Horsehead Resource Development v. Browner

Full title: HORSEHEAD RESOURCE DEVELOPMENT COMPANY, INC., PETITIONER, v. CAROL M…

Court: United States Court of Appeals, District of Columbia Circuit

Date published: Feb 22, 1994

16 F.3d 1246 (D.C. Cir. 1994)

Opinion

Nos. 91-1221 to 91-1223, 91-1230, 91-1234, 91-1237, 91-1238, 91-1240, 91-1241, 91-1244, 91-1245, 91-1572, 91-1575, 91-1576, 92-1392, 92-1459, 92-1601 and 92-1610.

Argued October 25, 1993.

Decided February 22, 1994.

Petitions for Review of an Order of the Environmental Protection Agency.

Richard G. Stoll argued the cause for industry petitioners and intervenors. Hunter L. Prillman and Gene A. Lucero argued the cause for industry petitioners. With Richard G. Stoll on the joint briefs were David P. Novello, John N. Hanson, Donald J. Patterson, Jr., Edward M. Green, Robert N. Steinwurtzel, Michael W. Steinberg, Arline M. Sheehan, David F. Zoll, Ronald Shipley, Christopher H. Marraro, Patricia G. Butler, Paul E. Gutermann, John N. Moore, Kevin A. Gaynor, Thomas R. Bartman, Joseph E. LeBlanc, Jr., David B. Graham, Lynn L. Bergeson, Kurt J. Olson, Howard B. Myers, Jeremiah J. Jewett, III, Lee B. Zeugin, G. William Frick, Robert L. Sullivan, and David B. Weinberg. Jeffrey S. Holik entered an appearance in 91-1222. Richard S. Wasserstrom entered an appearance in 91-1234. Jennifer L. Wurzbacher entered an appearance in 91-1238. Karl S. Bourdeau and Paul E. Shorb, III entered appearances in 91-1244. Samuel Ira Gutter and Gene A. Lucero entered appearances in 92-1392.

David R. Case argued the cause for petitioners Citizens for a Safe Environment, et al. With him on the briefs were Eli D. Eilbott, Charles W. Elliott, James E. Bradley, Mark A. Finkelstein, Turner R. Odell, Jr., and Douglas M. MacMillan. Charles W. Elliott, Jacqueline M. Warren, Joseph H. Guth and James E. Bradley entered appearances in 91-1240.

Steven E. Silverman, Attorney, United States Environmental Protection Agency and Christopher S. Vaden, Attorney, U.S. Dept. of Justice, argued the cause for respondents. With them on the briefs were Peter R. Steenland, Jr., Acting Deputy Assistant Attorney General, David A. Dana and Gretchen Slosser Pirasteh, Attorneys, United States Department of Justice.

William R. Weissman argued the cause for intervenors Edison Electric Institute, et al. With him on the brief were Douglas H. Green, Barton C. Green, Karl S. Bourdeau, John N. Hanson, Donald J. Patterson, Jr., Edward M. Green, Robert N. Steinwurtzel, David B. Weinberg, Kurt J. Olson, Richard G. Stoll, David P. Novello, Michael W. Steinberg, Arline M. Sheehan, Hunter L. Prillman, David F. Zoll, Ronald Shipley, Timothy Berg, Douglas E. McAllister, Paul E. Gutermann, John N. Moore, Neil J. King, Lynn L. Bergeson, Howard B. Myers, Jeremiah J. Jewett, III and Roderick T. Dwyer.

David R. Case, Eli D. Eilbott, Charles W. Elliott, James E. Bradley, Mark A. Finkelstein, Turner R. Odell, Jr. and Douglas M. MacMillan were on the brief for intervenors Citizens for a Safe Environment, et al.

Aaron H. Goldberg entered an appearance for intervenor International Mill Service in 91-1221.

Paul E. Shorb, III entered an appearance for intervenor American Iron and Steel Institute in 91-1221.

Jeffrey S. Holik entered an appearance for intervenor Association of Battery Recyclers, Inc. in 91-1221.

Norman L. Rave, Jr. entered an appearance for intervenor Chemical Waste Management, Inc. in 91-1221.

Susan M. Schmedes and G. William Frick entered appearances for intervenor American Petroleum Institute in 91-1221.

David R. Case entered an appearance for intervenor Hazardous Waste Treatment Council in 91-1221.

William F. Pedersen and Ira Dassa entered appearances for intervenor American Coke and Coal Chemicals Institute in 91-1221 and 92-1244.

Joseph E. LeBlanc, Jr. entered an appearance for intervenor Marine Shale Processors, Inc. in 91-1221.

William M. Guerry, Jr. and John L. Wittenborn entered appearances for intervenor Steel Manufacturers Association Specialty Steel Industry of the United States in 91-1221.

Lee B. Zeugin entered an appearance for intervenor Tennessee Eastman Company in 91-1221.

Richard S. Wasserstrom and Richard S. Moskowitz entered appearances for intervenor AKJ Industries, Inc. in 91-1240.

Aaron H. Goldberg entered an appearance for intervenors Dow Chemical Company, International Mill Service, Inc., American Iron and Steel Institute and American Mining Congress in 91-1572.

Edward M. Green and Roderick T. Dwyer entered appearances for American Mining Congress in 91-1572.

Before: WALD, EDWARDS and BUCKLEY, Circuit Judges.

Opinion for the Court filed PER CURIAM.

PER CURIAM:

This case involves challenges by industry and environmentalist petitioners to an Environmental Protection Agency ("EPA") rule interpreting section 3004(q) of the Resource Conservation and Recovery Act ("RCRA"), 42 U.S.C. § 692(q) (1988), which directs the agency to regulate facilities burning hazardous waste as fuel. The rule, entitled "Burning of Hazardous Waste in Boilers and Industrial Furnaces" ("BIF Rule"), 56 Fed.Reg. 7,134 (1991) (codified as amended by subsequent corrections and technical amendments at 40 C.F.R. Parts 260, 261, 264, 265, 266, 270, 271 (1992)), is principally designed to establish air emissions requirements for facilities burning hazardous waste as fuel. The BIF Rule also integrates the mandates of section 3004(q) and the Bevill Amendment, the latter of which is a RCRA provision that excludes certain types of waste from RCRA Subtitle C's hazardous waste management regime until the EPA completes several studies and submits their results to Congress. See 42 U.S.C. § 6921(b)(3)(A) (1988).

Petitioners attack the BIF Rule on three grounds.¹ First, both groups of petitioners contend that the BIF Rule fails to adhere to what they regard as Congress' clear expression of its intent in the Bevill Amendment. There are two types of "Bevill wastes" at issue in this case: cement kiln dust ("ckd"); and combustion residues produced when BIFs burn fuel consisting primarily of coal, oil, and other fossil fuels. Although the Bevill Amendment states that ckd and fossil fuel combustion residues are exempt from Subtitle C pending completion of the required studies, the BIF Rule provides that these Bevill wastes can under some circumstances be regulated as Subtitle C hazardous wastes when they are produced by BIFs burning fuel which contains hazardous waste. Industry petitioners contend that the Bevill Amendment exempts ckd and fossil fuel combustion residues from Subtitle C regardless of whether they are the product of hazardous waste fuels. Environmentalist petitioners, Citizens for a Safe Environment, et al. ("CASE petitioners"), make the opposite argument, contending that Bevill wastes produced by BIFs burning any quantum of hazardous waste are automatically subject to Subtitle C. We reject both of these challenges and uphold the BIF Rule as a permissible interpretation of RCRA § 3004(q) and the Bevill Amendment.

1 The two groups of petitioners originally offered twelve separate challenges to the BIF Rule; however, all but three of these issues were resolved by settlements between the parties on the eve of oral argument. We address only the three remaining challenges.

Second, we uphold the BIF Rule insofar as it regulates air emissions from BIFs burning mixtures of hazardous waste and non-waste fuel. Industry petitioners contend that the rule is invalid because non-waste fuels are outside the EPA's jurisdiction under RCRA. We hold, however, that RCRA § 3004(q) gives the EPA the authority to regulate facilities which burn hazardous waste, not merely the hazardous waste component of the fuel they burn. The EPA contends that it is not technologically feasible to distinguish air emissions generated solely by the hazardous waste component of mixed fuels. Because BIFs that burn hazardous waste are "treating" that waste as that term is defined in RCRA, we can find no error in the EPA's position that air emissions from the co-processing of mixed hazardous waste and non-waste fuels have a sufficiently close nexus to waste treatment to justify the regulation at issue. In addition, we hold that the agency's failure explicitly to respond to the specific alternatives proposed by Industry petitioners was not arbitrary and capricious because the EPA's general discussion of its decision to regulate non-waste fuels provided a sufficient response to Industry petitioners' proposals.

Finally, Industry petitioners challenge the EPA's purported regulation of products of incomplete combustion ("PICs") produced when BIFs burn hazardous waste fuel. We remand "Tier III" of the PIC provisions of the BIF Rule on the grounds that the regulation was promulgated without adequate notice and comment, and because the disputed rule lacks an adequate basis in the rulemaking record; however, we uphold the Tier I and Tier II standards.

I. BACKGROUND

Subtitle C of RCRA establishes a comprehensive "cradle-to-grave" regulatory program for the treatment, storage, and disposal of hazardous waste. United Technologies Corp. v. EPA, 821 F.2d 714, 716 (D.C. Cir. 1987). For RCRA purposes, burning hazardous waste constitutes "treatment" of it,² thus giving the EPA the authority to regulate this activity. The EPA promulgated the BIF Rule pursuant to the mandate of RCRA section 3004(q), which directs the agency to establish standards which will "protect human health and the environment" governing facilities that burn fuel containing hazardous waste. 42 U.S.C. § 6924(q) (1988).³

2 See 42 U.S.C. § 6903(34) (1988) ("treatment" means "any method, technique, or process, including neutralization, designed to change the physical, chemical, or biological character or composition of any hazardous waste. . . ."); 40 C.F.R. § 260.10 (1992) ("treatment" includes "any method, technique, or process. . . designed to . . . recover energy or material resources from [hazardous] waste. . . ."); see also Shell Oil Co. v. EPA, 950 F.2d 741, 752-56 (D.C. Cir. 1991) (upholding the EPA's decision to incorporate resource recovery within the regulatory definition of "treatment").

