unless it appears that net future actual emissions, adjusted to exclude those attributable to systemwide demand growth, will significantly increase.

The foregoing difficulties clearly suggest the need for some filing requirement on the occasion of a "physical change," whereby the utility would present to air regulators the utility's projected future actual emissions, the projected demand growth exclusion, and the principal assumptions [22 ELR 10209] underlying each projection. EPA also should consider installing a "feedback loop" whereby significant, unforecasted increases in emissions after a "physical change" may be remedied by requiring the installation of add-on controls or other effective measures.

The Demand Growth Exclusion

EPA makes a prima facie argument for the demand growth exclusion. In the preamble to the proposed rule, the agency explains that to be a modification, a physical change must "result in" increased emissions.⁶⁷ Thus, an emissions increase arising from independent factors, such as systemwide demand growth that would have occurred and affected the unit's operation with or without the physical change, does not result from the change and may be excluded from projected future emissions. If, on the other hand, increased utilization derives from changes undertaken to improve the cost-effectiveness of the unit, any increased emissions arising therefrom would be countable future emissions.⁶⁸

Such reasoning seems plausible at first glance, but it leads to further quandaries. The hypothesized distinction between a physical-change-to-accommodate-demand-growth and a physical-change-to-improve-efficiency collapses, or at least blurs, in practice. Many units are refurbished precisely to allow them to absorb demand growth more cost-effectively. Put another way, why would any rational utility spend the money to refurbish a power plant if the plant could maintain production and absorb systemwide demand growth without the refurbishment?

Then there is the problem of supervision and enforcement. The proposed rule offers utilities abundant opportunities to attribute all of a unit's increased operations to even nominal demand growth.⁶⁹ This opportunity arises because the test is self-judging, unless the utility seeks regulatory guidance or intervention. The utility, therefore, is not necessarily compelled to take consistent positions over time in allocating growth to various units.

The demand growth exclusion leads to a further, ironic consequence. It is evident that emissions increases attributable to demand growth have a greater impact on overall air quality than comparable increases arising from substitution generation, since the latter, by definition, will result in decreased emissions at some other location. Yet, under the proposed new rule, increased emissions attributable to systemwide demand growth would be sheltered from air quality impact review, while identical increases attributable to substitution generation would trigger the NSR.

Nor is there much logic in EPA's proposed restriction holding that "during the representative baseline period … the plant must have been able to accommodate the projected demand growth physically and legally even absent the particular change."⁷⁰ Most plants that are refurbished have entered, or are about to enter, a downward spiral of deteriorating performance. Otherwise, they would not be refurbished. Logically, the relevant question is not whether the plant could have carried the projected load during the "baseline" years preceding the change, but rather whether the unit, without refurbishment, could have met its projected share of growth cost-effectively in the future years when the demand growth is expected to occur. The answer to this question will almost always be no if the original decision to refurbish the plant was economically rational. By focusing on the physical capacity of the power plant at an irrelevant point in time, the proposed test overstates the demand growth exclusion and thereby understates "countable" future emissions.

EPA has previously opined that the modification rule set forth in the Clean Air Act reflects a congressional intention to focus the NSR at an "efficient and logical" point or, stated otherwise, "at the time substantial capital investments in, or other long-term decisions regarding, pollution generating facilities were made."⁷¹ EPA's proposed demand growth exclusion represents a sharp departure from this guiding principle of past practice in that EPA now proposes to permit a long-term decision to increase pollution without the NSR so long as the increase is attributable to projected demand growth. Under these circumstances, there is no clear and persuasive legal requirement, or policy justification, for the demand growth exclusion proposed by EPA.

The Presumption in Favor of Low-NO[x] Burners

Low-NO_x burners can reduce emissions of NO_x by approximately 20-60 percent (compared to 80-90 percent SO₂ reductions available from stack gas scrubbers) at a low cost of $ 8-$ 16 per kilowatt.⁷² More advanced NO_x reduction technologies, including selective noncatalytic reduction (SNCR) and selective catalytic reduction (SCR) devices, offer substantially greater reductions in NO_x emissions. EPA states that "SNR and SNCR are not in use in this country as retrofit technologies for coal-fired boilers, and … [c]urrent estimates of control costs for these technologies are much higher than for low-NO_x burners, especially in the context of retrofitting existing units."⁷³ Yet spokespersons for the SNCR industry claim that there have been "[e]xisting retrofits [of SNCR] on coal/oil/gas units in U.S. and Europe" at a cost of $ 3-$ 7 per kilowatt.⁷⁴ There clearly is a need to reconcile conflicting factual claims about the cost and capabilities of alternative technologies.

