This appeal challenges a December 31, 1974, order vacating approvals previously given for a project for the construction of a flood control dam (Pa-466), since the project violates the National Environmental Policy Act of 1969, 42 U.S.C. § 4321ff., and permanently enjoining defendants from taking any further action with respect to such construction until final approval by the appropriate Government officials after the filing and consideration of an environmental impact statement. As explained more fully below, an opinion filed on January 24, 1975, makes clear that the court concluded that there had been a substantive violation of § 5 of the Watershed Protection and Flood Prevention Act, 16 U.S.C. § 1005. The defendants appeal only from this conclusion that there has been a violation of 16 U.S.C. § 1005. We vacate the December 31, 1974, order only to the extent that it holds there has been a violation in the failure to apply to Dam Pa-466 only the benefits-costs determination required by 16 U.S.C. § 1005 and remand for further proceedings consistent with this opinion, including the filing and consideration of an environmental impact statement.

The area near Canadensis, Pa., is located at the confluence of several of the rivers and creeks that wind through the Pocono Mountain region of northeastern Pennsylvania, and, during recent decades, has been the scene of recurrent floods that have exacted a heavy toll in both human life and property. To mitigate the threat posed by the flooding, local governmental authorities and the Soil Conservation Service (SCS) of the Department of Agriculture (USDA), acting under authority of the Watershed Protection and Flood Prevention Act of 1954, 16 U.S.C. § 1001 et seq. (P.L. 566), developed a Work Plan for the construction of four floodwater detention dams north and west of Canadensis.1 This appeal from a final injunction restraining responsible SCS and USDA officials from awarding a contract for the construction of one of those dams (Dam Pa-466), presents as the principal issue whether § 5 of P.L. 566, 16 U.S.C. § 1001 et seq., as interpreted and applied, requires as a prerequisite to federal assistance that the benefits of Dam Pa-466 exceed its costs.

The floodwater detention dam at issue is a $2 million, 90-foot high, earthen structure proposed for construction along the Buck Hill Creek, a tributary of the Brodhead Creek. It is part of the four dam project originally envisioned for the Brodhead Creek area above Canadensis. A Work Plan prepared in March of 1961 described the project and provided, inter alia, that SCS would pay for the major portion of construction costs while the Monroe County Commissioners would acquire all necessary easements and rights of way (474a). The Work Plan also contained a cost/benefit analysis which ascribed to the project, evaluated as a whole, a benefit to cost ratio of 1.2 to 1. Pursuant to the Act, the 1961 Work Plan was transmitted to Congress where it was approved by resolutions of the Agriculture Committees of both Houses of Congress. See 16 U.S.C. §§ 1002, 1005(3).

Due primarily to Monroe County's failure to acquire certain rights of way from the Buck Hill Falls Co., little progress towards actual construction of the project was made during the years subsequent to 1961. In February of 1970, however, Buck Hill Falls Co. conveyed the necessary rights of way for a nominal consideration, subject to the condition that one of the four dams be deleted and Dam Pa-463 be modified as it was described in the 1961 Work Plan. These changes were incorporated into a Supplemental Watershed Work Plan issued in October of 1971. The Supplemental Work Plan contained an updated average annual cost/benefit analysis which fixed the benefit to cost ratio at 1.8 to 1. The project was again evaluated as a whole. SCS computed average annual cost by using a 3.25% discount rate.

All but one of the dams now remaining in the Brodhead Creek Watershed Project have either been completed or are now under construction. The third dam, however, Pa-466, has drawn the opposition of plaintiffs, who commenced this class action2 on December 10, 1974, seeking an injunction against the awarding of a contract for the construction of the dam. In addition to asserting a claim under the National Environmental Policy Act of 1969, 42 U.S.C. § 4321 et seq., plaintiffs alleged that SCS's decision to award the contract violated § 5 of P.L. 566, 16 U.S.C. § 1005(1), which authorizes federal participation in the construction of "works of improvement" only

"(a)t such time as the Secretary (of Agriculture) and the interested local organizations have agreed on a plan for works of improvement and the Secretary has determined that the benefits exceed the costs."

In support of their claim, plaintiffs contended that under § 5, as interpreted and applied by SCS, Dam Pa-466, viewed in isolation, must exhibit benefits exceeding its costs, regardless of the cost/benefit ratio of the Watershed Project evaluated as a whole. Plaintiffs contended that separately evaluated, the costs of Dam Pa-466 exceeded its benefits.