3 The relevant portion of RCRA § 3004(q) reads:
(q) Hazardous waste used as fuel
(1) Not later than two years after November 8, 1984, and after notice and opportunity for public hearing, the Administrator shall promulgate regulations establishing such —
(A) * * *

(B) standards applicable to the owners of facilities which burn, for purposes of energy recovery, any fuel [produced from or mixed with hazardous waste or] which otherwise contains any hazardous waste identified or listed under section 6921 of this title. . . .

Hazardous waste is burned for three purposes: to destroy the waste (incineration), to serve as fuel (energy recovery), and to recover usable materials such as metals (materials recovery). Hazardous waste often has a great deal of heat value when used as fuel; and this use also has the benefit of destroying or at least reducing the volume of the waste, thereby reducing reliance on landfilling. In its early attempts to implement Subtitle C, the EPA was particularly concerned that its regulations not discourage beneficial uses of hazardous wastes, such as energy recovery and recycling. See 45 Fed.Reg. 33,084 at 33,092-94 33,120 (1980). For this reason, in 1981, the EPA deferred regulating air emissions from BIFs burning hazardous waste as fuel or for materials recovery, but did adopt rules controlling emissions from hazardous waste incinerators, which burn waste primarily in order to destroy it. 46 Fed.Reg. 7,666, 7,678 (1981).

Exempting facilities that burned hazardous waste for energy recovery from Subtitle C's requirements created a regulatory "loophole" by means of which over half of the hazardous waste generated in the United States came to be burned in BIFs not subject to RCRA. H.R. Rep. No. 198, 98th Cong., 1st Sess., pt. 1, at 39 (1983), U.S. Code Cong. Admin. News 1984, p. 5576. Congress closed this loophole by enacting RCRA section 3004(q) as part of the Hazardous and Solid Waste Amendments of 1984, Pub.L. No. 98-616, § 204, 98 Stat. 3221. Section 3004(q) set a deadline of November 8, 1986 for the EPA to promulgate regulations governing the burning of hazardous waste for energy recovery.

In 1985, in response to section 3004(q), the EPA imposed its first controls on the marketing and burning of hazardous waste fuels. 50 Fed.Reg. 665 (1985); 50 Fed.Reg. 49,164 (1985). These regulations provide that hazardous waste or fuels containing hazardous waste are subject to transportation and storage controls under Subtitle C prior to being burned as fuel or being blended or processed for use as fuel.

In 1987, again pursuant to section 3004(q), the EPA published its proposed BIF Rule for public comment. 52 Fed.Reg. 16,982-17,050 (1987). The EPA issued a supplemental rulemaking proposal in 1989. 54 Fed.Reg. 43,718-63 (1989). The final BIF Rule, which we review in the instant case, was published in the Federal Register in 1991. 56 Fed.Reg. 7,134-240 (1991). The EPA subsequently promulgated a series of technical amendments and clarifications to the BIF Rule. See Corrections and Technical Amendments, 56 Fed.Reg. 32,688-852 (1991); Technical Amendments, 56 Fed.Reg. 42,504-17 (1991); Technical Clarification Amendments and Corrections, 57 Fed.Reg. 38,558-66 (1992).

II. DISCUSSION

A. Standard of Review

The standard of review in this case is the familiar two-part framework of Chevron U.S.A., Inc. v. Natural Resources Defense Council, 467 U.S. 837, 104 S.Ct. 2778, 81 L.Ed.2d 694 (1984). First, the court asks whether Congress "has directly spoken to the precise question at issue." Id. at 842, 104 S.Ct. at 2781. If so, then the intent of Congress is controlling. If, however, the meaning of the statute is ambiguous, then we will uphold the agency's interpretation so long as it is "based on a permissible construction of the statute." Id.

As this court explained in Natural Resources Defense Council v. Reilly, 983 F.2d 259, 266 (D.C. Cir. 1993):

[I]t is only legislative intent to delegate such authority that entitles an agency to advance its own statutory construction for review under the deferential second prong of Chevron. See Chevron, 467 U.S. at 843-44, 104 S.Ct. at 2782. "If Congress has explicitly left a gap for the agency to fill, there is an express delegation of authority . . . . Sometimes the legislative delegation to an agency on a particular question is implicit rather than explicit. In such a case, a court may not substitute its own construction of a statutory provision for a reasonable interpretation made by the administrator of an agency." Id.

CASE petitioners and Industry petitioners attack the BIF Rule's interpretation of the Bevill Amendment from opposite directions, both relying on Chevron step one to argue that Congress' clearly expressed intent requires the outcome they advocate. CASE petitioners argue that the EPA may not exempt any co-processed Bevill wastes from Subtitle C because Congress did not intend for the Bevill Amendment to cover combustion residues or ckd produced by fuel that contains hazardous waste. Industry petitioners argue that when Congress enacted the Bevill Amendment it was aware that Bevill devices and BIFs burned hazardous waste, and that the amendment was intended to exempt ckd and residues without regard to the type of fuel that produced them. The EPA contends, and we agree, that the language of the Bevill Amendment does not unambiguously address the issue of hazardous waste fuels burned in Bevill devices, and that its regulation is a permissible interpretation of the statute under Chevron step two.

B. Interpretation of the Bevill Amendment Under the BIF Rule 1. The Bevill Amendment

The Bevill Amendment⁴ to RCRA limits the authority of the EPA to regulate certain enumerated types of solid wastes until it completes studies concerning them and submits the results of those studies to Congress. The EPA has yet to complete the relevant studies, although the statutory deadlines expired in 1982 and 1983. See 42 U.S.C. § 6982(n), ( o) (1988). Two classes of Bevill wastes are relevant to the dispute before us: cement kiln dust or "ckd," which is the particulate matter emitted in the exhaust gases of cement kilns that is collected by emission control devices such as fabric filters or electrostatic precipitators; and combustion residues, such as ash and particulates, which are generated primarily from the combustion of coal or other fossil fuels.

4 42 U.S.C. § 6921(b)(3)(A) (1988):

[E]ach waste listed below shall, except as provided in subparagraph (B) of this paragraph, be subject only to regulation under other applicable provisions of Federal or State law in lieu of this subchapter until at least six months after the date of submission of the applicable study required to be conducted under subsection (f), (n), ( o), or (p) of section 6982 of this title and after promulgation of regulations in accordance with subparagraph (C) of this paragraph:

(i) Fly ash waste, bottom ash waste, slag waste, and flue gas emission control waste generated primarily from the combustion of coal or other fossil fuels.

(ii) Solid waste from the extraction, beneficiation, and processing of ores and minerals, including phosphate rock and overburden from the mining of uranium ore.

(iii) Cement kiln dust waste.

This court has described the origins of the Bevill Amendment at length in three previous decisions concerning the provision for mining wastes, 42 U.S.C. § 6924(b)(3)(A)(ii) (1988). See Solite Corp. v. EPA, 952 F.2d 473, 477-82 (D.C. Cir. 1991); Environmental Defense Fund v. EPA, 852 F.2d 1316, 1318-21 (D.C. Cir. 1988), cert. denied, 489 U.S. 1011, 109 S.Ct. 1120, 103 L.Ed.2d 183 (1989) (" EDF II"); Environmental Defense Fund v. EPA, 852 F.2d 1309, 1310-12 (D.C. Cir. 1988), cert. denied, 489 U.S. 1011, 109 S.Ct. 1120, 103 L.Ed.2d 183 (1989) (" EDF I"). Here we need only summarize our earlier holdings as to Congress' intent in enacting the Amendment.

In 1978, in its first proposal for regulation of hazardous waste under Subtitle C, the EPA proposed to create a category of "special wastes," among which were ckd and combustion residues produced by utilities, for which special, less stringent standards might be appropriate. See 43 Fed.Reg. 58,946, 58,992 (1978). The EPA had not yet fully studied these special wastes, but the agency believed they potentially presented unique problems because they were generated in very large volumes but posed relatively low hazards to human health. Because of the large volumes of these wastes, the EPA was concerned that it might be impractical to dispose of them in facilities that fully complied with Subtitle C of RCRA. However, the EPA eventually decided to abandon the concept of "special wastes." The agency pointed out that the definition of hazardous waste in its final regulations was less stringent than the one it had initially proposed (meaning fewer special wastes would qualify as hazardous wastes), and that the final waste management standards were more flexible than those it had originally proposed. Thus, it was felt that there was no need for a category of "special wastes." See 45 Fed.Reg. 33,084, 33,174-75 (1980); EDF II, 852 F.2d at 1319; EDF I, 852 F.2d at 1311.

One month before the 1980 Subtitle C regulations were to take effect, Congress enacted the Bevill Amendment as part of the Solid Waste Disposal Act Amendments of 1980, Pub.L. No. 96-482, 94 Stat. 2334, 2337 § 7. In response, the EPA amended its hazardous waste regulations in November 1980 to incorporate the Bevill Amendment's exemption of ckd, mining wastes and fossil fuel combustion residues from the requirements of Subtitle C. 45 Fed.Reg. 76,618 (1980).

