Union of Concerned Scientists v. Atomic Energy Commission and United States of America, Boston Edison Company, Intervenor

499 F.2d 1069, 163 U.S. App. D.C. 64, 4 Envtl. L. Rep. (Envtl. Law Inst.) 20605, 6 ERC (BNA) 1705, 1974 U.S. App. LEXIS 8212, 1974 WL 333570

• Court: Court of Appeals for the D.C. Circuit
• Decided: June 10, 1974
• Docket: 73-1099
• Status: Published

Opinion

McGOWAN, Circuit Judge:

The Union of Concerned Scientists (UCS) seeks review of an order of the Atomic Energy Commission (AEC) affirming the Initial Decision of an Atomie Safety and Licensing Board (ASLB) to authorize the issuance of an operating license to intervenor Boston Edison Company (Edison) for its Pilgrim Nuclear Power Station (Pilgrim). UCS assigns ten separate errors, which we deal with sequentially after an introductory section stating so much of the history of this case as is necessary to an understanding of the issues raised. For the reasons hereinafter set forth, we affirm.

I

The Atomic Energy Act of 1954, as amended, provides, that nuclear facilities must be the subject initially of a construction permit and thereafter of an operating license. 42 U.S.C. § 2235. Edison sought permission to build Pilgrim at Plymouth, Massachusetts in 1967, and its application was reviewed by the AEC staff and the Advisory Committee on Reactor Safeguards (ACRS). 1 It was approved, after' a hearing, by an ASLB. 2 The Commission issued the construction permit in August, 1968, and construction was begun. 3

In January, 1970, Edison initiated the process of obtaining an operating license by filing an application therefor, together with its Final Safety Analysis Report (FSAR). 4 The information and analy ses required for such a report is extensive, 10 C.F.R. § 50.34(b), as witness the fact that the report on Pilgrim ran to five volumes as originally submitted. The FSAR was reviewed by the Staff, and then by the ACRS, and additional information requested of and supplied by Edison. By this process of successive administrative reviews, the FSAR was refined and amended to the satisfaction of each body. 5 Each made its own safety report. 6

In April, 1971, after more than a year of administrative processing, the AEC noticed its consideration of Edison’s operating license application and the opportunity for affected persons to be heard. 36 F.R. 7696. The Commission received a number of requests for a hearing, including the joint petition of UCS and the Sierra Club, and accordingly scheduled a hearing before a designated ASLB at Plymouth, Massachusetts. 7

Pursuant to the regulation governing intervention, UCS filed its “detailed specifications” of matters it wished to have considered at the hearing, to which both Edison and the Staff filed answers. 8 Of the thirty-one radiological and safety contentions raised, 9 seventeen were rejected on grounds such as relevance, generality, and facial lack of merit ; three were waived; one was set down for hearings; and the remainder were held to be challenges to AEC regulations or criteria for emergency core cooling systems (ECCS) and reserved for possible reference to the Commission under the AEC’s Calvert Cliffs doctrine (see Part IV, infra). The ASLB set December 6, 1971, to begin hearing evidence on both the need for certification of rule challenges to the Commission and the remaining issue on the merits.

Meanwhile, on November 26, 1971, the Commission announced a rule making proceeding to begin in January, 1972 on the question of amending or making permanent the ECCS Interim Acceptance Criteria (IAC), a policy statement adopted without a hearing the previous June 25th. 10 Accordingly, the ASLB ruled that, since the IAC would be subject to general reconsideration in the rule making proceeding at which UCS could be heard, challenges thereto in a particular adjudication had been rendered moot, and would not be entertained for possible certification to the Commission (see Part III, infra). The December 6-8 hearing was limited, therefore, to the one contention held properly to be in issue.

An additional hearing on environmental matters, necessitated by this court’s decision in Calvert Cliffs Coordinating Committee v. AEC, 146 U.S.App.D.C. 33, 449 F.2d 1109 (1971), and the promulgation of new regulations intended to comply with that decision, 36 F.R. 18071 (Sept. 9, 1971), was held in June, 1972, to evaluate the Commission’s Final Environmental Statement. Completion of this proceeding, in which UCS again participated, closed the record and the ASLB issued an Initial Decision based thereon on September 13, 1972. Under AEC regulations, 10 C.F.R. § 2.764, an Initial Decision directing issuance of a license is effective immediately notwithstanding the filing of exceptions thereto. Thus, while Edison was issued an operating license for Pilgrim on September 15, UCS prosecuted ten exceptions before the Atomic Safety and Licensing Appeal Board, (ASLAB), which issued its decision of affirmance (ALAB-83) on December 4, 1972. Failing sua sponte review by the Commission itself, the decision became the final decision of the AEC twenty days later, 10 C.F.R. § 2.786, and UCS sought review in this court.

II

The most novel of UCS’s many exceptions is its threshold disagreement with the Commission over the proper role of an ASLB. UCS has contended since the beginning of the hearing that the ASLB must, despite the absence of controversy, review de novo and independently evaluate the evidence to determine whether the issuance of an operating license is consistent with “the health and safety of the public.” 11 This the ASLB declined to do, since the Commission Rules of Practice suggested it should not. The rule then in effect provided in part: 12

In contested proceedings, the board will determine controverted matters as well as decide whether the findings required by the Act and the Commission’s regulations should be made. As to matters which are not in controversy, boards are neither required nor expected to duplicate the review already performed by the regulatory staff and the ACRS and they are authorized to rely upon the testimony of the regulatory staff and the applicant, and the conclusions of the ACRS, which are not controverted by any party.

As UCS acknowledges, the substance of its argument goes to, and is inconsistent with, the validity of this rule. There is no merit, therefore, to Edison’s suggestion that the challenge to the rule is raised for the first time in this court. We think the rule a valid one, however, and turn to the framework within which it operates, and within which it must be evaluated, in order to show why.

The two-step procedure by which nuclear facilities are successively licensed for construction and for operation originated in the Atomic Energy Act of 1954. Before a construction permit could be issued, an applicant was required to submit to the Commission detailed design and safety information sufficient to enable it to find that a proposed facility would “provide adequate protection to the health and safety of the public.” Section 182, 42 U.S.C. § 2232(a). The application was reviewed by the Staff, and by the ACRS, whose report was not made public. If the Commission decided to approve the construction application, it was required to give public notice and to grant a hearing at the instance of “any person whose interest might be affected.” Again at the operating license stage, any interested person could require that a hearing be convened, and again, if no one so requested, no hearing would be held, even though “the definitive safety determination” would occur at the second stage. Power Reactor Development Co. v. International Union of Electrical Workers, 367 U.S. 396, 407, 81 S.Ct. 1529, 6 L.Ed.2d 924 (1961).

The Joint Committee on Atomic Energy initiated a study of Commission procedures in 1956. 13 As a result, 14 the Act was amended in 1957 to give the ACRS statutory status, require ACRS review of every license application and publication of its report, and make mandatory the holding of public hearings at both the construction and operating license stages. 71 Stat. 579 (1957), 42 U.S.C. §§ 2039, 2232(b), 2239(a).

By 1960 it had become apparent to both the AEC and the Joint Committee that the agency’s regulatory organization and procedures were in need of further refinement to meet problems associated with the combination of regulatory and promotional functions and the increasing workload arising from an accelerating number of license applications. Accordingly, each body conducted its own study and made recommendations, 15 and the Act was amended again in 1962. Pub.L.No. 87-615, 76 Stat. 409 (1962).

The Committee staff had recommended a number of changes in the reactor licensing process, including the creation of an “internal board with final licensing authority (not subject to the Commission review).” 16 The AEC, on the other hand, had recommended that any licensing board be subject to its review in order to avoid conflicts between AEC policy, as expressed in its rules, and results in particular adjudications 17 a consideration to which Congress was responsive.

Congress acted in 1962 to relieve the burdens on the Commission by authorizing it

to establish one or more atomic safety and licensing boards, each comprised of three members, one of whom shall be qualified in the conduct of administrative proceedings and two of whom shall have such technical or other qualifications as the Commission deems appropriate to the issues to be decided, to conduct such hearings as the Commission may direct and make such intermediate or final decisions as the Commission may authorize with respect to the granting, suspending, revoking or amending of any license or authorization ....

42 U.S.C. § 2241(a).

As conceived by Congress, the ASLB was to be “a flexible experiment in new administrative techniques,” enabling the Commission to draw on talents from outside the government and to delegate any of “a broad range of regulatory functions” while reserving to itself final authority in all matters. An ASLB was intended to “follow procedures appropriate to the resolutions of complicated technical and scientific questions, keeping in mind the necessity for preserving a suitable record for review,” but to emphasize informal procedures as much as possible. S.Rep.No. 1677, 87th Cong., 2d Sess. 2-7 (1962) (Joint Committee on Atomic Energy).

Consistent with this emphasis on reforming licensing procedures to reconcile considerations of practical administration with those of safety, the 1962 amendments dropped the mandatory two-hearing requirement that Congress had imposed in 1957. Section 189(a) of the Act, 42 U.S.C. § 2239(a), was amended to make a hearing mandatory only on the application for issuance of a construction permit.

As the Joint Committee explained:

At the “Radiation Safety and Regulation” hearings in June, 1961 and at the 1962 regulatory hearings, there was substantial unanimity of opinion that the mandatory hearing requirements of the Act with respect to power and testing facilities should be relaxed. The second hearing on the operating licenses was regarded, by most witnesses, as unnecessary and burdensome in the absence of bona fide intervention.