After conducting a five-day trial on the consolidated motions for preliminary and final injunctive relief, the district court, on December 31, 1974, entered an injunction carrying out its conclusions that filing of an environmental impact statement (EIS) was required and that the decision to construct Dam Pa-466 violated § 5 of P.L. 566.3 The district court expressly agreed with plaintiffs' contention that separate cost justification is required for each dam in a multi-dam project. The court also concluded, inter alia,4 that SCS had utilized an improperly low discount rate (3.25%) in violation of § 80 of the Water Resources Development Act, 42 U.S.C. § 1962d-17.5 This timely appeal followed.6

SCS does not challenge the district court's determination that an EIS is required under the circumstances of this case. With respect to P.L. 566, however, SCS contends that neither its cost/benefit determinations nor its decision to use a 3.25% discount is subject to judicial review. On the merits, SCS argues that neither § 5 of P.L. 566 nor its own project evaluation procedures require separate cost/benefit justifications for individual structures in a multi-dam watershed project. In SCS's view, all that was required under the circumstances of this case was that the benefits of the project, evaluated as a whole, exceed its costs.

Our analysis of the availability of judicial review begins with the question of whether plaintiffs have standing to sue and with § 10 of the Administrative Procedure Act (APA), 5 U.S.C. § 702.7 As interpreted by the Supreme Court, a person is "adversely affected or aggrieved within the meaning of a relevant statute" and hence has standing to sue under § 10 if he alleges (1) that he has or will sustain some actual or threatened injury in fact resulting from the challenged agency action, and (2) that the alleged injury is to an interest "arguably within the zone of interests to be protected or regulated by the statute in question." Data Processing Service Organizations, Inc. v. Camp, 397 U.S. 150, 153-54, 90 S. Ct. 827, 830, 25 L. Ed. 2d 184 (1970); see e. g., United States v. SCRAP, 412 U.S. 669, 686, 93 S. Ct. 2405, 37 L. Ed. 2d 254 (1973).

The plaintiffs in this litigation are owners of property and residences situated in the area immediately surrounding the proposed dam. They have alleged that construction of the dam will diminish the value of their properties and impair their enjoyment of the area's recreational and aesthetic resources.8 Such economic and conservational interests are clearly sufficient to satisfy the requirement of injury in fact. See, e. g., United States v. SCRAP, supra, 412 U.S. at 686, 93 S. Ct. 2405.

In addition, we believe that the various interests asserted by plaintiffs are "arguably" within the zone of interests to be protected by P.L. 566. Our view is based on the apparent purpose of the Act to benefit the residents of areas affected by flood dangers. And it is further supported by the language of § 1 of the Act, 16 U.S.C. § 1001, which identifies as one of the purposes of the statute the goal of "preserving, protecting and improving the Nation's land and water resources and the quality of the environment."9 Although plaintiffs might have had some difficulty satisfying the Data Processing test, if our analysis were focused exclusively on the purposes of the cost/benefit requirement of § 5, we do not believe the scope of our inquiry is so circumscribed. Davis v. Romney, 490 F.2d 1360, 1365 (3d Cir. 1974), makes clear that we are to examine "the statute, not the particular provision purportedly violated, to ascertain whether plaintiffs were 'aggrieved . . . within the meaning of a relevant statute.' " See 5 U.S.C. § 702.

Apart from the question of standing to sue, our inquiry into the availability of judicial review requires a separate examination of whether Congress has placed the agency's action beyond the reach of judicial cognizance. Section 10 of the APA provides for judicial review of agency action "except to the extent that (1) statutes preclude judicial review; or (2) agency action is committed to agency discretion by law." 5 U.S.C. § 701(a). As the party claiming the applicability of these exceptions to the general rule of reviewability, SCS bears the heavy burden of demonstrating by " clear and convincing evidence" that Congress intended to restrict access to the courts. Dunlop v. Bachowski, 421 U.S. 560, 567, 95 S. Ct. 1851, 41 L. Ed. 2d 377 (1975); Barlow v. Collins, 397 U.S. 159, 166-67, 90 S. Ct. 832, 25 L. Ed. 2d 192 (1970); A. O. Smith v. FTC, 530 F.2d 515, 521 (3d Cir. 1976). For the reasons set forth below, we believe SCS has failed to meet that burden with respect to either cost/benefit analyses under P.L. 566 or the Secretary's choice of discount rates under the Water Resources Development Act of 1974, 42 U.S.C. § 1962d-17.

Section 5 of the Watershed Protection and Flood Prevention Act, 16 U.S.C. § 1005(3), requires that a copy of any plan for "works of improvement" and a justification therefor be transmitted to Congress whenever, inter alia, the estimated federal contributions to construction costs exceeds $250,000. The purpose of this requirement is made clear in § 2 of the Act, 16 U.S.C. § 1002, which states:

"No appropriation shall be made for any plan involving an estimated Federal, contribution to construction costs in excess of $250,000, . . . unless such plan has been approved by resolutions adopted by the appropriate committees of the Senate and House of Representatives: . . .."

In SCS's view, this statutory scheme vests all oversight responsibilities in the congressional committees and in Congress as a whole through the appropriation process. Thus, the argument continues, the Watershed Protection and Flood Prevention Act is a statute that implicitly "preclude(s) judicial review." See 5 U.S.C. § 701(a)(1).

