Opinion for the Court filed by Circuit Judge GINSBURG.*

Dissenting opinion filed by Senior Circuit Judge BAZELON.

The authority of the Federal Energy Regulatory Commission (FERC or Commission 1) to license the construction of a hydroelectric plant on a body of water subject to federal jurisdiction is limited by two federal statutes: the Federal Power Act (FPA), 16 U.S.C. §§ 791a-823a (1976 & Supp. V 1981), and the National Environmental Policy Act (NEPA), 42 U.S.C. §§ 4321-4361 (1976). In this petition for review, Friends of the River and Dale Meyer (FOR or petitioners) challenge as inconsonant with both statutes FERC’s license to the Calaveras County Water District (CCWD) to operate a hydroelectric plant on the North Fork Stanislaus River in California. We find that in one respect FERC did not attend with due care to NEPA’s procedural requirements. Because we conclude that the Commission’s decision is in other respects unassailable, and because a remand to cure the procedural lapse would serve no sensible purpose, we affirm the Commission’s action.

I. Background

In November 1978, CCWD, a California county water district, applied to FERC for a license to construct a hydroelectric plant (the project or the Calaveras project) on the North Fork Stanislaus River in Alpine, Tuo*332lumne, and Calaveras Counties, California.2 The project, as described in CCWD’s application, involves the construction of a series of dams, the enlargement of an existing reservoir, and the construction of two new powerhouses. As planned, the project will produce an average annual energy yield of over 500 gigawatt-hours (Gwh), equivalent to the energy produced by roughly one million barrels of oil, and have a total power capacity of about 200 megawatts (MW).3

The principal ecological effect of the plant will stem from the enlargement of the Spicer River Reservoir, which at its maximum elevation will inundate approximately 1,780 acres of forest, meadow, and riparian habitat. While the reservoir will be moderate in size compared to nearby water projects, the loss of Gabbot Meadow, part of the inundated area, is likely to have a negative effect on deer and other wildlife populations. In its total dimension, the project will mean the loss of about 2,500 acres of wildlife habitat.

The power generated by the project is to be sold to the Northern California Power Agency (NCPA).4 NCPA is a consortium of twelve small utilities, each responsible for the retail distribution of electrical power within its service area. At the time of the 1978 application, the NCPA utilities did not generate any of the power they supplied to their customers; instead, they relied entirely on power purchased from Pacific Gas and Electric Company (PG & E), the largest electrical utility in California, and from the Central Valley Project (CVP), also located in California, of the federal Western Area *333Power Administration. NCPA, and a number of other utilities, including CVP, are bundled together with PG & E to compose a “PG & E area” for purposes of statewide planning by the California Energy Commission (CEC). Electricity Tomorrow 49. NCPA’s relationship with PG & E, however, is purely contractual. PG & E’s charter does not place it under a legal obligation to supply power to NCPA or to customers within NCPA’s service area. Rather, PG & E’s priority charge is to supply customers within its own service area as defined by its charter, an area distinct from NCPA’s.

NCPA intervened in the licensing proceedings before FERC immediately after CCWD filed the 1978 application. For NCPA, the project represents an important step towards energy self-sufficiency in the years to come. For CCWD, the project means an increase in the firm yield of water on the North Fork Stanislaus River; in addition, revenue generated by the project can be applied to the cost of the water CCWD provides, and can be used, in the future, to finance additional water projects in the district.

Litigation aimed at stopping the project began soon after CCWD filed the 1978 license application, when FOR filed suit in state court claiming that the environmental impact report filed by CCWD pursuant to the California Environmental Quality Act did not satisfy the requirements of that Act. In 1979, judgment was entered in California Superior Court rejecting all claims made by environmental plaintiffs.5

FERC commenced its investigation of the desirability of the Calaveras project in 1978. In late 1979, it issued a draft environmental impact statement (EIS) setting forth the environmental impacts of the project. It received extensive comments on the draft EIS from federal and state agencies as well as private parties, and issued a revised, final EIS in 1980. That document accorded only summary attention to the concern— raised before the agency — whether, in lieu of the project, NCPA’s power needs could be met by continued reliance on purchases from other utilities. The Commission did state in the EIS that power purchase contracts between NCPA members and PG & E “can be terminated on 6 to 24 months notice and cannot be considered as firm resources,” J.A. 187; it further stated that resort to other sources of energy, including purchased power, in preference to licensing the hydroelectric Calaveras project, “would consume nonrenewable fossil-fuel resources.” J.A. 328. FERC took testimony commenting on the final EIS at public hearings in Washington, D.C. and Angels Camp, California.

In February 1982, FERC issued an order and opinion granting CCWD a license to construct the Calaveras plant. It denied FOR’s petition for rehearing in an opinion issued in July 1982. The July order explained more cogently why the Commission rejected prospects for further reliance on purchased power as a ground for refusing the license. FERC observed that “PG & E’s [own] ability to meet load in the future is uncertain,” J.A. 960, and cited studies, in Electricity Tomorrow and a later (March 1981) CEC staff report, supporting that conclusion.

Petitioning for our review of the Commission’s grant of the license, FOR contends that the Commission has neglected its duties both under the FPA and under NEPA; we consider separately below FOR’s challenges under each statute.

II.' FPA Claims

Section 10(a) of the Federal Power Act provides that the Commission may issue *334a license for a hydroelectric plant only where

the project ... shall be such as in the judgment of the Commission will be best adapted to a comprehensive plan for improving or developing a waterway or waterways for the use or benefit of interstate or foreign commerce, for the improvement and utilization of waterpower development, and for other beneficial public uses, including recreational purposes.

Federal Power Act, 10(a), 16 U.S.C. § 803(a) (1976). Pursuant to this provision, the Commission has a multifaceted obligation to investigate, before approving any license, whether construction of the project is in the public interest. Udall v. FPC, 387 U.S. 428, 87 S.Ct. 1712, 18 L.Ed.2d 869 (1967). Once the Commission makes its investigation, however, its findings may be upset on judicial review only if they are not supported by substantial evidence. 16 U.S.C. § 8251 (b) (1976); see Scenic Hudson Preservation Conference v. FPC, 453 F.2d 463, 467-68 (2d Cir. 1971) (Udall did not overturn the traditional “substantial evidence” review standard), cert. denied, 407 U.S. 926, 92 S.Ct. 2453, 32 L.Ed.2d 813 (1972).

FOR asserts the inadequacy of FERC’s decisionmaking in multiple respects. Petitioners claim that need for the project has not been demonstrated; FERC, they contend, has overprojected demand for power in NCPA’s service area, and underprojected supply. Moreover, petitioners charge that FERC improperly rejected reasonable alternatives, including purchase of power from other utilities, and alternate means of constructing the project. FOR further argues that FERC misconceived the overall negative environmental effects of the project, and misjudged what is in California’s “public interest,” by overlooking state policy that attaches a low priority to new hydroelectric projects.

Our review securely indicates that only one of FOR’s claims warrants discussion — the claim that the project is unnecessary because NCPA can continue to purchase all the power it needs. FOR’s other contentions are insubstantial. The administrative record fully supports FERC’s appraisal of projected NCPA demand for power,6 and the Commission’s evaluation of future NCPA generating capacity.7 Substan*335tial evidence in the record also supports FERC’s rejection of the Utica-Union, Low Spicer Storage Complex alternative to building the CCWD project at its present location. Equally adequate evidence backs up FERC’s conclusion that the environmental impact mitigation measures negotiated by CCWD, the California Department of Fish and Game, and the U.S. Forest Service are reasonable and attainable. California’s announced policy concededly ranks large hydroelectric projects low among possible new power plant development, but ample evidence sustains FERC’s independent determination, as a federal agency, that the licensed hydroelectric plant is an appropriate development.8

FOR urges most strenuously that FERC should have declared the project unnecessary because NCPA’s power needs can be met reliably through continued purchases from PG & E, NCPA’s traditional source of electric power. FERC responded that

PG & E’s ability to meet load in the future is uncertain. Under adverse hydrological conditions, PG & E’s reserve margin in 1985 could be as little as 3 percent. Moreover, ... by 1992 the PG & E system will require, in addition to the [Calaveras] and another proposed hydroelectric project, an additional 493 MW of new capacity to be able to provide reliable electric service.