The Joint Committee determined that elimination of the second hearing requirement would not compromise the public interest in safety for two reasons: there would still be a mandatory hearing “at the critical point in reactor licensing — the construction permit stage —where the suitability of the site is to be judged;” and, even if no intervenor causes a hearing to be held at the operating license stage, “the decision would still be on the public record and if an important safety question was involved, could be made by the Board.” Id. at 7-8.

In implementation of the ASLB program authorized by Congress, the AEC promulgated new Rules of Practice to describe the functions and procedures of its Boards, 18 including the statement of general policy here challenged. That rule charges the ASLB with, in addition to resolving controverted matters, deciding “whether the findings required by the Act and the Commission's regulations should be made,” a responsibility it bears regardless of whether the proceeding is contested or uncontested. It is clear, however, that the ASLB was directed not to make the findings itself, but to determine “whether the application and the record of the proceeding contain sufficient information, and the review of the application by the Commission’s regulatory staff has been adequate, to support the findings proposed to be made by the Director of Regulation.” 19 An analogy is to the function of an appellate court applying the “substantial evidence” test, although it is imperfect because the ASLB looks not only to the information in the record, but also to the thoroughness of the review that the Staff, aided by the ACRS, has given -it.

In delegating a limited reviewing function to the ASLB, we have no doubt that the Commission has acted in a manner within the contemplation of the Act as amended in 1962. We noted in Siegel v. AEC, 130 U.S.App.D.C. 307, 400 F.2d 778, 783 (1968), where the Commission’s policy of excluding the possibility of enemy attack from its safety considerations at construction permit hearings was upheld, that the Atomic Energy Act erects “a regulatory scheme which is virtually unique in the degree to which broad responsibility is reposed in the administrative agency, free of close prescription in its charter as to how it shall proceed in achieving the statutory objectives.” Beyond this statement of general applicability, however, there are more specific congressional indications of a function short of de novo review.

Of first importance in this respect is the primary purpose of the 1962 amendments, namely, to unburden the Commission by authorizing it to delegate decisional authority and removing the necessity of holding unnecessary and duplicative hearings. The Staff’s safety analysis, aided by the ACRS, is reviewed prior to issuance of the construction permit, which the Joint Committee identified as “the critical point in reactor licensing.” No second hearing need be held at all “in the absence of bona fide intervention.” Yet it is contended that the mere assertion of a need — even a mere desire — for a further hearing by any of the many persons whose interests may be affected necessitates a second “independent evaluation” of all the evidence on safety considerations rather than only the matters put in issue. If this were the meaning of the 1962 amendments, one wonders why Congress would have bothered at all to relax the two-hearing requirement.

UCS argues, however, that reliance on the Staff’s analysis as to uncontested matters in effect vests final decisional authority in an adversary party, and that delegation to an ASLB of responsibility to make the definitive safety findings requires it to do so independently, receiving the Staff’s uncontested evidence only for the weight it deserves. 20 As a formal matter, of course, this practice would exceed the scope of the ASLB’s delegated authority. Whether the proceeding is uncontested or contested, its mandate is to review the sufficiency of the record and the adequacy of the analysis to support the necessary findings. Adding to that responsibility in a contested case the authority to resolve disputes in no way affects the ASLB’s authority respecting uncontested matters, nor makes the Staff a “party” as to those matters. As a practical matter, moreover, it would simply not be possible for the two technical members of the panel to evaluate in detail the totality of material relevant to safety matters that the Staff and ACRS have generated through many months of work. 21 This fact is so obvious that it borders on the ludicrous to suggest that Congress intended the ASLB so to function.

Petitioner places great reliance on three cases for the proposition that an ASLB has an affirmative duty to evaluate the correctness of, rather than sufficiency of support for, uncontested findings, lest an impossible burden be put on any who would oppose the issuance of a license. In Office of Communication, United Church of Christ v. FCC, 138 U.S.App.D.C. 112, 425 F.2d 543 (1969), we reversed a licensing decision supported by a record in which the burden of persuasion on the public interest question was shifted from the applicant, where by statute it is placed, to the intervenor in opposition to renewal of the license. “The Commissioners and the Examiners,” we said, “have an affirmative duty to assist in the development of a meaningful record” on which to evaluate the licensee's performance. Id. at 548. Petitioner’s reliance on this statement, however, bespeaks a fundamental misunderstanding of the AEC licensing process. The role of the ASLB is not to compile a record; it is to review a record already compiled by the Staff and ACRS, who have responsibility for the sufficiency of that record. In the Federal Communications Act, Congress provided for public participation in the development of the record. In the Atomic Energy Act, owing to the different nature of the issues involved, it excluded the public at that stage but, in the interest of public confidence in the thoroughness of the review process, S.Rep.No. 1677, supra at 9, invited public scrutiny at the later stage. Indeed, it authorized the Commission, in its discretion, to determine that certain applications, present no “significant hazards consideration” and to dispense with notice and publication of impending approval, excluding the public altogether. 22

The other two cases, Greene County Planning Board v. FPC, 455 F.2d 412 (2d Cir. 1972), and Calvert Cliffs Coordinating Committee, Inc. v. AEC, 146 U.S.App.D.C. 33, 449 F.2d 1109 (1971), each involve agency responsibility under Section 102 of the National Environmental Policy Act (NEPA). 42 U.S.C. § 4332. In Greene County it was held that agency reliance on a license applicant’s environmental impact statement was an abdication of the NEPA requirement that the agency do its own environmental analysis. Here, it is suggested that the applicant’s FSAR is equally likely to be too self-serving to justify ASLB reliance in making its definitive safety findings. However that may be, UCS's apprehension is unfounded; the FSAR is the very document on which Staff and ACRS scrutiny focuses. Only when those bodies are satisfied that it is accurate, usually after many changes and additions, does it go before the ASLB, with the accompanying reports on it. It is by then as much a document presented by the Staff as by the applicant and the danger perceived in Greene County is not operative.

This court’s decision in Calvert Cliffs appears only at first blush to be closer in point. The subject of review there was a regulation purportedly implementing NEPA but in fact inconsistent with it. The regulation prohibited an ASLB, in passing on a construction permit, from considering the environmental “balance among conflicting factors that is struck in the staff’s recommendation,” unless affirmatively challenged. Thus, the Commission’s responsibilities under NEPA were, in the words of the regulation, to be “carried out in toto outside the hearing process.” This singling out of environmental factors, where the ASLB automatically considers nonenvironmental matters, we held to be inconsistent with the NEPA requirement of agency consideration of the environment “to the fullest extent possible.”

In this case, environmental factors were independently evaluated by the ASLB in the supplemental operating license hearing held in June, 1972. Accordingly, USC’s grievance is the mirror image of that pressed in Calvert Cliffs, viz., that radiological and safety matters were singled out for laxer review than that given to environmental ones under the mandate of this court. NEPA and the Atomic Energy Act are two separate statutes, however, and while a lesser scrutiny of environmental issues is inherently inconsistent with consideration to the fullest extent possible, lesser scrutiny of nonenvironmental matters in an operating license proceeding need not logically be a similar infirmity under a distinct statute. Nor would any such disparity exist in this case but for an accident of timing.

Pilgrim’s construction permit was issued before enactment of NEPA in 1969. Its operating license hearing was scheduled after the effective date of NEPA and was to be conducted under the regulation later struck down in Calvert Cliffs. Acutely aware of the two-stage licensing procedure of the Atomic Energy Act and the fact that some construction permits issued pre-NEPA and others issued post-NEPA (but without adequate environmental consideration due to the faulty regulations) had not yet resulted in the issuance of operating licenses, this court ordered that fully adequate NEPA review be conducted before the operating licenses issued. 449 F.2d at 1128-1129 n. 43. The AEC thereupon promulgated new regulations to govern both all new license proceedings, and operating license proceedings in cases where there had been no environmental review at the construction stage or such review was inadequate. 23

The Supplemental Hearing on environmental matters in the case before us was called to comply with the requirement of Calvert Cliffs and the new regulations. UCS is simply wrong if it conceives that the environmental phase of operating license proceedings will continue to involve a standard of review different from that in safety matters once the backlog of reactors issued construction permits before Calvert Cliffs is exhausted. We expressly said in that opinion that full NEPA consideration “need not be duplicated, absent new information or new developments, at the operating stage.” 449 F.2d at 1128. To the extent there are new matters, such as alterations in the plans as evaluated at the construction stage, full NEPA consideration will of course be demanded.

Finally, we take this opportunity to illustrate what we said in Siegel about the “virtually unique” arrangement under which the AEC is to operate “free of close prescription in its charter as to how it shall proceed in achieving the statutory objectives,” but subject to oversight by a Joint Committee of Congress “in constant touch with what was happening in this unfolding area of industrial applications of atomic energy.” In 1967, that committee held extensive hearings on the AEC’s program for licensing and regulating nuclear reactors. Representative Holifield, then vice-chairman, explained in opening the hearings:

We on the committee regard the surveillance of this program as one of our primary responsibilities. Our last intensive review of this subject, in 1962, resulted in enactment of legislation which strengthened and streamlined the regulatory process.

