In sum, the EPA has recognized the problems that face trap-oxidizer technology and has identified plausible methods for seeking solutions to these problems. In so doing it has satisfied its burden of reasonableness. As we stated in NRDC v. EPA, 655 F.2d at 334:

The EPA is not obliged to provide detailed solutions to every engineering problem posed in the perfection of the trap-oxidizer. In the absence of theoretical objections to the technology, the agency need only identify the major steps necessary for development of the device, and give plausible reasons for its belief that the industry will be able to solve those problems in the time remaining. The EPA is not required to rebut all speculation that unspecified factors may hinder "real world" emission control.

IV. PROCEDURAL CHALLENGES

A. Must the Agency Give the Industry Four Years of Leadtime Before a NOx Revised Standard Takes Effect?

The EMA contests the agency's setting a NOx revised standard that, promulgated in 1985, is now scheduled to go into effect in 1988 rather than 1990.38 The EMA relies on the statutory prescription that a NOx revised standard "shall apply only for the period of three model years beginning four model years after the model year in which such revised standard is promulgated." 42 U.S.C. Sec. 7521(a)(3)(B). Although the first NOx revised standard is now scheduled for three years later than intended by Congress, the EMA insists that the separate requirement of a four year leadtime provision must be honored.

The EPA's justification in the NPRM for truncating the leadtime is as follows:

The Act required such standards to be promulgated, with four years of leadtime, for the 1985 model year. Clearly, while four years of leadtime could still be provided, it is no longer possible to have revised standards in place for vehicles or engines manufactured in 1985. In this situation, EPA believes it appropriate to implement revised standards with less than four years of leadtime, in order to have them take effect as soon as possible after 1985, provided that such standards allow adequate time for compliance by the manufacturers. This conclusion is based on Congress' clear desire to have standards in place by 1985 (which is now impossible), as well as its apparent intent to provide manufacturers adequate leadtime to meet any standard different from the 75 percent reduction standard. Thus, if a standard is developed which is clearly attainable in less than four years, it would be inappropriate and unnecessary to delay its implementation beyond the period actually needed for compliance.

This approach might at first appear to be an attempt to shift the burden of a failure by EPA to adopt standards in time for 1985 from the Agency to the manufacturers. This is, however, not the case. The manufacturers have already benefited from a two-year delay in any new standard. Their need for adequate leadtime will be fully considered in the standards and model years for implementation which EPA eventually adopts.

Fed.Reg. 40,258, 40,259 (1984), J.A. at 83

Although the EPA correctly pinpoints the statutory dilemma it faces between exacerbating its past failure to meet the 1985 deadline and engaging in a new violation of the four year leadtime requirement, it cannot cite us to any precedent allowing a court to ignore an explicit leadtime requirement.39 That requirement was enacted for the benefit of the manufacturers, to allow time for them to design and develop engines in compliance with newly promulgated standards. See, e.g., H.Rep. No. 294, 95th Cong., 1st Sess. 274-75 (1977) (discussing "abundant leadtime"), reprinted in 4 Leg.Hist. 2741-42, J.A. at 19-20. Whatever their complicity in or responsibility for the delay in meeting the 1985 starting date, we can find no precedent or basis for excusing the agency from the statutory command. Two wrongs do not make a right. Although fully cognizant of the frustration the drafters would have felt, could they have foreseen the course of events, we nonetheless find that they enacted a four year leadtime requirement and we have no alternative but to enforce it, unless or until Congress decrees otherwise.40 Here it is Congress itself that has refused to give the agency any leeway in adjusting deadlines, and even under Chevron, in a contest between clear congressional intent and a more penumbral congressional purpose, the clear meaning of the statute wins.41 We accordingly rule that the 6.0 g/BHP-hr NOx standard may not go into effect until the 1990 model year.

B. May the Agency Promulgate Revised NOx Standards for 1988 and 1991 at the Same Time?

The EMA again argues that the plain language of the statute--"Such standard shall apply only for the period of three model years beginning four model years after the model year in which such revised standard is promulgated," 42 U.S.C. Sec. 7521(a)(3)(B)--dictates that standards may not be promulgated in advance of the seven year framework. Here of course it is not manufacturers' leadtime that is inovlved, but rather the notion that the Administrator should not be making up his mind so far ahead, lest emerging technology make changes necessary. Fortunately, in this instance we need not decide the merits of the debate, because no party raised this objection before the agency. Indeed, parties commented extensively on the feasibility and level of a 1991 NOx standard. We follow our discussion of 42 U.S.C. Sec. 7607(d)(7)(B) in part II.C., supra, and dismiss the challenge against the EPA's promulgation of a 1991 revised NOx standard due to the parties' failure to raise this claim before the agency.

C. Did the Agency Provide Adequate Notice That It Would Promulgate a 1994 PM Standard?

The EMA contends that the EPA violated the notice requirement of 42 U.S.C. Sec. 7607(d)(3) by promulgating a 1994 PM standard after discussing in the NPRM only 1987 and 1990 PM standards. The EPA attempts to show (1) that the parties actually were on notice that a 1994 PM standard might be promulgated, (2) that the 1994 standard was a "logical outgrowth" of the proposed 1987 and 1990 standards and therefore legitimate under precedent of this court, and (3) that even if the agency erred this error was nonprejudicial under 42 U.S.C. Sec. 7607(d)(8), which provides that a rule may be invalidated due to procedural error only if the error was sufficiently great and relevant that the rule would likely have been changed absent the error.

The purposes of requiring notice of any proposed standard have been described by this court:

First, notice improves the quality of agency rulemaking by ensuring that agency regulations will be "tested by exposure to diverse public comment." ... Second, notice and the opportunity to be heard are an essential component of "fairness to affected parties." ... Third, by giving affected parties an opportunity to develop evidence in the record to support their objections to a rule, notice enhances the quality of judicial review....

Against these values, we must balance the public interest in expedition and finality. The notice requirement should not force an agency endlessly to repropose a rule because of minor changes, nor should a court vacate and remand an otherwise reasonable rule because of a minor procedural flaw....

Small Refiner Lead Phase-Down Task Force v. EPA, 705 F.2d 506, 547 (D.C.Cir.1983). This elucidation directs us to the following inquiry: were the actual parties to the rulemaking afforded genuine opportunity to present views and information relevant to the agency's decision to promulgate this 1994 PM standard?