Order Denying Rehearing, J.A. 959, 960. FERC has also noted that the NCPA-PG & E purchase contracts are terminable by PG & E on six to twenty-four months notice, and that PG & E’s priority responsibility is to customers within its own service area. J.A. 187, 411, 443. These findings, FERC argues, more than adequately justify its conclusion that NCPA cannot safely rely on PG & E power, and should be permitted to develop the capacity to stand on its own feet in the years to come.

We examine in turn FOR’s assertions that (1) substantial evidence does not underpin FERC’s conclusion that PG & E’s own resources will be insufficient to support NCPA’s growing needs, and (2) even if PG & E’s resources, considered in isolation, are inadequate to support NCPA, FERC should have extended its examination to the energy resources of utilities beyond PG & E, resources that could be made available to PG & E “during emergency conditions.” Brief for Environmental Petitioners at 20.

A. Extent of PG & E’s Own Resources

The most up-to-date set of energy forecasts in the agency record indicates that peak demand9 in the PG & E planning area will rise to 19,407 MW by 1992. Electricity Tomorrow 62. To meet a peak demand increase of this magnitude, PG & E will have to add 4,376 MW of capacity over and above what it now has. Id. Further, PG & E’s current reserve margin is “well below the optimum,” Id. at 22; to bring its reserve margin up to a safe level, PG & E will have to add an additional 1,394 MW by 1992. Id. at 67. Finally, PG & E cannot securely rely on all the power it is receiving now; some of its generating plant is outmoded, and all of its long-term contracts for purchase of firm capacity from other utilities are due to expire in the 1980s. Even assuming renewal of two long-term contracts with Pacific Northwest utilities, PG & E will have to *336add 1,754 MW of capacity to make up for retirements and contract expirations. Id. at 67-68. In sum, based on forecasts in the record, PG & E, to stay safe, must add a minimum of 7,524 MW of capacity by 1992 —a figure equivalent to about half of PG & E’s current demand.

To meet its energy needs, PG & E has developed a comprehensive plan to build coal-fired, hydroelectric, and cogeneration plants. The plan is dominated by additions of nuclear power plants in the early 1980s and coal-fired capacity in later years. Electricity Tomorrow 328. It appears that PG & E — and CEC — have included the Calaveras project in their resource planning.10

Even including the Calaveras project in the estimation, however, PG & E faces a shortfall in capacity additions reasonably expected to be operating by 1992.11 Just how PG & E is to make up this shortfall is not clear. One CEC staff report asserts that additional efforts along lines such as conservation and further geothermal generation will be sufficient. J.A. 987-92. The extent of PG & E’s future capacity needs, however, together with the utility’s heavy reliance on nuclear power generation, may place PG & E in an uncomfortable position in the next few years. Electricity Tomorrow reports that “if Diablo Canyon [nuclear power plant] is not licensed and operated, the PG & E area may require assistance from other utilities if unscheduled very large power plant outages are experienced.” Electricity Tomorrow 316. CEC studies further indicate that in a worst-case analysis, viewing the PG & E planning area as an isolated system, PG & E’s reserve margin could drop as low as 2.7% by 1985, more than 10% below optimum levels. Id. at 317. While other utilities could come to PG & E’s rescue in emergency conditions, the prospect remains that PG & E will end up in the next few years not as a seller of surplus energy but, to meet its own needs, as a purchaser of power from other systems with surplus supplies. See Electricity Today 331 (“The potential delay in obtaining the Diablo Canyon operating licenses has prompted PG & E to seek purchases of short-term capacity.... ”). This state of affairs has persisted. See Power Shortage — Or Sham?, Newsweek, June 13, 1983, at 64 (“PG & E remains confident it can help fill its growing needs by purchasing power from other systems with surplus supplies.”).

CEC’s apparent inclusion of the Calaveras project in its resource planning indicates the vulnerability of POR’s assertions concerning the adequacy of PG & E’s resources. FOR maintains that NCPA can rely on power from PG & E. FOR cites in support of its assertions, however, CEC studies that assume PG & E’s reliance on power from NCPA. CEC has found, moreover, that even if Calaveras is built, PG & E, if it is to bolster its reserve margin, will still need extensive capacity additions. FOR can thus tenably argue only that, if the Calaveras project is not built, some other project somewhere else perhaps could be built to take up the slack. But this possibility hardly undercuts the substantial record support for FERC’s finding that PG & E’s resources do not obviate the need for power the Calaveras project will generate.

B. Purchases from Beyond PG & E

FOR next argues that even if the Commission acceptably analyzed PG & E’s own *337resources, FERC did not go far enough. The Commission should have looked beyond California, FOR maintains; it should have extended its supply-demand analysis to the Pacific Northwest. FERC indicated that it regarded interconnections with utilities outside California as a speculative matter; it said in response to comments on the draft EIS:

An additional intertie with the Pacific Northwest would not provide the dependable capacity equivalent to the proposed project. The amount of power available from the Pacific Northwest depends on the magnitude of the power loads in the Pacific Northwest and the amount of water available for generation. At times, there is no power available for transfer from the Pacific Northwest, and, as future loads in that area increase, less power will be available to California.

J.A. 415. Concededly, FERC did not undertake detailed examination of resources beyond PG & E’s. It did not attempt to estimate what PG & E might buy from Oregon, Washington, British Columbia, and Manitoba, and sell to NCPA. We conclude that FERC was not required to enlarge its inquiry as FOR proposed.

Regional power planning is difficult and uncertain; it becomes increasingly difficult and uncertain as its scope is extended in time and space.12 One scholar has characterized regional power planning as “a task that confronts a degree of uncertainty and risk without historical precedent.”13 As to the Pacific Northwest, the same author stated: “[T]he best one can do” in that region with current planning methods still entails “major risks” of faulty prediction.14

FERC initially analyzed projected supply and demand in NCPA’s service area and found a need for new generating capacity. FOR pointed to the supply source next door, PG & E. FERC examined PG & E’s projected reserve margin and found it low; PG & E, it is reasonably anticipated, will need new sources of power even on the assumption that CCWD’s project is built. FOR contends FERC should have proceeded to the next door, a very large one, the Pacific Northwest.15 FERC’s terse response is not surprising. No one, apparently, has made projections that would justify *338safe reliance on Northwest power. California itself subdivides the state into distinct “planning areas”; CEC has noted that one of the state’s objectives is to meet demand while “not substantially increasing power purchases from utilities outside California.” Electricity Tomorrow 354.16

We are in no position to require FERC, in the circumstances of this case, to survey power availability more broadly than it did. How far in time and space a power projection investigation ought to extend is a question involving practical considerations of feasibility and a balancing of costs against expected useful results. Resolving these issues requires technical expertise and is properly left to the informed discretion of the responsible federal agency. Cf. Kleppe v. Sierra Club, 427 U.S. 390, 412, 96 S.Ct. 2718, 2731, 49 L.Ed.2d 576 (1976). It is impossible to say, on review of the record before us, that FERC failed to exercise its discretion intelligently.

Udall v. FPC, 387 U.S. 428, 87 S.Ct. 1712, 18 L.Ed.2d 869 (1967), a lead decision in this area, does not counsel a different result. It involved a controversy between the Secretary of the Interior and the FPC. Interior initially took the position that a consortium of four private utilities should not be licensed to develop a hydroelectric project on the Snake River at the Idaho-Oregon border until means of protecting anadromous fish (principally salmon and steelhead) had been studied. Later, Interior added its view that the project should be constructed by the federal government, not by private utilities; further, Interior questioned the immediate need for the plant. The FPC nevertheless approved the utilities’ proposal. The Court, per Justice Douglas, overturned that decision, and remanded the matter to the FPC.

The Court identified as the primary ground for remand the agency’s failure to take evidence on whether development of the river basin should proceed under federal rather than private, state, or municipal auspices. Id. at 431-32, 87 S.Ct. at 1714. It observed that the issue of federal, as distinguished from private or municipal, control “[might] conceivably make a vast difference in the functioning of the vast river complex” in question. Id. at 435, 87 S.Ct. at 1716. “Since the cases [had to] be remanded to the Commission,” the Court thought it appropriate to instruct the agency to consider as well the alternative of no development at all. Id. at 436, 87 S.Ct. at 1716.