Now, after 5 years of further evolution of the industry ... we believe another reevaluation is called for.

The committee heard detailed accounts of the way in which an ASLB operates. It was fully informed about the nature of its inquiry as to sufficiency of information and adequacy of Staff review. Hearings on Licensing and Regulation of Nuclear Reactors, supra note 21.

No legislation was then pending on the licensing process, but Mr. Holifield emphasized that some would be forthcoming if the hearings revealed the need. We do not, from the absence of any legislation, conclude that everything the Committee learned it approved. But we do think it relevant that the administrative procedure here challenged has withstood careful scrutiny without eliciting expressions of dismay from the congressmen who questioned agency representatives and ASLB members. Petitioner is not for the first time exposing to sunlight, which Justice Brandéis so aptly called the best of disinfectants, 24 an administrative process made suspect by neglect and inattention. 25

Ill

Petitioner next contends that, in granting a full-power operating license for Pilgrim without first resolving its challenges to the Interim Acceptance Criteria (IAC) for emergency core cooling systems, the Commission deprived UCS of due process, violated its own regulations, and abused its discretion.

The IAC had been published as an interim policy statement on June 29, 1971 in order to give immediate effect to standards developed on the basis of new information arising in the course of industry design changes in, and AEC semi-scale testing of, ECCS. 36 F.R. 12247. Both new and existing reactors were required to meet the four criteria as determined by .use of an acceptable evaluation model, of which examples were provided. 26 Edison submitted data demonstrating to the satisfaction of the regulatory Staff that the Pilgrim ECCS met the new criteria, which UCS purported to dispute. The ASLB determined that the challenges in question went to the IAC themselves, rather than to Pilgrim’s ability to comply with them, and then scheduled them for a hearing looking toward reference of substantial questions to the AEC. But UCS’s challenges were not resolved in this proceeding, it will be recalled, because the ASLB ruled that the AEC’s intervening announcement of a rule making to reexamine the criteria made it inappropriate to entertain duplicative challenges in an adjudicatory context.

After completion of the December hearing on radiological and safety matters, the AEC issued a Supplemental Notice concerning the impending rule making, 37 F.R. 288 (Jan. 8, 1972), which provided in part that the simultaneous conduct of a rule making would “not affect the orderly resolution, under the Commission’s existing regulations, of the matter of emergency core cooling, in hearings for light water-cooled power reactors pending before [ASLBs].” After the environmental hearing was completed in June, 1972, therefore, the ASLB proceeded to recommend issuance of a full power license for Pilgrim. It is UCS’s position that the agency should not have issued that license but either deferred action pending the outcome of rule making, or licensed operations only at a power level sufficiently below capacity to eliminate fully the asserted danger of the loss of coolant accident (LO CA) postulated by USC if thé IAC were in fact deficient. 27

Before entertaining this contention, we are obliged to consider whether the completion of rule making, and revision of the criteria, moots the issue UCS raises. 28 The new criteria became effective January 28, 1974. As petitioner points out, however, licensees are afforded six months from that date to demonstrate compliance, and the allowable time may be extended for good cause shown. Until Pilgrim is brought into compliance, therefore, its continued full-power operation depends upon the validity of its 1972 license. Pilgrim may, as earnestly pressed upon us, soon achieve compliance, or it may, in the nature of things, be unable to do so. We cannot say that this aspect of UCS’s challenge has been mooted before the occurrence of an event of speculative proximity.

A. Due Process.

We find petitioner’s due process theory unavailing. Due process does indeed require that, where a right to be heard exists, it must be accommodated “at a meaningful time and in a meaningful manner.” Armstrong v. Manzo, 380 U.S. 545, 552, 85 S.Ct. 1187, 1191, 14 L.Ed.2d 62 (1965); Goldberg v. Kelly, 397 U.S. 254, 267, 90 S.Ct. 1011, 25 L.Ed.2d 287 (1970). Petitioner here was deprived of an opportunity to be heard on certain issues before issuance of Pilgrim’s license.- In being shunted over to the simultaneous rule making, where it was heard, it was not, in the circumstances of this case, deprived of a timely forum by constitutional standards. Administrative action taken prior to a full hearing has always been permissible when the state’s interest in acting promptly to promote the general welfare, including economic well-being, outweighs the individual’s interest in having an opportunity to be heard before the state acts, perhaps in error, in ways that may cause him significant injury. 29 Surely, then, it is not unconstitutional where, as here, the sole responsibility for the public health and safety is firmly vested in the agency, Power Reactor Development Co., supra, the possibility of harm from the agency’s action is seen only by petitioner in its sole judgment, and that judgment rests upon its view of numerous complex technical questions within the competence of the agency to evaluate.

The Supreme Court recently addressed the due process requirement in the closely related context of agency enforcement. In Weinberger v. Hynson, Westcott & Dunning, Inc., 412 U.S. 609, 624-625, 93 S.Ct. 2469, 2480, 37 L.Ed.2d 207 (1973), it is said:

Regulatory agencies have by the requirements of particular statutes usually proceeded on a case-by-case basis, giving each person subject to regulation separate hearings. But there is not always a constitutional reason why that must be done. . To require separate judicial proceedings to be brought against each [regulatee], as if each were the owner of a Black Acre being condemned, would be to create delay where in the interests of public health there should be prompt action. A single administrative proceeding in which each [regulatee] may be heard is constitutionally permissible measured by the requirement of procedural due process.

In the instant case, where challenges to the agency’s standards rather than agency enforcement of those standards are consolidated into a single administrative proceeding, the same considerations are operative. If the agency could not consolidate the challenges to its rules into rule making, and meanwhile proceed with adjudications, UCS and other intervenors in other cases would effectively be able to impose a moratorium on licensing, despite the Commission’s judgment that it is prompt action that is called for. 30 It is no answer that Pilgrim might have been licensed at less than full power to obviate the hazard as seen by UCS; the considerations are the same, pro tanto.

B. Regulations.

1. Calvert Cliffs Memorandum.

Petitioner contends that this case comes within the well-settled rule that an agency’s failure to follow its own regulations is fatal to the deviant action. See, e. g., Service v. Dulles, 354 U.S. 363, 77 S.Ct. 1152, 1 L.Ed.2d 1403 (1957). The rule assertedly violated here is the AEC’s Calvert Cliffs doctrine that substantial questions as to the validity of a Commission safety standard should be certified to the Commission for resolution before an ASLB renders an initial decision. This, of course, was not done here because the Supplemental Notice of rule making called for completion of pending license proceedings under the IAC.

The Calvert Cliffs doctrine, announced by the AEC in 1969, may have been a unique experiment in administrative procedures. 31 It contemplated at once a procedure for adjusting general rules, and particularly radiological safety standards, to the needs of particular cases, without delegating the power of amendment and thereby sacrificing centralization of policymaking. This might be a desirable experiment in many areas of administrative law, but it is obviously so consuming of agency resources to entertain multiple challenges to the same regulation that only an agency deciding the gravest of questions could even consider it.

Whether the experiment was noble or not, 32 however, petitioner’s grievance that it was abandoned in this case is subject to a fatal flaw: Calvert Cliffs was not a rule or regulation. It was a notice to the ASLBs that the Commission would entertain certain questions, certified by the ASLBs. It was initiated in the course of an opinion in a particular case, and qualified in a directive of equal or greater dignity. No rights in its continued use, whether or not it was relied upon, vested in petitioner or anyone else. See C. H. Guenther & Son, Inc. v. NLRB, 427 F.2d 983 (5th Cir. 1970).

2. NEPA Regulations.

The Commission’s regulations implementing the National Environmental Policy Act require the Director of Regulation to prepare a final detailed environmental impact statement (FES) prior to the issuance of a license. The decision on issuance is thereby to be informed by the results of the FES cost-benefit analysis “weighing the environmental, economic, technical and other benefits against environmental costs and considering available alternatives.” 10 C.F.R. Part 50, App. D, § A.8. In this case, the required analysis assumed that Pilgrim would operate at full power.

One possible consequence of applying more stringent acceptance criteria to Pilgrim, however, could be to require its derating to something less than full-power operation. Accordingly, UCS argues, a complete FES would consider the balance of costs and benefits at less than capacity operation, under which circumstances the benefits would presumably be reduced proportionately. Failure to make this consideration is assigned as a violation of the quoted regulation, rendering the license fatally defective. The consequence, we are told, is that a subsequent reduction in operations and benefits will not give rise to a new environmental analysis, even though operation at reduced levels might not be cost-justified. 33 Alternatively, it is argued that the FES did not consider a simple deferral of full-power licensing pending the outcome of rule making. 34

This is a not inconsiderable objection to the FES, and Edison’s answer, that a lower power level of operation could only result in a lesser impact on the environment, is not adequately responsive to it. It would be adequate only if the aggregate “environmental, economic, technical and other benefits” of operating this particular nuclear reactor were known to decrease at a rate such that, for any given decrease in power rating, these benefits would always exceed associated environmental costs. This is not only unknown but unlikely. If, for example, a significant derating lowered the temperature of water pumped through the station from 69° F to 62° F — and we do not know that it would have this or a similar effect, but merely offer a hypothetical in aid of analysis — then the volume of marine life killed by exposure to temperatures in excess of 60° F would not change, leaving a decreased economic benefit against a constant environmental cost. In a closed system containing only power output and marine life, the balance could indeed be tipped by derating.

