EDF v EPA (PCB case)

EDF vs. EPA

(The "PCB Case")

For use in ENVL417-Environmental Law
Markups and annotations added
All other users refer to original source at:: 636 F.2D 1267, 205 U.S.APP.D.C. 139


ENVIRONMENTAL DEFENSE FUND, INC., Petitioner,
v.
ENVIRONMENTAL PROTECTION AGENCY, Respondent,
Ad Hoc Committee on Liquid Dielectrics of the Electronic Industries Association
et al., Joy Manufacturing Company, Edison Electric Institute et al., and
Aluminum Company of America, Intervenors.
Nos. 79-1580, 79-1811 and 79-1816.
United States Court of Appeals,
District of Columbia Circuit.
Argued June 6, 1980.
Decided Oct. 30, 1980.

A review regulations of promulgated by the Administrator of the United States Environmental Protection Agency. The Court of Appeals held that:
(1) no substantial evidence supported the administrative determination to classify certain polychlorinated biphenyl uses as "totally enclosed" and therefore exempt;
(2) no substantial evidence supported the administrative decision to exclude from regulation all materials containing concentrations of PCBs below 50 ppm;
(3) substantial evidence supported the administrative determination to allow continued use of eleven nontotally enclosed uses.


EDWARDS, Circuit Judge:

In this case the Environmental Defense Fund (EDF) petitions for review of regulations, issued by the U. S. Environmental Protection Agency (EPA), that implement section 6(e) of the Toxic Substances Control Act (TSCA).(FOOTNOTE 1 ) That section of the Act provides broad rules governing the disposal, marking, manufacture, processing, distribution, and use of a class of chemicals called polychlorinated biphenyls (PCBs).(FOOTNOTE 2)

EDF seeks review of three aspects of the regulations.(FOOTNOTE 3) First, it challenges the determination by EPA that certain commercial uses of PCBs are "totally enclosed," a designation that exempts those uses from regulation under the Act. Second, it claims that the EPA acted contrary to law when it limited the applicability of the regulations to materials containing concentrations of PCBs greater than fifty parts per million (ppm). Third, EDF challenges the decision by EPA to authorize the continued use of eleven non-totally enclosed uses of PCBs.


Footnote 3. EDF challenges the regulations, promulgated under s 6(e)(2) and (3) of the Act, 15 U.S.C. s 2605(e)(2) & (3) (1976), governing the manufacture, processing, distribution, and use of PCBs the so-called Ban Regulations. See 44 Fed.Reg. 31,542-58 (1979) (to be codified in 40 C.F.R. pt. 761). EDF does not challenge the Disposal Regulations, issued pursuant to s 6(e)(1) of the Act. EPA issued the final Disposal Regulations over a year before the Ban Regulations. See 40 C.F.R. pt. 761 (1978), 43 Fed.Reg. 7,156-64 (1978). EPA later reissued the Disposal Regulations with slight modification along with the final Ban Regulations. See 44 Fed.Reg. 31,542-58 (1979).

From our examination of the record, we find that there is no substantial evidence to support the EPA determination to classify certain PCB uses as "totally enclosed." We also find that there is no substantial evidence in the record to support the EPA decision to exclude from regulation all materials containing concentrations of PCBs below fifty ppm. Accordingly, on these first two points, we hold unlawful and set aside the challenged regulations, and remand to EPA for further proceedings consistent with this opinion. We find, however, that there is substantial evidence in the record to support the EPA determination to allow continued use of the eleven non-totally enclosed uses. Accordingly, on this third point, we uphold the EPA regulations.


I. BACKGROUND

A. Polychlorinated Biphenyls Polychlorinated biphenyls (PCBs) have been manufactured and used commercially for fifty years for their chemical stability, fire resistance, and electrical resistance properties. They are frequently used in electrical transformers and capacitors. However, PCBs are extremely toxic to humans and wildlife. The extent of their toxicity is made clear in the EPA Support Document (FOOTNOTE 4) accompanying the final regulations, in which the EPA Office of Toxic Substances identified several adverse effects resulting from human and wildlife exposure to PCBs. Epidemiological data and experiments on laboratory animals indicate that exposure to PCBs pose carcinogenic and other risks to humans. Experimental animals developed tumors after eating diets that included concentrations of PCBs as low as 100 parts per million (ppm). Experiments on monkeys indicate that diets with PCB concentrations of less than ten ppm reduce fertility and cause still births and birth defects. Other data show that PCBs may adversely affect enzyme production, thereby interfering with the treatment of diseases in humans. Support Document, supra note 4, at 9-18. EPA has found that PCBs will adversely affect wildlife as well as humans. Concentrations below one ppb (part per billion) are believed to impair reproductivity of aquatic invertebrates and fish. Some birds suffered "severe reproductive failure" when fed diets containing concentrations of only ten ppm of PCBs. Id. at 19. Because PCBs collect in waterways and bioaccumulate in fish,(FOOTNOTE 5) fish-eating mammals run a special risk of adverse effects. Such mammals may have "significantly higher concentrations of PCBs in their tissues than the aquatic forms they feed on." Id. at 36.


Footnote 5. One typical study reported that fathead minnows bioconcentrated PCBs by a factor of 230,000. Support Document, supra note 4, at 35.