In the wide-ranging discussion in the second part of the Udall v. FPC opinion, the Court conveyed this message: where a project may have major adverse environmental effects, the Commission should carefully consider the need for the project, in light of its multiple impacts on commerce, recreation, wilderness and wildlife preservation, before granting a license. A determination of need, the Court said,, required more than affirmation that the project would be “beneficial to the licensee” combined with ascertainment that the region would “be able to use the additional power.” Id. at 450, 87 S.Ct. at 1724.

The project proposed in Udall v. FPC risked “destroying] the river as a waterway,” yet the agency had not addressed “the desirability of [the river's] demise.” Id. The Court declared that the Commission should have explored alternate sources of power supplies, and the regional and even national need for the recreational and commercial resources, including fish and other wildlife, threatened by the project in question. Id. at 440, 442, 444, 450, 87 S.Ct. at 1718, 1719, 1720, 1724.17 The public in*339terest determination under section 10(a) of the Federal Power Act,18 in sum, should be overarching, the Court said; it should encompass the effects of the proposed project on navigation, water supply, recreation, and other public uses of a waterway, and not simply regional power demand.

In the case at hand, FERC’s performance is hardly comparable to the FPC’s in Udall v. FPC. The Commission did not restrict its inquiry to the benefits of the project to CCWD and NCPA. Nor did it stop at the region’s demand for additional power. It took into account alternate sources of power within the region, and the interest in preserving the region’s wilderness areas and wildlife. It cannot fairly be faulted, as the Court indicated the agency could in Udall v. FPC, for wholesale neglect of issues relevant to the public interest.

One portion of Justice Douglas’ opinion in Udall v. FPC, it is true, broadly discusses the entire world’s future supplies of power, and suggests that hydroelectric projects should be evaluated'in light of the fact that “My the end of the century ‘nuclear energy may account for about one-third of our total energy consumption.’ ” Id. at 446-47, 87 S.Ct. at 1721-23 (quoting Brown, The Next Hundred Years 109 (1957)). But this part of the decision does not instruct the Commission on how far afield it should go in considering alternate power sources. The opinion quotes a long passage from Interior’s petition to intervene in the FPC proceedings, in which Interior included a list of Northwest power resources inside and outside the utilities’ service areas. 387 U.S. at 444-46, 87 S.Ct. at 1720-21. As far as one can tell from the opinion, however, the only alternate sources of power the Court noted as immediately relevant to the FPC’s decision were projects physically located within the service area of the utilities the FPC had licensed. Id. at 448, 87 S.Ct. at 1723. All one can conclude from this less than crystalline part of Udall v. FPC is that the Commission must look beyond any single utility’s or consortium of utilities’ independent resources. FERC did that in this case.

We hold that the Commission did not abuse its discretion in declining to pursue an investigation into the amount of power PG & E might be able to purchase from Pacific Northwest utilities. FERC’s decision to limit close consideration of demand and resources to meet the demand to the NCPA and PG & E areas, and to avoid an excursion into more distant regions, was reasonable and comfortably within the Commission’s authority.19

III. NEPA Claims

FOR reasserts under a NEPA heading several of the contentions we found *340insubstantial when raised under an FPC rubric. See supra pp. 97-99. We find them no more worthy of discussion when presented under a different label.20 We can deal summarily as well with FOR’s contention that the EIS should have examined “water supply” aspects. CCWD’s plans in this regard were not sufficiently concrete to warrant evaluation in conjunction with the Calaveras power project. See Kleppe v. Sierra Club, 427 U.S. 390, 401-02, 96 S.Ct. 2718, 2726, 49 L.Ed.2d 576 (1976).21

FOR’s claim concerning the prospect of purchasing power in lieu of constructing a hydroelectric plant, however, bears separate discussion in the NEPA context. To the extent that FOR challenges the size of the region considered by FERC — the dimensions of the Commission’s investigation— our discussion relating to the FPA largely answers FOR’s criticism and we need not write many more words here. FOR raises a graver concern with respect to the timing of the investigation the Commission did undertake; FOR’s final contention concerning the need for a supplemental EIS on the other hand, does not require extended discussion.

A. The Dimensions of FERC’s Investigation

Section 102(2)(E) of NEPA calls upon each federal agency to “study, develop, and describe appropriate alternatives to recommended courses of action in any proposal which involves unresolved conflicts concerning alternative uses of available resources.” 42 U.S.C. § 4332(2)(E) (1976). This provision parallels the general EIS requirement of discussion of “alternatives to the proposed action.” Id. § 4332(2)(iii). FOR contended that the Commission was obliged to look to PG & E and beyond to satisfy the command in Udall v. FPC, supra, that the agency, in carrying out its mission under the FPA, explore “alternative sources of power.” We have explained above why the examination FERC conducted satisfied the FPA requirement and was consistent with the Court’s instructions in Udall. See supra pp. 100-103. FOR makes the same arguments in contending that purchasing power was an “alternative” the Commission should have considered expansively to carry out its obligations under NEPA.22

We note initially that it is not altogether apparent why power purchasing should rank as a course “alternative” to power generation. Purchasing prospects are perhaps more appropriately considered a component of need. Power purchasing is feasible and desirable only if a producer in the relevant region has excess supplies to sell, supplies adequate to obviate the need for new production in the area. FOR’s dominant claim is not that FERC failed to con*341sider “alternative” courses for production of power, but that there was no need for additional power generation.23 However, this court has once before accepted categorization of purchased power as an “alternative” within the meaning of NEPA,24 and we will adhere to that view.

Just as we ruled in considering FPA strictures that FERC was not obliged to survey power availability more broadly than it did, so we conclude here that the Commission properly confined its consideration of demand and available power supplies to the NCPA and PG & E areas. And for the same reasons. We add only the Supreme Court’s admonition in Vermont Yankee Nuclear Power Corp. v. NRDC, 435 U.S. 519, 98 S.Ct. 1197, 55 L.Ed.2d 460 (1978), that “the concept of alternatives [under NEPA] must be bounded by some notion of feasibility.” Id. at 551, 98 S.Ct. at 1215.25 FERC’s decision as to the space dimensions of its demand and supply exploration in this case was properly bounded by reasonable considerations of what could be forecast with a fair degree of reliability and with the energy, research, and time resources available to the agency. See NRDC v. Morton, 458 F.2d 827, 837 (D.C.Cir. 1972).26

B. The Timing of FERC’s Investigation

We are secure in the conclusion that the Commission’s investigation of power purchasing as a means to satisfy NCPA’s need, or as an “alternative” course, was substantively unassailable. However, an important question remains whether the license should be set aside and the matter returned to FERC because the agency did not present its analysis adequately in the EIS itself; instead, the Commission delayed adequate statement until its July 1982 order denying FOR’s rehearing petition.

The crux of the problem here lies in the difference between typical “substantial evidence” review of agency decisionmaking on the one hand, and NEPA review on the other. “Substantial evidence” review essentially concerns the merits of the decision ultimately reached by the agency — is that decision sustained by the evidence in the record taken as a whole? See Universal Camera Corp. v. NLRB, 340 U.S. 474, 71 S.Ct. 456, 95 L.Ed. 456 (1951); American Public Gas Association v. FPC, 567 F.2d 1016, 1028-29 (D.C.Cir. 1977), cert. denied, 435 U.S. 907, 98 S.Ct. 1456, 55 L.Ed.2d 499 (1978). Substantial evidence there must be, but that evidence may be scattered throughout the record. NEPA, on the other hand, establishes an • “essentially procedural” requirement. Vermont Yankee Nuclear Power Corp. v. NRDC, 435 U.S. 519, 558, 98 S.Ct. 1197, 1219, 55 L.Ed.2d 460 (1978). It calls upon agencies to present *342evidence and discussion relevant to their environmental decisionmaking in one comprehensive document — the environmental impact statement.