The existence of conflicting technology claims raises further questions about the validity of EPA's proposed presumption. Title I of the Clean Air Act expressly mandates case-by-case review of BACT requirements.⁷⁵ EPA attempts to harmonize the proposed rule with the statute by explaining in the preamble that the presumption

does not purport to relieve the permitting authority of the obligation to weigh the statutory factors in reaching BACT determinations. Rather, it reflects an exercise of policy judgment by EPA where it is the permitting [22 ELR 10210] authority that in most cases a BACT analysis … would lead to the selection of low-NO_x burners …⁷⁶

The use of "would" in this passage suggests, however, that BACT analysis, in most cases, would not be carried out in practice. Without rigorous BACT analysis there would be no basis for rebutting the presumption. A presumption thus shielded from rebuttal would be, effectively, a rule exempting decisions to install low-NO_x burners from case-by-case BACT review.⁷⁷

As new technologies come on line and established ones reveal more clearly their relative commercial merits, any presumption in favor of a particular technology likely will come increasingly into question. Logic would indicate that any presumption so heavily dependent on transient circumstances should have a sunset provision or, at a minimum, an arrangement for a periodic review to test its continuing viability in light of new facts.

The preamble to the proposed rule acknowledges that the statutory BACT provision reflects congressional intent to "create a procedural methodology for stimulating the widespread use of effective pollution control technologies."⁷⁸ Yet the proposed rule would appear to create a virtually irrebuttable presumption of indefinite duration against promising new NO_x-control technologies. The foregoing suggests that in preparing the final rule, EPA would be well advised to either remove the presumption, more clearly justify it, or temper it with a clearly stated methodology and procedure whereby the presumption could be rebutted in practice.

Conclusion

The WEPCo "fix" cannot be understood in isolation. It reflects a considered policy decision, taken at the instance of the DOE and White House staff, to soften or relax the modification rule as applied to electric utilities. In effect, the DOE and the White House have sought to reduce the cost of controlling power plant pollution by allowing operators maximum flexibility to choose which plants would reduce emissions under Title IV, while minimizing utilities' exposure to new source requirements under Title I. In their pursuit of utility cost savings, however, they have caused EPA to promulgate interpretations of Title I that limit the reach of the NSR and the NSPS to the minimum, or perhaps less than the minimum, called for in the Clean Air Act.

Enactment of Title IV represents a giant step towards the reduction of nationwide SO₂ emissions. However, reducing nationwide atmospheric loading of SO₂ is one, but only one, policy objective of the Clean Air Act. Others, such as protecting public health, stimulating technological innovation, and assuring sufficient air resources to support future economic growth, remain objectives that are served by Title I's approach to local air quality protection and are not supplanted by any nationwide, market-based mechanism of Title IV. Indeed, Congress made this point explicit and legally binding in Title IV itself. It would seem this provision received short shrift in the bureaucratic rough and tumble that produced the WEPCo proposal.

If these issues were confined only to the WEPCo situation, they would be of national significance, given the large impact of electric power production on our nation's air quality. In view of the growing interest in market-based regulation of pollutants within the administration and Congress, however, their importance is magnified.

Taking a longer term perspective, it seems clear that "efficiency" is not enhanced by a policy that allows an individual source to increase emissions without restriction until the air quality standard for the entire surrounding region has been violated or the remaining air quality increment consumed. Such a rule is a growth-restricting, not a growth-enhancing, measure. Moreover, as a nation we must be careful that in pursuing more "efficient" regulations, we do not shortchange the public health and environmental objectives these regulations are meant to serve.

