Opinion Per Curiam.

PER CURIAM:

This is a petition for review of an order of the Environmental Protection Agency implementing the 1990 amendments to § 112 of the Clean Air Act. Petitioners are General Electric Company and four trade associations: (1) National Mining Association, which represents companies that produce metal, coal, and minerals, and that manufacture mining equipment; (2) American Forest and Paper Association, which represents companies that make pulp, paper, paperboard, and solid wood; (3) Chemical Manufacturers Association, which represents companies that manufacture industrial chemicals; and (4) American Petroleum Institute, which represents companies engaged in the petroleum industry. We deny the petition for review with respect to the issues raised by General Electric, National Mining Association, and American Forest and Paper Association, but grant it with respect to Chemical Manufacturers Association and American Petroleum Institute’s challenge.

In 1990, as part of its comprehensive overhaul of the Clean Air Act, Pub.L. No. 91-604, *1353 84 Stat. 1676 (1970), Congress revised § 112 of the Act, which regulates emissions of hazardous air pollutants. Pub.L. No. 101-549, 104 Stat. 2399, 2531-84 (1990). Dissatisfied with EPA’s health-based regulation of hazardous air pollutants under the 1970 program, 1 S.Rep. No. 228, 101st Cong., 1st Sess. 128 (1989), U.S.Code Cong. & Admin.News 1990, 3385, Congress replaced this approach with a detailed, technology-based regulatory scheme. The 1990 amendments to § 112 establish an initial list, which EPA may periodically revise, of 189 hazardous air pollutants. 42 U.S.C. § 7412(b)(l)-(3). EPA must publish a list of “categories and subcategories” of “major sources” and certain “area sources” that emit these pollutants. 42 U.S.C. § 7412(e). For each listed “category or subcategory of major sources and area sources” of hazardous air pollutants, § 112(d) of the Act directs EPA to promulgate emission standards. 42 U.S.C. § 7412(d)(1).

Under the Act, “major sources” of hazardous air pollutants are potentially subject to stricter regulatory control than are “area sources.” 2 For example, major sources must comply with technology-based emission standards requiring the maximum degree of reduction in emissions EPA deems achievable, often referred to as “maximum achievable control technology” or MACT standards. 3 42 U.S.C. § 7412(d)(1) — (2). In order to obtain an operating permit under title Y of the Act, §§ 501-507, major sources must comply with extensive monitoring, reporting and record-keeping requirements. 42 U.S.C. §§ 7661-76611 Further, § 112(g) generally conditions the modification, construction or reconstruction of a major source on the source’s meeting MACT emission limitations. 42 U.S.C. § 7412(g).

“Area sources” of hazardous air pollutants are not necessarily subject to such stringent regulation. EPA need not list all “categories and subcategories” of area sources, 42 U.S.C. § 7412(e)(3), 4 and it does not have to establish emission standards for unlisted area sources, 42 U.S.C. § 7412(d)(1). For listed area sources, EPA may choose to promulgate emission standards requiring only “generally available control technologies or management practices.” 42 U.S.C. § 7412(d)(5). These standards can be less rigorous than those required for major sources under 42 U.S.C. *1354 § 7412(d)(1). S.Rep. No. 228, supra, at 172. Area sources are not subject to title V permitting requirements, or to § 112(g)’s restrictions on modification, construction and reconstruction of their facilities.

In July 1992, pursuant to § 112(c)(1), EPA published an initial list of categories of sources that emit hazardous air pollutants, 57 Fed.Reg. 31,576 (1992), and almost seventeen months later, it published a schedule for promulgation of emission standards for these listed source categories, as required by § 112(e), 58 Fed.Reg. 63,941 (1993). In August 1993, in order to “eliminate the need to repeat general information and requirements within each [emission] standard,” EPA proposed a rule codifying the “procedures and criteria needed to implement” emission standards for hazardous air pollutants. 58 Fed. Reg. 42,760, 42,760 (1993). It promulgated a final rule, which is the subject of this dispute, adopting these general provisions on March 16, 1994. 59 Fed.Reg. 12,408 (1994).

Among other things, the general provisions rule implements § 112(a)(l)’s definition of “major source.” The rule defines “major source” in terms nearly identical to those in § 112(a)(1) of the Clean Air Act:

Major source means any stationary source or group of stationary sources located within a contiguous area and under common control that emits or has the potential to emit considering controls, in the aggregate, 10 tons per year or more of any hazardous air pollutant or 25 tons per year or more of any combination of hazardous air pollutants, unless the Administrator establishes a lesser quantity, or in the case of radionuclides, different criteria from those specified in this sentence.

59 Fed.Reg. at 12,433-34 (to be codified at 40 C.F.R. § 63.2). A “stationary source” is “any building, structure, facility or installation which emits or may emit any air pollutant.” Id. An “area source [is] any stationary source ... that is not a major source.” Id. The preambles to the proposed and final rules, and other definitions adopted in the final rule explain in greater detail how EPA plans to identify major sources.

Petitioners challenge three aspects of EPA’s implementation of the definition of “major source.” First, National Mining Association and American Forest and Paper Association (collectively referred to as “National Mining Association”) and General Electric question EPA’s requiring the aggregation of all hazardous air emissions within a plant site — instead of only those emissions from equipment in similar industrial categories — in a § 112 major source determination. Second, National Mining Association challenges EPA’s requiring the inclusion of “fugitive emissions” in a source’s aggregate emissions in determining whether the source is major. Third, Chemical Manufacturers Association and American Petroleum Institute (collectively referred to as “Chemical Manufacturers Association”) contend that EPA overstepped its regulatory authority by permitting a source to reduce its “potential to emit” only with “federally enforceable” emission controls and limitations.