Section 102(2) of the Act

requires that responsible officials of all agencies prepare a “detailed statement” covering the impact of particular actions on the environment, the environmental costs which might be avoided, and alternative measures which might alter the cost-benefit equation. The apparent purpose of the “detailed statement” is to aid in the agencies’ own decision making process and to advise other interested agencies and the public of the environmental consequences of planned federal action.... Moreover, by compelling a formal “detailed statement” ..., NEPA provides evidence that the mandated decision making process has in fact taken place and, most importantly, allows those removed from the initial process to evaluate and balance the factors on their own.

Calvert Cliffs' Coordinating Committee v. AEC, 449 F.2d 1109, 1114 (D.C.Cir. 1971). The EIS serves as an assurance that the agency has indeed adverted closely and carefully to the environmental impacts of its actions. By “gathering] in one place a discussion of the relative environmental impact of alternatives,” NRDC v. Morton, 458 F.2d 827, 834 (D.C.Cir. 1972), the EIS makes it possible for the public and reviewing courts to consider conveniently how and why the agency made its final choices. See generally Grazing Fields Farm v. Goldschmidt, 626 F.2d 1068 (1st Cir. 1980). And of particular importance, the EIS requirement inhibits post hoc rationalizations of inadequate environmental decisionmaking.

FERC, we are impelled to say, did not measure up to NEPA’s command. Most of FERC’s analysis of the power purchase issue appears in the Commission’s July 1982 order, not in the EIS. Were we confronted with the slim references to the question in the EIS itself, unsupplemented by the discussion in the July 1982 order, we might consider a remand unavoidable. Once power purchasing is ranked an “alternative” course for NEPA purposes, see supra p. 104, a reviewing court appears bound to. conclude that what FERC reported in the EIS was inadequate.

The special circumstances this case presents, however, deter us from ordering a remand so that FERC can compose a new EIS. This court has recognized that “[t]he EIS ... is not an end in itself, but rather a means toward the goal of better decision-making.” North Slope Borough v. Andrus, 642 F.2d 589, 599-600 (D.C.Cir. 1980). A remand here, we believe, would treat the EIS as “an end in itself” and would not meaningfully serve NEPA’s goals.

Were we to return the case, the Commission would be instructed to evaluate prospects for NCPA’s continued reliance on purchasing power rather than generating it, and to write up the evaluation in the form of an EIS. But well before the start of this review proceeding, the Commission did make such an investigation, after receiving and responding to extensive comments from interested persons outside the agency, and FERC incorporated its findings in an opinion accessible to the public — the July 1982 order. Thus a remand here is not needed to “aid in the agencies’ own [environmental] decision making process,” Calvert Cliffs' Coordinating Committee v. AEC, 449 F.2d at 1114, for that process has already occurred and we have set out our reasons for finding it bona fide and sufficiently extensive. See supra pp. 99-103. The Commission’s July 1982 order, read in conjunction with the EIS, adequately “advise[s] other interested agencies and the public” of FERC’s reasoning and the information on which it rests. Id.27 We are not left to rely on post hoc rationalizations, see Motor Vehicle Manufacturers Association v. *343State Farm Mutual Automobile Insurance Co., - U.S. -, 103 S.Ct. 2856, 2870, 77 L.Ed.2d 443 (1983); we have before us FERC’s assessment, embodied in an opinion composed after due investigation and before the matter was brought to court.28 Sending the matter back “to teach the agency a lesson” would be an essentially punitive measure; we can discern no benefit to the public in such a course, and no genuine service to the policies NEPA advances.

The result we reach, and the reasons for it, are not novel. We note particularly Scenic Hudson Preservation Conference v. FPC, 453 F.2d 463 (2d Cir. 1971), cert. denied, 407 U.S. 926, 92 S.Ct. 2453, 32 L.Ed.2d 813 (1972), involving a hydroelectric plant, licensed by the Commission, with no document on file titled an EIS. The Commission had rendered a comprehensive opinion, however, which included considerable discussion of environmental impacts. Though NEPA governed, the court saw no purpose in a remand so that the agency’s discussion could be circulated for comment under a corrected label. It sufficed that the Commission’s opinion demonstrated “full compliance with the National Environment Policy.” Id. at 48129

A more conspicuous default is described in Warm Springs Dam Task Force v. Gribble, 621 F.2d 1017 (9th Cir. 1980). There, new information about earthquake fault lines matured the Army Corps of Engineers’ obligation to supplement an EIS it had filed prior to beginning construction of a dam in Sonoma County, California. The district court, ruling on a suit filed by environmentalist groups to enjoin further construction, found that the Corps had complied with NEPA even though it had taken no action to investigate the earthquake safety problem. The Ninth Circuit disagreed — the Corps had unlawfully ignored its obligation to investigate. The appellate court noted, however, that subsequent to the start of the district court hearing the Corps had finally ordered studies; and those studies largely obviated seismic safety concerns. It therefore declined to remand; it would be pointless, the Ninth Circuit said, to “order the Corps to conduct studies already completed to answer questions the Corps has already answered on a basis that could not be successfully challenged.” Id. at 1026.30

Finally, in Swinomish Tribal Community v. FERC, 627 F.2d 499 (D.C.Cir. 1980), our court confronted a situation similar to the one at hand. The Commission’s EIS, covering construction of a hydroelectric plant, failed to discuss at all the alternative of building a dam lower than the one proposed. There was nothing facially unreasonable about building the lower dam. The Commission eventually discussed the lower dam alternative at reasonable length; as in this case, much of the discussion appeared in the Commission’s order denying rehearing. We quoted from FERC’s order denying rehearing, concluded that the Commission had given the lower dam alternative *344the requisite consideration, id. at 514, and affirmed the Commission’s action.31

We appreciate the need for caution in cases in which an agency slights the EIS requirement. A reviewing court must be certain that the basic policies of NEPA are satisfied, and that a remand would indeed be pointless, before it rejects that course. Arizona Public Service Co. v. FPC, 483 F.2d 1275 (D.C.Cir. 1973), is illustrative. That case involved the Commission’s denial of a license to transport natural gas, an action that arguably had a significant effect on the environment. But no EIS was filed, nor did any agency statement account for the omission. The court was not positioned to say whether the agency in fact engaged in sound environmental decisionmaking, for nothing in the record demonstrated that it had. Remand is essential in such a case.

We recognize too that language can be found in decisions suggesting that pleas of “futility” will never do to justify noncompliance with NEPA. Invariably, such statements appear when there is no cogent demonstration, of the kind made here, that the agency gave due consideration to the relevant environmental factors and made that consideration manifest in an accessible,32 intelligible33 form. Understandably, courts have low tolerance for futility arguments pressed by parties who seek release from, an obligation to engage in serious environmental decisionmaking on the ground such an enterprise would not change the party’s mind about a planned development. See, e.g., Jones v. Lynn, 477 F.2d 885 (1st Cir. 1973); Hanly v. Mitchell, 460 F.2d 640, 648 (2d Cir.), cert. denied, 409 U.S. 990, 93 S.Ct. 313, 34 L.Ed.2d 256 (1972); Pennsylvania v. Morton, 381 F.Supp. 293, 300 n. 14 (D.D.C. 1974); City of New York v. United States, 337 F.Supp. 150 (E.D.N.Y. 1972).

But insistence on form when we have before us a case in which the agency did in fact “compl[y] with the National Environment Policy,” Scenic Hudson Preservation Conference v. FPC, 453 F.2d at 481, would jeopardize NEPA’s “lofty declarations.”34 Remands in such cases would inevitably breed cynicism about court commands; they would likely yield going-through-the-motions responses on the part of those told to attend to the court’s costly, resource-consuming instruction to redo, under the proper heading, what has already been done effectively.