EPA estimates that by 1975 up to 400 million pounds of PCBs had entered the environment. Approximately twenty-five to thirty percent of this amount is considered "free," meaning that it is a direct source of contamination for wildlife and humans. The rest, "mostly in the form of industrial waste and discarded end use products, is believed to be in landfill sites and thus constitutes a potential source of new free PCBs." Id. at 33-34.(FOOTNOTE 6) Other significant sources of PCBs include atmospheric fallout and spills associated with the use or transportation of PCBs. Id. at 29.


Footnote 6. However, the average non-landfill concentration of PCBs is below the level of detection. Id. at 34. One study indicated that only 0.1 percent of the soil samples analyzed had detectable levels of PCBs. Id.

EPA concluded in the Support Document that "the additional release of PCBs" into the environment would result in widespread distribution of the PCBs and "will eventually expose large populations of wildlife and man to PCBs." Id. at 36-37. EPA concluded further that: As a practical matter, it is not possible to determine a "safe" level of exposure to these chemicals. Because PCBs are already widely distributed throughout the biosphere, they currently pose a significant risk to the health of man as well as that of numerous other living things. As a consequence, any further increase in levels of PCBs in the biosphere is deemed undesirable by EPA. Id. at 38. Because "PCBs released anywhere into the environment will eventually enter the biosphere ... EPA has determined that any such release of PCBs must be considered 'significant.' " Id. In 1972, Monsanto, the major American manufacturer of PCBs, limited its sales of PCBs to manufacturers of transformers and capacitors. It ceased all manufacture of PCBs in 1977 and shipped the last of its inventory before the end of that year. Today, PCBs are produced in this country only as incidental byproducts of industrial chemical processes. There are no known natural sources of PCBs. Id. at 2.
B. Congressional Response

Responding to the dangers associated with the use of PCBs and other toxic chemicals, Congress in 1976 enacted the Toxic Substances Control Act (TSCA), Pub.L.No.94-469, 90 Stat. 2003 (1976). Although the Act is generally designed to cover the regulation of all chemical substances, section 6(e) refers solely to the disposal, manufacture, processing, distribution, and use of PCBs. No other section of the Act addresses the regulation of a single class of chemicals. The special attention accorded to PCBs in the Toxic Substances Control Act resulted from the recognized seriousness of the threat that PCBs pose to the environment and human health. During the debate over the Senate version of the Act, Senator Nelson, the author of the amendment adding the PCB subsection to the Senate bill, noted that PCBs were widespread in the environment and that they posed significant potential dangers to human health and to wildlife.(FOOTNOTE 7) In the House of Representatives, Congressman Dingell introduced a similar amendment to the House version of the Toxic Substances Control Act,(FOOTNOTE 8) not only because PCBs posed great dangers to the natural and human environments, but also because "the history of EPA is not one of vigorous and quick action." (FOOTNOTE 9)


Footnote 7. See 122 Cong.Rec. 8291-94 (1976), reprinted in House Comm. on Interstate and Foreign Commerce, Legislative History of the Toxic Substances Control Act, 94th Cong., 2d Sess., at 233-40 (1976) (hereinafter referred to as "Legislative History"). Section 6(e) was added to both the Senate and House bills during floor debate.
As enacted, section 6(e) of the Act sets forth a detailed scheme to dispose of PCBs, to phase out the manufacture, processing, and distribution of PCBs, and to limit the use of PCBs. Specifically, section 6(e) provides that, within six months of the effective date of the Act (January 1, 1977), EPA must prescribe methods to dispose of PCBs and to require that PCB containers be marked with appropriate warnings. 15 U.S.C. s 2605(e)(1). One year after the effective date of the Act, PCBs can be manufactured, processed, distributed, and used only in a "totally enclosed manner." Id. s 2605(e)(2)(A). One year later, all manufacture of PCBs is prohibited. Id. s 2605(e)(3)(A)(i). Six months after that (i. e. two and one-half years after the effective date of the Act), all processing and distribution of PCBs in commerce is prohibited. Id. s 2605(e)(3)(A)(ii). Thus, today, except for the specified authorizations and exemptions described below, the Act permits PCBs to be used only in a totally enclosed manner, and it completely prohibits the manufacture, processing, and distribution of PCBs. . . . (statutory provisions omitted) The statute sets forth only limited exceptions to these broad prohibitions.
C. EPA's Implementation of section 6(e)
EPA sought to implement section 6(e) through two sets of regulations. The first set of regulations the so-called Disposal Regulations set forth specific rules governing the disposal and marking of PCBs. The Disposal Regulations covered not only pure PCB compounds, but also materials contaminated with at least 500 ppm of PCBs. EPA chose this regulatory cutoff in order to regulate "disposal of most PCB's ... as soon as possible." Preamble to Final Disposal Regulations, 43 Fed.Reg. 7,151 (1978). EPA warned that it was considering a new cutoff "possibly in the range of 50 ppm or below" for the Proposed Ban Regulations. Id. In June 1978 EPA issued proposed Ban Regulations that would implement the prohibitions mandated in subsections 6(e)(2) and (3), define "totally enclosed manner," authorize several non-totally enclosed uses, and set forth the procedures for obtaining exemptions from the prohibitions. See Proposed Ban Regulations, 43 Fed.Reg. 24,801 (1978). As foreshadowed in the final Disposal Regulations, the proposed and final Ban Regulations (issued May 31, 1979) set fifty ppm as the cutoff. See 43 Fed.Reg. 24,813 (1978), 44 Fed.Reg. 31,543 (1979) (to be codified in 40 C.F.R. s 761.1(b)). The final regulations defined all electrical capacitors, electromagnets, and non-railroad transformers as totally enclosed,(FOOTNOTE 12) thus automatically exempting them from regulation under the Act. In the final regulations the Administrator authorized eleven non-totally enclosed uses to continue, including the servicing of totally enclosed uses,(FOOTNOTE 13) based on his consideration of the health and environmental effects of PCBs, the exposure to PCBs resulting from these activities, the availability of substitutes for the PCBs, and the economic impact of restricting those uses. See Support Document, supra note 4, at v. Most authorizations continue until July 1, 1984, with some authorizations expiring sooner. Only the authorization for carbonless copy paper continues indefinitely.
Footnote12. The complete list of totally enclosed uses is as follows: distribution and use, except servicing, of intact, non-leaking, non- railroad PCB transformers, PCB-contaminated transformers (defined to include transformers with PCB concentrations in the dielectric (or insulating) fluid of between 50 and 500 ppm), electromagnets, and capacitors; and processing, distribution in commerce, and use of PCB equipment containing intact, non-leaking capacitors. See Final Ban Regulations, 44 Fed.Reg. 31,549 (1979) (to be codified in 40 C.F.R. s 761.30). Footnote 13. The authorizations are: servicing of non-railroad transformers; use and servicing of railroad transformers; use and servicing of mining equipment; use in heat transfer systems; use in hydraulic systems; use in existing stocks of carbonless copy paper; use in pigments; servicing of electromagnets; use in natural gas pipeline compressors; use in small quantities for research and development; and use as a microscopy mounting medium. See id. at 31,549-51 (1979) (to be codified in 40 C.F.R. s 761.31). Detailed conditions attach to each of the authorizations.
II. JURISDICTION