However, again we escape, because the EMA's failure to petition for reconsideration of the 1994 PM standard bars it from raising the question for the first time on appeal. Section 7607(d)(7)(B) provides a mechanism whereby a party raising a claim whose grounds arose after the comment period can petition the agency for reconsideration:

If the person raising an objection can demonstrate to the Administrator that it was impracticable to raise such an objection within such time or if the grounds for such objection arose after the period for public comment (but within the time specified for judicial review) and if such objection is of central relevance to the outcome of the rule, the Administrator shall convene a proceeding for reconsideration of the rule and provide the same procedural rights as would have been afforded had the information been available at the time the rule was proposed. If the Administrator refuses to convene such a proceeding, such person may seek review of such refusal in the United States court of appeals for the appropriate circuit....

Assuredly, the EMA could not have raised the objection that the agency failed to give notice that it would promulgate a 1994 PM standard until after the final rules were published. But thereafter it should have moved under Sec. 7607 to ask the Administrator to convene a new proceeding to consider this objection and possibly any objections or comments it had to a 1994 rule. Certainly the failure to give notice that a rule will be promulgated is "of central relevance to the outcome of the rule," since notice is a predicate for all rules under the 1977 amendments. 42 U.S.C. Sec. 7607(d)(5). Our review of the record indicates that the EMA never made such a move. That failure dooms its challenge here.

This court has previously refused to consider the merits of an objection to a postcomment period agency action because the party failed to petition for reconsideration. American Petroleum Institute v. Costle, 665 F.2d 1176, 1190-91 (D.C.Cir.1981), cert. denied, 455 U.S. 1034, 102 S. Ct. 1737, 72 L. Ed. 2d 152 (1982). We follow suit in this case.

V. CALIFORNIA'S CHALLENGE TO THE 1988 LDT NOX STANDARDS

Intervenor California challenges the EPA's promulgation of NOx emission reduction standards for light duty trucks (LDTs), to take effect in 1988. It argues, first, that the agency should have set a leader-based standard, and second, that the standard should have been based on the lower polluting gasoline engine. With regard to LDTs that are classified as HDVs, see note 17, supra, our conclusions with regard to HDVs are equally applicable here. We have rejected the contention that the statute requires standards set to the capability of the technological leader, see part II.A., supra, and we have rejected the argument that NOx standards must be based on the gasoline engine's capabilities, due to the parties' failure to raise this claim before the agency, see part II.C., supra. However, because we reverse and remand the 1988 HDV NOx standard for failure to give statutorily mandated leadtime, we must also reverse and remand the 1988 LDT NOx standard as it pertains to LDTs that are also HDVs, see note 17, supra, for the same reason.

With regard to the LDTs that are too light to be classified as HDVs, see note 17, supra, California's arguments under Sec. 7521(a)(3)(B) are inapposite, since that section only applies to HDVs. Lighter LDTs can be regulated only under Sec. 7521(a)(1), see NRDC v. EPA, 655 F.2d at 338, which is not the subject of any arguments in this case.

We accordingly reject California's challenges to the 1988 LDT NOx standards. We do, however, rule that the 1.7 grams per mile LDT NOx standard as it pertains to LDTs that are also HDVs may not go into effect until the 1990 model year.

CONCLUSION

A congressional scheme for complex regulation, like the Clean Air Act, ultimately requires much adjusting and finetuning. Its numerous mandates are rarely specific enough to be put in place automatically or followed mechanically. More often its purposes must be implemented through means and compromises chosen by the agency. So long as the main thrust of agency implementation appears to advance the congressional purpose, courts should be wary of interfering. Where congressional intent as to means is not clear, our role is merely to ensure that the agency's means are reasonable ones.

In this case, petitioners have offered a number of theories, often plausible, of how Congress intended the Clean Air Act Amendments of 1977 to be carried out. Because in most instances, however, where their claims were cognizable by the Court, the EPA was able to adduce an equally reasonable interpretation of the law it was assigned to execute, we must defer to the agency. Accordingly, the petitions for review are

Denied, except for the effective date of the near-term NOx standards for HDVs, including LDTs that are HDVs.

Of the United States District Court for the Northern District of Illinois, sitting by designation pursuant to 28 U.S.C. Sec. 294(d)

Clean Air Act Amendments of 1965, Pub.L. No. 89-272 Sec. 202(a), 84 Stat. 992 (codified as amended at 42 U.S.C. Sec. 7521(a)(1))

Clean Air Act Amendments of 1977, Pub.L. No. 95-95, Title II, Secs. 214, 224, Title IV, Sec. 401(d), 91 Stat. 765, 767, 791 (codified at 42 U.S.C. Secs. 7521(a), 7525(g))

"Heavy duty vehicle" is defined as a truck, bus, or other vehicle weighing more than 6,000 pounds. 42 U.S.C. Sec. 7521(b)(3)(C). "Light duty vehicle" (LDV) is defined as any "passenger car or passenger car derivative capable of seating 12 passengers or less." 40 C.F.R. Sec. 86.082-2 (1985). The issues in this case relate to HDVs only.

NOx emissions form pollutants that create special hazards to people with respiratory ailments and to children. The pollutants, nitrogen dioxide and ozone, can harm lung functions and damage lung tissue. Additionally, it is believed that these pollutants increase susceptibility to respiratory infections. See EPA, Retention of the National Ambient Air Quality Standards for Nitrogen Dioxide, 50 Fed.Reg. 25,532, 25,539 (1985); EPA, National Primary and Secondary Air Quality Standards, 44 Fed.Reg. 8,202, 8,211, 8,215 (1979)

With regard to PM:

[I]nhalable particulates are small enough so that they are not as readily prevented by the natural body defenses from reaching the lower respiratory tract, as would coarser particles. Fine particulate matter can penetrate to the alveoli, or deepest recesses of the lungs, where the oxygen/carbon dioxide exchange takes place with the circulatory system. The body requires months or years to clear foreign matter from the alveolar region, as opposed to hours or days to clear the upper respiratory system.

The most obvious non-cancer health effect of an inhalable particulate, such as that produced by diesels, is injury to the surfaces of the respiratory system, which could result in reduced lung function, bronchitis or chronic respiratory symptoms. The hazardous chemicals that may be associated with particulate matter (e.g., organic compounds, lead, antimony, etc.) can either react with lung tissue or be transported to other parts of the body by the circulatory system. Particulate matter may also weaken the resistance of the body to infection and there are indications that it reacts adversely in conjunction with other atmospheric pollutants.