EPA promulgated the rule challenged here in accordance with the special rulemaking provisions of 42 U.S.C. § 7607(d). EPA rules that are “arbitrary, capricious, an abuse of discretion, or otherwise not in accordance with law” or in excess of EPA’s “statutory jurisdiction, authority, or limitations” must be set aside. 42 U.S.C. § 7607(d)(9). With respect to alleged procedural errors in EPA’s promulgation of a rule, these will invalidate the rule only “if the errors were so serious and related to matters of such central relevance to the rule that there is a substantial likelihood that the rule would have been significantly changed if such errors had not been made.” 42 U.S.C. § 7607(d)(8).

General Electric and National Mining Association have similar arguments against the final rule’s implementation of § 112(a)(1). Both maintain that EPA may not, in determining whether a site is a major source, include emissions from all facilities on a contiguous plant site under common control. These petitioners assert that, for purposes of major source determinations, EPA may aggregate emissions from different facilities on *1355 a contiguous plant site under eon. non control only when the facilities fall within a similar industrial classification. General Electric says EPA must aggregate emissions on a “source category” basis; 5 National Mining Association contends that EPA may combine emissions only if the emitting facilities fall within the same two-digit Standard Industrial Classification (SIC) Code. 6

In the preamble to the final rule, 7 EPA made clear that in determining whether a source is major, emissions from all sources of hazardous air pollutants within a plant site must be aggregated, so long as the sources are geographically adjacent and under common control. 59 Fed.Reg. at 12,412. As a result, if the total annual emissions of hazardous air pollutants from a plant site exceed the designated thresholds, each source emitting pollutants at the site must comply with the stricter MACT emission standards applicable to sources under § 112(d)(2), and with other requirements applicable to major sources.

Petitioners read § 112(a)(1) more restrictively. In their view, EPA’s approach will impermissibly regulate “minor facilities” that happen to be located at an industrial site with annual emissions of hazardous air pollutants that, in the aggregate, exceed the major source thresholds. See Brief for General Electric at 19. They contend that EPA may require aggregation of emissions from sources only if those sources fall within a single source category — General Electric’s argument, or the same two-digit SIC Code— National Mining Association’s contention. It follows, according to petitioners, that a source must comply with regulatory requirements applicable to major sources only if it belongs to some group of sources at an industrial site emitting, in the aggregate, more than the major source threshold. Under petitioners’ theories, it is possible that only some of a site’s sources would have to comply with the regulatory requirements applicable to major sources, including the stricter emission limitations of § 112(d)(2). Other sources of hazardous air pollutants would be regulated as area sources, possibly subject to less stringent emission standards or to none at all. 42 U.S.C. § 7412(e)(5).

EPA rejected petitioners’ methods of implementing “major source.” With respect to General Electric’s source category definition, EPA acknowledged that “[m]ore than one source category on the EPA’s source category list may be represented within a plant that is a major source” of hazardous air pollutants, as is the case for a large chemical manufacturing complex. 59 Fed.Reg. at 12,-411; see also 57 Fed.Reg. at 31,578 (“a large plant ... would clearly be a ‘major source,’ but would also comprise multiple source categories”). Congress intended, according to EPA, “that all portions of a major source be subject to MACT [emission standards] regardless of the number of source categories into which the facility is divided.” 59 Fed. Reg. at 12,411. “Thus, the EPA will set one or more MACT standards for a major source, and sources within that major source will be covered by the standard(s), regardless of whether, when standing alone, each one of *1356 those regulated sources would be major.” Id. EPA also rejected the SIC Code approach to implementing “major source,” advanced here by National Mining Association. Because § 112(a)(1) does not refer to SIC Codes, EPA reasoned that Congress intended major sources of hazardous air pollutants to “encompass entire contiguous ... plant sites without being subdivided according to industrial classifications.” 59 Fed.Reg. at 12,412. A separation of emission sources by SIC Codes “would be an artificial division of sources that, in reality, all contribute to public exposure around a plant site.” Id.

If § 112(a)(1) is viewed in isolation, EPA’s reading of the provision is not simply consistent with the provision; it is nearly compelled by the statutory language. Section 112(a)(1) states that a “group of stationary sources” need meet only three conditions to be termed a “major source”: (1) sources within the group must be “located within a contiguous area”; (2) they must be “under common control”; and (3) in the aggregate, they must emit or, considering controls, have the potential to emit 10 or more tons per year of a single hazardous air pollutant or 25 or more tons per year of any combination of hazardous air pollutants. Section 112(a)(1) says nothing about combining emissions only from sources within the same source categories or SIC Codes. In this respect, EPA’s definition of “major source,” set forth in the preamble to the final rule, is faithful to the language of § 112(a)(1).

Petitioners ask us to look beyond the language of § 112(a)(1). In the first of several loosely connected arguments, General Electric recites fragments from § 112’s other provisions, including: (1) § 112(e)(1), which directs EPA to publish “a list of all categories and subcategories of major sources and area sources”; (2) § 112(d)(1), which directs EPA to establish emission standards “for each category or subcategory of major sources and area sources”; and (3) § 112(j)(2), which describes what an operator of a “major source in [a] category” must do if EPA does not promulgate an emission standard for that “category of major sources.” From these portions of § 112, General Electric leaps to the conclusion that “major source must be defined “with reference to’ (and cannot be broader than) the source category defined by EPA for § 112 regulation.” Brief for General Electric at 16.