More to the point is the Commission’s practical observation that consideration in the FES of derating to an unknowable level as the result of then as yet unformulated criteria would have been speculative to the point of uselessness. If we were to assume that the AEC could foresee that the criteria eventually to emerge might require derating Pilgrim for a transitional period pending compliance, a rather specific rule would have to be hypothesized in order to yield an operating level from which to reevaluate the environmental balance. 35 Lacking a specific rule, the cost-benefit analysis could have been performed at a number of assumed operating levels, e. g., 95%, 75%, and 60%. But this exercise would not have made the FES a more complete consideration of Pilgrim’s environmental impact or a more adequate basis of decision. It would simply have incorporated an attempt, on highly speculative assumptions, to anticipate its environmental effect under changed circumstances. Consideration of deferral suffers from the same defect. Compliance with impact statement procedures always involves some forecasting, and perhaps more can be expected where “irretrievable commitments” are being made, but the agency cannot be required to “foresee the unforeseeable.” Scientists’ Institute for Public Information, Inc. v. AEC, 156 U.S.App.D.C. 395, 481 F.2d 1079, 1092 (1973).

Clearly, the more appropriate time at which to reevaluate Pilgrim’s cost-benefit ratio would be when the new criteria are actually applied to the reactor. Then, if derating of a given magnitude is indeed required, the change in benefits will be known, and the change, if any, in environmental costs will be fairly ascertainable. Petitioner has maintained that this second look will never be undertaken, but we think in this it errs.

If Pilgrim can comply with the new criteria, by definition, no derating will be required and there will be no need to reexamine costs and benefits after all. If Pilgrim cannot comply and derating is indicated, its license will be amended accordingly. Regardless of whether UCS or anyone else causes a hearing to be held at that stage, 36 a proposed amendment significantly to derate reactor operations, with its above-described potential for upsetting the environmental cost-benefit ratio, would itself so clearly be a major federal decision within the strictures of NEPA that AEC reevaluation could not be avoided. 37 This follows from the cases holding that a major federal action cannot be taken out of NEPA’s reach by fragmenting the project into a series of seemingly non-major component decisions to go forward. It applies whether the initial decisions predated the effective date of NEPA, Calvert Cliffs, supra, straddled it, Jones v. Lynn, 477 F.2d 885 (1st Cir. 1973), or were made subsequently. See Students Challenging Regulatory Agency Procedures v. United States, 346 F.Supp. 189, 200 (D.D.C.1972) (three judge court), preliminary injunction reversed on other grounds, 412 U.S. 669, 93 S.Ct. 2405, 37 L.Ed.2d 254 (1973); Council on Environmental Quality, Statements on Proposed Federal Actions Affecting the Environment: Guidelines, § 5, 36 F.R. 7724 (April 23, 1971), and Proposed Guidelines, §§ 5, 6, 38 F.R. 10857 (May 2, 1973). These cases and guidelines were concerned with decisions to expand and extend incrementally, but precisely the same hazard attends a decision to curtail operation of a project that was justified initially on the assumption it would operate to capacity, as the legislative history, as well as the logic, of NEPA suggest. 38 Accordingly, petitioner’s concern that the cost-benefit analysis required to implement NEPA “to the fullest extent possible” will never be based on actual costs and benefits in the event of derating is at best premature.

C. Abuse of Discretion.

Even if the AEC had the power to refer the rule challenges to rule making and proceed with the issuance of Pilgrim’s license under the IAC, UCS argues it was an abuse of discretion to do so. The basis for this argument is that the IAC were promulgated without a prior hearing, 5 U.S.C. § 553(b), to govern an extremely important safety system, and then shielded from pre-licensing challenge despite “a substantial body of information available to [the AEC] which indicated that the [IAC] did not provide an adequate basis for licensing.”

That the IAC were issued without notice and hearing is of no significance whatever if they were validly within an exception to that requirement, and there is no suggestion to the contrary. They were accompanied by a statement of reasons clearly showing why any delay would be contrary to the public interest; comments were solicited “with the view to possible amendments” or holding an informal rule making hearing, cf. Independent Broker-Dealers’ Trade Ass’n v. SEC, 142 U.S.App.D.C. 384, 442 F.2d 132, 144-45 (1971), and UCS submitted its views by letter, but did not seek judicial review. In providing for promulgation of rules of immediate effect under appropriate conditions, Congress obviously contemplated that they would be as effective as any other rules, and UCS can hardly impugn them on that ground.

As to the importance of ECCS in the event of a LOCA, we have no doubt. 39 But neither have we any basis to doubt the improbability of a LOCA occurring, or the validity of AEC’s general “defense-in-depth” concept in which the ECCS is but the last line of defense. Nonetheless, were UCS on sounder ground in claiming that the AEC had before it substantial evidence calling the IAC into question, we would have to consider deeply whether its discretion was abused in proceeding with licensure. For, if this were so, we might well join petitioner in drawing darker inferences from the process by which the IAC were announced and applied. No such showing has been made, however.

UCS directs our attention to a December, 1971 letter from a “research group” at the AEC’s Oak Ridge National Laboratory (ORNL) to the Director of Regulation allegedly expressing the view that the IAC’s 2300°F limit on the calculated maximum cladding (fuel casing) temperature under LOCA conditions would be inadequate to prevent excessive embrittlement (which could lead to frag mentation of the cladding) 40 But, as UCS itself brings out, the validity of this position and the identity of a more appropriate standard were the subject of disagreement among the staff and ORNL witnesses at the subsequent rule making. 41

The other item identified as significant in this regard is a February, 1971 memorandum from the Director of the Division of Reactor Development and Technology to the General Manager of the AEC on a research proposal by the General Electric Corporation, a supplier of ECCSs to utility companies. The statement on which UCS focuses is that, “[although test information is available on the response of simulated fuel pin bundles to a range of emergency coolant flow conditions, no assurance is yet available that emergency coolant can be delivered at the rates intended and in time period prior to clad and subsequent fuel melting to decay heat generation.” There are problems with assigning any significance, let alone an abuse of discretion, to the Commission’s proceeding to license Pilgrim after the date of this memorandum. First, the memorandum predates the IAC, as well as Pilgrim licensure. In the statement accompanying the IAC, the Commission expressly recognized that it still lacked “analytical methods capable of detailed realistic prediction of all phenomena known or suspected to occur during a [LOCA], supported in every aspect by definitive experiments directly applicable to the accident.” It continued:

In the absence of such perfection, adequate assurance of safety can be obtained from an appropriately conservative analysis based on available experimental information. In areas of incomplete knowledge, conservative assumptions or procedures must be applied. When further experimental information or improved calculational techniques become available, the conservatisms presently imposed will be reevaluated and a more realistic approach will be taken.

In order to hold the AEC abused its discretion, therefore, it would be necessary to find that the lack of unanimity among researchers and the absence of direct, as opposed to extrapolated, empirical support for the adequacy of ECCSs meeting the IAC clearly required the withholding of full-power licenses. That would effectively deprive the Commission of any discretion. Since unanimity among scientists, let alone empirical certainty, is rarely to be expected, it would substitute not even our, but the Union of Concerned Scientists’, judgment for that of the AEC.

IV

Five particular rulings of the ASLB limiting the scope of its inquiry by excluding or restricting UCS’s safety contentions are presented for review. The correctness of each ruling depends on whether the underlying contention challenged Pilgrim’s ability to comply with AEC regulations, as petitioner claims, or questioned the validity of the regulations themselves, as the AEC maintains. The target of a particular challenge may be a question because of the peculiar nature of certain AEC regulations, incorporating as they do a variety of factual assumptions. Since the regulation putatively under challenge in these rulings is the IAC, we preface our particularized analysis with a short explanation of the IAC.

There are four interim general criteria applicable to all light-water power reactors, and one evaluation criterion for determining whether a specific reac tor meets them. These criteria in their entirety provide:

IY. ... A. Criteria for all light-water power reactors.

These general requirements have been the basis of AEC safety review for some time. On the basis of today’s knowledge, the performance of the emergency core cooling system is judged to be acceptable if the calculated course of the loss-of-coolant accident is limited as follows:

1. The calculated maximum fuel element cladding temperature does not exceed 2,300°F. This limit has been chosen on the basis of available data on embrittlement and possible subsequent shattering of the cladding. The results of further detailed experiments could be the basis for future revision of this limit.

2. The amount of fuel element cladding that reacts chemically with water or steam does not exceed 1 percent of the total amount of .cladding in the reactor.

3. The clad temperature transient is terminated at a time when the core geometry is still amenable to cooling, and before the cladding is so embrittled as to fail during or after quenching.

4. The core temperature is reduced and decay heat is removed for an extended period of time, as required by the long-lived radioactivity remaining in the core.