At the outset one of the intervenors, Ad Hoc Committee on Liquid Dielectrics of the Electronic Industries Association (EIA), challenges the jurisdiction of this court to consider EDF's petition. EIA contends that under the authority of Environmental Defense Fund, Inc. v. Environmental Protection Agency, 598 F.2d 62 (D.C.Cir.1978), and subsection 20(a)(2) of TSCA,(FOOTNOTE 14) a federal district court is the proper forum in the first instance to hear EDF's claims. Upon considered review of this jurisdictional question, we find EIA's position to be without merit, and we hold that the matters at issue in this case are properly before this court for review. ... As explained in the 1978 EDF case, a federal district court is the appropriate forum whenever a court must gather and evaluate additional evidence. 598 F.2d at 91. Here no trial is necessary for there is a complete administrative record for this court to review. A review of the jurisdictional provisions of the Toxic Substances Control Act demonstrates that this court, and not the district court, has jurisdiction to consider EDF's petition.
III. USE AUTHORIZATIONS

A. Criteria for the "Unreasonable Risk" Determination ... Using the criteria set forth in subsection 6(c)(1),(FOOTNOTE 18) the Administrator found that eleven non-totally enclosed uses did not present an unreasonable risk. On the basis of these findings, EPA authorized the continued use of the eleven non-totally enclosed uses here in dispute.(FOOTNOTE 19) In attacking these use authorizations, EDF claims that the Administrator employed the wrong criteria in making his determinations concerning "unreasonable risk." (FOOTNOTE 20) ... Because there is no compelling evidence to support the position advanced by EDF, and because of the deference that we must accord an agency's reasonable interpretation of its statute, [citations omitted], we conclude that the statute does not preclude the EPA from using the subsection 6(c)(1) criteria in making the unreasonable risk determinations under 6(e)(2) (B). Moreover, because the expression "unreasonable risk of injury to health or the environment" is left undefined in section 6(e), the Administrator was required to give some meaning to it. Since the 6(c)(1) criteria obviously pertain to factors of "unreasonable risk," it was entirely appropriate for EPA to consider such criteria in ascribing a meaning to the use authorization provision in 6(e)(2)(B). EDF has shown nothing to indicate otherwise. In fact, EDF does not really contest use of the first three criteria in 6(c)(1) i. e. the effects on health and on the environment, and the availability of substitutes. Rather, EDF's primary focus is on the fourth criterion in 6(c) (1), relating to the economic consequences of the authorization. Yet, EDF's objections to the "economic consequences" criterion cannot stand in the face of section 2(c) of the Act, which expressly requires the Administrator to consider such factors. (FOOTNOTE 23) Furthermore, the particular economic factors that EPA took into account were plainly reasonable.(FOOTNOTE 24) The Administrator did not simply propose to consider the effect of the ban on industry, but also the effects on "the national economy, small business, technological innovation, the environment, and public health." See note 18, supra. This formulation, which considers a broad range of benefits and costs of the ban and use authorization, is entirely consistent with the section 2(c) requirement that the Administrator consider the economic and social impact on his actions. Because the 6(c)(1) criteria fulfill an express mandate of the statute and reflect a reasonable interpretation of an ambiguous phrase, we conclude that the Administrator did not err in choosing those criteria to make the unreasonable risk determinations under 6(e)(2)(B).(FOOTNOTE 25)