EPA, Regulatory Impact Analysis, Oxides of Nitrogen Pollutant Specific Study and Summary and Analysis of Comments [hereinafter Final RIA] at 4-46 to -47, Joint Appendix (J.A.) at 375-76. Additionally, PM "probably poses an additional risk of acquiring lung cancer." Final RIA at 4-47, J.A. at 376.

"Subparagraph (E)" gives the agency the power to change any gaseous emission standard prescribed in subparagraph (A)(ii) as a result of a pollutant specific study that the agency may conduct. 42 U.S.C. Sec. 7521(a)(3)(E). This provision is not at issue in this case

"Paragraph (1) of this subsection" provides the general authority for motor vehicle emissions regulation. 42 U.S.C. Sec. 7521(a)(1).

"Class or category thereof" refers to the agency's power to establish classes of vehicles based on various factors. 42 U.S.C. Sec. 7521(a)(3)(A)(iv). See note 18, infra.

Additionally, the Act requires that in order to revise gaseous emissions standards the agency must first find that the specified standards of subsection (A)(ii) cannot be met given the technology reasonably expected to be available without an incommensurate cost or fuel economy effect. 42 U.S.C. Sec. 7521(a)(3)(C)(i). Also, the agency must find that the National Academy of Sciences has not issued a report substantially to the contrary of these findings. 42 U.S.C. Sec. 7521(a)(3)(C)(ii)

If the agency does utilize its revision power under subsection (B), it must submit a report to Congress fulfilling various requirements. 42 U.S.C. Sec. 7521(a)(3)(D).

Neither the findings requirement nor the report requirement is at issue in this case.

The court also ordered the EPA to promulgate NCPs, 21 Env't Rep.Cas. at 1957-58, which the Act required be set by August 7, 1978, 42 U.S.C. Sec. 7525(g)(3), but which the agency had failed to issue. The agency finally promulgated the NCPs on the court-ordered schedule. 50 Fed.Reg. 35,374 (1985), J.A. at 213; 50 Fed.Reg. 35,454 (1985), J.A. at 225. These regulations are not at issue in this case

The requirement that the specified NOx standard be keyed to a baseline of emissions from "heavy-duty gasoline-fueled vehicles or engines," 42 U.S.C. Sec. 7521(a)(3)(A)(ii) (emphasis added), is not inconsistent with the agency's issuing a revised standard because diesel engines could not meet the specified standard. The specified, nonrevised standard provision only uses the gasoline baseline as a point from which to establish a figure. Then all engines, gasoline or diesel, must meet this number. The issue of whether it was appropriate for the EPA to set the revised NOx standard for both gasoline and diesel engines at the level attainable by the higher polluting diesels is discussed in part II.C., infra

A grams per brake horsepower-hour standard measures the pollution output of an engine in terms of the amount of work the engine is doing. Unlike the unit of measurement used for automobiles and light duty trucks (grams per mile), g/BHP-hr measurements are self-correcting for the weight of a HDV and the load on the engine. See 45 Fed.Reg. 4,136, 4,137 (1980), J.A. at 41. The NRDC relies on this weight and load neutral quality of the g/BHP-hr measurement as part of its rebuttal to the EPA's argument that standards based on the capability of the technological leader would be unworkable because of the diversity of engines. See note 18, infra

The statute almost always uses "vehicle" and "engine" in the disjunctive. Permitting engines rather than vehicles to be certified is a peculiarity of HDV regulation. As the agency explained in its Draft Regulatory Impact Analysis:

Unlike the light-duty vehicle industry where the engine and vehicle manufacturers are typically one and the same, heavy-duty diesel vehicles and the diesel engines used in them are often manufactured by independent companies. The engine/vehicle interconnections are so marked, in fact, that three of the major heavy-duty diesel engine manufacturers sell engines to every major heavy-duty diesel vehicle manufacturer. Because of this characteristic, as well as because of the logic of basing heavy-duty emissions on a useful engine work basis, [i.e., g/BHP-hr, see note 8, supra ] EPA requires heavy-duty engine certification rather than vehicle certification. [Sec. 7525, the certification provision, actually permits either engines or vehicles to be certified. The agency's regulations are no different. See 40 C.F.R. Secs. 86.088-21, 86.091-21 (1985).] This has facilitated the performance of certification requirements by the engine manufacturer and avoided the situation where many vehicle manufacturers might certify the very same engine resulting in a duplication of effort.

EPA, Draft Regulatory Impact Analysis and Oxides of Nitrogen Pollutant Specific Study [hereinafter Draft RIA] at 26 (emphasis added), J.A. at 233.

A trap-oxidizer consists of a trap (or filter) that catches particulate matter in the exhaust gas stream, and a mechanism to oxidize (burn) the collected particulate matter in order to regenerate the trap and thereby maintain its collection efficiency. Such devices are (or soon will be) widely used on LDVs, but are only in the developmental stage for HDVs. See part III, infra

The agency describes emissions averaging as follows:

A participating manufacturer would be required to determine emission limits (subject to ceilings ...) for each heavy-duty engine family to be produced in a given model year. This family emission limit would serve as the effective standard by which EPA would determine compliance of all engines within the family.

The participating manufacturer's year end sales and power-weighted average ... of all engine families' emissions would then be required to comply with the applicable ... standard.