General Electric’s logic is hard to grasp. Rather than supporting General Electric, the provisions the company invokes, read in full and in context, tend to support EPA’s implementation of “major source” without reference to source categories. Section 112 directs EPA to perform certain tasks on a category-wide basis — it is to identify categories of major and area sources of hazardous air pollutants (§ 112(c)(1)), and it must promulgate category-wide emission standards for these sources (§ 112(d)(2)). It by no means follows that because the statute in several provisions uses the terms “major source” and “category” in the same sentence — which is all General Electric’s argument amounts to — EPA must read a source category restriction into § 112(a)(l)’s definition of “major source.” Nor does § 112(c) somehow prohibit EPA from applying § 112(d)’s MACT emission limitations “to minor sources in a listed category of major sources without complying with the statutory requirements for listing area sources.” See Brief for General Electric at 18. Section 112(c) simply requires the listing of all major sources and those area sources presenting adverse health or environmental effects. 42 U.S.C. § 7412(c)(1), (3). Neither § 112(c) nor § 112(d) says anything about EPA’s including “minor sources” in a “listed category of major sources.” 8

Taking its argument one step further, General Electric contends that EPA’s definition *1357 of “major source” will lead to “anomalous and unreasonable results” when other parts of §112 are implemented. Brief for General Electric at 19. General Electric sees a tension between the final rule’s definition of “major source” and § 112(g), which describes requirements for constructing, reconstructing or modifying a major source. 42 U.S.C. § 7412(g). The idea is that under § 112(g), an operator of a small emissions unit at a large facility might have to install MACT in order to modify or reconstruct the unit, even though the unit would not be subject to a MACT standard under § 112(d). We agree with EPA that General Electric’s argument reflects a misreading of § 112, which draws no distinction between “major source” for purposes of § 112(d) and § 112(g). If a small emissions unit is a “major source” because it is located at a plant that emits or, considering controls, has the potential to emit 10 or more tons per year of a single hazardous air pollutant or 25 or more tons per year of any combination of hazardous air pollutants, it is subject to all the regulatory requirements imposed on major sources, including those of § 112(d) and § 112(g). 9

In addition to alleging inconsistencies between EPA’s definition of “major source” and § 112’s other provisions, General Electric insists that EPA’s implementation of § 112(a)(1) is at odds with other aspects of the Clean Air Act. It points to § 112(a)(3)’s provision that “stationary source” has the same meaning as it does in § 111 of the Act, 42 U.S.C. § 7411, which deals with performance standards applicable to new sources. In Alabama Power Co. v. Costle, 636 F.2d 323, 395-96 (D.C.Cir.1979), General Electric notes, this court rejected EPA’s definition of “stationary source” as used in § 111(a)(3) as a “combination” of “buildings, structures, facilities, or installations.” 10 To be sure, Alabama Power struck down EPA’s defining “source” for purposes of its preventing significant deterioration in air quality program (established pursuant to 42 U.S.C. §§ 7470-7479) 11 as any “structure, building, facility equipment, installation or operation (or combination thereof).” 636 F.2d at 394, 395-96 (italics added). But that was because EPA had unlawfully expanded § 111(a)(3), which defines “stationary source” as “any building, structure, facility or installation which emits or may emit any air pollutant,” without reference to combinations of these things. 636 F.2d at 395. That is not the case here. Section 112(a)(1) expressly provides that a “major source” is “any stationary source or group of stationary sources ” with emissions exceeding certain limits. 42 U.S.C. § 7412(a)(1) (italics added). Indeed, one could infer from § 112(a)(1) a congressional intent, in the context of hazardous air pollution regulation, to override Alabama Power. 12

We also reject General Electric’s final argument that EPA’s implementation of § 112(a)(1) was procedurally flawed. EPA sufficiently addressed the effect of its definition of “major source” on various aspects of the hazardous air pollution program, and it was explicit enough about binding effect of the final rule’s definition on future actions. In both the proposed and final rules, EPA *1358 extensively discussed the implications of the general provisions, including its definition of “major source,” on other aspects of § 112. See 59 Fed.Reg. at 12,414-18; 58 Fed.Reg. at 42,764-68. The agency further explained that it intended the general provisions to be “the minimum generic requirements necessary for implementation” of emission standards for hazardous air pollutants. 59 Fed. Reg. at 12,415. To the extent EPA finds it appropriate to override specific aspects of the final rule in future rulemakings for specific source categories, “EPA will describe in the [new] subpart exactly which requirements of the General Provisions are applicable to the specific source category and which have been overridden.” Id. at 12,408-09. Even if General Electric were correct about EPA’s alleged procedural defects, the company has not shown why these supposed mistakes were so serious that, had they not been made, there is a “substantial likelihood that the rule would have been significantly changed.” See 42 U.S.C. § 7607(d)(8).

National Mining Association takes a somewhat different tack in its challenge to EPA’s definition of “major source.” For serious, severe and extreme ozone nonattainment areas, § 182(c)-(e) (42 U.S.C. § 7511a(e)-(e)), and for the title V permitting program, § 501(2) (42 U.S.C. § 7661(2)), the Act defines the terms “major source” or “major stationary source” in language very similar to that of § 112(a)(1). See, e.g., 42 U.S.C. § 7511a(c) (defining “major source” as “any stationary source or group of sources located within a contiguous area and under common control” with emissions exceeding 50 tons per year of volatile organic compounds). As EPA acknowledges, to some extent it uses a SIC Code approach to defining “major source” in these programs. See, e.g., 40 C.F.R. § 70.2 (emissions to be aggregated by SIC Code for title V major source determination). 13 National Mining Association reasons that EPA is required to use the same approach in defining “major source” pursuant to § 112(a)(1).

This argument warrants little discussion. As EPA explained in the preamble to the final rule, “because of the different objectives of section 112 and title V ... and because section 112 contains its own definition,” it would define “major source” for purposes of § 112 without reference to SIC Codes. 59 Fed.Reg. at 12,412. The explanation is reasonable. Different programs have different objectives and structures. EPA is not bound to any one definition of “major source.” See Alabama Power, 636 F.2d at 397-98 (“EPA has latitude to adopt definitions of ... ‘source’ that are different ... from those [used in other programs].”); of. Mobil Oil Corp. v. EPA, 871 F.2d 149, 153 (D.C.Cir.1989) (“This court has previously upheld the agency’s decision to employ different definitions of the term ‘facility’ in construing different portions of RCRA.”). The § 112 and § 182 (42 U.S.C. § 7511a) programs are plainly distinguishable. Section 112 is directed at limiting nationwide emissions of hazardous air pollutants; § 182 addresses emissions of volatile organic compounds in serious, severe and extreme ozone nonattainment areas. Apart from this very obvious difference, in § 182 the term “major source” is used mainly to determine the sort of control technology that must be installed by an operator wishing to modify a facility in a way that increases emissions of volatile organic compounds. See, e.g., 42 U.S.C. § 7511a(e)(7). Whether a source is major for purposes of § 112 governs not only modifications, § 112(g), but also emission limitations applicable to new and existing sources of hazardous air pollutants, § 112(d). The title V program also differs fundamentally from § 112. It is essentially procedural — other provisions in the Act supply the substantive requirements applicable to sources subject to the permitting program.