In EDF II, this court held that "it is clear that Congress intended the Bevill exclusion to encapsulate the `special waste' concept articulated by the EPA in 1978." EDF II, 852 F.2d at 1329. That case held that the EPA was not permitted simply to exempt from Subtitle C all mining wastes potentially within the scope of the Bevill Amendment, but rather was required to comply with Congress' clearly expressed intent that the agency exempt only those mining wastes that would have fallen within the category of "special wastes." EDF II thus required the EPA to develop criteria to determine whether potentially exempt mining wastes were "high volume, low hazard" wastes, such as the EPA had proposed to include in the "special waste" category in 1978. Id. at 1331. Accordingly, this court also held that the EPA could not continue to grant a Bevill exemption from Subtitle C to six types of smelter wastes that the agency had found to be low volume and high hazard, because such wastes were outside the scope of the Bevill Amendment. Id. at 1329-30.

In Solite we reviewed the EPA's regulatory definitions of the Bevill Amendment's high volume/low hazard criteria for mining wastes, which the agency had promulgated in response to our decision in EDF II. Solite reaffirmed EDF II's holding that Congress intended the Bevill Amendment to encapsulate the special waste concept. See Solite, 952 F.2d at 482-83 (quoting EDF II, 852 F.2d at 1329). This court went on to hold, however, that Congress had not directly addressed "the appropriate delineation of the `special waste' concept, and the attendant `high volume' and `low hazard' criteria. . . ." Id. at 482 (emphasis added). Our review of the EPA's volume and hazard criteria for mining waste thus was limited to the question of whether the agency's interpretation was a permissible construction of the Bevill Amendment. Id. (citing Chevron, 467 U.S. at 843, 104 S.Ct. at 2782).

2. The Integration of the Bevill Amendment and Section 3004(q) Mandates Under the BIF Rule

The EPA has adopted a two-part definition of "hazardous wastes" that are subject to Subtitle C. "Characteristic hazardous wastes" are substances exhibiting one or more of four hazardous characteristics: ignitability, corrosivity, reactivity and toxicity. See 40 C.F.R. Part 261, Subpart C (1992). "Listed hazardous wastes" are those substances the EPA has enumerated in published lists of specified wastes. See id. at Subpart D. By regulation, a listed waste continues to be regulated as a hazardous waste after it is mixed with other wastes — a provision known as the "mixture rule." 40 C.F.R. § 261.3(a)(2)(iv) (1992). A listed waste also remains a hazardous waste for Subtitle C purposes even after it is "treated" in some fashion — the so-called "derived-from" rule. 40 C.F.R. § 261.3(c)(2)(1), (d)(2) (1992). As noted above, for RCRA purposes, the burning of a hazardous waste for any purpose, whether for fuel, materials recovery or destruction, constitutes "treatment." 42 U.S.C. § 6903(34) (1988).⁵

5 In 1991, in Shell Oil Co. v. EPA, 950 F.2d 741 (D.C. Cir. 1991), this court vacated the mixture and derived-from rules for failure to give adequate opportunity for notice and comment. The Shell Oil decision suggested, however, that the EPA repromulgate the rules on an interim basis under the APA's good cause exemption, id. at 752, and the agency did so on March 3, 1992. 57 Fed.Reg. 7,628 (1992). The interim rule has been challenged in Mobil Oil Corp. v. EPA, No. 92-1211, 1993 WL 411447, currently scheduled for oral argument before this court on March 8, 1994.

All parties to this litigation agree that ckd and combustion residues from Bevill devices burning hazardous waste fuel would be subject to Subtitle C by operation of the derived-from rule absent the Bevill Amendment; this is so because they would contain residues from the burning of hazardous waste. The question before this court is the extent to which the Bevill Amendment does or does not exempt ckd and combustion residues produced by burning fossil fuels mixed with hazardous waste from the operation of the derived-from rule.

The BIF Rule integrates the mandates of the Bevill Amendment and section 3004(q) by providing both quantitative and qualitative requirements which Bevill wastes must satisfy in order to retain their exemption from Subtitle C. 40 C.F.R. § 266.112 (1992). The quantitative aspect of the rule requires that in order for ckd to retain its Bevill exemption, the kiln producing it must process at least fifty percent by weight of ordinary, non-hazardous-waste raw materials. Id. § 266.112(a)(3). The EPA noted in its statement accompanying the final BIF Rule that although the Bevill Amendment refers to devices burning "primarily" fossil fuels, it does not explicitly so limit the Subtitle C exclusion for ckd. The agency determined, however, that, to retain its Bevill exclusion, a cement kiln must burn chiefly non-hazardous waste materials. See 56 Fed.Reg. 7,134, 7,197-98 (1991). This provision apparently is intended to prevent cement kilns from effectively becoming hazardous waste incinerators by burning far more waste than actually necessary to fuel cement making. The EPA interpreted the Bevill Amendment's exclusion for residues "primarily from the combustion of coal or other fossil fuels" as exempting only residues from BIFs burning at least fifty percent coal.⁶ Residues from BIFs burning oil or gas mixed with any quantity of hazardous waste are not covered by the Bevill Amendment because oil and gas generally produce very little combustion residue, so any residue from the co-processing of these fuels with hazardous waste would necessarily be "significantly affected" by the hazardous waste. Id. at 7,198 n. 87.

6 The percentage of coal burned in a BIF may be measured on a "total heat input or mass input basis, whichever results in the greater mass feed rate of coal." 40 C.F.R. § 266.112(a)(1) (1992).

The BIF Rule's qualitative requirements provide that Bevill wastes which meet the foregoing quantitative criteria remain exempt from Subtitle C so long as they have not been "significantly affected" by the burning of hazardous waste. The EPA has established a test to be performed on a case-by-case basis at each facility to determine whether co-processed Bevill wastes have been "significantly affected." See 56 Fed.Reg. 7,196-99 (1991). The agency considered adopting uniform baseline levels rather than requiring site-specific testing, but determined that it did not have sufficient data to do so. See id. at 7,198; 54 Fed.Reg. 43,718 at 43,734 43,735-36 (1989).

The "significantly affected" test consists of two parts. First, a Bevill device operator may opt to determine whether concentrations of specified toxic compounds in co-processed residue are significantly higher than in residues produced from burning solely non-hazardous-waste fuels. 40 C.F.R. § 266.112(b)(1) (1992). If the co-processed Bevill wastes are sufficiently similar to ordinary Bevill wastes, the wastes remain exempt from Subtitle C. Second, the operator may determine whether toxic compounds are present in the waste in levels that exceed so-called "health based limits" — levels that "could pose a significant risk to human health." 56 Fed.Reg. 7,134 at 7,198 7,199 (1992); see 40 C.F.R. § 266.112(b)(2) (1992). The first part of the qualitative test need not be conducted if the residue does not exceed the health-based limits. 40 C.F.R. § 266.112(b) (1992). If, however, the residue exceeds the health-based limits, it is exempted from Subtitle C only if it is sufficiently similar to normal residue to pass the first test.

The EPA noted in its statement accompanying the final BIF Rule that it believed that it had struck a "reasonable balance between the terms of the Bevill amendment and other provisions and regulations relating to hazardous waste management." 56 Fed.Reg. 7,134, 7,197 (1991). The agency observed that to disqualify co-processed residues from the Bevill exclusion in cases in which the residue was not "significantly affected" would

exalt form over substance by barring from Bevill eligibility a residue that was not discernably affected by burning hazardous waste. Given that such material could be exactly the high volume/low hazard residue that Congress told the Agency to study before regulating, EPA does not agree with an interpretation that automatically forecloses it from Bevill status. In addition, use of Bevill devices provides needed treatment capacity for a number of hazardous wastes, and the Agency would be reluctant to adopt an interpretation that discouraged safe processing of hazardous waste. . . .

Id. (footnote omitted). The EPA also pointed out that air emissions from Bevill devices, as opposed to combustion residues, would be subject to the requirements of the BIF Rule in all cases; and that the facilities at which Bevill devices were located would be subject to Subtitle C's corrective action requirements in the event hazardous constituents were released into the environment in ways or quantities not contemplated by the facilities' RCRA permits. Id.; see also American Iron Steel Inst. v. EPA, 886 F.2d 390, 393-96 (D.C. Cir. 1989) (holding Bevill wastes are not exempt from RCRA corrective action requirements because those requirements apply to hazardous constituents, not merely to hazardous wastes), cert. denied, 497 U.S. 1003, 110 S.Ct. 3237, 111 L.Ed.2d 748 (1990).

3. Industry Petitioners' Challenge to the BIF Rule's Interpretation of the Bevill Amendment

Industry petitioners contend that Congress intended that Bevill wastes be exempt from Subtitle C without regard to the materials burned by the device that produced them.⁷ We hold that such an interpretation of the Bevill Amendment would be both irrational and contrary to the law of this circuit as defined by Solite and EDF II, and accordingly we deny the petitions for review on this point.

7 The attack on the regulation of ckd was submitted by Industry petitioners Cement Kiln Recycling Coalition; Giant Group, Ltd.; Lafarge Corporation and Solite Corporation. Industry petitioners' challenge to the regulation of combustion residues was submitted on behalf of Chemical Manufacturers Association and Tennessee Eastman Company.

Industry petitioners make much of the fact that RCRA section 3004(q) specifically provides that: "Nothing in this subsection shall be construed to affect or impair the provisions of [the Bevill Amendment]." 42 U.S.C. § 6924(q)(1)(C) (1988). According to Industry petitioners, this clause evinces a congressional intent to exclude from Subtitle C all Bevill wastes that are the product of burning hazardous waste fuels. This argument begs the question, however. We agree that section 3004(q) plainly does not overrule or modify the Bevill Amendment; the question before us is how to integrate the two provisions. We are concerned in this case not with whether the Bevill Amendment retains its vitality, but with what it means.