1. Clean Air Act (CAA) §§ 110, 111, 165, 42 U.S.C. §§ 7410, 7411, 7475 (West. Supp. 1991), ELR STAT. CAA 19, 25, 68.

2. See 40 C.F.R. § 51.165 and Appendix S, and §§ 52.21, 60.14, and 60.15 (1990), which describe in detail the rules governing the applicability of NSR and NSPS modifications at existing facilities.

3. Letter from Don R. Clay, Acting EPA Ass't Administrator for Air and Radiation, to David A. Kee, EPA Region V, Dir. Air and Radiation Div. (Sept. 9, 1988) (EPA Docket No. A-90-06, II-B-4) (relating EPA's position on applicability of NSPS and NSR requirements to WEPCo's proposed renovation plans).

4. Wisconsin Electric Power Co. v. Reilly, 893 F.2d 901, 20 ELR 20414 (7th Cir. 1990).

5. Drafters of the 1990 Clean Air Act Amendments found the issue too hot to handle and left it for the regulators to decide. CLEAN AIR ACT AMENDMENTS OF 1990, JOINT EXPLANATORY STATEMENT OF THE COMMITTEE OF THE CONFERENCE, H.R. REP. NO. 952, 101st Cong., 2d Sess. 344 (1990). ("… conferees urge a quick resolution of the WEPCo matter by EPA as appropriate.").

6. See Notice of Proposed Rule — Requirements for Preparation, Adoption, and Submittal of Implementation Plans; Approval and Promulgation of Implementation Plans; Standards of Performance for New Stationary Sources, 56 Fed. Reg. 27630 (1991) (proposed June 14, 1991).

7. See Letter from Don R. Clay to David A. Kee, supra note 3.

8. See Clean Air Act Implementation Special Rules for Utilities: Hearings Before the Subcomm. on Health and the Environment of the House Comm. on Energy and Commerce, 102d Cong., 1st Sess. (1991). Congressman Waxman (D-Cal.) alleged that the proposed rule was improperly influenced by political pressure from the White House. In addition, the Natural Resources Defense Council (NRDC) has threatened to sue. See Letter from David G. Hawkins, Senior Attorney, NRDC, to approximately 90 utility companies (June 21, 1991) (on file with the NRDC).

9. See Acid Rain Program: Permits, Allowance System, Continuous Emissions Monitoring, and Excess Emissions, 56 Fed. Reg. 63002, 63004 (proposed Dec. 3, 1991).

10. CAA § 110, 42 U.S.C. § 7410 (West Supp. 1991), ELR STAT. CAA 19.

11. CAA §§ 160-169, 42 U.S.C. §§ 7470-7479 (West Supp. 1991), ELR STAT. CAA 66-71.

12. CAA § 111, 42 U.S.C. § 7411(a)(1)(C) (West Supp. 1991), ELR STAT. CAA 25. For utilities, this typically means "stack gas scrubbers" for SO₂, "low-NO_x burners" for NO_x, and often, enhanced measures to control soot. Scrubbers can add up to a penny per kilowatt hour; low-NO_x burners and particulate controls are less expensive, but still costly.

13. The degree of sophistication of the technology depends on the program applicable to the project. The PSD program applies in areas that achieved national ambient air quality standards (NAAQS), while the nonattainment program applies in areas that have failed to attain these standards. In nonattainment areas, where there is a need to reduce existing levels of pollution, the law requires technology to meet the "lowest achievable emission rate," which is equal to the rate achievable by best technology required by any plan or used in practice by any similar source. In PSD areas, where the goal is simply to prevent deterioration of local air quality, the standard is "best available control technology," a term not further elucidated by the statute. In practice, it generally incorporates NSPS requirements. See CAA § 169, 42 U.S.C. § 7479, ELR STAT. CAA 70.

14. CAA §§ 160-169, 170-173, 42 U.S.C. §§ 7470-7479, 7501-7503 (West Supp. 1991), ELR STAT. CAA 70; 40 C.F.R. pts. 51, 52 (1990).