In a second argument, National Mining Association contends that EPA’s definition of “major source” is inconsistent with the legislative history of the 1990 amendments to the Clean Air Act. It bases this conclusion on an excerpt from a House Report discussing the “major source” definitional language added *1359 to the ozone nonattainment provision in § 182 of the Act, 42 U.S.C. § 7511a:

The definition of “major source” [in the ozone nonattainment area] and elsewhere in the bill uses the term “group of sources located within a contiguous area and under common control.” The Committee understands this to mean a group of sources with a common industrial grouping, i.e., the same two-digit SIC code. It is the approach followed by EPA as a result of the Alabama Power litigation. It avoids the possibility that dissimilar sources, like a power plant and an adjacent coal mine, will be considered as the same “source” because of common ownership.

H.R.Rep. No. 490(1), 101st Cong., 2d Sess. 236-37 (1990) (italics added). Seizing upon the italicized language, National Mining Association urges that “Congress intended that a source-based limitation, such as the SIC limitation, apply to all of the definitions” of “major source.” Brief for National Mining Association at 21.

We agree with EPA that the legislative history of the Clean Air Act does not compel a conclusion that Congress intended to limit “major source” in the way National Mining Association suggests. The section of the Report that specifically discusses § 112(a)(l)’s definition of “major source” contradicts the Association’s interpretation of the provision:

For purposes of the definition [of “major source”], all emissions of listed pollutants are counted from a group of sources within a plant boundary (contiguous property under common ownership). This is to assure that emissions from the facility as a whole are adequately controlled.

H.R.Rep. No. 490(1), supra, at 324 (italics added). The Senate Report contains similar language: 14

The definition of “major source” [in § 112] also includes provisions to assure that stationary sources which would otherwise be subject to the emissions standards, are not excluded from control requirements as the result of arbitrary subdivision or description of the source. A stationary source potentially subject to an emissions standard because it emits a listed air pollutant is to be defined to include “all emission points and units of such source located within a contiguous area and under common control”. This language will prevent a facility from avoiding control by subdividing its operations into separate emissions points, product lines or units at one site with individual emissions rates less than the 10 or 25 ton per year thresholds.

S.Rep. No. 228, supra, at 151, U.S.Code Cong. & Admin.News 1990, at 3536 (italics added). These explicit references to § 112(a)(l)’s definition of “major source” support EPA’s view. At most, the legislative history leaves unresolved whether the agency must define “major source” with reference to SIC Codes. In such circumstances, “it is enough that the Agency’s construction is reasonable.” Natural Resources Defense Council v. EPA, 22 F.3d 1125, 1141 (D.C.Cir.1994) (citing Ohio v. EPA, 997 F.2d 1520, 1527 (D.C.Cir.1993)). EPA’s interpretation of “major source” satisfies that standard. 15

National Mining Association also thinks EPA erred in deciding to count “fugitive emissions” of hazardous air pollutants in determining whether a “source” is a “major source,” without first conducting a rulemaking pursuant to § 302(j), 42 U.S.C. § 7602(j). “Fugitive emissions” are defined in the final rule as:

those emissions from a stationary source that could not reasonably pass through a stack, chimney, vent or other functionally *1360 equivalent opening. Under section 112 of the Act, all fugitive emissions are to be considered in determining whether a stationary source is a major source.

59 Fed.Reg. at 12,433 (to be codified at 40 C.F.R. § 63.2). 16

Section 302(j) of the Act, as interpreted in Alabama Power v. Costle, 636 F.2d at 369-70, requires EPA to conduct a separate rule-making to achieve this result, so the Association contends. This provision, fully quoted in the margin, 17 states that, “[ejxcept as otherwise expressly provided,” a “major stationary source” or “major emitting facility” is any stationary source of air pollutants that “directly emits, or has the potential to emit” at least 100 tons per year of any air pollutant, including “any major ... source of fugitive emissions ... as determined by rule by the Administrator.” Alabama Power held that EPA could not, without rulemaking, include fugitive emissions of air pollutants in a facility’s aggregate emissions for purposes of determining whether the facility was a “major emitting facility” within § 169(1), 42 U.S.C. § 7479. 636 F.2d at 368-70. For the prevention of significant deterioration in air quality program (42 U.S.C. §§ 7470-7479), § 169(1) defines “major emitting facility” as any of 28 categories of sources that emit 100 or more tons per year of any air pollutant. 42 U.S.C. § 7479(1). For “any other source,” the threshold is 250 or more tons per year. Id. Although § 169(1) did “expressly make a substantial modification in the 302(j) definition of ‘major,’ ” it had “no ‘express’ provision modifying section 302(j)’s ‘rule’ requirement as to fugitive emissions.” Alabama Power, 636 F.2d at 370. “Therefore under section 169(1) controlled in this respect by section 302(j), the calculation of the threshold quantity emissions may include fugitive emissions only as determined by rule by the Administrator.” 18 636 F.2d at 370.