Fed.Reg. 10,606, 10,633 (1985), J.A. at 136. See part II.B., infra

More than 150 different manufacturers, environmental groups, state and local governmental units, and individual citizens submitted material to the public docket during the comment period. Not surprisingly, the manufacturers tended to observe that the proposed levels were too high, or that not enough leadtime had been granted; the environmental groups and state and local governmental units tended to remark that the standards were not high enough. The comments and the agency's responses to them, with regard to all of the elements of the NPRM, can be found at 50 Fed.Reg. 10,606, 10,609-44 (1985), J.A. at 112-47

The agency's shift from a 4.0 to a 5.0 g/BHP-hr 1991 NOx standard was based primarily on data received from manufacturers, kept confidential because they revealed future engine designs and production timetables, that "represent the best estimate currently available of heavy-duty diesel engine technology in the 1991 time frame," and that "suggest that 5.0 g/BHP-hr is the lowest feasible standard for 1991." 50 Fed.Reg. 10,606, 10,624 (1985), J.A. at 127; Final RIA at 2-51, J.A. at 340. The NRDC and California both object to this use of confidential data, the latter claiming that 42 U.S.C. Sec. 7607(d)(6)(A)(ii), which provides that a "promulgated rule shall be accompanied by ... an explanation of the reasons for any major changes in the promulgated rule from the proposed rule," disallows the use of confidential data, because it cannot be judicially reviewed. The EPA did, however, combine the data from the confidential reports of the various manufacturers and plot them on a graph that was made part of the public record. The agency then discussed the plotted data at some length, analyzing both the limits of the projected technology and the expected tradeoff between NOx reductions and PM reductions (what reduces one may increase the other, see note 27, infra ). See Final RIA at 2-51 to -55, J.A. at 340-44. Although we harbor some misgivings about the confidential status of the data relied upon by the EPA in changing its 1991 NOx standard, we are, on balance, satisfied that the agency, by compiling the data into a composite graph and then explaining the significance of the numbers plotted, has provided a reasoned explanation for moving from a 4.0 to 5.0 long-term NOx standard

Assuming that its statutory "technological leader" argument is correct, see part II.A., infra, the NRDC challenges at some length both the dates and the levels of the 1988 and 1991 NOx and PM standards. It contends that the agency in fact concluded that the technological leader could meet more stringent standards on an expedited schedule. In light of our conclusion that the statute does not require leader-based standards, see part II.A., infra, we need not enter the fray regarding what standards the technological leader could actually have met and when it could have met them

In No. 85-1294, the NRDC is petitioning on behalf of itself and the American Lung Association, the Sierra Club, the Coalition for Clean Air, the West Michigan Environmental Action Council, the Group Against Smog and Pollution, the Delaware Valley Citizens' Council for Clean Air, Ida M. Callan, Christine J. Geymer, Katherine K. Kemp, and Brandt C. Mannchen. The state of California intervened in support of the NRDC, and the Engine Manufacturers Association, the Motor Vehicle Manufacturers Association, and International Harvester Co. intervened in support of the EPA. American Trucking Associations, Inc., filed an amicus curiae brief in support of the EPA

In No. 85-1296, the EMA is petitioning on behalf of Caterpillar Tractor Co., International Harvester Co., IVECO Trucks of North America, Mack Trucks, Inc., and White Engines, Inc. International Harvester Co. intervened in support of the EMA, and the NRDC, on its own behalf and on behalf of the groups and individuals it represents in No. 85-1294, intervened in support of the EPA.

The NRDC claims that the statute requires standards based on the capability of the leading (that is, most technologically advanced) engine, as opposed to the leading manufacturer. The EPA argues that this is inconsistent with diversity among engines, and with the fact that an engine that effectively reduces NOx emissions often will increase PM emissions, and vice versa. Since we reject the NRDC's argument that the statute requires leader-based standards, see part II.A., infra, we need not decide whether those standards would be engine-based or manufacturer-based. For a discussion of the diversity dispute, see note 18, infra. For a discussion of the inverse relationship between NOx and PM reductions, see note 27, infra

California also challenges the 1988 NOx standards for light duty trucks (LDTs). LDTs between 6,000 and 8,500 pounds are classified as HDVs, 40 C.F.R. Secs. 86.081-2(a), 86.080-2 (1985), but lighter LDTs are classified neither as HDVs nor, because they are not passenger cars, see note 2, supra, as LDVs. The EPA proposed a 1987 1.2 grams per mile NOx standard for the lighter LDTs, 49 Fed.Reg. 40,258, 40,269 (1984), J.A. at 93, under 42 U.S.C. Sec. 7521(a)(1), which reads:

The Administrator shall by regulation prescribe (and from time to time revise) in accordance with the provisions of this section, standards applicable to the emission of any air pollutant from any class or classes of new motor vehicles or new motor vehicle engines, which in his judgment cause, or contribute to, air pollution which may reasonably be anticipated to endanger public health or welfare.

Additionally, the agency proposed a 1987 1.7 grams per mile NOx standard for the heavier LDTs under 42 U.S.C. Sec. 7521(a)(3)(B). 49 Fed.Reg. 40,258, 40,269 (1984), J.A. at 93. In the final rulemaking, the agency kept these standards but pushed back the implementation date from 1987 to 1988. 50 Fed.Reg. 10,606, 10,621 (1985), J.A. at 124.

The EPA buttresses its statutory construction with a technological argument about the diversity of engines:

As is well known, [heavy-duty diesel engines] are extremely diverse in terms of design, operating characteristics and durability requirements, reflecting the many different productive needs they fulfill.... Emission controls developed for one engine type are not necessarily transferable to all others; they may be transferable only with great difficulty or not at all. Manufacturers likewise vary in the number and types of engines they market. As a result, the ability of one manufacturer to meet a particular standard, especially if it has a limited product line, may not establish that engines outside its line can obtain the same control. What is possible for the "leading manufacturer" may be impossible or extremely difficult for other manufacturers, regardless of how aggressively they have pursued advanced control technology.

EPA Br. at 43 (footnotes omitted). To require standards based on the capability of a technological leader might make sense in a field with a homogenous product, maintains the EPA, but not in a field such as heavy duty engines that is characterized by heterogeneity.

The NRDC does not dispute the fact of engine diversity, but it does suggest ways that the agency could deal with such diversity within a technological leader framework. First, it maintains that the g/BHP-hr measurement unit serves as a least common denominator, making diverse engines comparable and making a leader-based mandate sensible. See note 8, supra. Second, it contends that the agency can account for engine diversity through dividing engines into different classes under 42 U.S.C. Sec. 7521(a)(3)(A)(iv), and then by setting a different leader-based standard for each class of engines.

If the issue to be determined were whether leader-based standards are feasible, then this debate would be relevant. However, since we ultimately decide that leader-based standards are not required by the statute, we need not determine whether the NRDC's rebuttal to the EPA's diversity argument suffices as the foundation for the feasibility of leader-based standards. We can say this much, however: the passage from the fact of engine diversity, through the neutralizing force of the g/BHP-hr measurement and the "solve-the-problem-by-dividing-it-into-parts" answer of the classification power, to the possibility of leader-based standards, is a complex one. It is highly unlikely that Congress would require the EPA to transverse such an obstacle course without even mentioning it. In other words, the very complexity of the technological feat EPA would have to achieve to apply a technological leader standard to an industry of diverse engines underscores the need for a clearer signal that Congress wanted it done that way.