15. In 1972, EPA approved a number of state implementation plans with sulfur oxide control strategies based on dispersing pollution through tall stacks rather than measures that would have reduced emissions, such as fuel switching or pollution control devices. EPA's approval of one such plan for Georgia was challenged by environmental organizations and rejected in Natural Resources Defense Council, Inc. v. EPA, 489 F.2d 390, 4 ELR 20204 (5th Cir.), modified, 421 U.S. 60, 5 ELR 20264 (1974). The strongly worded opinion held that the Clean Air Act prohibited the use of dispersion devices, such as tall stacks, in preference to measures that would reduce emissions. This point was not reviewed by the Supreme Court and was further supported in two other cases, Big Rivers Electric Corp. v. EPA, 523 F.2d 16, 5 ELR 20532 (6th Cir. 1975), cert. denied, 425 U.S. 934 (1976), and Kennecott Copper Corp. v. Train, 526 F.2d 1149, 6 ELR 20102 (9th Cir. 1975), cert. denied, 425 U.S. 934 (1976).

16. Edison Electric Institute, Power Statistics Database, Data File UDESIGN-1 (1989) (prepared by Utility Data Institute, Washington, D.C.).

17. EPA, NATIONAL AIR POLLUTANT EMISSIONS ESTIMATES 1940-1985, at 29-30 (Jan. 1987) [hereinafter EPA ESTIMATES]. It should be noted that utility use of coal more than doubled over this period. Id. at 55. Thus, although utilities as a group reduced their average emissions rates, the persistence of old plants precluded progress sufficient to offset the trend towards increasing reliance on coal.

18. Letter from Subcomm. on Oversight and Investigations of the House Comm. on Energy and Commerce, U.S. House of Representatives, to Secretary Watkins (DOE), Administrator Reilly (EPA), and Mr. Boskin (Council of Economic Advisers) 2 (Oct. 9, 1990) (EPA Docket No. A-90-06, II-I-20).

19. The first major warning about acid rain came in a paper prepared by the Swedish government for the first United Nations Environment Program conference in Stockholm in 1972. In the United States, a report of the National Academy of Sciences (NAS) focused public attention to the acid rain problem. NATIONAL RESEARCH COUNCIL, NAS, COMM. ON ATMOSPHERE & THE BIOSPHERE, ATMOSPHERE-BIOSPHERE INTERACTIONS: TOWARD A BETTER UNDERSTANDING OF THE ECOLOGICAL CONSEQUENCES OF FOSSIL FUEL COMBUSTION (1981).

20. By the early 1970s, scientists were raising concerns about the health effects from exposure to the pervasive urban concentrations of sulfur compounds. An example was a report prepared by EPA in 1974. EPA, REP. No. 650/1-74-004, HEALTH CONSEQUENCES OF SULFUR OXIDES: A REPORT FROM CHESS, 1970-1971 (May 1974). At the time, the CHESS report was highly controversial, but similar conclusions are now widely accepted. See, e.g., Clean Air Act Amendments of 1987, Part I, Hearings Before the Subcomm. on Environmental Protection of the Comm. on Environment and Public Works, 100th Cong., 1st Sess. 487 (1987) (testimony of Haluk Ozkaynak).

21. NATIONAL RESEARCH COUNCIL, NAS, COMM. ON ATMOSPHERIC TRANSP. & CHEM. TRANSFORMATION IN ACID DEPOSITION, ACID DEPOSITION: ATMOSPHERIC PROCESSES IN EASTERN NORTH AMERICA (1983).

22. The idea of allowing utilities to trade their emission reduction obligations under the acid rain program in order to achieve reductions in regionwide atmospheric loading was not new in 1990; it had been a feature of a number of prior acid rain bills stretching back to 1982. See, e.g., S. 1894, 100th Cong., 1st Sess. (1987) (introduced by Sen. George Mitchell and reported out of the Senate Comm. on Env't and Pub. Works); S. 145, 98th Cong., 1st Sess. (1983) (introduced by Sen. Mitchell); S. 2001, 98th Cong., 1st Sess. (1983) (introduced by Sen. Robert Stafford and reported out of the Senate Comm. on Env't and Pub. Works); S. 3041, 97th Cong., 2d Sess. (1982) (introduced by Sen. Stafford and reported out of the Senate Comm. on Env't and Pub. Works); H.R. 144, 101st Cong., 1st Sess. (1989) (introduced by Rep. Jim Cooper); H.R. 4331, 100th Cong., 2d Sess. (1988) (introduced by Rep. Cooper); H.R. 3400, 98th Cong., 1st Sess. (1983) (introduced by Rep. Gerry Sikorski).