Finding this case indistinguishable from Alabama Power, National Mining Association spins out the following argument: § 112(a)(1) defines “major source” in terms of a “stationary source or group of stationary sources”; a § 112 major source is thus, “by definition, a ‘major stationary source,’ ” subject to the requirements of § 302(j); § 112(a)(1) does not expressly exempt a major source from § 302(j)’s fugitive emissions rulemaking requirement; therefore, EPA may not require a source to include fugitive emissions of hazardous air pollutants in the source’s total emissions without a § 302(j) rulemaking. 19

The Association’s argument is not very persuasive. Alabama Power was decided in the pre-Chevron age. Moreover, there is a notable difference between § 302(j) and § 112(a)(1). Section 302(j) speaks of sources that “directly” emit air pollutants, on the one hand, and fugitive emissions on the other, thus suggesting that emissions of the fugitive variety are not direct. By contrast, § 112(a)(1) does not contain the modifier “di *1361 reetly,” and it does not mention fugitive emissions as a separate category of emissions. Furthermore, one cannot say that § 302(j) supplies “quantitative terms” for § 112(a)(l)’s definition of “major source,” as it did for § 169(1), one of the provisions at issue in Alabama Power, 636 F.2d at 369. When it comes to hazardous air pollutants, the quantities — 10 tons of any one kind per year or 25 tons of a combination — are specified in § 112(a)(1), not § 302(j), which has a 100-ton threshold.

While the Association’s argument is thin, EPA’s counterargument is hardly overwhelming. EPA thinks its best point is that § 112(a)(1) defines “major source” whereas § 302(j) defines “major stationary source” and “major emitting facility”; since the phrases are different so must be the meanings with respect to counting fugitive emissions. The problem with this argument is that at a critical juncture § 302(j) also uses the phrase “major source” — a “major stationary source” may include a “major ... source of fugitive emissions” if EPA so decides in a rulemaking. And as the Association notes, other provisions of the Act unrelated to § 112 or § 302(j) refer to “major source” and “major stationary source” interchangeably, see 42 U.S.C. §§ 7426(a)(1)-(2), 7511a, 7513a. If EPA’s point related to differences between major stationary sources and major mobile sources, the omission of “stationary” in § 112(a)(1) and its inclusion in § 302(j) might be significant, but that of course is not EPA’s point.

EPA fares better when it tells us that title V of the Act explicitly draws a distinction between the nomenclature of § 112 and that of § 302, defining “major source” for permitting purposes as either a “major source as defined in section 7412 [§ 112]” or a “major stationary source as defined in section 7602 [§ 302] ... or part D of subchapter I [nonattainment program].” 42 U.S.C. § 7661(2). And on EPA’s side is the Senate committee report stating that the definition of “major source” in § 112 “will only apply in the context of this section and should not be confused with other meanings of the term ‘major source’ in parts C (prevention of significant deterioration) or D (non-attainment) of the Act.” S.Rep. No. 228, supra, at 150-51, U.S.Code Cong. & Admin.News 1990, at 3535-3536.

We conclude that EPA may require the inclusion of fugitive emissions in a site’s aggregate emissions without conducting any special rulemaking, even if “major source” and “major stationary source” mean the same thing. Section 112(a)(1) expressly provides that a “major source” is any stationary source or group of stationary sources “located within a contiguous area and under common control” and emitting more than 10 tons per year of a single hazardous air pollutant or 25 tons per year of such pollutants combined. An emission may be fugitive, but it is still an emission from a stationary source. And so the italicized language certainly may be read as EPA reads it — that all emissions are to be counted in determining whether a source is major, subject only to the qualification that they emanate from a contiguous site under common control. So read, § 112(a)(1) satisfies § 302(j)’s “[e]xcept as otherwise expressly provided” clause such that fugitive emissions may be counted in a source’s aggregate emissions without a special rulemaking.

As noted above, in determining whether a source is to be categorized as a “major source” of emissions (or by default an “area” source), EPA was directed by Congress to calculate the amount of hazardous air pollutants a stationary source “emits or has the potential to emit considering controls.” Clean Air Act § 112(a)(1), 42 U.S.C. § 7412(a)(1) (emphasis added). In its final rule, EPA defined a source’s “potential to emit” as its “maximum capacity ... to emit a pollutant under its physical and operational design.” 59 Fed.Reg. at 12,434. To comply with the statutory directive to “consider[] controls” while determining emissions capacities, the rule also provides:

Any physical or operational limitation on the capacity of the stationary source to emit a pollutant, including air pollution control equipment and restrictions on hours of operation or on the type or amount or material combusted, stored, or *1362 processed, shall be treated as part of its design if the limitation or the effect it would have on emissions is federally enforceable.

Id. (emphasis added). Under the rule, a control is deemed to be “federally enforceable” if it is “enforceable by the Administrator and citizens under the Act or ... under other statutes administered by the Administrator.” Id. at 12,433.

Petitioner Chemical Manufacturers Association argues that this restrictive definition— which disregards emissions limitations imposed by state or local regulations not deemed “federally enforceable” — is contrary to the language of § 112(a)(1) of the Act. The government contends that since the word “controls” is not defined in the statute, it was open to EPA under Chevron to define the term, and it has done so reasonably. See Chevron U.S.A. Inc. v. Natural Resources Defense Council, Inc., 467 U.S. 837, 843, 104 S.Ct. 2778, 2782, 81 L.Ed.2d 694 (1984). According to petitioners, even if Chevron Step II is to be reached — because the statute does not reveal a specific congressional intent — we should conclude that EPA’s construction of “controls” is impermissible.

It is common ground that Congress meant the word “controls” to refer to governmental regulations and not, for instance, operational restrictions that an owner might voluntarily adopt. (We note, however, that the word could be read that broadly, which certainly supports the government’s position that the term is not clear on its face.) Petitioners further conceded at oral argument — quite properly, we believe — that Congress intended the term to stand for effective controls. EPA clearly is not obliged to take into account controls that are only chimeras and do not really restrain an operator from emitting pollution. Nevertheless, petitioners claim that EPA has imposed the federal enforceability requirement in pursuit of policy objectives unrelated to concerns about the effectiveness of controls imposed at the state and local level. EPA is accused of interpreting the statute so as to pressure states — through the ministrations of sources eager to have local controls counted in determining their capacity to emit under § 112 — to seek EPA approval of state emissions policies. This objective, petitioners claim, is no part of § 112’s requirement that controls be considered in determining whether a facility qualifies as a major source. It is an impermissible interpretation since it subordinates the effectiveness of controls to other considerations not approved by Congress.