Neither the Senate nor the Conference Report mentions "technological leader." The Conference Report begins with the standard "The Senate concurs in the House provision ..." followed by a number of exceptions to this broad concurrence. H.R.Conf.Rep. No. 564, 95th Cong., 1st Sess. 163 (1977), U.S.Code Cong. & Admin.News 1977, pp. 1077, 1543, reprinted in 3 U.S. Library of Congress, A Legislative History of the Clean Air Act Amendments of 1977, at 543 [hereinafter Leg. Hist.], J.A. at 14. The Conference Report does discuss, however, the reason for the NOx revision provision (the conferees were uncertain whether the specified standard would be obtainable and wanted to leave the agency room to adjust), again without mentioning "technological leader."

In using the same term which the Court used in the International Harvester case--i.e., "technological leader," the Committee intends to make clear that revised standards are to be based on the emissions control capability of the best technology projected to be available for production. However, emissions performance while the primary, is not the sole consideration in determining the best means reasonably expected to be available: noise, safety, and cost of compliance within the time available must also be considered

Indeed, the apparently separate Chevron searches for congressional clarity and agency reasonableness are often collapsible into one; a petitioner will lose on both Chevron strands when the agency rebuts a reasonable reading with a reasonable reading of its own. See, e.g., Japan Whaling Ass'n v. American Cetacean Soc'y, --- U.S. ----, 106 S. Ct. 2860, 2867, 92 L. Ed. 2d 166 (1986) ("The language of the [relevant statutes] might reasonably be construed in [the manner suggested by respondents], but the Secretary's construction ... is also a reasonable construction of the language used in [the relevant statutes]."); Arrow Air, Inc. v. Dole, 784 F.2d 1118, 1126 (D.C.Cir.1986) ("[W]here ... two reasonable interpretations are possible, the agency's view must prevail."); Office of Consumers' Counsel v. FERC, 783 F.2d 206, 221 (D.C.Cir.1986) ("We agree with the protestors that there are quibbles with the Commission's choice.... But ... we cannot say that the Commission's definition is inconsistent with congressional intent.")

International Harvester involved a challenge to the EPA's failure to suspend LDV emissions standards. In the course of reversing the agency for failure to rebut adequately the motor vehicle manufacturers' showing that the suspension was necessary because they could not comply in time with the statutory standards, the court discussed at some length the "irony" of a system where the inability of the lagging manufacturers to meet emissions standards could lead to a suspension of those standards and a resulting benefit to those lagging manufacturers. See 478 F.2d at 637-38. The EPA rightly points out that the limited number of manufacturers being compared in International Harvester (Ford and General Motors) makes any extrapolation from that case to the multi-manufacturer HDV industry problematical. What the agency overlooks, though, is that we must examine International Harvester as a reference in the House Report; that is, what matters here about International Harvester is not what the case actually stands for, but rather what the House in 1977 thought it stood for. In any case, even this hermeneutic maneuver does not elevate the value of International Harvester to evidence of clear congressional intent to mandate standards based on the capability of the technological leader. See discussion in text

NRDC v. EPA, 655 F.2d 318 (D.C.Cir.), cert. denied, 454 U.S. 1017, 102 S. Ct. 552, 70 L. Ed. 2d 415 (1981), provides support for both the NRDC and the EPA on the technological leader issue. Among other things, the court upheld a worst performing diesel PM standard with regard to LDVs and LDTs, under 42 U.S.C. Sec. 7521(a)(1). Recognizing that some LDTs are HDVs (those between 6,000 and 8,500 pounds, see note 17, supra ), and therefore that the HDV PM provision of 42 U.S.C. Sec. 7521(a)(3)(A)(iii) would come into play, the court held that "[w]e have already upheld the worst-performing-diesel standard for light-duty vehicles as within the EPA's regulatory discretion, and we find that the similar standard for light-duty trucks is equally so." 655 F.2d at 340. This case is not helpful, however, with regard to NOx standards, because of the differences in legislative history between the NOx and PM provisions. See part II.A.2., infra

In its treatment of LDT standards, the court discussed in a footnote the origin of NCPs, first quoting footnote 79 from International Harvester (quoted in text supra after note 21) and then opining that "[c]iting International Harvester, Congress enacted such a scheme of 'nonconformance penalties' for heavy-duty vehicles in section 224 of the 1977 Clean Air Act Amendments...." 655 F.2d at 339 n. 35. Although this discussion tends to support the NRDC's theory of the impetus for NCPs, it does not advance the NRDC's case for a leader-based mandate.

As pointed out in note 22, supra, NRDC v. EPA, 655 F.2d 318 (D.C.Cir.), cert. denied, 454 U.S. 1017, 102 S. Ct. 552, 70 L. Ed. 2d 415 (1981), concluded that PM standards for HDVs need not be based on the capability of the technological leader. The NRDC is correct, though, to indicate that the NRDC v. EPA court gave this issue extremely brief treatment, and we too are hesitant to accord this case dispositive weight with regard to whether the NRDC's interpretation of leader-based PM standards are required for HDVs. However, our discussion of the significant differences in the legislative history between NOx and PM standards, supra, is in accord with the conclusion reached in NRDC v. EPA that leader-based PM standards are not statutorily mandated for HDVs

Although it was not raised by any party before the agency, and accordingly cannot be dispositive here, see 42 U.S.C. Sec. 7607(d)(7)(B), there is an additional argument against emissions averaging. The Act's testing and certification provision, 42 U.S.C. Sec. 7525, speaks of "any," "a," or "such" motor vehicle or engine being tested and certified. With averaging, some vehicles or engines would not be required to comply with the standards and would not be subject to NCPs for failing to so comply. This practice appears inconsistent with the requirement that "any," "a," or "such" vehicle or engine be tested and required to comply with emissions standards. This linguistic analysis is, though, somewhat ambiguous, because the testing and certification provisions do give the agency the power to test new vehicles in such manner as it deems appropriate, so as to conform to the prescribed standards, 42 U.S.C. Sec. 7525(a)(1), leaving room for the counterargument that the manner of testing deemed appropriate or the content of the standards themselves is within the discretion of the agency

However, any argument running counter to the engine specific thrust of the testing and compliance provisions must overcome other legislative history of the 1970 amendments to the Clean Air Act. The House Report, referring to the then new testing and compliance provision, Sec. 7525, states that

[t]he legislation, therefore, provides for more stringent testing of automobiles. Such testing is not limited, as heretofore, to the testing of prototypes. Such testing will continue but the tests should require each prototype rather than the average of prototypes to comply with regulations establishing emission standards.