B. Criteria for specific reactors. Each reactor shall be evaluated in accordance with the general criteria of section IV.A., and using a suitable evaluation model. Examples of acceptable evaluation models are der scribed in Appendix A. These evaluation models are acceptable to the Commission but their use is not mandatory. Other evaluation models may be proposed by applicants for review in individual cases. [Footnotes omitted]

The evaluation models referred to in paragraph IV.B. involve numerous assumptions that the AEC deems adequately conservative. In addition, the AEC specified certain variations from the models generated by reactor vendors. Thus, the General Electric Evaluation Model applicable to Pilgrim is subject to six qualifications, a typical one of which is, “4. It should be assumed that channel wetting does not occur until 60 seconds following the wetting time calculated using the Yamanouehi analysis.” 42

In promulgating criteria and evaluation model standards, the AEC has, in effect, made two statements — first, that a reactor whose ECCS meets the criteria will control a LOCA and is, therefore, safe for operation; and second, that it will, in fact ■ meet the criteria if its design features stand up under evaluation of a particular sort 43 A challenge that in form states that Pilgrim does not comply with the IAC, therefore, may in fact be any of three different types of statements, depending upon the reason given for the challenge. A simple example based on the IAC may be useful.

Consider these statements in a hypothetical regulation:

1. A reactor the maximum fuel element cladding temperature of which does not exceed 2300 °F will not be susceptible to dangerous embrittlement; and

2. A reactor of design XYZ will not generate a cladding temperature in excess of 2300 °F.

The general contention that the Pilgrim reactor is incapable of preventing excessive embrittlement may be based on any of three sorts of reasons:

1. The reactor does not conform to design specifications XYZ;

2. The reactor, although it is of XYZ design, cannot in practice prevent embrittlement, i. e., XYZ design will not prevent temperature in excess of 2300 °F; and

3. The reactor, although of XYZ design, and although able to keep the temperature from exceeding 2300 °F, is liable to embrittlement, i. e., a temperature of 2300°F is too high to prevent embrittlement.

A contention based on reasons of type 1 is indeed an objection to Pilgrim’s ability to comply with the regulation. Since the regulation incorporates the factual statements contradicted in reasons of type 2 and 3, however, they constitute challenges to the validity of the regulation itself.

Under the AEC’s Calvert Cliffs doctrine, the distinction illustrated above was determinative of whether an ASLB would hear a particular contention or certify it to the Commission. In the setting of this case, of course, the latter alternative was not certification but simply rejection, with the reminder that the rule challenge would be heard in the rule making forum. We turn now to the question of whether certain of UCS’s contentions were properly eliminated from the ASLB hearing on the ground that they are challenges to the IAC.

A. Contentions No. 2(a) and 16. Contention No. 16 of the intervention petition was first conditionally accepted as a matter to be heard, depending upon whether it was to be pursued as a compliance or rule objection. As reiterated in petitioner’s initial detailed specification (IDS) of issues and re-numbered 2(a), it states:

2. The Applicant has misanalyzed the proportions of the maximum hypothetical accident that can be assumed to occur for the Applicant’s proposed station.

(a) The maximum hypothetical accident set forth in the Pilgrim Station FSAR is a LOCA. The LOCA assumes that the [ECCS] is of a size and capacity sufficient to prevent an uncontrolled meltdown of the majority of nuclear fuel. There is no such reasonable assurance that the [ECCS] can function as quickly as necessary or with the reliability that is demanded if such a meltdown is to be avoided.

This contention is obviously quite general, since no reasons for the ECCS’s inadequacy are alleged; hence the conditional ruling. Another contention in the IDS challenged the ECCS’s adequacy in somewhat more specific terms. Contention No. 16 states:

16.a. The maximum cladding temperature that would be reached during the design basis LOCA at Pilgrim Station will exceed 2300° F.

b. The amount of fuel cladding reacting with steam during the design basis LOCA at Pilgrim Station will exceed one (1 %) per cent.

c. The clad temperature transient during the design basis LOCA at Pilgrim Station will not be terminated while core geometry is still amenable to cooling.

UCS repeatedly stated that these contentions challenged Pilgrim’s ability to comply with three of the IAC, and that UCS questioned the IAC themselves only in the alternative. The only proof UCS offered, however, concerned a study, conducted by one of its members, of a boiling water reactor test procedure “used in substantial support of the Applicant’s position that the Pilgrim nuclear power station meets the Interim Acceptance Criteria.” JA 560. The test procedure assigned as erroneous, however, was part of the General Electric Evaluation Model. As such, any attempt to demonstrate its invalidity was correctly held to be an attack on the IAC and not on Pilgrim’s compliance therewith. Indeed, no evidence whatsoever specific to Pilgrim was offered in support of these contentions. 44

B. Contention No. 32(a). After the above-referenced contentions had been ruled out of the hearing, UCS moved to file nine additional detailed specifications allegedly challenging Edison’s application of the General Electric Evaluation Model as improper. The motion was denied in its entirety because the ASLB viewed the additional contentions as being duplicative of the prior challenges to the IAC and improper for the same reason. Only one of the rejected contentions is set before this court in terms, but it is said to be illustrative. That contention questions “whether the evaluation model used was precisely what the Commission was referring to in Appendix A, Part 2, of the Interim Policy Statement” announcing the IAC.

In the July, 1971, letter transmitting its report on the Pilgrim ECCS evaluation, Edison had represented that “[t]he Pilgrim ECCS performance has been evaluated in strict accordance with the model described in Appendix A, Part 2 of the AEC policy statement.” Petitioner’s objection seems to be that the model does not provide an adequate technical description of the methods it uses, that Edison’s submission was accepted because the results were within an acceptable range, and that these results cannot be meaningfully analyzed without more methodological information.

Whatever deficiency the model may suffer from because of generality, the merits of the model are not an issue in this court. The only issue is whether the demand for greater specificity calls in question the model itself or Edison’s compliance therewith. We have no basis on which to doubt the Commission’s uncontradicted representation that application of the model requires the use of certain analytical techniques, but does not require any particular method of computation. Taking the model to allow a degree of computational flexibility, we cannot but agree that insistence on greater specificity constitutes a challenge to the model — and thus the IAC— rather than to Edison’s compliance. The question is a close one because, on the unelaborated record before us, the distinction between analytical techniques and computational methods appears to be a fine one. It is the agency’s task to resolve close questions in a reasoned manner, however, and we are far from able to say that the AEC’s resolution of this one was unreasonable.

C. Contention No. 2(b). In the FSAR on Pilgrim, the maximum hypothetical accident analyzed is a controlled LOCA. In contention 2(b), petitioner urged that the “true maximum hypothetical accident that can be assumed for this reactor [is] a meltdown of the entire fuel core, with subsequent breaching of the containment due to penetration of the melted fuel and subsequent interaction with ground water, releasing radioactivity in quantities many orders of magnitude above the release set forth in the [FSAR].”

This contention was rejected as a challenge to the IAC for the simple reason that the accident UCS would postulate could only occur upon failure of the ECCS. If the criteria are met, the ECCS is presumed to be effective, in which case a LOCA would not escalate into a meltdown of the fuel core.

Petitioner points to the AEC’s Reactor Site Criteria, 10 C.F.R. 100.11(a), in support of its view. That regulation provides that, in evaluating a proposed site in terms of potential accident damage from radiation, “an applicant should assume a fission product release” such as would result from the maximum credible accident occurring. “Such accidents,” it is explained in a footnote, “have generally been assumed to result in substantial meltdown of the core with subsequent release of appreciable quantities of fission products.”

Again we think petitioner misapprehends the issue here open to review, namely, the ruling that its contention was a rule-challenge beyond the ASLB’s authority to entertain on the merits. Petitioner is correct in noting that conformance with the IAC does not establish conformance with the Reactor Site Criteria. But conformance with the latter regulation is clearly found in the Staff Safety Evaluation, which states (JA 503):

In calculating the potential consequences of the postulated [LOCA], to provide a conservative estimate we have assumed that in spite of the operation of the [ECCS], large amounts of fission products would be released from the reactor fuel.

It was open to petitioner to challenge the manner in which the Staff performed this site criteria analysis, but that would raise a question quite different from the maximum credible accident for purposes of analyzing ECCS performance. The AEC has chosen to employ a most conservative (drastic) assumption in determining site suitability because site selection is the most critical decision. Once a site has been approved, however, it is entitled to indulge more realistic assumptions, such as the assumption that an ECCS meeting the IAC will work effectively. That assumption may be wrong, but the forum for challenging it was correctly held not to be the licensing hearing but the rule making.

D. Contention No. 6. This contention challenged Edison’s computation of the maximum containment pressure expected to obtain during a LOCA: Edison used “equilibrium assumptions concerning the thermodynamic behavior of the coolant, whereas non-equilibrium assumptions in the calculation — that indicate much higher containment pressure in the event of blowdown — should have been used.”

Petitioner objects to elimination of this contention on procedural grounds. The ASLB “initially regarded the contention as a discrete issue concerning the assumptions underlying engineering design of a safety system,” but Edison argued it was a challenge to the IAC on the ground that equilibrium assumptions were sanctioned by the evaluation model deemed acceptable therein. The ASLB therefore reserved the issue and allowed Edison to establish the factual basis for its position. 45

At the prehearing conference held a few days thereafter, counsel for Edison circulated what petitioner disparagingly refers to as “unsworn letter signed by him containing various factual allegations purporting to show that Edison has complied with the [IAC] in calculating the maximum containment pressure.” The ASLB thereafter ruled that Contention No. 6 was a rule-challenge.