C. Supplementation of the EIS

FOR urges, referring to submissions annexed to its brief, and its Application for Leave to Adduce Additional Evidence before FERC and Motion for an Order Establishing Appropriate Procedures, see supra note 6, that we direct FERC to compile a supplemental EIS, thereby taking into account the most recent information relevant to PG & E’s ability to provide energy to NCPA, NCPA’s own future electrical capacity, and future electrical demand in the NCPA service area.35 General*345ly, the initial decision whether a supplemental EIS is required should be made by the agency, not by a reviewing court. People Against Nuclear Energy v. NRC, 678 F.2d 222, 233-34 (D.C.Cir. 1982), rev’d on other grounds sub nom. Metropolitan Edison Co. v. People Against Nuclear Energy, - U.S. -, 103 S.Ct. 1556, 75 L.Ed.2d 534 (1983). The record does not indicate that FOR ever presented a request to the Commission asking for supplementation of the EIS, although arguably FOR could have done so even after the close of administrative proceedings with respect to the license. See Michigan Consolidated Gas Co. v. FPC, 283 F.2d 204, 225 (D.C.Cir. 1960). We need not consider the effect of FOR’s apparent failure to approach the Commission first, however, since we reject FOR’s request on its merits.

Under the Council on Environmental Quality’s NEPA Guidelines, an agency should prepare a supplemental EIS if there exists “significant new .. . information relevant to environmental concerns and bearing on the proposed action or its impacts.” 40 C.F.R. § 1502.9(c) (1981).36 This guideline does not mean that an agency must release and circulate a formal supplemental EIS, or a formal document explaining why the agency believes a supplemental EIS is unnecessary, every time some new information comes to light. Rather, a reasonableness standard governs. See California v. Watt, 683 F.2d 1253, 1268 (9th Cir. 1982), cert. granted, - U.S. -, 103 S.Ct. 2083, 77 L.Ed.2d 295 (1983); Warm Springs Dam Task Force v. Gribble, 621 F.2d 1017, 1024 (9th Cir. 1980); People Against Nuclear Energy v. NRC, 678 F.2d at 234 (expressly adopting the Warm Springs analysis); Humane Society of the United States v. Watt, 551 F.Supp. 1310, 1322 (D.D.C. 1982), aff’d mem., 713 F.2d 865 (D.C.Cir. 1983).

FOR’s new information, as we indicated earlier, see supra notes 6 & 7, appears to be of questionable value. Even if the information were somewhat more weighty, however, we would hesitate to command FERC to attend to it. CEC, as part of its continuing task of surveying California’s energy picture, issues new forecasts and predictions every few months, each soon to be superseded by the next. Were we to order the Commission to reassess its decisions every time new forecasts were released, we would risk immobilizing the agency. As the Supreme Court has admonished:

Administrative consideration of evidence ... always creates a gap between the time the record is closed and the time the administrative decision is promulgated [and, we might add, the time the decision is judicially reviewed],... If upon the coming down of the order litigants might demand rehearings as a matter of law because some new circumstance has arisen, some new trend has been observed, or some new fact discovered, there would be little hope that the administrative process could ever be consummated in an order that would not be subject to reopening.

Vermont Yankee Nuclear Power Corp. v. NRDC, 435 U.S. 519, 554-55, 98 S.Ct. 1197, 1217, 55 L.Ed.2d 460 (1978) (quoting ICC v. City of Jersey City, 322 U.S. 503, 514, 64 S.Ct. 1129, 1134, 88 L.Ed. 1420 (1944)). After prolonged consideration, FERC issued the CCWD license in 1982; review and reassessment of that decision, absent sufficient reason, should not continue indefinitely.

*346Conclusion

Our decision accords FERC no license to ignore requirements Congress has established for its decisionmaking process. We must hold the Commission, like all agencies, to a “strict standard of compliance” with NEPA. Calvert Cliffs' Coordinating Commission v. AEC, 449 F.2d 1109, 1112 (D.C.Cir. 1971). But in this case, while FERC fell short of a formal requirement, its decision-making adequately responded to the concerns underlying NEPA. See Scenic Hudson Preservation Conference v. FPC, supra. We therefore decline to order a remand that would cause added delay and added cost to the public without genuinely advancing any purpose served by NEPA. Accordingly, the order of the Commission is

Affirmed.

dissenting:

This petition for review vividly illustrates the tension between two of this country’s most pressing concerns: the need to assure adequate supplies of energy and the need to protect the environment. In a hydroelectric licensing proceeding, FERC is responsible for balancing these often competing goals. Two statutes give content to this task: the Federal Power Act (FPA) and the National Environmental Policy Act (NEPA). The FPA, which sets forth a public interest licensing standard, requires that FERC take a comprehensive, regional approach in investigating the need for and alternatives to a proposed hydroelectric facility. NEPA, an “essentially procedural” mechanism, sets forth a number of requirements designed to ensure that the agency has taken a “hard look”1 at the reasonable alternatives to its proposed action.

In the instant proceeding, environmental groups, private citizens, and California agencies have persistently and forcefully argued that the need for the proposed North Fork Stanislaus project’s power is greatly exaggerated, and that the alleged need can easily be met through the reasonable alternative of purchasing power from intrastate or interstate suppliers. In response to these arguments, however, FERC devoted a total of two scattered and misleading sentences to the issue in its Environmental Impact Statement (EIS). Regardless of adamant public criticism of its draft EIS, the agency made no revisions. Turning to the licensing order, FERC did not address the alternative of purchased power at all. In its Order Denying Rehearing, FERC finally gave explicit treatment to the role of PG & E and other utilities in supplying NCPA’s needs. This treatment, however, was limited to two points: first, the fact that the issue “was not new to this Commission”; and second, a statistic in the California Energy Commission’s (CEC) 1981 Electricity Tomorrow (ET)2 Report concerning PG & E’s “reserve margin.” FERC’s treatment of this statistic, however, was patently capricious.

The court today holds that this shoddy performance satisfies the exacting statutory requirements of the FPA and NEPA. A careful analysis of precedent and the licensing record, however, demonstrates that FERC has violated both the letter and the spirit of these statutes. I therefore respectfully dissent.3

I. Background

FERC has presented both a “purpose” and a “need” for the North Fork Stanislaus hydro project. Separation of the two is essential to understanding the instant dispute.

The proffered purpose of the project is for the Calaveras County Water District (CCWD), the project applicant, to make money to finance water-development projects. Section 2.1 of the EIS states that *347“[t]he sale of the power generated by this project would reduce the cost of water to the residents of Calaveras County.”4 This section indicates that power will be sold, not only to NCPA, but “throughout the Northern California area power systems.”5

A licensee’s bare “purpose” is not enough under either NEPA or the FPA; FERC must also determine that a “need” for the power exists.6 Despite the statement in section 2.1 of the EIS that power will be sold “throughout the Northern California area,” it has not focused on statewide energy needs. Instead, it has consistently sought to demonstrate a need for the power solely by reference to the NCPA area. NCPA’s needs, painted in apocryphal terms, formed the basis both of the EIS and of FERC’s licensing order.7

The reasoning presented is straightforward: based on (sharply-contested) demand projections, NCPA will have certain capacity and energy requirements in future years. Purchases from the Western Area Power Administration (WAPA) and NCPA’s development of its own geothermal facilities will meet a portion of this need. After accounting for these projected supplies, however, a deficit will remain.8 Although eontradict*348ing section 2.1 of the EIS, section 1.1 states that “all of the electric power produced by the project will be sold to NCPA.”9 This power will reduce NCPA’s deficit.

Until now, “almost all” of NCPA’s needs have been met through purchase contracts with WAPA and PG & E.10 FERC did not, however, consider continued power purchasing (other than the WAPA contracts) to be a feasible resource for meeting NCPA’s needs, and therefore did not include this alternative in the alternatives section of the EIS.11 In the entire EIS, FERC gave a one-sentence explanation of why PG & E could no longer be relied upon:

The contracts for the purchase of power between individual members of NCPA and PGE can be terminated on 6 to 24 months notice and cannot be considered as firm resources to meet NCPA’s load in 1985, when the proposed project is scheduled to be operational.12

The only other discussion of the power-purchase alternative consists of one sentence that appears one hundred forty-one pages later in the EIS, asserting that denial of the license would force NCPA “to secure or develop other means to meet its electrical capacity and energy requirements, which would consume nonrenewable fossil-fuel resources.” 13

Public comments in response to the Draft EIS sharply criticized FERC’s failure to assess power purchasing as an alternative. For example, the California Resources Agency, parent agency of the CEC, asserted that this failure, among “several other major deficiencies,” required that “a new complete draft ... be prepared and made available for public comment.”14 The Resources Agency stated:

The need for the energy is not demonstrated. The report merely states that the Northern California Power Agency (NCPA) wishes to have their own generation facilities instead of continuing to buy energy from the Western Area Power Administration (WAPA) and Pacific Gas and Electric (PGandE). There is no showing whatsoever that WAPA or PGandE have insufficient power for NCPA areas. Most, if not all, of this capacity already exists in WAPA and PGandE facilities; nor does the applicant make any kind of convincing argument that a controlled utility such as PGandE could' arbitrarily deny municipalities needed, available energy.15

*349FERC’s staff responded with three words of commentary: “See Section 1”16 — a reference to the one-sentence analysis that prompted the Resources Agency’s criticisms.