B. Application of the Criteria

EDF's final attack on the use authorizations is that the Administrator did not properly apply his own criteria in making the unreasonable risk determinations. Here, too, we reject EDF's position.(FOOTNOTE 26) Petitioner's reply brief, p. 33 n.1. Because use in microscopy mounting medium, research and development, and pigments also involves small amounts of PCBs or short authorization periods, see Preamble to Final Ban Regulations, 44 Fed.Reg. 31,535-37 (1979), we assume that EDF does not challenge these authorizations. We note, however, that our review of the record finds substantial evidence supporting the Administrator's decision to authorize those uses. As a result, we restrict our discussion in the text to the servicing of transformers and electromagnets, and the use and servicing of railroad transformers. EPA's use authorizations for servicing non-railroad transformers and electromagnets are predicated in part on its determination that non- railroad transformers and electromagnets are totally enclosed. See id. at 31,530, 31,536. Because of our disposition of EDF's challenge to the Administrator's decision finding that these classes of uses are totally enclosed, see section V, infra, the present discussion applies only to transformers and electromagnets that are actually totally enclosed. The standard of judicial review for rules promulgated under section 6(e) is expressly set forth in subsection 19(c)(1)(B)(i): "the court shall hold unlawful and set aside such rule if the court finds that the rule is not supported by substantial evidence in the rulemaking record." 15 U.S.C. s 2618(c)(1)(B)(i). Evidence includes "any matter in the rulemaking record." Id. The substantial evidence standard (FOOTNOTE 27) mandated by the Act is generally considered to be more rigorous than the arbitrary and capricious standard normally applied to informal rulemaking.(FOOTNOTE 28) Under the substantial evidence standard, a reviewing court must give careful scrutiny to agency findings and, at the same time, accord appropriate deference to administrative decisions that are based on agency experience and expertise. Because administrative decisions often involve judgments based on incomplete or even conflicting scientific data, the agency "may have to fill gaps in knowledge with policy considerations." AFL-CIO v. Marshall, 617 F.2d 636, 651 (D.C.Cir.1979). Consequently, reviewing courts "must examine both factual evidence and the agency's policy considerations set forth in the record." Id. The court's role in reviewing regulations is to ensure public accountability "by requiring the agency to identify relevant factual evidence, to explain the logic and the policies underlying any legislative choice, to state candidly any assumptions on which it relies, and to present its reasons for rejecting significant contrary evidence and argument." United Steelworkers of America v. Marshall, --- F.2d ----, ----, No. 79-1048 (D.C.Cir. Aug. 15, 1980). With these general guidelines in mind, we review EPA's PCB use authorizations. ... In an attempt to reduce the costs of compliance and the risks associated with exposure to PCBs, the Administrator created two categories for transformers PCB-contaminated transformers (containing PCB concentrations between fifty and 500 ppm) and PCB transformers (containing PCB concentrations greater than 500 ppm). Because the Administrator found that proper protective clothing and good management practices should reduce PCB exposure to "very low levels," the regulations permit routine servicing of PCB transformers and electromagnets. Preamble to Final Ban Regulations, 44 Fed.Reg. 31,531 (1979).(FOOTNOTE 29) Additionally, the Administrator heard uncontradicted evidence that a prohibition of routine servicing would significantly increase the chances of catastrophic transformer failure, presenting "far greater risks to health and the environment than that associated with the minimal PCB exposure during routine servicing." Id. However, the Administrator found that the rebuilding of PCB transformers and electromagnets (i. e. non-routine servicing) risks greater exposure to PCBs due to leaks, spillage, or volatilization of the dielectric. The Administrator also found that a prohibition against "any servicing (including rebuilding) of PCB Transformers that involves removing the coils from the casing ... will cost about $12 million the first year and steadily less each year thereafter." Id. Applying these findings to the 6(c) (1) criteria, the Administrator ruled that continued routine servicing, without rebuilding, and not involving the removal of coils, would present no unreasonable risk of injury.(FOOTNOTE 30) The uncontradicted evidence and the explication of policy considerations in the present record is sufficient to uphold the use authorization for routine servicing of PCB transformers and electromagnets. ... Because they contain lower concentrations of PCBs, PCB-contaminated transformers present correspondingly smaller risks associated with exposure. Accordingly, the Administrator found that routine servicing of PCB-contaminated transformers presents no unreasonable risk of injury. Furthermore, because ninety-nine percent of all large transformers are PCB-contaminated transformers, a prohibition on rebuilding could cause "extremely high" costs. Id. Balancing these factors, the Administrator concluded that there was no unreasonable risk associated with the rebuilding or other extensive servicing of PCB-contaminated transformers. Through these two classifications the Administrator has sought to encourage users to convert to PCB-contaminated transformers by draining their PCB transformers and refilling them with some other dielectric fluid. Only after converting the transformers can users rebuild their transformers, thereby reducing operating costs. Thus, the Administrator has created an incentive to dispose of PCBs without imposing extraordinary costs on industry. These policy considerations and findings reflect the criteria outlined in subsection 6(c) (1). Because the Administrator has carefully articulated his policy judgments, and because there is substantial evidence in the record to support his findings, we uphold the use authorization for PCB-contaminated transformers. The Administrator's authorization of the use and servicing of railroad transformers also reflects a proper balancing of the subsection 6(c)(1) criteria. Because of the strenuous conditions under which they operate, railroad transformers often leak PCBs onto railroad beds,(FOOTNOTE 31) risking exposure to the environment and to workers and other persons near rail lines. On the other hand, nearly 1,000 of these transformers are currently used in railroad engines. A flat prohibition of their use would produce "an unacceptably severe curtailment of railroad service." Id. at 31,532. In order to balance the social and economic impact of a prohibition against the risks to health and the environment, the Administrator sought a solution that would permit continued use while promoting conversion to non-PCB dielectric fluid. In reaching his solution, the Administrator considered the ninety million dollars in costs associated with immediate conversion to non-PCB dielectrics and the undermined safety risks associated with fire and explosion in using non-PCB dielectrics. Consequently, EPA issued regulations requiring railroads to reduce the concentration of PCBs in railroad transformers to six percent (60,000 ppm) by 1982, and to 1,000 ppm by 1984. This timetable, EPA believes, will give it sufficient time to evaluate the risks associated with use of non-PCB fluids, and will also substantially reduce the costs of compliance. Id. at 31,533. It is clear that the Administrator has properly applied the 6(c)(1) criteria in making the unreasonable risk determinations. Where scientific knowledge is incomplete, EPA has set forth specific policy considerations explaining the final regulations. Finding substantial evidence in the record to support the Administrator's findings, we uphold the authorizations for railroad transformers.(FOOTNOTE 32)