As we noted above, this court held in EDF II that the EPA was required to limit Bevill wastes excluded from Subtitle C to those wastes that are high volume/low hazard. In Solite we held that the EPA had discretion to define the "high volume" and "low hazard" criteria so long as its definitions were permissible interpretations of the Bevill Amendment. In the instant case a "high volume" criterion is not at issue. However, the two-part "significantly affected" test under the BIF Rule is plainly a means to limit the Bevill exclusion to wastes that are in fact low hazard; we therefore review it under Chevron step two.

Neither Industry petitioners nor CASE petitioners challenge the specific parameters the EPA has chosen for "health-based limits," or the test the agency prescribes to determine whether Bevill wastes from devices burning hazardous waste fuel are significantly different from those produced by burning ordinary fuel. Thus, we consider the permissibility of the "significantly affected" test in the absence of any specific challenges to the test itself. Both groups of petitioners argue only that the EPA may not impose such a test at all, not that the test the EPA has promulgated is flawed in any way.

Although the Solite and EDF II decisions involved only mining wastes under the Bevill Amendment, the analyses in those opinions are wholly applicable to the instant case as well. Industry petitioners suggest that our earlier decisions permitted the EPA to impose hazard and volume criteria only because such guidelines were necessary to implement the mining waste exemption under the Bevill Amendment, which did not specifically enumerate the wastes it covered, but simply referred to "[s]olid waste from the extraction, beneficiation, and processing of ores and minerals." 42 U.S.C. § 6921(b)(3)(A)(ii) (1988). However, we find nothing in our prior decisions to suggest that Congress did not intend to apply the "special waste" framework to all of the wastes listed in the Bevill Amendment — indeed, ckd and combustion residues generated by utilities were among the wastes the EPA suggested for the special waste category in its 1978 proposal. See EDF II, 852 F.2d at 1319.

Further, it simply makes no sense to permit Bevill devices to become inadequately regulated dumping grounds for hazardous materials. According to a source cited by CASE petitioners, cement kilns alone now burn more than twice as much waste as RCRA-permitted incinerators, a figure that Industry petitioners do not attempt to refute. See Brief for Petitioners Citizens for a Safe Environment, et al. at 18-19. Industry petitioners would have us hold that Congress intended that cement kilns and other Bevill devices may burn anything — even spent nuclear fuel, infectious medical waste, or discarded chemical weapons — without the resulting residues being subject to the hazardous waste regulation regime Congress created by enacting Subtitle C. This we are unwilling to do.

The EPA's interpretation of the Bevill Amendment is obviously a reasonable one. Under BIF Rule, Bevill wastes retain their exemption from Subtitle C so long as they are not "significantly affected" by the hazardous waste fuel burned to produce them, or do not present a hazard to human health and the environment. The BIF Rule is designed to distinguish combustion residues and ckd from materials that are, in effect, very different substances. Quite simply, if ckd from co-processing hazardous waste fuels is "significantly affected," it is no longer just ckd, but ckd plus the products of combustion of hazardous waste. We therefore hold that the BIF Rule is a permissible interpretation of the Bevill Amendment's low hazard criterion.

4. CASE Petitioners' Challenge to the BIF Rule's Interpretation of the Bevill Amendment a. CASE Petitioners' Standing

Industry intervenors contend that CASE petitioners lack standing to challenge the BIF Rule's interpretation of the Bevill Amendment.⁸ We need not linger long on this issue as we hold that CASE petitioners plainly do have standing.

8 This standing challenge was offered by Industry intervenors American Mining Congress, Cyprus Miami Mining Corporation, Edison Electric Institute, Solite Corporation and Tennessee Eastman Company. The EPA did not contest CASE petitioners' standing.

To support their standing argument, Industry intervenors attempt to rely on this court's decisions in Hazardous Waste Treatment Council v. EPA, 861 F.2d 270 (D.C. Cir. 1988) (" HWTC I"), and its companion case, Hazardous Waste Treatment Council v. EPA, 861 F.2d 277 (D.C. Cir. 1988), cert. denied, 490 U.S. 1106, 109 S.Ct. 3157, 104 L.Ed.2d 1020 (1989) (" HWTC II"). In HWTC II this court held that the Hazardous Waste Treatment Council ("HWTC"), a trade association of firms engaged in the treatment of hazardous waste and one of the CASE petitioners in this case, lacked standing to challenge the EPA's then-existing section 3004(q) regulations governing the burning of hazardous waste as fuel. HWTC contended in that case that it had standing because the alleged laxity of the EPA's regulatory scheme would injure its members by diminishing the market for their waste management services. Among the injuries HWTC alleged was that because the EPA at that time exempted all Bevill residues from Subtitle C, HWTC's members who operated hazardous waste incinerators were at a competitive disadvantage when selling waste incineration services because they were forced to bear the costs of disposing of their combustion residues in compliance with Subtitle C. See HWTC II, 861 F.2d at 281. We found these allegations inadequate to establish standing. See id. at 280-85.

In the instant case, Industry intervenors contend that HWTC attempts to redress the same injury found inadequate for standing in HWTC II. This argument is to no avail, however. Even assuming arguendo that Industry intervenors are correct as to HWTC, the environmental organizations who are also among the CASE petitioners clearly do have standing. As was the case in HWTC I, because we find that the environmental organizations among the CASE petitioners have standing, the question of HWTC's standing raises a moot issue. See HWTC I, 861 F.2d at 273 (citing Bowen v. Kendrick, 487 U.S. 589, 620 n. 15, 108 S.Ct. 2562, 2580, 101 L.Ed.2d 520 (1988)).

Industry intervenors argue that the environmental organizations among the CASE petitioners lack standing because the EPA's health-based limits are designed to ensure protection of human health and the environment, and so petitioners cannot allege an actual injury. We disagree. The Supreme Court recently held that "the irreducible constitutional minimum of standing contains three elements." Lujan v. Defenders of Wildlife, ___ U.S. ___, 112 S.Ct. 2130, 2136, 119 L.Ed.2d 351 (1992). First, a petitioner must have suffered an "injury in fact" that is both concrete and particularized, and actual or imminent. Second, the injury must be fairly traceable to the challenged action of the defendant. Third, "it must be likely, as opposed to merely speculative, that the injury will be redressed by a favorable decision." Id. (internal quotation omitted). CASE petitioners have alleged, and Industry intervenors do not dispute, that some of their members live in communities near Bevill devices of various types. In HWTC I we upheld the standing of an environmental group which asserted that the EPA's failure to regulate used oil as a Subtitle C hazardous waste constituted an injury-in-fact to its members living in communities potentially subject to incidents resulting from mismanagement of used oil. 861 F.2d at 273. Further, in that case this court also held that in the event the environmental group won its lawsuit, remanding the used oil regulations to the EPA for reconsideration of its subtitle C listing decision would be "likely to redress the injuries alleged." Id.

The instant case presents a standing inquiry identical in all relevant respects to the allegations we held adequate for standing in HWTC I. The environmental groups among the CASE petitioners allege that the BIF Rule is an unlawful interpretation of the Bevill Amendment which exposes their members to greater risks than they would face if all Bevill wastes were regulated under Subtitle C. If the EPA were required to regulate Bevill wastes as Subtitle C hazardous wastes, those wastes would be subject to a more stringent regulatory regime than the current BIF Rule imposes, providing greater protection to petitioners' members.

Because the environmental organizations among the CASE petitioners have standing, we hold that the CASE petitioners as a group also have standing to challenge the BIF Rule.

b. CASE Petitioners' Interpretation of the Bevill Amendment

CASE petitioners assert that the Bevill Amendment exempts from Subtitle C only Bevill wastes produced by facilities burning purely non-hazardous-waste fuel. As we discussed at length above, Congress intended the Bevill Amendment to enact the high volume/low hazard "special waste" concept. The plain language of RCRA section 3004(q) makes clear that Congress knew when it added that section that Bevill devices burned hazardous waste, and suggests that it expected they might be subject to a different regulatory regime than that imposed on other BIFs. Given these facts, we cannot hold that the legislature clearly intended automatically to remove the Bevill exemption for all Bevill wastes generated from co-processing hazardous wastes.

RCRA section 3004(q)(2)(C)(i) provides that "no fuel which contains any hazardous waste may be burned in any cement kiln which is located within" a city with a population of over five hundred thousand, unless the kiln complies with RCRA's incinerator regulations. 42 U.S.C. § 6924(q)(2)(C)(i) (1988). This provision is incontrovertible evidence that Congress knew cement kilns and other Bevill devices burned hazardous waste. Further, the fact that Congress singled out cement kilns in large cities, but did not address other types of boilers or industrial furnaces that burn hazardous waste, suggests that the legislature believed that cement kilns might be subject to a different regulatory regime than other BIFs, as there would be no reason to assume that if cement kilns were subject to the BIF Rule they would present hazards distinguishable from other devices burning similarly hazardous wastes.