The Resources Agency continued: “The alternatives section is substantially deficient. It does not include even all the obvious alternatives for energy production. Certainly the alternative of continued purchases from WAPA and PGandE appears reasonable. An additional intertie with the Pacific Northwest is another alternative for peaking power that should be analyzed.”17 The staff responded:

NCPA would continue to purchase power from the Western Area Power Administration and PGE, although the proposed project would reduce the amount of power that NCPA would be required to purchase from PGE. The power that PGE would not have to supply NCPA would come primarily from oil- and coal-burning generating units that PGE intends to acquire. An additional intertie with the Pacific Northwest would not provide the dependable capacity equivalent to the proposed project. The amount of power available from the Pacific Northwest depends on the magnitude of the power loads in the Pacific Northwest and the amount of power available for generation. At times, there is no power available for transfer from the Pacific Northwest, and, as future loads in that area increase, less power will be available to California.18

Friends of the River (FOR) also criticized the EIS’s needs analysis for failing to document the need for the power. In addition, FOR objected to FERC’s failure to take state-wide needs and policies into account.19 FERC’s staff responded, once again, by simply referring back to section 1 of the EIS.20

Comments received from Dale Meyer continued the theme of the Resources Agency’s and FOR’s criticisms: “The need for the project power is not documented .... The State of California Energy Commission data would lead one to conclude the peak demand for the NCPA servpce] area may be greatly overestimated.”21 Again the staff responded with the unhelpful observation, “See Section 1 for discussion of need for project power.”22 The staff added a surprising observation. Although the need for the project power was premised on the unavailability of additional firm supplies, the response to Meyer continued: “It is Staff’s position that NCPA should not be denied the option to develop its own generation resources even if a private supplier can supply NCPA’s needs at present.”23

Arguments that the power was not needed, and that NCPA could continue to secure energy from PG & E and other suppliers, also dominated the licensing proceeding.24 *350In its licensing order, however, FERC did not discuss power purchasing from PG & E or other utilities at all. Rather, it proceeded to use precisely the same analytic process as in the EIS: NCPA’s projected needs, minus WAPA and geothermal supplies, yielding a deficit that required construction of the North Fork Stanislaus facility.25

At the rehearing stage, the environmental petitioners renewed their criticisms of the EIS’s needs and alternatives analyses, arguing that CCWD had failed to meet its burden of establishing the project’s need.26 This time FERC explicitly addressed the environmental petitioner’s criticisms. First, it stated in its Order Denying Rehearing, the petitioner’s arguments “are not new to this Commission.”27 Second, the CEC’s 1981 ET Report “in fact demonstrates that PG & E’s ability to meet load in the future is uncertain. Under adverse hydrological conditions, PG & E’s reserve margin in 1985 could be as little as 3 percent.”28

II. FPA Claims

A. FERC’s Planning Duty

Section 10(a) of the FPA provides:

That the project adopted .. . shall be such as in the judgment of the Commission will be best adapted to a comprehensive plan for improving or developing a waterway or waterways for the use or benefit of interstate or foreign commerce, for the improvement and utilization of waterpower development, and for other beneficial public uses, including recreational purposes.29

In Udall v. Federal Power Commission,30 the Supreme Court held that this language imposes a comprehensive planning duty:

The grant of authority to the Commission to alienate federal water resources does not, of course, turn simply on whether the project will be beneficial to the licensee. Nor is the test solely whether the region will be able to use the additional power. The test is whether the project will be in the public interest. And that determination can be made only after an exploration of all issues relevant to the “public interest,” including future power demand and supply, alternate sources of power, the public interest in preserving reaches of wild rivers and wilderness areas, the preservation of anadromous fish for commercial and recreational purposes, and the protection of wildlife.31

This duty to assess needs and alternatives imposes a difficult burden, fraught with the uncertainties of prediction. But as the FPC (predecessor of FERC) noted shortly after Udall:

An applicant seeking to utilize the nation’s waterways for the development of power has an obligation to look beyond its own needs to those of the region in which it operates. Western Massachusetts Electric Co., 39 FPC 723, 734-35, 40 FPC 296 at 300. Far from being merely an unwelcome burden we are convinced that the reasonable regional planning required to be detailed by the exhibit will, in most cases, be beneficial to the prospective licensee, as well as the general public.32

*351One of the keys to comprehensive planning is the consideration of whether power purchasing is a feasible alternative to new construction. Power purchasing is a straightforward concept: utilities in need of energy and capacity contract with other utilities for the purchase and transfer of firm supplies. Transfer of this power is accomplished either through “interconnection” — use of transmission lines — or through “displacement” across power grids.33

Within the utility industry, interconnection through power-purchase contracts is an integral and basic tool. The record demonstrates that every California utility analyzed in the CEC’s 1981 ET Report relies, to some extent, on both intrastate and interstate interconnections for the delivery of firm, purchased capacity and energy.34

California’s state energy policy also encourages the use of interconnection for meeting future energy needs.35 The ET Report, in the chapter on Policies and Recommendations, states:

California should support the construction of additional transmission interconnections between the state and out-of-state utilities and between utilities within the state.

Studies by the Energy Commission, by the California Power Pool, and by the U.S. General Accounting Office have consistently shown that increased interconnections between Northwestern, California, and Southwestern utilities would provide substantial benefits to all utilities.36

At the federal level, it is clear that the FPA requires consideration of power purchasing in a licensing proceeding. In Udall, for example, there was evidence in the record indicating that purchased power from other sources was an alternative with “reasonable prospects of realization.”37 The Court held that the FPC’s failure to explore this alternative violated its statutory duty to examine “future power supply and demand” and “alternative sources of power.” 38 Even before Udall, federal courts had held that alternate sources of power, including interconnection, must be considered. In Scenic Hudson Preservation Conference v. Federal Power Commission,39 for example, the Second Circuit held with respect to the FPC’s licensing of a New York pumped-storage hydroelectric project:

The Commission neither investigated the use of interconnected power as a possible alternative to the Storm King project, nor required Consolidated Edison to supply such information. The record sets forth Consolidated Edison’s interconnection with a vast network of [New England] utilities .... There is no evidence in the record to indicate that either the Commission or Consolidated Edison ever seriously considered this alternative.... [T]his failure of the Commission ... cannot be reconciled with its planning re*352sponsibility under the Federal Power Act40

In addition to its general section 10(a) planning duties, FERC is also responsible for pursuing interconnection as a result of section 202(a) of the FPA41

For the purpose of assuring an abundant supply of electric energy throughout the United States ... the Commission is empowered to divide the country into regional districts for ... interconnection .... It shall be the duty of the Commission to promote and encourage such interconnection and coordination within each such district and between such districts.42

Pursuant to this provision, FERC has promoted transfer of interconnected power in the California area through the Pacific Northwest — Southwest system, a series of three major north-south transmission lines linking California with the Northwest and the Southwest.43

B. Analysis

The court holds that FERC has fully complied with the letter and the spirit of its planning function. In this case, however, the record fails to support this holding.

1. NCPA’s Relationship with PG & E

The court accepts without question FERC’s assertion that NCPA cannot rely on continued power purchases from PG & E because contracts between the two utilities are short-term in nature and not necessarily renewable.44 FERC has argued that PG & E is likely to cut off power to NCPA in the future.45 Yet there is no evidence in the record to substantiate this argument.