23. CAA §§ 401-416, 42 U.S.C. §§ 7651-7651o (West Supp. 1991), ELR STAT. CAA 148-167.

24. CAA §§ 404-405, 42 U.S.C. §§ 7651c-7651d (West Supp. 1991), ELR STAT. CAA 151-159.

25. CAA § 407, 42 U.S.C. § 7651f (West Supp. 1991), ELR STAT. CAA 159.

26. Annual particulate emissions from coal-fired utility power plants fell from 2.4 million tons in 1970 to 0.63 million tons in 1985, compared to roughly 15 million tons of SO₂ in both years. EPA ESTIMATES, supra note 17, at 28.

27. CAA §§ 401-416, 42 U.S.C. §§ 7651-7651o (West Supp. 1991), ELR STAT. CAA 148-167.

28. See Hunton & Williams Memorandum, The WEPCo Decision: An Opportunity to Apply the New Source Rules in a Manner That Advances Important Administration Policies (Feb. 6, 1990) (EPA Docket No. A-90-06, II-D-35) (alleging that applying WEPCo approach strictly to pollution control projects could "potentially quadruple the cost of the Administration's acid rain [program]."); Letter from Edison Electric Inst., American Pub. Power Ass'n, and National Rural Elec. Coop. Ass'n, to the DOE and EPA 1 (Apr. 25, 1991) (EPA Docket No. A-90-06, II-D-40) ("[I]t is absolutely critical that EPA's forthcoming WEPCo rule provide immediate clarification that utilities can begin to undertake the kind of pollution control projects called for under the 1990 Amendments to the Clean Air Act without fear of triggering new source review.").

29. Utility spokesmen have not relied on this argument in recent testimony. See, e.g., Statement of Henry V. Nickel on Behalf of the Utility Air Regulatory Group on EPA's Proposed Rule on New Source Review and Confirmation of Existing Law 2 (July 19, 1991) (EPA Docket No. A-90-06) [hereinafter UARG Testimony]. EPA likewise has eschewed the argument in explaining (in the preamble to its proposed rule) why the WEPCo fix is confined to utilities. One explanation may be that utilities do not wish to see the WEPCo rule confined to them as a special case. They have consistently advocated its general application as the only "proper" interpretation of existing law. Also the Title IV argument is legally weak. As seen, Title IV explicitly provides that compliance with Title IV does not exempt affected utilities from complying with the other requirements of the Act. CAA § 413, 42 U.S.C. § 7651l (West Supp. 1991), ELR STAT. CAA 164.

30. CAA § 413, 42 U.S.C. § 7651l (West Supp. 1991), ELR STAT. CAA 164.

31. Chemical Mfrs. Ass'n v. EPA, Docket No. 79-1112, at 4 (D.C. Cir. Feb. 22, 1982) (settlement agreement).

32. UARG Testimony, supra note 29 at 2.

33. CAA § 111(a) (4), 42 U.S.C. § 7411(a)(4), ELR STAT. CAA 25, which addresses the NSPS program, provides that a modification is "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…." See also CAA § 169(2)(C), 42 U.S.C. § 7479(2)(C), ELR STAT. CAA 70 (1988) (PSD program); CAA § 171(4), 42 U.S.C. § 7501(4), ELR STAT. CAA 74 (1988) (nonattainment program).

34. Current NSPS regulations also provide that the replacement of an existing pollution control system is not a "modification" unless EPA determines that the new system is less environmentally beneficial than the old system. 40 C.F.R. § 60.14(e)(5) (1990). The NSPS exclusion for pollution control projects reflects EPA's view that Congress would not wish to deter the installation or improvement of pollution control devices by the prospect of triggering NSPS rules requiring possibly more stringent technologies. See also Letter from David A. Kee, EPA Region V, Dir. Air and Radiation Division, to Timothy J. Method, Asst. Comm'r, Office of Air Management, Indiana Dep't of Envtl. Management 1 (Jan. 30, 1990) (EPA Docket No. A-90-06, II-C-17).