At the time Congress enacted section 3004(q) it was aware that Bevill devices burned hazardous waste. In that section it specifically reaffirmed the Bevill Amendment, while at the same time ordering the EPA to regulate facilities burning hazardous waste as fuel. Nevertheless, nowhere in section 3004(q), or anywhere else in RCRA, does Congress indicate how it wished the EPA to integrate the potentially conflicting provisions at issue in this case. Under these circumstances, we cannot hold that the legislature has "directly addressed the precise question at issue." Chevron, 467 U.S. at 843, 104 S.Ct. at 2782. The BIF Rule provides for the continued vitality of the Bevill Amendment, as section 3004(q) requires, by ensuring that high volume/low hazard criteria remain the standard for exemption of Bevill wastes from Subtitle C. At the same time, the rule recognizes that Bevill wastes that are not low hazard do come within section 3004(q)'s broad mandate that the EPA regulate the burning of hazardous waste. This result is a permissible interpretation of both section 3004(q) and the Bevill Amendment.

c. CASE Petitioners' Challenge to the EPA's Exemption of Bevill Wastes from RCRA's Land Disposal Restrictions

CASE petitioners also challenge the EPA's determination that Bevill wastes that pass the BIF Rule's "significantly affected" test are not subject to RCRA's land disposal restrictions ("LDR"). The LDR prohibit the landfilling of Subtitle C hazardous wastes unless the wastes are treated according to specified standards, or disposed of in a land disposal facility that meets RCRA's "no migration" standard. See generally Chemical Waste Management, Inc. v. EPA, 976 F.2d 2, 8-9 (D.C. Cir. 1992), cert. denied, ___ U.S. ___, 113 S.Ct. 1961, 123 L.Ed.2d 664 (1993). Because the residues from the incineration of hazardous wastes remain hazardous waste by virtue of the derived-from rule, they ordinarily must comply with the LDR. However, because the BIF Rule provides that Bevill wastes that are not "significantly affected" are exempt from Subtitle C, these wastes are not subject to the LDR by virtue of a 1986 EPA regulation which provides that wastes that are not Subtitle C hazardous wastes need not comply with RCRA's land disposal restrictions. See 40 C.F.R. § 268.1(b) (1992) (non-hazardous wastes not subject to LDR); id. § 261.4(b)(4), (7), (8) (Bevill wastes not hazardous waste unless "significantly affected").

CASE petitioners argue that the LDR attach irrevocably to hazardous wastes at the moment they are generated. For example, waste solvents that are hazardous wastes ordinarily must be disposed of in keeping with the LDR. CASE petitioners contend that these same solvents should not be able to evade the LDR by being burned in a cement kiln, thereby potentially becoming Bevill wastes exempt from Subtitle C rather than "treated" hazardous waste. In addition, CASE petitioners argue that they were not given adequate notice or opportunity to comment on the EPA's decision to exclude some co-processed Bevill wastes from the LDR.⁹ The EPA counters that the LDR issue was never at issue in the BIF rulemaking because the agency did not reopen the matter, but reopened only the question of whether the agency should exempt Bevill wastes generated by co-processing hazardous waste from Subtitle C. According to the EPA, Bevill wastes that are not "significantly affected" are simply exempt from the LDR by operation of regulations that were in place before the BIF Rule was even proposed. The agency notes further that it has initiated a separate rulemaking to deal with the question of the LDR's applicability to co-processed Bevill Wastes. See 56 Fed.Reg. 55,160, 55,166 (1991). Because we uphold the EPA's decision to exempt some Bevill wastes from Subtitle C, we also hold that those residues that are excluded are not subject to the LDR as those provisions by their plain language do not apply to materials that are not hazardous waste. CASE petitioners' argument that they were not provided adequate notice of, and opportunity to comment on, the LDR is thus moot, because the BIF Rule did not re-open or modify the LDR, but simply established the circumstances under which Bevill wastes would be subject to Subtitle C. The fact that the LDR do not apply to non-Subtitle C wastes is the result of the EPA's 1986 regulation, not a consequence of the BIF Rule that the EPA was required to address in the BIF rulemaking.

9 Chemical Waste Management also considered a challenge by the Natural Resources Defense Council ("NRDC"), one of the CASE petitioners, to a regulation providing that Bevill wastes were exempt from the LDR. The court vacated the LDR exemption on notice and comment grounds and remanded it for consideration in an ongoing rulemaking. 976 F.2d at 33 n. 20. In response to the vacation of the LDR exemption in Chemical Waste Management, the EPA asserted that its 1986 regulations, specifically 40 C.F.R. §§ 261.11(a) and 268.1(b), provided that wastes excluded from the definition of "solid waste" or "hazardous waste" were exempt from the LDR. Because the wastes exempted by these regulations included Bevill wastes, the EPA maintained that the vacation of the LDR exemption restored the status quo ante, which happened to dictate an identical result. Thus, despite NRDC's "victory" in Chemical Waste Management, Bevill wastes which are not Subtitle C hazardous wastes remained exempt from the LDR.

CASE petitioners base their argument that the EPA may not exempt hazardous wastes burned in Bevill devices from the LDR on this court's decision in American Petroleum Institute v. EPA, 906 F.2d 729, 740-42 (D.C. Cir. 1990) (" API"), which held that the EPA erred in assuming that RCRA required it to cease regulating metal slag as a hazardous waste when it arrived at a reclamation facility. In that case, the EPA had assumed that the slag became a raw material rather than a "solid waste" when it was used for materials recovery. API strongly suggested, but did not decide, that the EPA could not reconcile its decision to cease regulating slag at the point of metals reclamation with RCRA's "cradle-to-grave" regulatory structure. Id.

CASE petitioners assert that API requires the EPA unstintingly to regulate hazardous wastes under Subtitle C. We need not resolve this sweeping question in order to decide this case, however. Whatever RCRA might require in the case of other solid wastes, Congress has mandated that high volume/low hazard Bevill wastes have a special status pending completion of the Bevill studies. It is permissible for the EPA to find that one aspect of that special status is exemption from the LDR. The EPA's position on this issue is a simple syllogism: The LDR apply only to Subtitle C hazardous wastes; Bevill wastes that are not "significantly affected" are not hazardous wastes; therefore, Bevill wastes that are not hazardous waste are not subject to the LDR.¹⁰ Because we uphold the EPA's determination that Bevill wastes produced by co-processing hazardous waste are in some circumstances exempt from Subtitle C, we also uphold the agency's decision to exempt non-hazardous Bevill wastes from the LDR as a permissible construction of the Bevill Amendment and of Subtitle C.

10 We hold that this syllogism is a permissible interpretation of the Bevill Amendment, not that the EPA is required to so read the statute. In its notice of proposed rulemaking requesting comments on the possibility of subjecting Bevill wastes to the LDR, the EPA suggested that our holding in American Iron Steel Institute, 886 F.2d at 395-96, might permit the agency to apply the LDR even to Bevill wastes not subject to Subtitle C. 56 Fed.Reg. 55,160, 55,166 (1991). The permissibility of this alternative interpretation is not before us in this case, and we express no opinion as to that question.

C. The BIF Rule's Regulation of Air Emissions

Several of the Industry petitioners¹¹ contend that the BIF Rule exceeds the EPA's statutory authority under RCRA because it regulates materials that are not "hazardous waste." The BIF Rule establishes a complex set of air emissions controls for toxic metals, chlorine, and hydrogen chloride emitted from BIFs burning hazardous waste fuel. The rule creates a three-tiered system of regulation in which the levels of emissions and feed rates allowed are proportional to the degree of sampling and analysis a BIF operator performs on the BIF's emissions. See generally 56 Fed.Reg. 7171-80 (1991). An owner or operator has the option of satisfying the rule by demonstrating compliance with any one of the three tiers. The first tier of controls assumes that all of the regulated materials fed into the BIF are emitted into the air, without regard for any pollution control devices that may be present, and so provides stringent limits on the rate at which the metals, chlorine, and hydrogen chloride may be fed into the unit. The operator demonstrates compliance by analyzing the toxic content of the materials burned, rather than by measuring actual air emissions. Under the second tier, the BIF owner or operator demonstrates compliance through emissions testing. The third tier permits an owner or operator to conduct both emissions testing and site-specific dispersion modeling (which takes into account actual dispersion conditions at the facility) to show that the emissions do not exceed acceptable levels. The EPA expects most facilities to comply with the BIF Rule via the third tier. 56 Fed.Reg. 7173 (1991).

11 Chemical Manufacturers Association, Marine Shale Processors, Inc., Tennessee Eastman Company, and the American Petroleum Institute.

Industry petitioners challenge the three-tiered approach because it obligates operators to measure either the relevant constituents of, or the air emissions from, all fuels fed into the BIF and not just those which are hazardous waste. The EPA responds that RCRA enables the agency to regulate facilities that burn hazardous waste for fuel, not merely the hazardous waste components of the fuel itself. Because both the EPA and Industry petitioners base their arguments on the first prong of Chevron, arguing that Congress has directly spoken to the precise question at issue, we begin by examining the relevant statutory language. Section 3004(q) of RCRA authorizes the EPA to "promulgate regulations establishing . . . standards applicable to the owners and operators of facilities which burn, for purposes of energy recovery, any fuel [which is produced from hazardous waste or from hazardous waste and any other material] . . . or any fuel which otherwise contains any hazardous waste. . . ." 42 U.S.C. § 6924(q)(1)(B) (emphasis added). In addition, § 3004(a) of RCRA directs the EPA to "promulgate regulations establishing such performance standards, applicable to owners and operators of facilities for the treatment, storage, or disposal of hazardous waste . . . as may be necessary to protect human health and the environment." 42 U.S.C. § 6924(a) (emphasis added).

We agree with the EPA that the plain language of RCRA defeats Industry petitioners' contentions. First, § 3004(q) applies to facilities, not just fuels, and therefore the EPA may regulate BIF owners or operators who elect to burn "any fuel which . . . contains any hazardous waste." 42 U.S.C. § 6924(q)(1)(B). No BIF is required to burn hazardous waste fuel, but if it chooses to do so, § 3004(q) permits the EPA to monitor total feed rates and air emissions; if a BIF elects not to burn hazardous waste, then the BIF Rule would not apply. Second, § 3004(a) allows the EPA to regulate owners and operators of facilities that "treat" hazardous waste. As discussed above, supra at 13 n. 2, for RCRA purposes, burning hazardous waste is considered "treatment" of hazardous waste. Therefore, §§ 3004(a) and (q) allow the EPA to regulate facilities that burn non-waste fuels if those fuels are being burned with hazardous waste.