IV. THE FIFTY PPM REGULATORY CUTOFF

As a part of the regulatory scheme for PCBs under section 6(e), EPA limited application of the Disposal and Ban Regulations to materials containing concentrations of at least fifty ppm of PCBs. With one exception,(FOOTNOTE 33) materials with lower concentrations remain unregulated under the TSCA regulations.(FOOTNOTE 34) EDF contends that this limitation contravenes the statutory command in subsections 6(e)(2)(A) and 6(e)(3)(A) to regulate "any polychlorinated biphenyl." While we do not adopt all of EDF's reasoning, we find that, under the applicable standard for judicial review,(FOOTNOTE 35) there is no substantial evidence in the record to support the Administrator's decision to establish a regulatory cutoff at fifty ppm. Throughout the rulemaking proceedings for both the Disposal and Ban Regulations, EPA assumed that it would adopt some sort of regulatory cutoff. In the Disposal Regulations, EPA set the cutoff at 500 ppm, not because of health and environmental considerations, but in order to choose "a level at which regulated disposal of most PCB's can be implemented as soon as possible." Preamble to Final Disposal Regulations, 43 Fed.Reg. 7,151 (1978). EPA was reluctant to impose a lower cutoff since, from available information, the agency could not determine the "regulatory impact on commercial products" for lower levels. Subsequent to those proceedings, however, the agency acquired evidence that led it to believe that the "impact on commercial products of defining lower levels of contamination as 'PCB Mixtures' appears less than first believed .... (As a result), the Agency plans to propose a lower concentration of PCB's, possibly in the range of 50 ppm or below, to define PCB mixture in the forthcoming" Ban Regulations. Id. In the Proposed Ban Regulations, EPA listed four reasons for setting the regulatory cutoff at fifty ppm. First, EPA believed that a fifty ppm limit would "exclude from the rule municipal sludges and other mixtures containing low (less than 50 ppm) levels of PCB's whose presence is due to ambient levels of PCB present in the air or water." Preamble to Proposed Ban Regulations, 43 Fed.Reg. 24,804 (1978). As EPA develops in its brief, Congress did not design section 6(e) to regulate ambient sources of PCBs. Second, EPA believed that some industrial chemical processes inevitably produce traces of PCBs, and that careful controls could reduce the concentrations of PCBs only to fifty ppm. Third, EPA felt that it was impractical to regulate the "diffuse and extremely numerous PCB sources" with concentrations below fifty ppm. Id. EPA believed that the proposed cutoff would ensure maximum effectiveness of the regulation by focusing "Agency attention under TSCA upon the most significant and controllable sources of PCB exposure." Id. Fourth, the agency believed that other statutes were available to regulate low concentrations of PCBs, particularly municipal sludges and dredge soils. In the Final Ban Regulations, EPA adopted the proposed fifty ppm regulatory cutoff. Although industry favored a cutoff of 500 ppm in order to reduce the costs of complying with the regulations, EPA found that industry could comply with the more stringent standard. See Preamble to Final Ban Regulations, 44 Fed.Reg. 31,516 (1979). Furthermore, lowering the cutoff from 500 to fifty ppm would "result in substantially increased health and environmental protection." Id. A cutoff below fifty ppm, on the other hand, would "provide an additional degree of environmental protection but would have a grossly disproportionate effect on the economic impact and would have a serious technological impact on the organic chemicals industry." Id. While it did not have firm data, EPA believed that for some chemical processes, it was technically impossible to eliminate the inadvertent production of PCBs. EPA also feared that because of limited disposal facilities, a lower cutoff would increase disposal requirements and interfere with the disposal of high concentration wastes. In short, EPA believed that the fifty ppm cutoff "provides adequate protection for human health and the environment while defining a program that EPA can effectively implement." Id. (FOOTNOTE 36) Both EPA and EDF claim that the statutory language and legislative history support their positions on the regulatory cutoff. The statutory language is simple: "no person may ... use any polychlorinated biphenyl in any manner other than in a totally enclosed manner." 15 U.S.C. s 2605(e)(2)(A). Similarly, the prohibitions on manufacture, processing, and distribution refer to "any polychlorinated biphenyl." See id. s 2605(e)(3)(A). Taken literally, this language might require EPA to regulate every molecule of PCB. We are reluctant, however, to impose such an extreme interpretation absent support in the legislative history. The legislative history reveals that Congress was aware of existing environmental contamination by PCBs the so-called ambient sources of contamination. For example, during the Senate debate of the amendment adding section 6(e) to the TSCA bill, Senator Nelson, who introduced the amendment, read into the record reports showing the widespread environmental contamination by PCBs. ... EPA concluded, we believe correctly, that despite Congress' recognition that existing contamination of PCBs in the environment posed continuing risks to humans and wildlife, Congress did not design section 6(e) to regulate ambient sources of PCBs. Congressman Gude, who co-authored section 6(e), argued that it "will speedily eliminate the introduction of additional PCB's into the environment." Id. at 27186, reprinted in Legislative History, supra note 7, at 585 (emphasis added). Congressman Leggett noted that "PCB's cannot be removed from the environment" and that even if "PCB's were eliminated now," waterways will remain contaminated for years. Id. at 27187, reprinted in Legislative History, supra note 7, at 588. From these statements we conclude that section 6(e) was intended to regulate point sources of contamination. (FOOTNOTE 37)
FN37. From the sparse legislative history of s 6(e), it also appears that Congress focused its attention on the deliberate use, manufacture, and distribution of PCBs. Throughout the congressional debate, members of Congress referred to Monsanto Company as the sole producer of PCBs. See 122 Cong.Rec. 8294 (1976), reprinted in Legislative History, supra note 7, at 240 (Senator Tunney, speaking in support of the section, referred to Monsanto as the "sole domestic manufacturer of PCB's"); id. at 27187, reprinted in Legislative History, supra note 7, at 588 (Congressman Leggett, speaking in support of the corresponding section in the House bill, referred to Monsanto as "the only American manufacturer of PCB's"). Because Congress may not have been aware that other manufacturers produce PCBs as an incidental by-product of their manufacturing processes, it is possible that Congress did not intend to regulate the incidental manufacture of PCBs.