Although it is the regulations implementing the 1990 amendments to the Clean Air Act which are directly before us, this dispute had its genesis at least a decade earlier. Following the passage of the Clean Air Act Amendments of 1977, the agency took the position that the phrase “potential to emit” as used in the definition of “major emitting facilities” excluded even emissions-reducing equipment, such as scrubbers, filters, and other technologies. See 40 C.F.R. §§ 51.24(b)(3), 52.21(b)(3) (1978). We rejected that position in Alabama Power. See 636 F.2d at 353-55. In the wake of that case, EPA proposed a new definition of “potential to emit” that would have taken into account air pollution control equipment, but not operational restraints. See 44 Fed.Reg. 51,924 (1979). The final regulations issued in 1980, however, adopted the position that capacity calculations could factor in operational restraints — but only if they were “federally enforceable.” See 45 Fed.Reg. 52,676, 52,746 (1980). The regulations defined as “federally enforceable” those emissions restrictions that were “enforceable by the Administrator.” Id. at 52,737. The requirement of federal enforceability was, EPA explained, “necessary, as a practical matter, to ensure that sources will perform the proper operation and maintenance for the control equipment.” Id. at 52,688.

The 1980 rule was challenged in this court, but in a February 1982 settlement, EPA agreed to amend its position on federal enforceability. The proposal that followed would have taken into account emission limits “enforceable under federal, state or local law and discoverable by the Administrator and any other person.” 48 Fed.Reg. 38,742, 38,-748, 38,755 (1983). But by the time the final rule was issued, in 1989, the agency had apparently decided to abandon the terms of the settlement. The final regulations revert *1363 ed to the former position of requiring federal enforceability as the sine qua non for crediting operational restraints. “Federal enforceability” was still defined to reach only those limitations “enforceable by the Administrator,” but this term now included state constraints imposed under federally approved plans. See 54 Fed.Reg. 27,274, 27,285-86 (1989). New litigation followed but the cases were stayed (in our court) in anticipation of the 1990 amendments.

Congress thus acted in 1990 against a backdrop of over a decade of skirmishing between the agency and affected companies, during which the issue of whether and to what extent state and local controls were to be credited in calculating a source’s “potential to emit” was very much in the forefront. In drafting § 112 Congress specifically directed EPA to consider controls in determining which producers should be classified as “major sources,” but conspicuously did not limit controls to those that are federally enforceable. The government maintains that since Congress did not specify what kind of controls would qualify, EPA was free to answer that question. It permissibly did so by once again requiring that they had to be “federally enforceable,” a term which, in EPA’s final manifestation of the concept, applies to “all limitations and conditions that are enforceable by the Administrator and citizens under the Act or that are enforceable under other statutes administered by the Administrator.” 54 Fed.Reg. at 12,433.

As we have noted, it is certainly permissible for EPA to have refused to take into account ineffective controls (indeed, it is likely that a contrary interpretation would be impermissible). But is it also open to EPA under the statute to refuse to consider controls on grounds other than their lack of effectiveness? To qualify as “federally enforceable,” (as best we can determine) controls are required, in addition to being effective as a practical matter, to have been approved by EPA and integrated into the state implementation plan, or SIP, drawn up by each state to enforce substantive restrictions under the Clean Air Act and submitted to the Administrator for approval under § 110, 42 U.S.C. § 7410. Once included within the SIP, a state control becomes enforceable not only by the state which is its primary regulating authority, but also by the Administrator under § 113 of Act, 42 U.S.C. § 7413, and, in certain settings, by private citizens, who can bring suit for noncompliance with federal pollution control programs under § 304, 42 U.S.C. § 7604.

EPA has identified several state and local regulatory approaches through which states can impose restraints and have them deemed “federally enforceable.” Constraints imposed upon a source under a state operating permit, for example, will be deemed “federally enforceable” if the state program has been approved as a “federally enforceable state operating permit program,” or FESOPP, by EPA. A state permitting program cannot stand alone; it must be incorporated into the SIP, must impose upon sources a legal obligation to observe the permit constraints, must be enforceable as a practical matter— i.e., must be “effective” — must not be inconsistent with other requirements under the SIP or federal law, and must be issued pursuant to a public hearing process. See 54 Fed.Reg. 27,274,27,281-82 (1989). Other approaches are available as well. General, as opposed to source-specific, permits can also be issued under a FESOPP, or under a state general permitting program similarly approved for inclusion within the SIP. And a state can impose constraints by general prohibitionary or exclusionary rules, so long as they are included within its SIP. Finally, the SIP could be amended to reflect special, source-specific limitations. See Memorandum, Options for Limiting the Potential to Emit (PTE) of a Stationary Source Under Section 112 and Title V of the Clean Air Act, at 2-4 (EPA Jan. 25, 1995).

For each of these regulatory methods, however, EPA has proposed conditions for achieving “federal enforceability” that go beyond the mere effectiveness of particular constraint as a practical matter. Inclusion in the SIP, for example, is required in each instance even though EPA’s own approach suggests that it is a consideration independent of and in addition to the need that a constraint be effective for it to count towards reductions. There may, moreover, be regu *1364 latory techniques in addition to those that EPA deems susceptible to “federal enforceability” that are equally effective, and yet which are foreclosed as mechanisms for reducing a source’s capacity to emit as a result of EPA’s approach.

What EPA has not explained is how its refusal to consider limitations other than those that are “federally enforceable” serves the statute’s directive to “consider[] controls” when it results in a refusal to credit controls imposed by a state or locality even if they are unquestionably effective. Under EPA’s regime, even a state program of unassailable effectiveness would not qualify in computing a source’s capacity to emit unless it had been submitted not only for EPA approval, but also for inclusion in the SIP. In doing so, EPA would sacrifice a statutory objective in pursuit of ends that, at least as presented in argument to us, have not been justified, either in terms of § 112 or other provisions of the Act. EPA has not explained why it is essential that a control be included within a SIP. It is not apparent why a state’s or locality’s controls, when demonstrably effective, should not be credited in determining whether a source subject to those controls should be classified as a major or area source. 20

The government contends that its interpretation of § 112 accords with legislative history. The Senate version of the bill described controls as “installed and operating,” S.16B0, 101st Cong., 1st Sess. § 112(a) (1989); see also S.Rep. No. 228, swpra, at 151, and that version was rejected. But the term “installed and operating” controls presumably included controls imposed by an operator whether or not mandated by any regulations. The Senate had apparently wished to focus exclusively on the actual operating conditions. That Congress rejected that approach does not necessarily suggest that it implicitly delegated to EPA authority to limit the class of govemmentally imposed restraints that could be taken into account to those that are “federally enforceable.”