H.R.Rep. No. 1146, 91st Cong., 2d Sess. 6 (1970), U.S.Code Cong. & Admin.News 1970, pp. 5356, 5361, reprinted in 2 Leg. Hist. 896. Representative Paul Rogers of Florida, a co-sponsor of the bill and a conferee, stated during debate that

[w]e did not have an adequate testing program for automobiles, and used only four prototypes to determine if the standards were being met by a particular manufacturer. And, instead of taking each result on a separate basis, they were using an average on the four prototype automobiles. We are changing this and saying that each prototype automobile must meet the standard itself--not be averaged. We are also going to have the daily inspection of the assembly-line production so that we will get to this problem of cleaning up the auto emissions.

Cong. Rec. 19,210 (1970). The Conference subsequently adopted the House's testing provisions. The 1977 amendments to the Clean Air Act added the NCP provision to Sec. 7525, but otherwise left it as was. There is also no indication in either the statute or the legislative history that HDVs were to be treated differently from cars for the purpose of testing and compliance

We find the foregoing history troubling. However, since, as we stated above, this argument was never raised by any party, and therefore 42 U.S.C. Sec. 7607(d)(7)(B) mandates that it not be dispositive here, we set it out only for the agency's consideration and possible explanation in future proceedings.

See also note 7, supra, regarding the gasoline-based specified NOx standard, note 9, supra, regarding the interchangeability of "engine" and "vehicle," and note 27, infra, regarding the tradeoff between NOx and PM emissions reduction

Because the statute mandates a gasoline-based standard in the specific 1985 provision, 42 U.S.C. Sec. 7521(a)(3)(A)(ii), see note 7, supra, and is silent with regard to whether gasoline or diesel engines shall serve as the benchmark for revised standards, one could contend that while the agency is constrained in the specific 1985 standard, it has leeway to choose its benchmark for a revised standard. However, a two-step argument from the legislative history cuts the other way: First, both the House bill and the Senate bill mirrored the final law on this point; that is, both required a gasoline-based standard in the specified 1985 NOx provision, and both were silent on this issue in the revision provision. Second, as quoted in the text, both the House and the Senate Reports clearly contemplate a gasoline-based revised standard. The logical conclusion from this argument is that Congress simply forgot to include the specification of a gasoline-based standard in the revision provision

The EPA attempts to support many of its arguments in this case by referring to the "well-established" tradeoff for diesels between NOx and PM emissions control "using conventional control techniques." EPA Br. at 11. However, the agency has admitted elsewhere that "there are some techniques available which can reduce engine-out particulate emissions with[out] increasing NOx emissions." 46 Fed.Reg. 1,910, 1,913 (1981). Considering that the statute requires technology-forcing techniques, see part II.A., supra, and part II.D., infra, and not "conventional control techniques," and adding the fact that once a NOx standard is set a PM standard can be geared to it, and vice versa, the agency's heavy reliance on the tradeoff problem appears somewhat problematical

Other cases stating the principle that a party must first raise an argument before an agency include Specialty Equipment Mkt. Ass'n v. Ruckelshaus, 720 F.2d 124, 141 (D.C.Cir.1983), and American Petroleum Inst. v. Costle, 665 F.2d 1176, 1190-91 (D.C.Cir.1981), cert. denied, 455 U.S. 1034, 102 S. Ct. 1737, 72 L. Ed. 2d 152 (1982)

In Railroad Yardmasters of America v. Harris, 721 F.2d 1332, 1337-39 (D.C.Cir.1983), this court held that the contention that the National Mediation Board had no power to act at all at a time when there were two vacancies on the Board was not waived due to petitioner's failure to raise this argument before the Board. The court distinguished a Supreme Court case, United States v. L.A. Tucker Truck Lines, 344 U.S. 33, 37, 73 S. Ct. 67, 69, 97 L. Ed. 54 (1952), where the Court held that the appellee could not raise for the first time in federal court an objection to a hearing examiner's qualifications under the Administrative Procedure Act (APA), but rather should have raised this claim before the Interstate Commerce Commission

The basic jurisdictional nature of the challenge in Yardmasters--to the makeup of the agency itself--is not present in this case.

Intervenor International Harvester attempts to enlist Sierra Club v. Costle, 657 F.2d 298 (D.C.Cir.1981), to show that trap-oxidizer technology for HDVs is not adequately demonstrated. Indeed, Sierra Club did conclude, in a detailed footnote, that dry scrubbing could not be considered an adequately demonstrated technology, and some of the concerns of that court would be equally pressing here were we to confront squarely the issue of whether trap-oxidizer technology for HDVs is adequately demonstrated. However, because we hold, infra, that the agency properly set technology-forcing PM standards, we need not determine whether the current state of trap-oxidizer technology would suffice to meet a nontechnology-forcing standard

This analysis is based on an assumption that all parties appear to share, namely that a standard cannot both require adequately demonstrated technology and also be technology-forcing. The agency describes technology-forcing standards as those that "are to be based upon that technology which the Administrator determines will be available, and not necessarily that technology which is already available. The adoption of such standards helps to encourage and hasten the development of new technology." 49 Fed.Reg. 40,258, 40,258 (1984) (emphasis in original), J.A. at 82. On the other hand, an adequately demonstrated technology "is one which has been shown to be reasonably reliable, reasonably efficient, and which can reasonably be expected to serve the interests of pollution control without becoming exorbitantly costly in an economic or environmental way." Essex Chemical Corp. v. Ruckelshaus, 486 F.2d 427, 433 (D.C.Cir.1973) (emphasis added), cert. denied, 416 U.S. 969, 94 S. Ct. 1991, 40 L. Ed. 2d 558 (1974). That the provisions at issue in this case seek to promote technological advances while also accounting for cost does not detract from their categorization as technology-forcing standards. As we pointed out in NRDC v. EPA, 655 F.2d 318, 328 (D.C.Cir.), cert. denied, 454 U.S. 1017, 102 S. Ct. 552, 70 L. Ed. 2d 415 (1981):

The legislative history of both the 1970 and the 1977 amendments demonstrates that Congress intended the agency to project future advances in pollution control capability. It was "expected to press for the development and application of improved technology rather than be limited by that which exists today." S.Rep. No. 1196, 91st Cong., 2d Sess. 24 (1970), reprinted in 1 Legislative History 424....