Petitioner objects to the ASLB’s reliance on the unsworn letter in question, removing No. 6 from the hearing, with the result that “pre-trial” discovery, cross-examination and rebuttal was correctively restricted. Even the most casual reading of the “letter,” which Edison here characterizes as an amended response to Contention No. 6, defeats petitioner’s argument. The letter does no more than identify that portion of the evaluation model described in NEDO10329 on which Edison relied in using equilibrium assumptions. Since the model is incorporated by reference into the IAC, nothing would prevent the ASLB from noticing the cited section on its own. We are aware of no requirement, moreover, that counsel’s statements at a prehearing conference be under oath, and petitioner cites to none. Petitioner’s own counsel was present, but did not question the contents of the “letter.” Its objection now is monumentally unimpressive.

E. Scape' of cross-examination concerning Contention No. 9. This contention, dealing with the adequacy of means provided to keep the hydrogen that evolves during a LOCA at a sub-explosive level of concentration, was the sole contested issue before the ASLB. Petitioner did not offer evidence on the issue, intending to prove its case by cross-examination of Edison and Staff witnesses. When UCS set out to do so, a disagreement about the scope of No. 9 arose. The Staff, with whom Edison agreed, argued that the quanity of hydrogen in question must be the amount expected if the ECCS functions properly, since hypothesizing the greater amount would require the ASLB to assume that an ECCS meeting the I AC would not work. The Board agreed, as we think it was required to do, and limited the inquiry to the question of hydrogen control during a controlled LOCA. The logic of its position is manifest, even unaided by the deference normally due to agency evidentiary rulings.

V

Petitioner’s final objection warranting detailed comment concerns the adequacy of the AEC’s analysis of the environmental impact of Pilgrim’s operation. The regulatory framework within which nuclear power stations operate injects a novel element into the projection of actual environmental costs reasonably to be anticipated. The problem that concerns us here is the AEC’s selection, from a number of possibilities, of a standard for the accurate measure of liquid and gaseous radioactive effluents.

For the purposes of environmental analysis, the AEC projected the release of five curies (Ci)/year for liquid effluents, and 60,000 microcuries (.06 mCi)/second for gaseous effluents. 46 USC objects that these estimates represent “design objectives” for Pilgrim’s radwaste control system — the levels that would result if the reactor is operated at all times to achieve the full potential of the control system — and that experience makes reliance upon design objectives being achieved highly problematic. In support of this contention, petitioner points out that the technical specifications of Pilgrim’s license themselves permit, and thus presumably contemplate, significantly higher release rates. 47

We think it obvious that Pilgrim’s anticipated environmental effect from effluents depends upon its actual anticipated release rate, and not upon either its design objectives or maximally permissible rates. Further, the best basis from which to project actual effect is clearly any actual experience with similar reactors, adjusted to reflect material differences in Pilgrim’s design and circumstances. To the extent that new equipment or procedures are involved, of course, this comparative exercise becomes increasingly less practicable and resort must be had to the Staff’s evaluation of their efficacy.

The Staff’s analysis of gaseous effluents, for which comparative data were available, reflects precisely this order of preference. First, the FES describes Pilgrim’s initial gaseous treatment system in some detail, and then concludes, “On the basis of operating experience with power reactors of similar design, it is expected that the off-gas system described, above will keep releases of gaseous radioactive wastes well within the limits specified in 10 CFR Part 20.” 48 The specific estimate of 60,000 mCi/seeond, it continues, ■

is made on the basis of experience at other operating plants with similar gaseous waste systems and takes into consideration a probable lower than average load factor during early operation and few fuel failures at low power levels during start-up of the facility.

The relevance of similar experience was limited to projecting initial gaseous release because an augmented gaseous treatment system was expected to be operational by January 1, 1974. Apparently because the Staff lacked experience with a system comparable to the augmented system, estimated releases for 1974 and beyond were based upon its evaluation of the new equipment.

Petitioner’s objection that these analyses, insofar as they are not based on operating experience elsewhere, unrealistically assumed that control equipment would work at design objectives is not well-supported in the record. When the AEC reports, for example, that the liq-' uid waste treatment system “appears to be capable of reducing discharges to Cape Cod Bay to less than 5 Ci/yr,” its meaning seems to be that a 5 Ci release can in fact be expected, not that 5 Ci is the best that can be hoped for. While 5 Ci may in fact represent the full potential of the equipment, the AEC is not precluded from projecting realization of that potential in the absence of either an experimental or probabilistic basis for doubting its realization. Petitioner does not even suggest such a basis for departing from the 5 Ci estimate, and it would not be reasonable to require the AEC to pose implausible reasons for doing so only in order to show that they are implausible.

The objection based upon the discrepancy between the average estimates made in the FES and the maxima found in the technical specifications is weightier. As noted, supra note 48, 10 C.F.R. § 20.106 and Appendix B specify maximum allowable concentrations of radioactive materials. Another regulation provides that each operating license must include technical specifications requiring compliance with § 20.106, procedures for the control of effluents, use of a radiological waste system, and periodic reporting of the actual quantity of radio-nuclides released to unrestricted areas. 10 C.F.R. § 50.36a. On the basis of experience, the AEC expects compliance with these required technical specifications actually to “keep average annual releases of radioactive material in effluents at small percentages of the limits specified in § 20.106” and the operating license. Still, it does not expect release rates to be constant at every given moment. As the regulation explains:

[T]he licensee is permitted the flexibility of operation, consistent with considerations of health and safety, to assure that the public is provided a dependable source of power even under unusual operating conditions which may temporarily result in releases higher than [the small percentages of the limits specified in § 20.106 that would occur under normal operating conditions] but still within the limits specified in § 20.106 and the operating license. It is expected that in using this operational flexibility under unusual operating circumstances, the licensee will exert his best efforts to keep levels of radioactive material in effluents as low as practicable.

It is clear, both from the face of this regulation and experience as portrayed in the record, that it would be grossly unrealistic to project release rates based on continuous operation at the maximum rate permissible at any moment, i. e., the maximum set in the technical specifications, rather than at average rates realistically estimated. 49 But it might be equally unrealistic, depending upon experience elsewhere and factors peculiar to Pilgrim, to assume that Pilgrim would never encounter a continuing need of some duration to discharge radioactive waste at average rates in excess of normal operating levels at or near the maximum of its technical specifications. We are not able to say from the present FES, however, that this is not what the AEC has done, and indeed the contrary appears to be true.

On the one hand, the Commission’s stress on the need for expressing effluent release rates in terms of their average over a calendar quarter in order to take account of necessary variations over short periods suggests that some expected number of “operational occurrences” — including fluctuations up to the maxima — may have been factored into the derived estimated average. On the other hand, we are told that “[temporary (for an obviously unknown period of time) attainment of any . . . particular level [other than the average] is speculative and impractical to calculate, and hence need not be addressed.” This suggests that the probability of periods of continuing excess releases was not factored into the estimated average quarterly release rates.

Considering the unpredictable nature of some of the factors the Commission considers in deciding whether and for how long to tolerate, rather than require correction of, excess releases — the need for power, public health and safety, the licensee’s efforts at correction — it would actually be surprising if it had attempted to incorporate the probability of their occurrence into its estimate of environmental impact. Moreover, the ASLAB, in rejecting petitioner’s contention that the FES should reflect the impact of the maximum excess releases that the AEC intends to tolerate, stated that

if these circumstances change so that releases during normal operation would be significantly increased beyond the level expected by the staff, the staff must consider the environmental impact of such increases before authorizing them.

Petitioner is dissatisfied with this expression of intent to consider the impact of excess releases as they actually arise, rather than on a probabilistic basis at the operating license stage, because it does not, as they read it, comply with NEPA. Specifically, UCS complains that it does not require public disclosure and consideration of alternatives. We do not agree.

By disregarding the environmental impact of licensing Pilgrim insofar as that impact depends upon subsequent events and AEC decisions of a highly speculative nature, the Commission does not, nor could it, remove from NEPA’s reach such decisions as would otherwise come within the ambit of that Act. What we said with respect to the equally conjectural possibility of a subsequent derating decision, Part III, supra, applies with equal force to a later decision to authorize a significant increase in Pilgrim’s release of radioactive effluents. In addition, however, we think the AEC could discharge its obligation under NEPA by examining the environmental impact of increased release rates at any such time as the probability and magnitude thereof became sufficiently quantifiable to yield to meaningful analysis. That the facts were too speculative at the time of ■ licensing, and consideration could lawfully await the time when particular facts arise to require discrete decision, does not preclude making a single probabilistic assessment of environmental impact at an appropriate time, any more than such an assessment would have been in appropriate in the FES had the facts been less elusive. 50

YI

Petitioner’s other objections involve the ASLB’s admission of the FSAR into evidence, and the asserted failure of its Initial Decision to rule on each of petitioner’s proposed findings and rulings, as required by the Administrative Procedure Act, 5 U.S.C. § 557(c), and 42 U.S.C. § 2231. The evidentiary point is founded upon 10 C.F.R. § 2.43(c), which provides, inter alias that only “reliable” evidence be admitted. Because the professional qualifications of certain of the Edison and other non-agency employees who worked on preparation of the FSAR were not established in the record, UCS contends that no foundation for reliability was provided.

The argument is specious. By 10 C. F.R. § 50.34(b), the FSAR must be included in the license application. As then in force, 10 C.F.R. 2.743(g), 27 F. R. 377 (Jan. 13, 1962), provided that “[i]n any proceeding involving an application, there shall be offered in evidence the record of the application, including the application. . . . ” Even in the absence of such a provision, we do not see how reliability can be established prior to at least conditional admission in a proceeding in which reliability is the ultimate issue.