The government’s argument that Congress, when it enacted the amendments, implicitly ratified EPA’s past treatment of non-federally enforceable controls is not persuasive either. As we have indicated, throughout the 1980s EPA vacillated on the degree of federal involvement required. And when Congress voted in 1990, the 1989 regulations requiring federal enforceability as a prerequisite to recognition were the subject of litigation challenging their legality under the existing statute. Given this situation, Congress cannot be said to have ratified EPA’s position by introducing the phrase “considering controls” without an explicit “federal enforceability” limitation.

As presented in its briefs, EPA’s core justifications for its federal enforceability policy are the need to avoid the administrative burden that EPA would have to bear were it obligated to evaluate the effectiveness of state and local controls and the desirability of uniformity in environmental enforcement. These, of course, are not illegitimate agency objectives. Administrative problems, in particular, can under certain circumstances inform an agency’s construction of imprecise statutory language. See Drummond Coal Co. v. Hodel, 796 F.2d 503, 507 (D.C.Cir.1986). Here, however, EPA would have us accept a rather strained interpretation of the statute based on what appears to be only its unwillingness to evaluate any state or local controls that are not federalized. If there is a closer fit between the notion of “federal enforceability” and § 112’s concern with crediting effective controls it is not evident on this record.

As for national uniformity, the government contends that “one of Congress’ driving concerns in amending the hazardous air pollutants provision in the Act in 1990 was to *1365 remedy the haphazard state of air toxic regulations .... The states’ approaches to regulation varied widely,” creating “ ‘a patchwork of differing standards’ ” (citing H.R.Rep. No. 490(1), 101st Cong., 2d Sess. 232 (1990)). Just so; but the amendments do create a national substantive standard, namely categories of sources (major and area) and corresponding technological compliance measures. By no means does that suggest that Congress necessarily intended for state emissions controls to be disregarded in determining whether a source is classified as “major” or “area” under that national standard. Nor did Congress mandate that EPA assume the administration and enforcement of all governmental efforts at emissions limits. If such administration and enforcement is necessary to ensure that controls are effective in the context of the extant regulatory environment, EPA has certainly not made that case and has not indicated how that consideration supports its claim that its interpretation of the statute is reasonable.

In sum, EPA’s definition of “major source” without respect to source categories or two-digit SIC codes is reasonable, as is its requirement that fugitive emissions be included in a source’s aggregate emissions in determining whether the source is major. We therefore deny the petition for review with respect to these issues, advanced here by petitioners General Electric, National Mining Association and American Forest and Paper Association. However, EPA has not explained why the criteria for federal approval and the consequences of that approval are related to ensuring the practical effectiveness of state controls such that the set of controls considered under § 112 should be limited in that fashion. We therefore grant the petition for review with respect to the challenge raised by Chemical Manufacturers Association and American Petroleum Institute.

Notes

. The previous version of § 112 directed EPA to list those hazardous air pollutants that it intended to regulate because they might "cause, or contribute to, an increase in mortality or an increase in serious irreversible or incapacitating reversible, illness.” Pub.L. No. 91-604, 84 Stat. 1676, 1685 (1970) (codified as amended at 42 U.S.C. § 7412). For such pollutants, EPA was to institute emission standards that provided for "an ample margin of safety to protect the public health.” Id. Over 18 years, EPA listed only 8 pollutants as hazardous, S.Rep. No. 228, 101st Cong., 1st Sess. 131 (1989), and regulated only some sources of 7 of these chemicals, id. at 128; H.R.Rep. No. 490(1), 101st Cong., 2d Sess. 322 (1990).

. Section 112(a)(1) provides:

The term "major source" means any stationary source or group of stationary sources located within a contiguous area and under common control that emits or has the potential to emit considering controls, in the aggregate, 10 tons per year or more of any hazardous air pollutant or 25 tons per year or more of any combination of hazardous air pollutants. The Administrator may establish a lesser quantity ... for a major source ... on the basis of the potency of the air pollutant, persistence, potential for bioaccumulation, other characteristics of the air pollutant, or other relevant factors. 42 U.S.C. § 7412(a)(1). An "area source" is "any stationary source ... that is not a major source,” and does not include “motor vehicles or nonroad vehicles subject to regulation under [42 U.S.C. §§ 7521-7590].” 42 U.S.C. § 7412(a)(2).

. EPA develops such standards "taking into consideration the cost of achieving such emission reduction, and any non-air quality health and environmental impacts and energy requirements.” 42 U.S.C. § 7412(d)(2). For new sources, the maximum achievable reduction in emissions must be at least as stringent as the emission control achieved in practice by the best controlled similar source. 42 U.S.C. § 7412(d)(3). For existing sources in a category of 30 or more such sources, the maximum achievable reduction in emissions must be at least as stringent as the average emission limitation achieved by the 12 best-performing sources in that category. Id.

. EPA is directed to list only those "area sources" that present "a threat of adverse effects to human health or the environment.” 42 U.S.C. § 7412(c)(3). No later than five years after November 15, 1990, EPA shall have listed sufficient categories of area sources to ensure regulation of 90 percent of area source emissions of the 30 hazardous air pollutants presenting the greatest threat to public health in the largest number of urban areas. Id.', 42 U.S.C. § 7412(k).