First, there is nothing about short-term purchase contracts that per se excludes them from being considered in long-range forecasting. Utility industry practice is to the contrary. In its resource planning, for example, PG & E projects 600 MW of diversity exchanges with Bonneville Power Administration to continue indefinitely, even though the current contract is short-term and scheduled to expire in 1986.46

Second, the court has pointed to nothing in the record suggesting an actual possibility that PG & E would “refuse to serve the NCPA area in the reliably consistent manner it has for so many years.”47 NCPA’s members are customers of PG & E. As a regulated utility, PG & E bears responsibility for meeting the energy and capacity needs of its entire planning and service area. NCPA is part of this area.48 The PG & E area’s generation needs and resources are considered as a whole both by PG & E and by the CEC49 when establishing forecasts and considering plans for facilities. *353Both NCPA’s expected demands and the various possible resources that might meet them are accounted for by PG & E’s forecasts and plans.50

The only hard piece of evidence cited by FERC in support of its dismissal of future power purchases from PG & E concerns the worst-case “reserve margin”51 for NCPA. Citing Table III — 2 of the CEC’s 1981 ET Report, FERC argued in its Order Denying Rehearing that data on PG & E’s reserve margin demonstrates that “PG & E’s ability to meet NCPA’s load in the future is uncertain.” 52 FERC uses this uncertainty to justify its conclusion that PG & E might terminate service to the NCPA area. Table III — 2 shows that, under a “worst-case analysis,”53 PG & E’s reserve margin in 1985 could be as low as three percent. FERC’s use of this data is capricious.54

In setting forth PG & E’s “worst case scenario” reserve-margin projections, the CEC cautioned that:

The margins shown for the PG & E area are for that system operating in isolation; PG & E’s existing interconnections with other utilities and their likely support of PG & E during emergency conditions is ignored. Thus, if Diablo Canyon is not licensed and operated, the PG & E area may require assistance from other utilities if unscheduled very large power plant outages are experienced during the next few years. PG & E is well interconnected with other utilities and will become more so with its intended construction of the third 500 kV AC line to the Pacific Northwest. Thus, with continued cooperation among the state’s utilities and regulatory agencies in developing and implementing contingency measures as has occurred over the last few years, California will continue to have reliable electric supplies.55

In short, Table III — 2 calculates the reserve margin while explicitly ignoring the significance of interconnection. The table cannot then be used to show the inadequacy of such interconnection. Despite FOR’s persistent efforts, FERC has not responded to this important qualification at all. There is nothing in the record to suggest that the CEC’s analysis is erroneous. FERC’s cryptic reliance on the CEC reserve-margin forecast is misleading and should therefore fail.

2. NCPA’s Relationship with Suppliers Beyond PG & E

In response to criticism of its refusal to consider outside power-purchases, FERC’s staff stated: “At times, there is no power available for transfer from the Pacific-Northwest, and, as future loads in that area increase, less power will be available to California.”56 Based on nothing more than this assertion, FERC continued its refusal to expand the inquiry beyond PG & E. The court notes that FERC did not undertake a detailed examination of resources beyond *354PG & E’s. Yet, it holds that FERC was not required to enlarge its inquiry.57

a. Scope of Inquiry

Central to the court’s holding is the notion that the area of inquiry is within the agency’s discretion, not for judicial determination. Although this is not in dispute, the limits of the inquiry must still reflect a rational determination, supported by substantial evidence. FERC’s treatment of sources beyond PG & E must be consistent with its planning duty. The scope of the inquiry has to bear some resemblance to reality. Current utility practice, CEC policy, and judicial precedent all demonstrate that power purchasing from beyond PG & E is not at all “speculative.”

b. Availability of Power

Although FERC contends there is sometimes no power available for transfer, this assertion is squarely contradicted by the record. Power is available and is forecast to be available from other California utilities as well as the Pacific Northwest,58 Canada,59 the Southwest,60 other states in the western region,61 and Mexico.62 The CEC fully anticipates that other utilities will continue outside purchases.63 The utilities themselves incorporate their reliance on these power purchases in their internal forecasts.64

The utility industry in this country is highly dependent on power purchasing and interconnection. Both intrastate and interstate arrangements are common practice.65 PG & E relies heavily on power purchasing and intends to do so in the future.66 The CEC has vigorously promoted increased inter-utility reliance in its future policy recommendations.67 In light of these considerations, FERC’s refusal to consider power-purchasing beyond PG & E at all is unjustified.

c. Geographic Boundary of FERC’s Inquiry

While refusing to look any further than PG & E in its exploration of alternatives, FERC does look to WAPA,68 a source clearly outside of the geographic boundary FERC imposed on its inquiry. To set a geographic limit for its investigation of alternatives, yet arbitrarily to ignore it by considering purchases from WAPA, constitutes capricious action on the part of the Commission.

d. The CEC’s Posture

The court argues that the CEC has included the Calaveras project in its future energy projections. Yet, this assumption is not supported by the CEC’s ET Report. This report, on which the majority relies, contains two sets of energy projections. The first, prepared by PG & E, is entitled “Proposed Utility Resource Plan.”69 The second, prepared by the CEC itself, is entitled “Alternative Scenario.”70 PG & E’s “Proposed Utility Resource Plan” does include Calaveras in its forecast.71 The CEC’s “Alternative Scenario,” however, does not.72 The CEC explicitly states that the “Alternative Scenario” is its “Preferred Out*355look”73 and the forecast that it adopts in planning.74

Special weight should be given to the official state policy embodied in the CEC recommendation.75 This court should look to the “Preferred Outlook” which allows for California’s future energy needs absent power from the Calaveras project.

The majority emphasizes that California has set a goal of limiting out-of-state power purchases.76 Indeed, this statement is made once in the ET Report.77 It must, however, be viewed within the context of the entire report. The importance of future construction of additional transmission interconnections with out-of-state utilities and between utilities within the state receives far greater emphasis.78 The CEC actively promotes continued and increased interconnection in the future while ranking additional hydroelectric development low among its priorities.79

The California Resources Agency, the entity charged with registering the state’s reaction, has vigorously opposed the Calaveras project throughout all stages of the proceedings. It found FERC’s treatment of the power-purchasing alternative completely inadequate and without any basis.80 Curiously, the court ignores the Resources Agency’s substantial opposition to the project.

III. NEPA Claims

A. The Requirements of NEPA

NEPA requires that in all significant federal actions:

all agencies of the Federal Government shall ... study, develop, and describe appropriate alternatives to recommended courses of action in any proposal which involves unresolved conflicts concerning alternative uses of available resources.81

The agency must set forth its analysis in a “detailed statement,” the EIS. NEPA sets forth exacting procedural requirements governing exactly what must be included in the EIS. These requirements cut to the very heart of NEPA; they are the “linchpin” of the Act.82

Determining which alternatives are “appropriate,” therefore warranting detailed treatment, is a common-sense inquiry. An “EIS cannot be found wanting simply because the agency failed to include every alternative device and thought conceivable by the mind of man.”83 Thus “fanciful,”84 “remote,”85 “uncommon,”86 and “unknown”87 alternatives need not be scruti*356nized. On the other hand, “appropriate,”88 “reasonable”89 and “obvious”90 alternatives must be given detailed treatment in the EIS, even though ultimately rejected by the agency in favor of its proposed action. The standard of review is therefore governed by the “rule of reason.”91 In evaluating an agency’s choice of alternatives for inclusion in the EIS, a reviewing court gives content to the rule-of-reason standard by looking to, inter alia, agency and judicial precedents, comments received by the agency in response to the draft EIS, and the administrative record.92 The quest is to ensure that the agency has taken a “hard look” at alternatives; 93 the “detailed statement” requirement “prevents stubborn problems from being shielded from internal and external scrutiny.”94

In considering whether an agency’s discussion of alternatives complies with NEPA, a reviewing court must ultimately employ two criteria: it must decide (a) whether the discussion indicates that the agency in “good-faith objectivity” has taken the required “hard look” at alternatives; 95 and (b) whether the discussion is detailed enough to permit those who did not participate in its preparation to “understand and consider meaningfully” the reasoning, premises, and data relied upon, and to permit a “reasoned choice among different courses of action.”96

The Council on Environmental Quality (CEQ) has promulgated regulations governing the implementation of NEPA.97 These regulations are entitled to “substantial deference” as interpretations of NEPA.98 Moreover, FERC’s own regulations bind it to deal fully “with alternative courses of action to the proposal and, to the maximum extent practicable, the environmental effects of each alternative.”99

The CEQ regulations require the Commission to “[r]igorously explore and objectively evaluate all reasonable alternatives.” 100 It must “identify any methodologies used and shall make explicit reference ... to the scientific and other sources relied upon for conclusion in the statement.” 101 The agency must respond to comments received regarding the Draft EIS.102 If it determines that the comments do not warrant further agency response, an explanation must be given, “citing the sources, authorities, or reasons which support the agency’s position ....”103

*357B. Analysis

The FPA and NEPA impose many similar requirements on the agency. The procedural requirements of NEPA, however, are distinct from and far more stringent than the duties imposed by the FPA. Compliance with the FPA is no guarantee of compliance with NEPA.