35. Memorandum from Gerald A. Emison, Dir. EPA Office of Air Quality Planning and Standards, to Director, Air Management Division, Regions I, III, V, and IX, at 2 (July 7, 1986) (on file with EPA).

36. However, EPA has given particular projects a nonbinding "safe harbor" from the NSR in individual instances. See, e.g., Letter from William G. Rosenberg, EPA Ass't Administrator for Air and Radiation, to Patrick M. McCarter, Senior Vice President of Electric Operations, Public Service Company of Colorado (July 23, 1990) (EPA Docket No. A-90-06, II-C-7); Letter from W. Rosenberg to Andrew Aitken, Vice President, Dir. of Environmental Affairs, New England Power Service Co. (Mar. 25, 1991) (EPA Docket No. A-90-06, II-C-22).

37. EPA, DRAFT REVISED INTERPRETATIVE RULING 6 (June 20, 1990) (EPA Docket No. A-90-06, II-F-2).

38. Id. at 21.

39. See, e.g., Letter from Edison Electric Inst., supra note 28, at 1 ("at a minimum, no WEPCo rule should contain the 'rebuttable presumption' approach that EPA's 1990 draft interpretive rule proposed. Instead, the WEPCo fix should simply recognize a blanket pollution control exclusion."). Memorandum from Linda Stuntz, Deputy Under Secretary for Policy, Planning, and Analysis, U.S. DOE, to Don Elliott, EPA Gen. Counsel 3 (Mar. 15, 1991) (EPA Docket No. A-90-06, II-D-7).

40. Memorandum from L. Stuntz to D. Elliott 1 (Apr. 9, 1991) (EPA Docket No. A-90-06, II-D-9).

41. See 50 C.F.R. § 52.21(b)(2)(iii)(f) (1990).

42. 56 Fed. Reg. 27641-43 (1991).

43. Id. at 27635. The proposed new rule will also implement certain special rules for repowering projects and clean coal technology demonstration projects. Briefly, repowering projects that qualify for Phase II compliance extensions under the acid rain program will be exempt from NSPS requirements so long as the repowering does not increase the unit's hourly emissions over its baseline level. See CAA § 409(d), 42 U.S.C. § 7651h(d) (West Supp. 1991), ELR STAT. CAA 162; 56 Fed. Reg. 27637 (1991). Facilities participating in temporary (5 years or less) clean coal technology demonstration projects will be exempt from NSPS and NSR requirements during and after the project, so long as there are no emissions increases unrelated to the project. See CAA § 415(b)(2), 42 U.S.C. § 7651n(b)(2) (West Supp. 1991), ELR STAT. CAA 165; 56 Fed. Reg. 27637 (1991). Facilities participating in permanent DOE- or EPA-sponsored clean coal technology demonstration projects will be excluded from NSPS and PSD requirements so long as "potential" emissions from the unit do not increase as a result of the project, but will receive no special treatment under the nonattainment program. See CAA § 415(b)(2), (3), 42 U.S.C. § 7651n(b)(2), (3) (West Supp. 1991), ELR STAT. CAA 165; 56 Fed. Reg. 27637 (1991).

44. See, e.g., 40 C.F.R. § 51.165(a)(1)(v)(C)(5) (1990).

45. See, e.g., 56 Fed. Reg. 27640 (1991).

46. Id. at 27632.

47. Id. at 27631.

48. Id. at 27638.

49. Id. Under the proposed rule, a source operator may choose years prior to the five-year normal baseline period if EPA determines that these years are more representative of normal operations (e.g., the plant was shut down or operated at artificially low capacity for most of five-year period). Id. Utilities have said they would like to see the test further relaxed to provide an automatic look-back of more than five years in cases where a unit had ceased operations for a period prior to the change. See UARG Testimony, supra note 29, at 6.