As one commentator put it,

in the new legislation "Congress expressed a mood" that ought not to be lost upon administrators and reviewing courts. While the identical words "appropriate consideration to the cost" appear in both the original provision (Sec. [7521(a)(2) ] and the new one (Sec. [7521(a)(3)(A)(i) ], the context is no longer blandly neutral as to how the balance shall be struck. Congress was plainly convinced that standards set under the prior authority should be tightened, and the language it chose to express that conviction, modified by the new and compelling phrase "greatest ... reduction achievable," had a history of judicial construction indicating that costs must be well-nigh prohibitive to be decisive.

D. Currie, Air Pollution: Federal Law and Analysis Sec. 2.26, at 2-52 (1981) (footnotes omitted) (reference at end to stationary source cases).

Helpful, though not dispositive, here, too, is NRDC v. EPA, 655 F.2d 318 (D.C.Cir.), cert. denied, 454 U.S. 1017, 102 S. Ct. 552, 70 L. Ed. 2d 415 (1981). The court there upheld technology-forcing PM standards for LDTs that are also HDVs under 42 U.S.C. Sec. 7521(a)(3)(A)(iii). 655 F.2d at 338-40 (standard based on worst performing diesel can satisfy technology-forcing mandate); see also notes 22-23, supra. The discussion in NRDC v. EPA was quite cursory, though, and the court pointed out that it would not draw a distinction between the general motor vehicle emissions regulation authority, 42 U.S.C. Sec. 7521(a)(1), and the HDV PM emissions regulation authority, 42 U.S.C. Sec. 7521(a)(3)(A)(iii), "because petitioners' objections are too general to implicate the varying nuances of the separate statutory provisions." 655 F.2d at 638. The EMA's objections in this case are quite specific, and we have examined the issue anew in the text, supra

With regard to the first three areas, the EMA claims that the "EPA did not provide a reasonable basis for predicting that the fundamental barriers to heavy-duty trap technology can be overcome." EMA Br. at 20. With regard to the latter two areas, the EMA claims that the "EPA failed to support the standards with reasonable and reliable methodology." EMA Br. at 41. Because all five areas involve technological decisions by the EPA with regard to PM standards, we group them together for discussion

The standard for judicial review under the Clean Air Act does not include a "substantial evidence" test, but it now seems settled that the "arbitrary or capricious" test serves the same function. See Association of Data Processing Serv. Org. v. Board of Governors of the Fed. Reserve System, 745 F.2d 677, 681-86 (D.C.Cir.1984). Justice (then Judge) Scalia provided a helpful explanation:

[I]n their application to the requirement of factual support the substantial evidence test and the arbitrary or capricious test are one and the same. The former is only a specific application of the latter, separately recited in the APA not to establish a more rigorous standard of factual support but to emphasize that in the case of formal proceedings the factual support must be found in the closed record as opposed to elsewhere.

745 F.2d at 683.

New maintenance regulations promulgated along with the NOx and PM standards require that manufacturers produce traps that will function without maintenance for 150,000 miles or 4,500 hours. 50 Fed.Reg. 10,60 6, 10,663 (1985) (to be codified at 40 C.F.R. Sec. 86.088-25(a)(4)(iii)(D)), J.A. at 166

An independent study sponsored by the agency explains that (1) durability and reliability requirements would not be as strict for buses as for most other types of HDVs, (2) buses operate on a predictable cycle that produces a frequent occurrence of moderate high exhaust temperatures, which will greatly aid the oxidization of PM, and (3) transit buses are usually serviced regularly. Final RIA at 2-69, J.A. at 358

See, e.g., Environmental Defense Fund v. EPA, 598 F.2d 62, 83-84 (D.C.Cir.1978); Ethyl Corp. v. EPA, 541 F.2d 1, 28 (D.C.Cir.) (footnotes omitted), cert. denied, 426 U.S. 941, 96 S. Ct. 2662, 49 L. Ed. 2d 394 (1976):

... Where a statute is precautionary in nature, the evidence difficult to come by, uncertain, or conflicting because it is on the frontiers of scientific knowledge, the regulations designed to protect the public health, and the decision that of an expert administrator, we will not demand rigorous step-by-step proof of cause and effect.... The Administrator may apply his expertise to draw conclusions from suspected, but not completely substantiated, relationships between facts, from trends among facts, from theoretical projections from imperfect data, from probative preliminary data not yet certifiable as "fact," and the like.

Although the EMA argued in its original brief for a 1989 date, in its reply brief it accedes to intervenor International Harvester's argument that a 1990 date would actually constitute "four model years after the model year in which such revised standard is promulgated." 42 U.S.C. Sec. 7521(a)(3)(B). The agency itself apparently reads the statute similarly; in its brief the EPA agrees that a standard promulgated during the 1980 window period would have been effective starting with the 1985 model year. EPA Br. at 76

The EPA and the EMA dispute whether precedent exists to support an agency's circumvention of an explicit statutory mandate. None of the cases cited is apposite

The EPA relies on Citizens to Save Spencer County v. EPA, 600 F.2d 844 (D.C.Cir.1979), where this court held that the agency had broad latitude to harmonize two Clean Air Act provisions that facially dealt with the same issue differently. In that case, though, as the EMA points out, "the Court was confronted with a conflict between two statutory provisions which were irreconcilable when enacted." EMA Reply Br. at 17. Here, of course, no such facial inconsistency exists.

The EMA cites, first, SEC v. Sloan, 436 U.S. 103, 98 S. Ct. 1702, 56 L. Ed. 2d 148 (1978), where the Court rejected an attempt by the SEC to circumvent an express statutory time limit governing trading suspension orders on the basis of the SEC's declaration that its action was mandated by an important public interest. In Sloan, however, the countervailing reason for the agency's action was its view of the public interest, not its attempt to mitigate the effects of its failure to meet a specified deadline.