Although they make a Chevron step I argument, Industry petitioners do not rely on RCRA's statutory language. Instead, they focus on this court's decision in American Mining Congress v. EPA, 824 F.2d 1177 (D.C. Cir. 1987) [hereinafter AMC I], interpreting RCRA's definitional section, 42 U.S.C. § 6903. Section 6903 defines "hazardous waste" as a subset of "solid waste." 42 U.S.C. § 6903(27). AMC I held that the EPA does not have jurisdiction under RCRA to regulate materials reused for materials recovery within an ongoing production process because such materials were never discarded and thus are not "solid waste." Industry petitioners reason that the EPA lacks jurisdiction under RCRA to monitor the constituents of non-waste fuels burned with hazardous waste because these co-processed fuels too are not "hazardous waste." However, the holding of AMC I is not dispositive here. AMC I involved an altogether different facet of waste disposal governed by a different statutory section, i.e., the scope of the RCRA term "solid waste," and not the EPA's right under §§ 3004(a) and (q) to regulate facilities handling non-waste materials when the materials are co-processed with fuel that is unquestionably "hazardous waste" subject to RCRA. As discussed above, the plain language of §§ 3004(a) and (q) clearly authorizes the EPA to regulate facilities that burn a mixture of hazardous waste and non-waste.

Even assuming that the language of RCRA does not unambiguously permit regulation of non-waste fuels when they are co-processed in a facility burning hazardous waste, the EPA's decision to regulate such non-waste is a permissible construction of the statute under Chevron step II. While the EPA is not free to regulate non-hazardous raw materials at will, in this case the agency has demonstrated an adequate nexus between the three-tiered controls and the management of hazardous waste. See generally In the Matter of BP Chemicals America, Inc., 1991 RCRA LEXIS 23, [*]9 (Aug. 20, 1991). The EPA persuasively argues that the BIF Rule's regulation of the content of non-waste fuels burned in a BIF along with hazardous waste is vindicated by the inherent difficulty in differentiating between the molecules of emissions generated by the hazardous waste components of BIF fuel and those generated by the non-waste components. See 56 Fed.Reg. 7144. The health and environmental effects of BIF emissions from the mix are the same whatever the source of an individual molecule. The EPA also notes that because compliance with the BIF Rule is often demonstrated by monitoring emissions during a trial burn, feed rates of all inputs to the unit must be continuously controlled to assure that the results of the trial burn are representative of actual operations. Were the EPA to allow BIF operators to demonstrate compliance on the basis of one trial burn alone, the operators might burn hazardous waste along with less toxic raw materials during the trial burn and then increase the toxicity of the raw materials in later burns. See 56 Fed.Reg. 7144 (1991). Finally, the EPA claims that limitations must be placed on non-waste feeds because co-processing of hazardous waste and raw materials can make the commingled emissions more toxic than the sum total of the individual emissions from the hazardous and non-hazardous materials alone. See 56 Fed.Reg. 7157 (1991); see also 56 Fed.Reg. 7176 (1991) (BIF Rule limits metal feed rate for both hazardous waste and non-waste fuels because when added to emissions from hazardous waste, noncarcinogenic materials can cause a maximum exposed individual ("MEI") concentration that exceeds threshold level for protecting human health). Thus the EPA's requirement that BIF owners and operators monitor the constituents of all their fuel inputs when they burn both hazardous waste and non-waste is a reasonable application of its RCRA authority under §§ 3004(a) and (q).

Industry petitioners argue, however, that the EPA failed to address precisely the alternatives they posed to non-waste monitoring. A rulemaking will be overturned as arbitrary and capricious if "the EPA has failed to respond to specific challenges that are sufficiently central to its decision." International Fabricare Inst. v. EPA, 972 F.2d 384, 389 (D.C. Cir. 1992) (citing American Mining Congress v. EPA, 907 F.2d 1179, 1191 (D.C. Cir. 1990)); accord Motor Vehicles Mfrs. Ass'n v. State Farm Mut. Life Ins. Co., 463 U.S. 29, 48, 103 S.Ct. 2856, 2869, 77 L.Ed.2d 443 (1983) (alternative ways of achieving objectives of statute should be addressed and reasons given for abandonment). During the BIF rulemaking, Industry petitioners suggested the following three ways to limit regulation of a BIF's air emissions to those attributable to the burning of hazardous waste: (1) perform a test burn using non-waste fuel to determine the removal efficiencies of the BIF unit's air pollution control devices (APCDs), and then apply those removal efficiencies to the emissions from the burning of hazardous waste; (2) allow a BIF owner or operator to rely on the conservative removal efficiencies already specified by the EPA for various types of APCDs and then calculate backwards to determine the allowable metals content for the hazardous waste feed; and (3) conduct baseline tests to characterize the relationship between waste feed composition ( i.e., proportion of hazardous waste in the fuel) and removal efficiency. See Comments of the Chemical Manufacturers Ass'n on EPA's Supplement to the Proposed BIF Rule, reprinted in Joint Appendix ("J.A.") at 667-68. These proposals, however, occupied only three sentences of a voluminous notice and comment record and were not supported by any evidence regarding their effectiveness at protecting human health and the environment.

Although the EPA did not explicitly respond to the specific suggestions of Industry petitioners, the agency did, as a general matter, explain the reasons underlying its decision to regulate non-waste fuels when they are co-processed with hazardous waste. As discussed above, the BIF Rule set forth three justifications for regulating non-waste fuels when they are burned with hazardous waste fuels: (1) the inadequacy for compliance purposes of test burns that do not impose limitations on the toxicity content of non-waste fuels, (2) the difficulty of differentiating between toxic air emissions from the burning of non-waste and hazardous waste, and (3) the phenomenon that co-processing non-waste and hazardous waste in a BIF can result in more toxic emissions. See 56 Fed.Reg. 7144 (1991). Implicit in the EPA's discussion of the problems embedded in regulating hazardous waste alone is a rejection of Industry petitioners' alternative proposals. In light of the sketchy nature of these alternative proposals, we find that this portion of the BIF rulemaking (establishing the three-tiered system of air emissions controls) survives arbitrary and capricious review.

The BIF Rule's regulation of non-waste fuels is well within RCRA's grant of authority to the EPA to regulate facilities burning hazardous waste. Due to the difficulties associated with regulating emissions from hazardous waste fuels only, the EPA has reasonably decided to monitor the constituents of, and the air emissions from, all fuels burned in a BIF facility that burns hazardous waste. Although it would have been instructive for the EPA to have addressed more precisely the three proposals suggested by Industry petitioners, we conclude that the agency's general discussion of its decision to regulate non-waste adequately set forth its reasons for rejecting the proposals.

D. Regulation of Products of Incomplete Combustion 1. Background

Under RCRA, the EPA must develop standards to control emissions from burning hazardous waste fuels "as may be necessary to protect human health and the environment." RCRA § 3004(q)(1); 42 U.S.C. § 6924(q)(1). In response to this mandate, the EPA proposed standards for emissions of toxic organic compounds in the BIF rulemaking. The BIF Rule, which was issued on February 21, 1991, imposes three requirements on BIFs burning toxic organics: (1) a 99.99% destruction removal efficiency ("DRE"); (2) good combustion conditions in kilns and other regulated BIFs so as to control products of incomplete combustion ("PICs"); and (3) in certain cases, health risk assessments. See 56 Fed.Reg. 7,134, 7,146-71 (1991) ("Final Rule").

Here, Industry petitioners Lone Star Industries, Inc., Holnam, Inc., Safety-Kleen Corp., and the Cement Kiln Recycling Coalition ("cement kiln petitioners") challenge the second of these requirements as applied to wet process kilns. Cylindrical in shape, a wet kiln rotates on its own axis and is slightly inclined so that raw materials introduced at the top end travel slowly to the lower end. The raw materials (such as clay, shale, limestone, and marl) are mixed in a slurry form and calcined, i.e., heated to a high temperature, as the slurry slides down the kiln towards the furnace at the lower, or "hot" end. Significantly, these raw materials contain widely varying amounts of organic material that are driven off as the slurry mixture calcines.

The furnace at the hot end of the wet kiln requires large amounts of fuel, which can be powdered coal or other fossil fuel or, alternately, high-BTU hazardous waste. One benefit of using hazardous waste in this manner is immediately obvious: It takes the place of scarce fossil fuels. Another benefit is that the high furnace temperatures (upwards of 2,500 degrees Fahrenheit) destroy the hazardous organic compounds contained in the waste. The drawback is that the burning of hazardous waste results in the creation of incompletely burned organic compounds (i.e., PICs) that are emitted from the upper end of the kiln.

A kiln's utility as a means of destroying hazardous wastes turns on its ability to fully destroy them. In practice, destruction of hazardous wastes in the fuel is a function of the combustion efficiency of the kiln: Under poor conditions of efficiency, the principal organic hazardous constituents ("POHCs") of the toxic organic compounds contained in the hazardous waste fuel will be only partially broken down, thereby increasing the production of PICs. The health effects of PIC emissions are unclear — some are known carcinogens, others may be, and others are undetectable. See Final Rule, 56 Fed.Reg. 7,150, col. 1. This uncertainty led the EPA to regulate PICs in the BIF Rule, for "[g]iven the limited information about the hazards that PIC emissions may pose, EPA believes it is prudent to require that boilers and industrial furnaces operate at a high combustion efficiency to minimize PIC emissions." Id.

The EPA chose to monitor carbon monoxide ("CO") and total hydrocarbon ("THC") levels as indicators of combustion efficiency and required that one of three emissions standards be met: (1) a stack or by-pass duct CO level not to exceed 100 ppmv, 40 C.F.R. § 266.104(b) ("Tier I standard"); (2) a stack or by-pass duct CO level above 100 ppmv so long as the stack or by-pass duct levels of THC are below 20 ppmv, § 266.104(c) ("Tier II standard"); or (3) for those industrial furnaces not having a by-pass duct (i.e., wet kilns), meeting an alternative THC limit that ensures that flue gas hydrocarbon ("HC") and CO concentrations when burning hazardous waste fuels are not greater than when not burning hazardous waste, § 266.104(f) ("Tier III standard").