However, given Congress' deep concern with the dangers associated with PCBs, prompting it to include a special section governing PCBs in an otherwise general act, we cannot find that Congress intended to exclude incidentally manufactured PCBs from the Act. Congress' express concern with widespread contamination and worker exposure, the chemical's toxicity at extremely low levels, and the statutory language of "any polychlorinated biphenyl in any manner" lead us to conclude that Congress intended to include all commercial sources within the EPA regulations. As EPA noted in the Preamble to the final Ban Regulations: the prohibition applies to the manufacture of any substance or mixture that contains PCB at 50 ppm or greater, including PCB that is an intermediate or "impurity" or "by-product" .... While the production of PCBs under such circumstances may not be intentional and may have no independent commercial value, section 6(e) of TSCA applies to any production of PCBs. Preamble to Final Ban Regulations, 44 Fed.Reg. 31,527 (1979) (emphasis in original). Because it is reasonable, we defer to this construction of the statute by EPA. See Power Reactor Dev. Co. v. Int'l Union of Elec., Radio and Machine Workers, 367 U.S. 396, 408 (1961). Partly in order to incorporate congressional intent, the Administrator chose a regulatory cutoff at a level that he felt would exclude the ambient sources from regulation.(FOOTNOTE 38) We are troubled by this regulation, however, since the purpose of section 6(e) is to prevent the "introduction of additional PCB's into the environment." The selection of a cutoff undermines the congressional intent to regulate non-ambient sources of PCBs if non-ambient sources of contamination remain unregulated. It is equally troubling that the Administrator apparently is not aware of the amount of PCBs excluded from regulation by the fifty ppm or other possible cutoffs. Particularly because the Administrator has found that any exposure to PCBs may have adverse effects, (FOOTNOTE 39) the Administrator's flat exclusion of some industrial sources of contamination must undergo careful scrutiny.(FOOTNOTE 40) While some cutoff may be appropriate, we note that the Administrator did not explain why the regulation could not be designed expressly to exclude ambient sources, thus directly fulfilling congressional intent, rather than achieve that goal indirectly with a cutoff, thereby partly contravening congressional intent. (FOOTNOTE 41) Thus, a desire to exclude ambient sources of contamination, without more, cannot support the regulatory cutoff. EPA also seeks to justify the regulatory cutoff on the basis of the serious impact a lower cutoff would have on industries that inadvertently produce PCBs during the manufacturing process. See Preamble to Final Ban Regulations, 44 Fed.Reg. 31,516 (1979). As EPA readily concedes, however, the inadvertent commercial production of PCBs is to be regulated under the Act. See note 37, supra. By providing a blanket exemption for concentrations below fifty ppm, the Administrator has circumvented the authorizations and exemptions requirements provided in the statute. EPA made no finding that the cutoff would involve no unreasonable risk to health or the environment.(FOOTNOTE 42) As the EPA noted in its Support Document for the final Ban Regulations, justifying a fifty rather than a 500 ppm cutoff, "the authorization and exemption processes are the most effective way to deal with any difficulties. The authorization and exemption processes allow the Agency to tailor the compliance requirements and to be informed as to which companies are having problems and how they are disposing of their waste streams." Support Document, supra note 4, at 93-94. We agree with EPA. Consequently, the burdens faced by industries cannot be the basis for the fifty ppm cutoff.(FOOTNOTE 43) One of the intervenors, Edison Electric Institute (EEI), and EPA have attempted to justify the fifty ppm cutoff as an administratively created exemption to the Act.(FOOTNOTE 44) See Alabama Power Co. v. Costle, 636 F.2d 323, No. 78-1006 (D.C.Cir. Dec. 14, 1979). Under the heading of "administrative necessity," this court has recognized that an agency may depart from the requirements of a regulatory statute. See id. at 357-360 (opinion for the court by Leventhal, J.). While the court in Alabama Power emphasized that "(c)ategorical exemptions from the clear commands of a regulatory statute, though sometimes permitted, are not favored," id. at 358-360, it also noted that there is "substantive authority (for an agency) to take appropriate action to cope with the administrative impossibility of applying the commands of the substantive statute." Id. at 358-359. However, "(t)he agency's burden of justification in such a case is especially heavy." Id. at 359. Considerations such as the availability of enforcement resources are relevant to the administrative necessity exemption. It appears, however, that EPA is not even aware of the amount of PCBs left unregulated by the cutoff. Having made no showing that it cannot carry out the statutory commands for concentrations of PCBs below fifty ppm, EPA fails to meet its heavy burden. Thus, administrative need, on this record, provides no basis for the fifty ppm cutoff. EEI also seeks to justify the regulatory cutoff under a second principle, the "de minimis " exception to statutory commands. In Alabama Power, this court found that an agency has the power, "inherent in most statutory schemes, to overlook circumstances that in context may fairly be considered de minimis." Id. at 360. That power "is not an ability to depart from the statute, but rather a tool to be used in implementing the legislative design." Id. at 360. As the Alabama Power court emphasized, de minimis authority may be available "when the burdens of regulation yield a gain of trivial or no value." Id. It is not sufficient that the agency may believe that the costs outweigh the benefits, for Congress has already made the judgment that the benefits of regulation are sufficient. The record in the present case is replete with findings and data that PCBs are toxic to wildlife in concentrations well below fifty ppm. Furthermore, the record shows that PCBs bioaccumulate in animals, concentrating as they move up the foodchain. Most importantly, EPA expressly found that any exposure of PCBs to the environment or humans could cause adverse effects. These findings leave us unable to say that the Administrator could rationally conclude that the benefits of regulating concentrations below fifty ppm are of no value.(FOOTNOTE 45) Consequently, we conclude that the de minimis exception to the Act is not available to justify the fifty ppm cutoff.(FOOTNOTE 46) We reemphasize that the Administrator has other, more appropriate means providing him with flexibility to avoid disproportionate impacts on industries or on health and the environment. Those tools are the authorization and exemption provisions in subsections 6(e)(2)(B) and 6(e)(3)(B). The standards enunciated therein, requiring findings of no "unreasonable risk of injury to health and the environment" and, in the case of exemptions, good faith efforts to find substitutes, reflect a plain congressional intention that cannot be ignored. For if there is an unreasonable risk of injury, as there may be given EPA's findings, surely Congress did not intend to permit the continued use, manufacture, processing or distribution of PCBs in concentrations below fifty ppm. EPA's ad hoc consideration of economic impact and disposal requirements, leading to a conclusion that the fifty ppm cutoff "provides adequate protection for human health and the environment," Preamble to Final Ban Regulations, 44 Fed.Reg. 31,516 (1979), is neither as rigorous nor as strict as the statutorily required unreasonable risk determination based on the subsection 6(c)(1) criteria.(FOOTNOTE 47) Thus, we remand this part of the record to EPA for further proceedings.
V. TOTALLY ENCLOSED USES