1. Power Purchasing from Beyond PG & E

The court recognizes that power purchasing from sources beyond PG & E received no treatment in the EIS.104 Yet because it finds purchasing to be “speculative,”105 it holds that this alternative does not require detailed consideration in the EIS. At least five factors demonstrate that, under the “rule of reason,” power purchasing is an “obvious,” “reasonable,” and completely “feasible” alternative that must receive “detailed” treatment in the EIS.106

a. Public comments. Numerous commentators noted that the Draft EIS had failed absolutely to provide any information why FERC believed that, with the exception of WAPA sales, NCPA could no longer rely on contractual deliveries of firm capacity and energy. They noted the alternative, pursued by numerous other California utilities, of interconnection with the Northwest.107

b. Industry practice. The power-purchase contract is an integral and basic tool in the utility industry. The record demonstrates that every California utility analyzed in the CEC’s 1981 ET Report relies, to some extent, on both intrastate and interstate interconnections for the delivery of firm, purchased capacity and energy.108

c. California policy. California ranks power-pooling and interconnections among utilities near the top of its list of “Preferred Options for Meeting Electricity Needs”— and much higher than conventional hydro facilities.109 State policy actively encourages these practices.

d. NCPA’s previous practice. Until now, NCPA has always met its power needs exclusively through power purchasing from other suppliers. These contractual deliveries have apparently been sufficient to meet NCPA’s needs, even in adverse hydro years. This history demonstrates the reasonableness of continued purchasing as an “alternative” to be considered under NEPA.

e. Precedent. Previous cases that have examined this issue have held that power purchasing is a reasonable alternative requiring detailed treatment in an EIS. In National Wildlife Federation v. Andrus,110 for example, the Department of the Interi- or had authorized a 28 MW powerplant on the San Juan River in New Mexico. In its EIS, the department dismissed the alternative of “obtaining the necessary power from alternative power sources already extant in the area” in a three-sentence discussion strikingly similar to FERC’s cryptic analysis in the instant case.111 The district court *358held this analysis “lacking in the detail which is clearly required by the statute itself and in the ‘rigorous exploration and objective evaluation of the environmental impacts of all reasonable alternative actions, particularly those that might . .. avoid some or all of the adverse environmental effects.’ ”112 Moreover, the alternative of “obtaining power from the uncommitted reserves stored in the Colorado River Storage Project system” was brought to the agency’s attention, but not treated at all in the EIS. The court held that this alternative was “not unreasonable ... and [it] should have been discussed in the FES.”113

The failure to consider power purchasing in National Wildlife contrasts with Mason County Medical Association v. Knebel.114 There the issue was whether the Rural Electrification Administration, in authorizing a coal-fired steam generating facility, had failed in its EIS “ ‘to adequately discuss or seriously consider the alternative of purchasing winter peak power’ from nearby ‘summer-peaking’ utilities” in neighboring states.115 The district court and the court of appeals held that this alternative had received sufficient consideration in the EIS. The EIS itself set forth seven distinct reasons why power purchasing could not meet projected needs. These reasons were supported by citation in the EIS to various studies, letters, and affidavits. The EIS clearly showed that other utilities had been contacted about power purchasing; it set forth in a coherent, comprehensive manner why those utilities were unable or unwilling to supply the projected, needs. Based on this treatment of the power-purchasing alternative, the Sixth Circuit held that the EIS “presents sufficient information for a reasoned choice of alternatives.”116

The power-purchasing issue also was presented in Swinomish Tribal Community v. FERC.117 There FERC had authorized the raising of the Ross Dam, in Seattle, Washington, to provide additional power for Seattle’s projected growth in energy demand. The D.C. Circuit found that power purchasing, along with other alternatives, had been “considered in detail” in the EIS: The annual costs of power purchasing were “established and compared” with the costs of alternative energy sources, and the “beneficial and detrimental environmental effects” of purchased power were “carefully considered” and explicated in the EIS.118 Moreover, it was “unchallenged” that Seattle’s major supplier had already given the city written notice that, within several years, it would be unable to meet Seattle’s growth in energy demand.119

2. Rehabilitation of the EIS

The court holds that although FERC violated NEPA by failing to include its discussion of power purchasing in the EIS, the violation was cured by the Commission’s July 1982 Order Denying Rehearing. I agree that “the EIS is not an end in itself, but rather a means toward the goal of better decision-making.”120 FERC, however, failed to meet this goal through either the EIS, the Order Denying Rehearing, or both documents read together.

The Order Denying Rehearing did not contain the detailed consideration of alternatives lacking in the EIS.121 Moreover, its “analysis” was limited to Table III — 2122 of the ET Report which completely ignores *359interconnection. Reliance on this table’s statistics in isolation results in inaccurate treatment of PG & E.123

Of even greater significance, the Order Denying Rehearing did not discuss reliance on utilities beyond PG & E at all; the discussion was limited to PG & E’s role. Power purchasing from beyond PG & E was a reasonable alternative requiring detailed treatment in the EIS. The Order Denying Rehearing did nothing to correct the EIS’s failure to do so.

Even if the Order Denying Rehearing had contained the discussion lacking in the EIS, however, it would not cure FERC’s NEPA violation. In analyzing the adequacy of an EIS, the reviewing court may look to the entire administrative record only for certain narrowly defined reasons. First, the EIS may, for the sake of convenience, have cited to data, studies, and other materials in the administrative record to support its analyses of alternatives.124 Second, “[sjtudy of the administrative record by the court helps to assess the degree of discussion any particular alternative deserves ....”125 Dissection of the record cannot, however, cure an otherwise-defective EIS. “NEPA expressly places the burden of compiling information on the agency so that the public and interested government departments can conveniently monitor and criticize the agency’s action.”126 A court “can use the administrative record to set the standard for how much discussion within the EIS a particular alternative merits, but cannot deem the unincorporated record to satisfy that standard.”127

In the instant case the Order Denying Rehearing was not circulated for comment. There was no opportunity for the public and other agencies to raise objections to the order through comments. To say that the reading of the Order Denying Rehearing together with the EIS satisfies the statute is to undermine the very foundation of NEPA.

The court expresses the fear that a remand of this ease would breed cynicism towards the law.128 I strongly disagree. NEPA’s strict procedural requirements are not a “pointless technicality.”129 They are the procedural means Congress and the CEQ have specified “to vindicate the substantive goals of NEPA.”130 This court is faced with an instance in which FERC has palpably failed to carry out its statutorily mandated investigation. It is the function of a reviewing court to determine whether an agency has complied with its congressional mandate, not to decide when a noncomplying agency is “close enough.” Failure to order a remand, in light of blatant statutory violations, can only breed far greater cynicism towards the law.

IV. Conclusion

The court today holds that power purchasing from beyond the PG & E area need not be considered under either the FPA or NEPA. As the court notes, predictions of this type confront a degree of uncertainty *360and risk without historical precedent.131 It is precisely because of such uncertainty and risk, however, that FERC’s myopic approach' must be rejected. As uncertainty increases, so does the need for comprehensive and intelligent planning. “Despite the explosion of uncertainty — -indeed because of it — it is more urgent than ever to plan and to share risks regionally.”132

The FPA mandates regional planning, the promotion of interconnection, and the consideration of alternate sources of power. ■ NEPA requires the rigorous consideration of alternatives in a form which will facilitate public comment. Because FERC has abdicated these statutory duties, I respectfully dissent.