The EPA's rationale for monitoring CO as an indicator of combustion efficiency is based on the following understanding of the mechanics of the combustion process:

In the first stage of the combustion of hazardous waste fuel, the POHCs thermally decompose in the flame to form other, usually smaller, compounds termed products if [sic] incomplete combustion. In this first stage of combustion, these PICs also decompose to form CO.

The second stage of combustion involves the oxidation of CO to CO[2] (carbon dioxide). The CO to CO[2] step is the slowest (rate-controlling) step in the combustion process because CO is considered to be more thermally stable (difficult to oxidize) than other intermediate products of the combustion of hazardous waste constituents. . . .

Thus, in the waste combustion process, the "destruction" of POHCs is independent of flue gas CO levels. CO flue gas levels cannot be correlated with DREs for POHCs, and may also not correlate well with PIC destruction. Although some emissions data indicate a weak correlation between CO and PICs, the data generally indicate that there is a relationship between the two parameters: When CO is low, PIC emissions are relatively low. The converse may not hold: when CO is high, PICs may or may not be high.

Low CO is an indicator of the status of the CO to CO[2] conversion process, the last rate-limiting oxidation process. Because oxidation of CO to CO[2] occurs after the destruction of a POHC and its (other) intermediates (PICs) the absence of CO is a useful indication of POHC and PIC destruction.

Final Rule, 56 Fed.Reg. 7,150, col. 2 (emphasis in original).

The choice of CO as an indicator of combustion efficiency creates a technical problem, that of restricting CO measurement to the CO emitted by the hazardous waste fuel. This is well-nigh impossible in wet kilns, for as the slurry slides down the kiln, the organic materials in the slurry combust and CO is created ("slurry off-gases"). These CO emissions are unrelated to the CO produced by the combustion of hazardous waste at the lower end of the kiln ("combustion off-gases"). Moreover, there is no way of isolating the two chemical processes because, unlike dry kilns, which have a by-pass duct that permits the diversion of the kiln off-gas into a separate flue before it is united with the emissions generated by the raw materials, in a wet kiln the combustion off-gases combine with emissions from the slurry, and both exit from the upper end of the kiln. Thus background emissions of CO attributable to the oxidation of the raw materials will enter into the measurement of CO emissions and may prevent wet kilns from meeting the Tier I standards regardless of their combustion efficiency.

Recognizing this problem, on October 26, 1989, the EPA proposed to add a second alternative emissions standard. See 54 Fed.Reg. 43,718, 43,722-27 (1989) (EPA's first supplement to the proposed BIF Rule ("First Supplement")). The CO standard was to be retained (the Tier I standard), but an alternative standard was to be added (the Tier II standard). The Tier II standard was to be either technology-based (limits on THC emissions) or health-based (the facility would have to demonstrate that PIC emissions posed an acceptable risk to the maximally exposed individual). See 54 Fed.Reg. 43,722, col. 2. Under the technology-based standard, THC emissions were not to exceed 20 ppmv. 54 Fed.Reg. 43,723, col. 3; 40 C.F.R. § 266.104(c). This standard was chosen on the basis of the data for hazardous waste incinerators and BIFs operating under good combustion conditions, as well as the EPA's use of risk assessment methodology to show a 20 ppmv level was protective of human health. 54 Fed.Reg. 43,723, col. 3.

Unfortunately, wet process kilns may be incapable of meeting the technology-based Tier II standards for the same reason that prevents them from meeting the Tier I standard. Indeed, oxidation of the organic materials contained in the slurry mixture produces not only CO, but also HC. These background emissions are independent of the emissions from the combustion of hazardous wastes at the lower end of the kiln. First Supplement, 54 Fed.Reg. 43,724, col. 1. Thus neither CO nor THC emissions are an accurate indicator of the combustion efficiency of wet kilns due to the impossibility of isolating the two combustion processes occurring within the kiln. Reconsidering this problem, the EPA requested comments on the possibility of different THC limits as well as the feasibility of a third standard — a site-specific limit "where THC levels when burning hazardous waste would be limited to baseline THC levels without burning hazardous waste." 54 Fed.Reg. 43,724, col. 1. The EPA continued discussions with the affected industry well after the end of the comment period. Final Rule, 56 Fed.Reg. 7,155 n. 28.

In the final BIF Rule, the EPA adopted the Tier I and the technology-based Tier II standards and also promulgated a Tier III standard. This last was aimed at the "6 to 10 [wet process kilns that] may not be able to comply with the HC limit of 20 ppmv even though they generate minimal HC from sources other than raw materials." 56 Fed.Reg. 7,155, col. 3. The Tier III standard is essentially a waiver provision that allows the EPA to formulate alternative CO and THC limits on a case-by-case basis if operators of wet kilns can meet certain conditions. 40 C.F.R. § 266.104(f)(1). To obtain a Tier III waiver, the wet kiln operator must, inter alia, (1) demonstrate that flue gas levels of CO and THC do not increase over baselines measured when the kiln is fired by non-hazardous fuels in a manner designed to minimize hydrocarbon emissions, (2) identify POHCs in the kiln's emissions, and (3) conduct a site-specific risk assessment to demonstrate that the maximum annual average ground level concentrations do not exceed listed levels. 40 C.F.R. § 266.104(f)(3). It is this Tier III standard that bears the brunt of the cement kiln petitioners' challenges to the PIC standards. They argue both that the EPA gave insufficient notice and opportunity for comment on the Tier III standard and that each of the three tiers is arbitrary and capricious as applied to wet process kilns.

2. Did the EPA Provide Adequate Notice and Opportunity for Comment on the Tier III Hydrocarbon Standard?

The cement kiln petitioners contend that the EPA failed to accord adequate notice and opportunity for comment on the Tier III standard. The validity of this challenge turns on the relationship between the proposed regulations and the Tier III standard contained in the final BIF Rule. See Shell Oil Co. v. EPA, 950 F.2d 741, 747 (D.C. Cir. 1991) ("The relationship between the proposed regulation and the final rule determines the adequacy of notice."). As we stated in Shell Oil, adequate notice is given when "the final rule is a `logical outgrowth' of the one proposed." Id. at 747. Nonetheless, the "EPA undoubtedly has authority to promulgate a final rule that differs in some particulars from its proposed rule." Small Ref. Lead Phase-Down Task Force v. EPA, 705 F.2d 506, 546 (D.C. Cir. 1983).

The EPA certainly raised the possibility of a specific standard for wet kilns. In the First Supplement, the agency explicitly recognized that "[p]reheater and precalciner cement kilns, for example, may not be able to readily achieve such a low THC concentration for the same reason that they typically cannot achieve CO levels below 100 ppmv." 54 Fed.Reg. 43,724, col. 1. It thus called for comments on

(1) The types of industrial furnaces for which a THC level of 20 ppmv is representative of good combustion conditions; (2) whether alternative THC limits may be more appropriate for certain industrial furnaces; and (3) whether an approach to identify a site-specific THC limit representative of good operating practices may be feasible (e.g., where THC limits when burning hazardous waste would be limited to baseline THC levels without burning hazardous waste.). In support of comments, we request data on emissions of CO and THC under baseline and hazardous waste burning conditions, including characterization of the type and concentration of individual organic compounds omitted.

Id. This placed the cement kiln petitioners on notice that a non-hazardous waste fuel THC baseline standard might be adopted for wet kilns. Moreover, the EPA held repeated meetings with these petitioners as well as individual cement kiln operators to discuss an alternative hydrocarbon limit for cement kilns burning hazardous waste fuel. 56 Fed.Reg. 7,155 n. 28.

While the cement kiln petitioners may have had notice of the possibility of a THC standard for wet kilns, we are hard pressed to find any precursor to the dual-baseline/health-based emissions testing standard actually embodied in the Tier III provision in any of the three proposed regulations preceding the final rule. See Proposed Rule, 52 Fed.Reg. 16,982, 16,997-17,000 (1987); First Supplement, 54 Fed.Reg. 43,718, 43,721-28; second supplement to the proposed rule, 55 Fed.Reg. 17,862, 17,880-89 (1990) ("Second Supplement"). The EPA contends that the rulemaking proposals disclosed the central elements of the final Tier III standard and attempts to predicate notice of the Tier III standard on mentions of its component parts.

To this end, the EPA cites its determination in the First Supplement that THC could serve as a good indicator of combustion efficiency, 54 Fed.Reg. 43,722-24, which it reiterated in the Second Supplement, 55 Fed.Reg. 17,884-86. The EPA also points to its call for comments on the possibility of a site-specific limit of THC levels to those experienced when burning non-hazardous waste fuel, 54 Fed.Reg. 43,724, col. 1, reiterated in the Second Supplement, 55 Fed.Reg. 17,885, col. 3. It also cites the determination that both CO and THC could be measured on the basis of hourly rolling averages, 54 Fed.Reg. 43,726-27, and the suggestion that a THC standard might be based on generic risk assessment. 54 Fed.Reg. 43,723, col. 1.

Nowhere in the proposed rulemakings, however, does the agency indicate it is contemplating the possibility of dual CO and THC baselines. This omission is critical because notice of individual parts of a proposed rule is not necessarily notice of the whole. One purpose of notice is to promote informed decisionmaking, and comments addressed to one specific component part of the standard do not necessarily bear on the viability of Tier III as a whole. See Fertilizer Inst. v. EPA, 935 F.2d 1303, 1312 (D.C. Cir. 1991) ("Because comments appropriate to a determination of the hazards of radionuclides would not necessarily be appropriate to a determination of the hazards of entities using or producing radionuclides, we conclude that the notice published by the EPA did not provide interested parties with an adequate opportunity to comment."). Here, more notice should have been given of the form of the Tier III standard, for while the "logical outgrowth" standard does not require the agency to assiduously lay out every detail of a proposed rule for comment, it does require that the "agency . . . publish notice of either the substance of a proposed rule or a `description of the subjects and issues' covered by a proposed rule." Id. at 1310-11 (quoting 5 U.S.C. § 553(b)(3)). Such a description must "provide sufficient detail and rationale for the rule to permit interested parties to participate meaningfully." Id. at 1311 (internal quotation marks and citations omitted).