EDF also petitions for review of the Administrator's decision to list several uses, including non-railroad transformers, capacitors, and electromagnets, (FOOTNOTE 48) as totally enclosed uses and therefore exempt from the regulations promulgated under section 6(e). Because we find no substantial evidence in the record to support the Administrator's classifications, we remand this part of the record for further proceedings.... ... First, Congress left to the Administrator the task of deciding which uses were to be deemed totally enclosed. The statute delegates to the Administrator the duty of ensuring that human and environmental exposure is "insignificant," a word that he must define. Second, given Congress' enactment of a special section for regulating PCBs, we cannot believe that Congress meant to leave unregulated leaking transformers, capacitors, and electromagnets. Congress could not have intended to designate, whether explicitly or implicitly, all transformers, capacitors, and electromagnets as totally enclosed. Third, references in the congressional debates to "closed uses," do not necessarily refer to "totally enclosed" uses. Put simply, closed systems develop leaks.(FOOTNOTE 52)
Footnote 52. No doubt transformers, capacitors, and electromagnets are "closed uses." Otherwise, the liquid PCBs would spill out. The problem is identifying those that leak, whether due to design defect, accident, or some other cause. Senator Nelson emphasized this point when, after discussing "closed" and "non-enclosed" uses, he separately defined "totally enclosed manner" as ensuring that " 'any leakage of a PCB from its enclosure will be insignificant.' " 122 Cong.Rec. 8292 (1976), reprinted in Legislative History, supra note 7, at 234.
EPA argues briefly that the record contains substantial evidence supporting the agency's classification of transformers, capacitors, and electromagnets as totally enclosed. In fact, there is no substantial evidence. To begin with, we have found no evidence in the record discussing the probabilities or magnitudes of leaks from capacitors. This plainly does not amount to substantial evidence. As such, capacitors cannot be classified as totally enclosed uses. Furthermore, EPA points to no evidence describing electromagnets as totally enclosed. The only evidence in the record that we have found is one statement by Commonwealth Edison Co. that they had one electromagnet, and that during normal operation there is no exposure to humans or the environment. See J.A. tab Y. From this statement, it is not clear what constitutes "normal operation." Perhaps leakage occurs during abnormal operation or through human error. Perhaps abnormal operations occur frequently. We cannot know the answers to these and other questions, for EPA failed to collect evidence on them. Furthermore, this sort of evidence, regarding a single electromagnet especially in light of the severe consequences associated with even small leaks cannot constitute substantial evidence that electromagnets are totally enclosed. Finally, the EPA's evidence that transformers do not leak comprises three general statements by industry representatives to the effect that totally enclosed containers do not leak. One commentator from the Nebraska Power Industry Committee stated that use of transformers "generally involves no release of PCBs." J.A. tab O. This statement does not amount to substantial evidence that the regulations "ensure" that there is no exposure. Furthermore, there is evidence in the record directly contradicting the industry's statements. One commentator stated that transformers "occasionally blow up and occasionally are mishandled." J.A. tab S. Another commentator, from the Michigan Department of Natural Resources, testified that "we recognized that environmental losses can occur through accidental rupture or leakage." J.A. tab L. He concluded that "we still believe that a high risk of exposure and environmental losses likely exist in such facilities." Id. In light of the record in this case, we find that there is no substantial evidence that the regulations concerning totally enclosed uses "will ensure that any exposure of human beings or the environment to a polychlorinated biphenyl will be insignificant." 15 U.S.C. s 2605(e)(2)(C) (emphasis added). This lack of substantial evidence calls into question EPA's implicit finding that it can designate entire classes of uses, rather than individual containers, as totally enclosed. Of course, we are not directing EPA to apply the "totally enclosed" proviso on a more individualized basis. On the present record, however, EPA's findings as to which PCB uses may be classified as totally enclosed cannot stand. Accordingly, we remand the record to the EPA for further proceedings consistent with this opinion.
VI. CONCLUSION