The EMA also cites two cases for the proposition that an agency cannot justify its circumvention of an express statutory mandate by pointing to a past pattern of violating that mandate. Sloan, 436 U.S. at 119, 98 S.Ct. at 1712; Wilderness Society v. Morton, 479 F.2d 842, 865 (D.C.Cir.), cert. denied, 411 U.S. 917, 93 S. Ct. 1550, 36 L. Ed. 2d 309 (1973). Here the agency is not advancing any such argument.

Finally, the EMA cites a case where this court rejected the EPA's argument that it could promulgate a rule without the prior notice and comment period required by the APA in order to meet a deadline. New Jersey v. EPA, 626 F.2d 1038 (D.C.Cir.1980). However, failing to comply with a general APA requirement that exists independently of the statutory deadline at issue is quite different from failing to comply with a requirement that was enacted as part of the same specific regulatory scheme as the statutory deadline at issue. It is not wholly unreasonable to assume in the latter but not the former instance that Congress intended the two requirements to go hand in hand, and that the desuetude of the one could lead to the undoing of the other.

We realize that pushing back the implementation of the 1988 NOx standard to 1990 will leave that 6.0 g/BHP-hr standard in place for only one year, for we have affirmed the 1991 5.0 g/BHP-hr NOx standard, see part IV.B., infra. This might be thought to conflict with the statutory language that mandates that a revised NOx standard "apply only for the period of three model years ...," 42 U.S.C. Sec. 7521(a)(3)(B). However, while "only for the period of three model years" certainly rules out a standard remaining in place for more than three years, the language leaves open the possibility of a standard's remaining in place for less than three years. Indeed, the House, which originated the NOx revised standards authority, summarized the heavy duty standards in its committee report by explaining that "[a]dditional revisions of up to 3 years each could be granted...." H.R.Rep. No. 294, 95th Cong., 1st Sess. 19 (1977), U.S.Code Cong. & Admin.News 1977, p. 1097, reprinted in 4 Leg. Hist. 2486, J.A. at 17, see also H.R. Conf. Rep. No. 564, 95th Cong, 1st Sess. 162 (1977) (summarizing House bill as allowing revisions "of up to 3 years each"), reprinted in 3 Leg.Hist. 542, J.A. at 13

In sum, our holdings today approve a 6.0 g/BHP-hr NOx standard for 1990 model year HDVs and a 5.0 g/BHP-hr NOx standard for 1991 model year HDVs. See also part V, infra, for the effect of our discussion in this part on LDT NOx emissions standards.

We reach this result with some discomfort, unable to locate any precedent for ignoring the plain language of the statute in a situation like this. Courts have countenanced an agency's disregard of plain meaning generally for two reasons, neither present here. One reason is typified by National Congress of Hispanic Am. Citizens v. Usery, 554 F.2d 1196 (1977) [hereinafter Hispanic I ]. In that case, the Secretary of Labor argued that a specific timetable for promulgation of OSHA standards was not meant to be mandatory, and cited in support a separate provision in the Act granting the Secretary wide discretion in determining priority for establishing different standards. See also National Congress of Hispanic Am. Citizens (El Congreso) v. Marshall, 626 F.2d 882 (D.C.Cir.1979) (following Hispanic I ). The 1977 amendments that we construe today provide no such broadly worded "escape hatch."

A second reason is represented by Citizens to Save Spencer County v. EPA, 600 F.2d 844 (D.C.Cir.1979), discussed at note 39, supra. In Spencer County, the agency was confronted with a true "conflict" in that adhering to the plain meaning of one part of the statute would render another part absurd. That situation is not presented here. Compliance with the four year leadtime requirement will not create a direct conflict with any other statutory provision.

The "conflict" in the case before us is more subtle; by requiring the agency to adhere to the four year leadtime provision, we implicate Congress' paramount goal of putting the standards into place as soon as possible. Having run afoul of the mandate that a NOx standard be in place by 1985, the EPA has attempted to strike a balance by providing a standard with less leadtime than the statute provides but that can admittedly be met within the shortened leadtime. Thus, the presumed purpose of the leadtime requirement--to give manufacturers time to meet the standard--has been fulfilled.

Here, the EPA quite carefully did not set its 1988 revised NOx standard at the level it would otherwise have set a 1990 standard (which would have accorded four years leadtime), but ratcheted the standard down to a level attainable within the shorter time. 49 Fed.Reg. 40,258, 40,267 (1984), J.A. at 91 ("This standard [6.0 g/BHP-hr for 1988] represents a modest decrease from current engine emissions ... which should be attainable in the short term with techniques compatible with the proposed 0.60 g/BHP-hr particulate standard."). The agency has thus met Congress' initial purpose for requiring leadtime by purposely fitting the standard to the shortened leadtime.

We make a second observation. Unchallenged in this litigation is another part of the NOx revised standards provision that, on its face, is also being violated. That provision reads "[d]uring the period of ... June 1 through December 31, 1980, in the case of oxides of nitrogen, and during each period of June 1 through December 31 of each third year thereafter, the Administrator may [promulgate revised NOx standards]." 42 U.S.C. Sec. 7521(a)(3)(B) (emphasis added). If we were to adhere strictly to this language, we would permit revised NOx standards to be promulgated only in 1980, 1983, 1986, 1989, and so on. Indeed, the standards at issue in this case, promulgated in 1985, would be invalidated on this ground. But most would agree such invalidation would frustrate congressional intent beyond toleration. A common sense reading of Sec. 7521(a) tells us that the requirement that a NOx revised standard can only be promulgated in "each third year" after 1980 is part of a scheme that provides for either a specified NOx standard in place by 1985 or a revised NOx standard promulgated in 1980 and in place by 1985. When the starting point shifted, the rest of the statutory dates had to follow or the entire time scheme would have been emasculated. It might not then be illogical to conclude similarly that the four year leadtime requirement must be viewed in the context of the rules scheme as a whole, and that a regrettably late start allows some accommodation in getting the program on line so long as it does not violate the interests Congress meant to protect by its leadtime requirements.

However, ultimately we come down on the side that courts are not entitled to substitute their broader reading of congressional intent for the plain language of the statute unless adherence to that plain language would bring about a patently absurd result.