The APA objection is no more helpful to petitioner. Reasoned findings, “which alone make effective judicial review possible,” Baltimore & O. R. R. v. Aberdeen & Rockfish R. R., 393 U.S. 87, 92, 89 S.Ct. 280, 283, 21 L.Ed.2d 219 (1968), are not lacking here. We stop only to note that petitioner identifies not a single finding or conclusion the basis for which it finds unclear, but relies instead upon the ASLB’s statement that all proposed findings and conclusions not incorporated in its decision were rejected for want of “reliable, probative and substantial” support, or as being unnecessary. Without pointing “to any material issue, or to any group of minor matters that may have cumulative significance” about which “the parties and the court should not be left to guess,” Radio Station KFH Co. v. FCC, 101 U.S.App.D.C. 164, 247 F.2d 570, 572 (1957), petitioner cannot make out a violation of the APA.

It is patent from the foregoing that, despite the procedural nature of the issues raised, we have had to deal with subject matter ranging far beyond the normal ken of judges. This case is only one among many such these days, and, thus, suggests the extraordinary tasks that the seemingly endless statutory proliferation of judicial review of agency action is imposing on the federal courts. How meaningful judicial review on this scale can be over any sustained period of time, at least if not invariably invoked with an acute sense of responsibility and a willingness to recognize that many of these more esoteric battles must largely be won or lost at the agency level, is problematic. In this instance, at any rate, we cannot say that the agency fell below the legal standards that it was required to meet. The petition for review is, accordingly, denied.

It is so ordered.

1 . The ACRS is a body of up to fifteen experts drawn from without the government for the purpose of advising the Commission on “the hazards of proposed or existing reactors and the adequacy of proposed reactor safety standards.” 42 U.S.C. § 2039. The ACRS must review and report on each application for a commercial or industrial construction permit or operating license, and may be asked by the AEC to review and report on any amendments thereto or on licenses for facilities to engage in medical therapy or research and development. 42 U.S.C. § 2232(b). Its reports are made part of the record of the application and are made available to the public except to the extent that a security classification intervenes.

2 . An ASLB is “comprised of three members, one of whom shall be qualified in the conduct of administrative proceedings and two of whom shall have such technical or other qualifications as the Commission deems appropriate to the issues to be decided.” For each application an ASLB is appointed from a panel of qualified members to conduct hearings, render decisions, and perform such other regulatory functions as the Commission delegates to it. 42 U.S.C. § 2241(a). Board members may be appointed from within or without the Commission. Id., at § (b).

3 . Before issuing a construction permit, the AEC must find, on the basis of specified submissions dealing with safety features and analysis, that “there is reasonable assurance that . . . the proposed facility can be constructed and operated at the proposed location without undue risk to the health and safety of the public.” 10 C.F.R. § 50.35(a) (1973).

4 . 10 C.F.R. § 50.30(d). A preliminary safety analysis report had accompanied the application for a construction permit, pursuant to 10 C.F.R. § 50.34(a).

5 . Neither Staff nor ACRS approval is required by law for the AEC to issue a license, but in practice it is very unlikely that an applicant would persist in going before an ASLB over their objection. Murphy, Atomic Safety and Licensing Boards: An Experiment in Administrative Decision Making on Safety Questions, 33 Law & Contemp. Prob. 566, 569, 579 n. 56 (1968). For a case in which a construction permit was issued despite an adverse ACRS report, see Power Reactor Development Co. v. Int’l Union of Elec. Workers, 367 U.S. 396, 400, 81 S.Ct. 1529, 6 L.Ed.2d 924 (1961).

6 . These reports, together with the FSAR, were submitted to state and local bodies of apposite jurisdiction pursuant to 42 U.S.C. § 2232(b) and 10 C.F.R. § 2.102(c).

7 . 36 F.R. 13287 (July 17, 1971). UCS and Sierra were admitted as a party (Joint Intervenors), as was the Commonwealth of Massachusetts. Only UCS is before us claiming error.

8 . 10 C.F.R. § 2.714(a), (c) ; see 36 F.R. 13287 (July 17, 1971) (Notice of Hearing).

9 . Contentions relating to liquid and gaseous effluents and environmental considerations were raised separately and their consideration deferred pending receipt of additional information from Edison and the Staff’s Detailed Statement of Environmental Considerations. ASLB Memorandum, Oct. 28, 1971, JA 142, 160.

10 . 36 F.R. 22774 (Nov. 30, 1971). See Criteria for Emergency Core Cooling Systems for Light-Water Power Reactors, 36 F.R. 12247 (June 29, 1971) (IAC).

11 . Section 103(a) of the Act, 42 U.S.C. § '2133(a), authorizes the Commission to issue licenses “subject to such conditions as the Commission may by rule or regulation establish . . . . ” Section 103(d) adds as one “limitation” on this authority:

In any event, no license may be issued to any person within the United States if, in the opinion of the Commission, the issuance of a license to such person would be inimical to the common defense and security or to the health and safety of the public.

See 42 U.S.C. § 2232(a). The regulation governing the issuance of an operating license, 10 C.F.R. § 50.57 (a), provides that the Commission will first find that

(3) There is reasonable assurance (i) that the activities authorized by the operating license can be conducted without endangering the health and safety of the public

(6) The issuance of the license will not be inimical to the common defense and security or to the health and safety of the public.

12 . 10 C.F.R. Part 2, App. A (Statement of General Policy) VI(d), as added Sept. 23, 1966 (31 F.R. 12774) (emphasis added). The rules have since been amended specifically to provide that “[i]n an operating license proceeding, the board will determine only the matters in controversy among the parties as the issues to be decided,” id. at VIII (b), as amended, 37 F.R. 15127 (July 21, 1972) ; and “[t]he board, in an operating license proceeding, will make findings only on the matters in controversy.” Id. at VIII (c) (emphasis added).

13 . The Joint Committee was established in 1954 to oversee developments under the 1954 Act for Congress. 42 U.S.C. §§ 2251-2257. The Committee’s concern was aroused in 1956 by the AEC’s issuance, despite the (non-public) ACRS report, of the construction permit litigated in Power Reactor, supra. See W. Berman & L. Hydeman, The Atomic Energy Commission and Regulating Nuclear Facilities in 2 Joint Comm, on Atomic Energy, 87th Cong., 1st Sess., Improving the AEC Regulatory Process 425, 447-454 (Jt. Comm. Print 1961) (hereinafter cited as JCAE Staff Study).

14 . See Staff of the Joint Comm, on Atomic Energy, 85th Cong., 1st Sess., A Study of AEC Procedures and Organization in the Licensing of Reactor Facilities (Jt. Comm. Print 1957).

15 . The committee staff’s report appears in 1 JCAE Staff Study, supra note 13; that of the AEC, The Regulatory Program of the Atomic Energy Commission in 2 JCAE Staff Study, supra note 13, at 87-393; and AEC, A Report on the Regulatory Program of the Atomic Energy Commission, id. at 395-422. See Cavers, Administrative Decisionmaking in Nuclear Facilities Licensing, 110 U.Pa.L.Rev. 330, 332-34 (1962).

16 . 1 JCAE Staff Study, supra note 13, at 69.

17 . 2 JCAE Staff Study, supra note 13, at 419.

18 . 10 C.F.R. Part 2 (1962), as amended, 10 C.F.R. Part 2 and App. A. (1966).

19 . 10 C.F.R. Part 2, App. A. IY(c), as amended, 10 C.F.R. Part 2, App. A. VI (c) (1973).

20 . UCS is apparently under the misapprehension that appointment of an ASLB to make the safety findings required before issuance of a license occurs only in contested cases. Petitioner’s Reply Brief at 7. In fact, however, the Rules of Practice for ASLBs contemplate that many cases in which an ASLB is convened will be uncontested. See 10 C.P.R. Part 2, App. A (Introduction), as amended, 10 C.P.R. Part 2, App. A.

21 . Hearings on Licensing and Regulation of Nuclear Reactors Before the Joint Comm, on Atomic Energy, 90th Cong., 1st Sess., pt. 1, at 151 (testimony of Dr. David B. Hall, Los Alamos Scientific Laboratory), 157 (statement of Warren E. Nyer, Vice Chmn., ASLB Panel), 167 (statement of Samuel W. Jensch, ASLB Panel) (1967).

22 . 42 U.S.C. § 2239 and 10 C.F.R. § 50.58 appear to limit this authority to amendments to construction and operating licenses. The Senate Report, however, unequivocally contemplates the same discretion in the “issuanee of an operating license” as well. In any event, it is a power to be “exercised with great care.” S.Rep.No. 1677, supra at 8.

23 . See 10 C.F.R. Part 60, App. D (1973).

24 . L. Brandéis, Other People’s Money 92 (1914).

25 . See also Hearings on AEC Licensing Procedure and Related Legislation Before the Subcomm. on Legislation of the Joint Comm, on Atomic Energy, 92d Cong., 1st Sess., pt. I at 468-69 (comments by the Environmental Defense Fund, Inc., with respect to proposed amendments to 10 C.F.R. Part 2, Appendix A, before the AEC), <& passim; pt. 4 at 1478-1501 (reprinting Murphy, supra note 5) (1971).