On the basis of the foregoing, we find that there is substantial evidence in the record to support the use authorizations; therefore, we uphold those regulations. However, because we find no substantial evidence in the record to support either the fifty ppm cutoff or the EPA classification of certain PCB uses as totally enclosed, these latter two regulations cannot be upheld. Consequently, we set aside the regulations dealing with the fifty ppm cutoff and the classification of certain PCB uses as totally enclosed, and remand those portions of the record for further proceedings consistent with this opinion.

[my favorite part!] We feel constrained to add one final note to emphasize our concern in this case. Human beings have finally come to recognize that they must eliminate or control life threatening chemicals, such as PCBs, if the miracle of life is to continue and if earth is to remain a living planet. This is precisely what Congress sought to do when it enacted section 6(e) of the Toxic Substances Control Act. Yet, we find that forty-six months after the effective date of an act designed to either totally ban or closely control the use of PCBs, 99% of the PCBs that were in use when the Act was passed are still in use in the United States.(FOOTNOTE 53) With information such as this in hand, timid souls have good reason to question the prospects for our continued survival, and cynics have just cause to sneer at the effectiveness of governmental regulation.
The EPA regulations can hardly be viewed as a bold step forward in the battle against life threatening chemicals. There is no substantial evidence in the record to support certain of the EPA regulatory enactments, and portions of the regulations are plainly contrary to law. Thus, the effort by EPA has, in certain respects, fallen far short of the mark set by the congressional mandate found in section 6(e) of the Toxic Substances Control Act. On remand, we trust that EPA will act with a sense of urgency to find effective solutions to enforce the Act. We are not so naive as to assume or suggest that hasty responses will ensure effective regulations. However, we are well able to see, from the plain text of the Act, that the deadlines for the enactment of regulations to enforce section 6(e) have passed. We therefore believe that EPA should act with expedition to complete the important task assigned to it by Congress. So ordered.