It is the possibility of meaningful participation that is lacking here, for while the EPA proposed individual elements of the Tier III standard separately, the component parts were never collected together in such a fashion as to enable the parties to anticipate and adequately comment on the ultimate Tier III standard. The agency did request data on "emissions of CO and THC under baseline and hazardous waste burning conditions," First Supplement, 54 Fed.Reg. 43,724, col. 1, but this gives no indication that a dual baseline was contemplated and could just as easily have been a call for data to support the Tier I and Tier II standards. Notice of these two standards was exhaustive, while the cement kiln petitioners were only on notice that a site-specific THC standard was contemplated for wet process kilns.

As we have stated before, general notice that a new standard will be adopted affords the parties scant opportunity for comment. Small Ref. Lead Phase-Down Task Force, 705 F.2d at 549. The agency's obligation is more demanding — it must "describe the range of alternatives being considered with reasonable specificity. Otherwise, interested parties will not know what to comment on, and notice will not lead to better-informed agency decision-making." Id. Here, the "range of alternatives" described by the EPA did not include the ultimate standard; nor can the ultimate standard be considered a "logical outgrowth" of the individual component parts cited by the EPA. Ultimately, the EPA simply failed to give interested parties sufficient notice of the form that the Tier III standard might take, undermining the aims of meaningful participation and informed decisionmaking. Id. at 547 ("[N]otice improves the quality of agency rulemaking by ensuring that agency regulations will be tested by exposure to diverse public comment." (internal quotation marks omitted)).

The EPA makes much of the comments submitted on issues that were to become critical parts of the final rule, as well as the meetings it held with industry. While we have noted that insightful comments may be reflective of notice and may be adduced as evidence of its adequacy, see, e.g., Shell Oil, 950 F.2d at 751, we have rejected bootstrap arguments predicating notice on public comments alone. Ultimately, notice is the agency's duty because "comments by members of the public would not in themselves constitute adequate notice. Under the standards of the APA, notice necessarily must come — if at all — from the Agency." Id. (internal quotation marks and citations omitted). Here, moreover, the comments submitted by the parties undercut the EPA's argument for they only go to the individual component parts of the final Tier III standard. Not one contemplates anything more than a standard measuring the amount of THC emissions over a non-hazardous fuel baseline combined with backup risk assessment. No mention is made of a dual CO/THC baseline in any of the comments, and at oral argument counsel for the EPA conceded that this indicated that no notice was provided of a possible CO baseline standard for wet kilns. In sum, the EPA cannot base notice of the Tier III standard on either submitted comments or proposals concerning individual parts of the ultimate standard. The Tier III standard not being a logical outgrowth of any precursor proposal, it is vacated and remanded for further consideration.

3. Are the EPA's Rules to Control PIC Emissions Reasonable?

As discussed above, the EPA enacted a three-tiered protective standard to control emissions from the burning of hazardous fuels. The cement kiln petitioners attack the rationale underlying these standards, contending that each is arbitrary and capricious. Under our familiar rubric, "[t]he scope of judicial review of agency decisionmaking under the arbitrary-and-capricious standard is narrow. Nonetheless, the agency must examine the relevant data and articulate a satisfactory explanation for its action including a rational connection between the facts found and the choice made." Natural Resources Defense Council v. EPA, 859 F.2d 156, 209 (D.C. Cir. 1988) (footnote and internal quotation marks omitted).

a. The Tier III Standard

These petitioners contend that the Tier III standard has no rational basis. First, they argue that the EPA articulated no basis for its conclusion that kilns could reliably quantify both CO and THC baselines when burning non-hazardous fuel. In the BIF Rule, the EPA adduced two pieces of support: "commenters['] [assertions] that when hazardous waste is burned, hydrocarbon levels do not increase and often decrease," 56 Fed.Reg. 7,157, col. 1, and the results of test runs on a single wet process cement kiln located at Hannibal, Missouri. 56 Fed.Reg. 7,157 col. 1, 7,163 col. 2. The results of test runs allegedly demonstrated that emissions of THC are quantifiable and actually decrease when hazardous fuels are burned (in comparison to a non-hazardous waste burning baseline). Id.; Emissions Testing of a Wet Cement Kiln at Hannibal, Missouri, BBSP-S0147, reprinted in J.A. at 218, Table 4-13 at 4-22. Yet Table 4-13 reveals the very problem raised by the cement kiln petitioners: that the inhomogeneity of the organic material in the raw materials precludes the establishment of a reliable baseline, and thus the baselines are not a valid indicator of good combustion. At oral argument, moreover, counsel for the EPA disavowed reliance on the Hannibal test runs, admitting that they were conducted for a different purpose and could not be relied on to support the Tier III standard.

Thus the supportive comments cited by the EPA are the sole evidence of the feasibility of the Tier III standard. These comments, however, refer only to the possibility of implementing a THC baseline and make no reference to the possibility of simultaneously quantifying CO and THC baselines. The agency thus had no information on this issue and was relying on pure speculation when it decided that a standard of no increase of CO and THC over quantifiable CO and THC baselines was achievable. See National Gypsum Co. v. EPA, 968 F.2d 40, 43-44 (D.C. Cir. 1992) (agency cannot "infer" facts not in the record); Natural Resources Defense Council, 859 F.2d at 210 (agency actions based upon speculation are arbitrary and capricious). Such speculation is an inadequate replacement for the agency's duty to undertake an examination of the relevant data and reasoned analysis; thus the EPA's action in promulgating the Tier III standard was arbitrary and capricious. See Specialty Equip. Mkt. Ass'n v. Ruckelshaus, 720 F.2d 124, 137 (D.C. Cir. 1983) (agency's lack of justification demonstrates action was arbitrary and capricious).

We note that, at oral argument, counsel for the cement kiln petitioners maintained that they had proposed three alternative methods by which the objectives of the Tier III standards could be met and that the EPA had failed to consider those comments. The EPA, of course, is obliged to respond to "relevant" and "significant" comments. Home Box Office, Inc. v. FCC, 567 F.2d 9, 35 n. 58 (D.C. Cir. 1977). We need not consider this argument as grounds for remand, however, because the cement kiln petitioners failed to raise it prior to oral argument; thus we deem it waived.

b. Tier I and Tier II Standards

All BIFs are governed by the three-tiered standard; thus, even though we have vacated the Tier III standard, wet process kilns remain subject to the first two. The cement kiln petitioners do not contend that they were provided with inadequate notice or opportunity for comment on the Tier I and Tier II standards, but they do argue that these standards are "fundamentally irrational" as applied to wet kilns and must be vacated as arbitrary and capricious.

The EPA argues that the three tiers implement its statutory duty, and that if wet kilns cannot comply with one of these standards, they should simply stop burning hazardous wastes as fuel. The EPA attests, and the cement kiln petitioners do not dispute, that there are dangers, known and unknown, associated with burning hazardous wastes. Ultimately, it is PIC emissions that are being regulated, and from the standpoint of health and safety, there is no reason to distinguish between PICs emitted from the combustion of hazardous wastes and those emitted from the combustion of the slurry. The EPA has determined that

hazardous waste burning may affect the type and concentration of organic compounds emitted from an industrial furnace that has elevated HC concentrations attributable to raw materials. For example, the chlorine in the hazardous waste may result in higher concentrations of chlorinated organic compounds.

56 Fed.Reg. 7,157, col. 3. The EPA candidly admits that the use of CO and THC as indicators of combustion efficiency is problematic as applied to wet kilns, yet it contends that the existing PIC standards are the best application of existing scientific knowledge. "This court does not demand certainty when there is none[,] and the Agency here may apply its expertise to draw conclusions from . . . probative preliminary data not yet certifiable as fact and the like." Solite Corp. v. EPA, 952 F.2d 473, 490 (D.C. Cir. 1991) (internal quotation marks and citation omitted). As we have stated before, "[w]e cannot of course substitute our judgment for that of the agency. What we do require is that the [agency] come to grips with the obvious ramifications of its approach and address them in a reasoned fashion." Natural Resources Defense Council, 859 F.2d at 209-10 (internal quotation marks and footnotes omitted).

In the instant case, the EPA has met this standard — it recognized the effects that the Tier I and Tier II standards would have on wet kilns and attempted to accommodate them in the Tier III standard. The lack of notice and a reasoned basis for the Tier III standard necessitates that it be vacated, but the inapplicability of the other two to wet kilns, standing alone, does not persuade us that their promulgation was arbitrary and capricious. The EPA issued those regulations as protective of human health and the environment; and if it is technologically impossible for wet process kilns to meet them, they can simply stop using hazardous waste as fuel. See id. at 206-09 (EPA has no duty to provide for variances from water quality standards that may be technologically impossible to achieve).

We thus vacate and remand the Tier III standard while affirming the Tier I and Tier II standards. Nonetheless, we note that the EPA has recognized that the Tier I and Tier II standards are inappropriate for wet process cement kilns. The agency may therefore wish to consider establishing an interim replacement standard for Tier III under the "good cause" exemption of 5 U.S.C. § 553(b)(3)(B) pending full notice and opportunity for comment. See, e.g., Shell Oil, 950 F.2d at 752.

III. CONCLUSION

For the reasons given above, the BIF Rule is upheld in part, and the Tier III PIC standard is remanded to the EPA for reconsideration consistent with the terms of this opinion.

So Ordered.

* * * * * *

as may be necessary to protect human health and the environment. 42 U.S.C. § 6924(q) (1988).