The Atomic Energy Act was again amended in 1972 to add a new Section 192, 42 U. S.C. § 2242, authorizing the issuance of temporary operating licenses under specified circumstances to alleviate power shortages without awaiting the outcome of often protracted adversary hearings. See H.Rep.No. 1027, 92d Cong., 2d Sess. (1972). The report indicates not only awareness, but implicit approval, of the AEO’s requirement that intervenors raise specific issues for resolution at the operating hearing. Id. at 5-6. That requirement is largely inconsistent with the notion of de novo review by the ASLB.

26 . An evaluation model is a computational procedure incorporating certain assumptions. AEC acceptance of a given model necessarily entails its endorsement of the assumptions involved as reasonably conservative.

27 . Loss of coolant accidents are defined as “those accidents that result from the loss of reactor coolant at a rate in excess of the capability of the reactor coolant makeup system” as a result of certain breaks. 10 C.F.R. Part 60, App. A. (Definitions and Explanations) (1973).

28 . See Acceptance Criteria for [ECCS] Light Water-Cooled Nuclear Power Reactors, 10 C.F.R. Part 50 (1974); 39 F.R. 1001 (Jan. 4, 1974). By order dated Jan. 16, 1974, this court directed the parties and intervenor to submit memoranda on the question of the effect of the AEC decision on the pendency of this case.

29 . See Freedman, Summary Action by Administrative Agencies, 40 U.Chi.L.Rev. 1 (1972) ; see Arnett v. Kennedy, 416 U.S. 134, 94 S.Ct. 1633, 40 L.Ed.2d 15 (1974).

30 . According to Freeman, The Emerging Role of Rule Making in the AEC Decisional Process (What We Can Learn from the ECOS Rule Making), 14 Atomic Energy L.J. 144, 147 (1972), in 1971 “the same AEC regulations were being challenged on the same ground in numerous AEC licensing cases, often without the presence of witnesses and parties necessary to their meaningful resolution and often in an atmosphere of chaos and disorder.” At the instance of the industry and environmentalists alike, the Commission turned to rule making as an alternative to this mounting duplication and delay. As Judge Friendly recently had occasion to remark, such resort to generic proceedings “would seem to benefit all parties, particularly the poorly-financed environmental groups.” Ecology Action v. AEC, 492 F.2d 998, 1002 (2d Cir. 1974).

31 . Apparently the experiment has not been an unqualified success. See note 30, supra.

32 . The AEO has since promulgated a regulation in lieu of Calvert Cliffs in order to provide a procedure for requesting a waiver or exception to a rule or regulation that “would not [in the particular proceeding] serve the purposes for which [it] was adopted.” 10 C.F.R. § 2.758 (1973).

33 . This is not to say that operations will not be cost-justified from the standpoint of Edison, which could result in abandonment, but that taking account of costs “externalized” by Edison, particularly environmental costs, it might not be cost-justified from a societal point of view, which should result in the denial of a license.

34 . UCS also objects that the EES did not present responsible opposing scientific views of the expected environmental impact of licensing Pilgrim, ■ as required by Committee for Nuclear Responsibility, Inc. v. Seaborg, 149 U.S.App.D.C. 380, 463 F.2d 783, 787 (1971). Accord, Environmental Defense Fund, Inc. v. Corps of Engineers, U.S. Army, 325 F.Supp. 749, 759-760 (E.D.Ark. 1971); Environmental Defense Fund, Inc. v. Corps of Engineers, U.S. Army, 348 F.Supp. 916, 933 (N.D.Miss.1972); see Monroe County Conservation Council, Inc. v. Volpe, 472 F.2d 693, 698 (2d Cir. 1972). The views in question are said to be those of “at least some knowledgeable experts within the AEC [who] believed that the [IAC] did not provide the degree of emergency cooling assurance deemed necessary by the AEC.” Petr.R.Br. at 30.

Petitioner misreads the “responsible opposing views” requirement, and then misapplies it. The requirement that opposing views be aired is designed to implement environmental full disclosure by requiring publication of the sometimes embarrassing comments that must, under NEPA, be solicited from relevant federal, state, and local agencies, 42 U.S.C. § 4332(C), as each of the cited cases illustrates. In this way, “the officials making the ultimate decision, whether within or outside the agency” are aided in making a responsible choice. Seaborg, supra, 463 F.2d at 787. That policy would also extend, we think, to responsible views presented by a member of the public, see Environmental Defense Fund, supra, 325 F.Supp. at 759, as in the environmental hearings held in this case. But it does not mean that the agency staff cannot speak with one voice before the decision-maker, any more than each outside agency would have to acknowledge differences of opinion within itself in commenting on the action under consideration. Each remains free to do so, of course, but an opportunity for a “minority report” from dissidents from each agency is neither required nor practical in an executive organization.

Moreover, we are unconvinced that the views here referred to by petitioner are in fact environmental views. Saying that the project won’t work is not the same as saying that it will have greater environmental impact than the FES suggests,, even if such failure would have environmental consequences. The basic objection is to the feasibility, not the environmental impact.

35 . Even now, with the criteria in final form, the AEC can only estimate an average derating of 5 per cent and somewhat less for boiling water reactors such as Pilgrim. AEC Director of Regulation, Press Conference, May 9, 1973, as cited by AEC Br. at 27. n. 26.

36 . An amendment can be made without opportunity for a hearing if the AEC determines that it “involves no significant hazards consideration.” 42 U.S.C. § 2239(a). Whether a derating amendment would normally come within that exception to the hearing requirement is of no moment here because the AEC has positively represented to this court that “there will be an opportunity for a hearing in connection with changes in [Pilgrim’s] operating license conditions necessary to effectuate compliance with the new standards.” AEC Supp. Mem., supra note 28, at 3.

37 . “Compliance [with NEPA demands] that environmental issues be considered at every important stage in the decision making process ... where alterations might be made in the proposed action to minimize environmental costs.” Calvert Cliffs, supra, 449 F.2d at 1118. In the context of a derating license amendment, the proposal is to license operations at a lower level than theretofore. An alternative to be considered is complete abandonment of the project, just as it was at both the construction and full-power operating license stages. EES at 78, JA 721. As at those stages, sunk costs are not appropriately considered costs of abandonment, although replacement costs may be if construction of a substitute facility could reasonably be expected as a consequence of abandonment.

38 . The Senate Report speaks expressly of both “expansion or revision- of ongoing programs.” S.Rep.No. 296, 91st Cong., 1st Sess. 20 (1969) (emphasis added).

39 . For a useful and intelligible explanation of the role and operation of EOOS, see Lapp, Nuclear Salvation or Nuclear Folly?, N.Y. Times, Feb. 10, 1974, § 6 (Magazine), at 12.

40 . Neither this nor the following document relied upon are in the record of this case.

41 . Under the new criteria the maximum has been lowered to 2200° F, applied to the hottest region of the hottest fuel rod, and a new limit imposed on the maximum allowed local oxidation, to the same end. AEC RM-50-1, Opinion at 25-26, 28.

42 . Loss-of-Coolant Accident and Emergency Core Cooling Models for General Electric Boiling Water Reactors, NED0-10329. See 36 E.R. 12219, IAC App. A, Pt. 2 (June 29, 1971).

43 . Each statement is made with a high degree of confidence, but neither has yet proved true in practice — i. e., without assumptions — because there has never been a real LOCA.

44 . UCS’s offer of proof involved describing a test that consisted of heating “a simulated fuel bundle of the kind that is used in the Pilgrim nuclear power station,” and initiat ing spray cooling “simulating the kind . . . that would be available in the Pilgram reactor in the event of a postulated [LOCA].” JA 566. The effect of the cooling mechanism on the pressure in the pressure vessel at the time of spray initiation and thereafter was observed to be different from the effect expected. Hence, the “anomalous” results “were found by the investigators not to be satisfactory for use in heat transfer calculations in the General Electric analytical model.” JA 567. Earlier tests used in formulating the procedure of the General Electric model, therefore, “did not predict correctly.”

45 . ASLB Memorandum and Order of Oct. 28,1971, at 12, JA 153.

46 . The quantity of gaseous discharge was projected to decrease by at least a factor of forty upon completion of certain modifications, expected to occur by January 1, 1974.

47 . The maximum release rate of radioactive liquid effluents is set at 10 Oi/quarter, and that of gross gaseous activity at .10 mCi/second when averaged over a calendar quarter. Technical Specifications 3.8.A.2, 3.-8.B.3, Supp.No. 1 to Staff Safety Evaluation at 177, 179, JA 645, 647. The figures used in the FES are thus 12.5 per cent and 60 per cent respectively of the maximum found in the technical specifications.

48 . FES at 34, JA 696. Appendix B, Table II of the regulation cited in the text specifies the maximum concentration of a host of radioactive materials allowable in any unrestricted area. An unrestricted area is defined as any area not controlled by the licensee for purposes of protecting individuals from exposure to radiation, and any area used for residential quarters. 10 C.F.R. § 20.3(a) (17).

49 . UCS has argued that the BBS should either assume continuous release at the technical specification maxima, or else that those maxima should be reduced to the levels used in the FES. The latter suggestion involves relief not available in this forum.

50 . Were the AEG to issue rules, for example, governing the conditions under and upon which excess releases will be authorized, analysis of the expected environmental impact thereof would aggregate impact of the anticipated'dndividual decisions.