50. First, the preamble states the baseline "shall be calculated using the highest hourly emissions rate achievable at any time during the 5 years prior to the change." 56 Fed. Reg. 27638 (1991)(emphasis added). A few sentences later, however, the preamble states "[t]he proposed change will allow utilities to demonstrate that an earlier, higher capacity was more representative of the unit's maximum hourly emissions rate." Id.

51. David G. Hawkins, NRDC, Statement Before EPA Regarding Proposed Change to EPA New and Modified Source Rules 6 (July 19, 1991) [hereinafter Hawkins Statement].

52. 56 Fed. Reg. 27636 (1991).

53. Id.

54. The baseline figure was normally derived from actual average yearly emissions over the two years preceding the change, but this could be adjusted to reflect average emissions over any two consecutive years in the previous five years, if EPA determined that such years were more "representative" of the plant's prechange performance. 40 C.F.R. § 52.21(b)(21)(ii) (1990).

55. The "actual-potential" test applied whenever the unit in question had not yet "begun operations" post-change. 40 C.F.R. § 52.21(b)(21)(iv) (1990). Since prudence requires utilities to seek EPA approval prior to making any significant changes in their plant, EPA expected that the actual-potential test would almost always be applied in the absence of a federally enforceable agreement on plant use. 56 Fed. Reg. 27632 (1991).

56. Wisconsin Electric Power Co. v. Reilly, 893 F.2d 901, 916-18, 20 ELR 20414, 20421-22 (7th Cir. 1991).

57. EPA's offer of a federally enforceable use limitation as a means to avoid the actual-potential test was not considered a sufficient palliative because, according to the court, electric utilities cannot predict the precise shape of future electricity demand and therefore cannot promise never to increase annual emissions above any particular level. Wisconsin Electric Power Co., 893 F.2d at 917 n.13, 20 ELR at 20422.

58. Memorandum from L. Stuntz to W. Rosenberg 2 (Apr. 29, 1991) (EPA Docket No. A-90-06, II-F-8).

59. Memorandum from William G. Rosenberg to Roger B. Porter, Asst. to the President for Economic and Domestic Policy, Executive Office of the President 1 (Mar. 30, 1990) (EPA Docket No. A-90-06, II-F-24).

60. 56 Fed.Reg. 27642 (1991) (adding new 40 C.F.R. § 52.21(b)(21)(v) (1990)). "Replacements" are defined in the proposed rule to be the same as "reconstructions" under current NSPS regulations. That is, they are subject to the NSR. See 56 Fed. Reg. 27636 (1991), 40 C.F.R. § 60.15 (1990).

61. 56 Fed. Reg. 27636 (1991).

62. The projection will cover the two-year period after the change, or a different consecutive 2-year period within 10 years after the change, if the reviewing authority determines that such a period is more representative of normal source operations. 56 Fed. Reg. 27637 (1991).

63. Id. (emphasis added). Again, demand growth will normally be projected over the next 2 years but may extend up to 10 years if the reviewing authority permits.

64. Id. at 27638.

65. Id. at 27634.

66. Hawkins Statement, supra note 51, at 12-16.

67. 56 Fed. Reg. 27630 (1991).

68. Id. at 27637.

69. Hawkins Statement, supra note 51, at 12.

70. 56 Fed. Reg. 27637 (1991).

71. EPA, DRAFT REVISED INTERPRETATIVE RULING 6 (June 20, 1990) (EPA Docket No. A-90-06, II-F-2).

72. 56 Fed. Reg. 27638 (1991).

73. Id.

74. Carlo Castaldini, Acurex Corp., Selective Noncatalytic Reduction, Paper presented to the NO_x Subcomm., EPA Acid Rain Advisory Comm. (Aug. 27, 1991).

75. CAA § 169(3), 42 U.S.C. § 7479(3) (West Supp. 1991), ELR STAT. CAA 70.

76. 56 Fed. Reg. 27639 (1991).

77. See Alabama Power Co. v. Costle, 636 F.2d 323, 355, 10 ELR 20001, 20008 (D.C. Cir. 1979) (EPA generally lacks authority to expand congressionally created exemptions or to create new exemptions except as justified by administrative necessity or de minimis circumstances).

78. 56 Fed. Reg. 27639 (1991).