. For purposes of § 112, a "category” of sources "is a group of sources having some common features suggesting that they should be regulated in the same way and on the same schedule.” 57 Fed.Reg. at 31,578. For example, in the area of waste treatment and disposal, some of the categories EPA listed pursuant to § 112(c)(1) include “hazardous waste incineration” and "sewage sludge incineration.” 57 Fed.Reg. at 31,591. Source categories in the area of polymers and resin production include "cellophane production” and "boat manufacturing.” Id. at 31,592.

. Standard Industrial Classification Code refers to the nomenclature used to categorize industries. Facilities may be categorized into major groups (2-digit SIC Code), industry groups (3-digit SIC Code), or industry codes (4-digit SIC Code), depending on the level of detail appropriate. Standard Industrial Classification Manual (Exec. Office of the President, Office of Management and Budget) 12 (1987).

.We consider EPA's preamble to the final rule in construing its definition of "major source.” Cf. United States v. Exxon Corp., 773 F.2d 1240, 1266 (D.C.Cir.1985), ce rt. denied, 474 U.S. 1105, 106 S.Ct. 892, 88 L.Ed.2d 926 (1986); Wiggins Bros. v. Department of Energy, 667 F.2d 77, 88 (Temp.Emer.Ct.App.1981), cert. denied, 456 U.S. 905, 102 S.Ct. 1749, 72 L.Ed.2d 161 (1982) ("It is well settled by decisions of this Court that the preamble to a regulation of DOE ... should be considered in construing the regulation and determining the meaning of the regulation.”) (citations omitted).

. According to General Electric, in the agency's notice listing initial source categories pursuant to § 112(c) EPA “explained” that § 112(d) standards did not apply to "minor sources" within a "listed category of major sources.” Brief for General Electric at 18 n. 48 (referring to 57 Fed.Reg. at 31,583). General Electric mischaracterizes EPA’s statement. After setting forth a definition of “major source” virtually identical to that in the final rule, EPA simply noted that some sources in the listed categories might be area sources. 57 Fed.Reg. at 31,583. Consistent with § 112, EPA stated that only major sources and those area sources that threaten human health or the environment must comply with emission limitations. Id.

. According to General Electric, EPA’s definition of "major source" also conflicts with Congress' attempt to "simplify and prioritize” regulation of hazardous air pollutants. Brief for General Electric at 20-21. This argument has no basis in the statute, and it ignores the statutory mechanisms that address these concerns. See, e.g., 42 U.S.C. § 7412(e) (setting out schedule under which EPA must promulgate emission standards); 42 U.S.C. § 7412(d)(1) (permitting EPA to “distinguish among classes, types and sizes of sources within a category or subcategory in establishing [emission] standards”).

. National Mining Association makes essentially the same argument, Brief for National Mining Association at 20, which we also reject for the reasons set forth here.

. Because "source” was not defined in the section of the Act that deals with preventing significant deterioration in air quality, § 169 (42 U.S.C. § 7479), the Alabama Power court held that the term had the same meaning as in § 111. 636 F.2d at 395.

. General Electric also points out that under § 112(c)(1), the source categories EPA lists shall, "to the extent practicable ... be consistent with the list of source categories established pursuant to” § 111. Brief for General Electric at 15. We do not see how this supports a conclusion that EPA must define "major source" in terms of source categories.

. In August 1994, however, EPA proposed changing the definition of "major source" in the title V permitting rule so that it conformed with the definition of that term under the § 112 general provisions. 59 Fed.Reg. 44,460, 44,514 (1994).

. The version of the bill discussed in the Senate committee report would have defined "major source” as "any stationary source (including all emission points and units of such source located within a contiguous area and under common control).” S.Rep. No. 228, supra, at 513.

. National Mining Association also contends that EPA’s definition of "major source" without reference to SIC Codes may result in the inclusion of fugitive emissions in a source's aggregate emissions without the rulemaking required by § 302(3), 42 U.S.C. § 7602(3). Because we hold that, for purposes of a § 112 major source determination, fugitive emissions may be included in a source's aggregate emissions without a § 302(j) rulemaking, infra pp. 1359-61, we reject this argument.

. In contrast to fugitive emissions, emissions emanating from a stack, chimney or vent are often called “point source” emissions. See, e.g., Alabama Power, 636 F.2d at 368.

. Section 302(j), 42 U.S.C. § 7602(j), provides:

Except as otherwise expressly provided, the terms "major stationary source” and "major emitting facility" mean any stationary facility or source of air pollutants which directly emits, or has the potential to emit, one hundred tons per year or more of any air pollutant (including any major emitting facility or source of fugitive emissions of any such pollutant, as determined by rule by the Administrator).

. In a § 302(j) rulemaking, EPA will include fugitive emissions in a source's aggregate emissions if the source has "the potential to degrade air quality significantly" and “no unreasonable socioeconomic impacts relative to the benefits would result from subjecting the sources to the relevant [programs].” 49 Fed.Reg. 43,202, 43,208 (1984).

.The proposed definition of “fugitive emissions” did not specify that such emissions were to be considered in determining whether a source is major. 58 Fed.Reg. at 42,785. Therefore, the preamble to the final rule does not discuss the issue raised here. EPA addressed the fugitive emissions question, however, in its proposal of conforming changes to the title V permit regulations. 59 Fed.Reg. at 44,514-15. EPA said that fugitive emissions should be counted in § 112 major source determinations because the term “major source” is distinguishable from the § 302(j) “major stationary source” definition, and because "section 112 establishes a new regulatory program the focus of which is specific hazardous air pollutants at source categories to be determined by EPA.” 59 Fed.Reg. at 44,514.

. Major source requirements also apply to those sources with emissions that actually exceed the major source thresholds. For a source in compliance with emissions limitations — whether federal, state or local- — "potential to emit” will exceed actual emissions, and the “potential to emit” figure will determine whether the source is major. However, should a source claim to have lowered its emissions below major source levels, but fail to conform to that claim, it will nonetheless be a major source if its actual emissions exceed the designated thresholds. A major source that fails to observe applicable requirements is subject to sanctions under § 113 of the Act, 42 U.S.C. § 7413.