The critical difficulty with SCS's argument is that it rests on an invalid major premise that the purpose of the legislative scheme is to give Congress exclusive jurisdiction to enforce the substantive requirements of the Act. SCS has pointed to nothing on the face of the statute or in its legislative history that indicates Congress viewed itself as the final arbiter of whether a project plan complied with the Act's mandate that benefits exceed costs. The Act does not require the Secretary to prepare a written cost/benefit analysis. And the sections requiring committee approvals as a prerequisite to appropriations state only that the "plan and the justification therefor" must be transmitted to Congress. If Congress had intended to assume the duty of ensuring that benefits exceeded costs, we believe it would have explicitly required preparation and transmission to Congress of a written economic analysis. Further, review of economic data can be quite time-consuming and we are reluctant to impose the burden of such review on Congress where Congress may well have viewed the scheme it established as a means of freeing itself from the onus of individually authorizing watershed projects.10 Particularly because the approval requirement applies only to the bigger and more expensive projects, we believe Congress viewed the committee approval process essentially as a budget control measure, rather than as the exclusive means of ensuring SCS compliance with the substantive mandates of the Act. Cf. Merriam v. Kunzig, 476 F.2d 1233 (3d Cir.), cert. denied, 414 U.S. 911, 94 S. Ct. 233, 38 L. Ed. 2d 149 (1973); 40 U.S.C. § 606(a).

Our conclusion is in harmony with the series of cases on which SCS principally relies,11 all of which arose under the Flood Control Act of 1936, 33 U.S.C. § 701a et seq. In each of those cases the courts refused to examine cost/benefit analyses prepared by the Corps of Engineers because Congress had specifically authorized the projects in question after receiving and reviewing the reports. Unlike P.L. 566, no substantive provision of the Flood Control Act of 1936 required an administrative official to determine that benefits exceeded costs before proceeding. More significantly, the individual projects in the cases cited to us were specifically authorized by Congress after consideration of their economic merits. P.L. 566, in contrast, does not provide for individual project authorization and no such authorization has been passed.12

We have also concluded that the Secretary's choice of discount rates under § 80 of the Water Resources Development Act of 1974, 42 U.S.C. § 1962d-17, is subject to judicial review under appropriately limited standards.13 Section 80 authorizes use of the 3.25% discount rate employed by SCS in the October 1971 Supplemental Work Plan only if "the appropriate non-Federal interests (gave) . . . satisfactory assurances to pay the required non-Federal share of project costs . . . " prior to December 31, 1969.14 Relying on Akers v. Resor, 339 F. Supp. 1375 (W.D.Tenn.1972), SCS contends that the "extent" of the Secretary's discretion under this statute is so great that the determination of whether any "assurances" are "satisfactory" has been "committed to agency discretion by law" and hence is unreviewable. 5 U.S.C. § 701(a)(2).

Cases falling within this "very narrow exception" are "rare" for the exception applies only "where 'statutes are drawn in such broad terms that in a given case there is no law to apply.' " Citizens to Preserve Overton Park v. Volpe, 401 U.S. 402, 410, 91 S. Ct. 814, 821, 28 L. Ed. 2d 136 (1971). In our view, this is not such a case. See Western Addition Community Organization v. Weaver, 294 F. Supp. 433 (N.D.Cal.1968). The statute indisputably vests a broad discretion in the Secretary. But that discretion is not wholly without judicially discernible limits. Although the administrative and legislative history of the Act does not define the phrase "satisfactory assurances," it does furnish at least some "law"15 against which a court can measure the Secretary's determination. Furthermore, the subject matter requires no special expertise and is not of such a nature that judicial consideration is for any other reason impractical or inappropriate.

The principal issue on the merits of this appeal is whether each individual dam in a multi-dam watershed project must have a benefit to cost ratio of greater than 1 to 1 before federal assistance in the construction of the dam is permissible. For the reasons set forth below, we conclude that multi-dam watershed projects may be cost analyzed as a single unit and that, under the circumstances shown in this record, it was not "arbitrary, capricious, an abuse of discretion, or otherwise not in accordance with law,"5 U.S.C. § 706(2)(A), to treat the dams in the Brodhead Creek project as a single unit.

It is clear at the outset that nothing in the Act itself requires that each individual "work of improvement" proposed in a multi-unit watershed project have a favorable cost/benefit ratio. Section 5 of the Act provides in pertinent part:

"At such time as the Secretary and the interested local organizations have agreed on a plan for works of improvement, and the Secretary has determined that the benefits exceed costs . . . the Secretary is authorized . . . to participate in the installation of such works of improvement in accordance with the plan."

16 U.S.C. § 1005(1). As we understand it, § 5 requires only that the benefits of every "plan" exceed the costs associated with implementing the plan; not that every segment of the plan be cost-justified when viewed in isolation. Furthermore, § 2 of the Act explicitly states that "(a) number of subwatersheds when they are component parts of a larger watershed may be planned together when the local sponsoring organizations so desire." 16 U.S.C. § 1002.

Although the Act may thus be read to permit the cost justification as a single unit of a group of dams that have been planned together (67a), plaintiffs contend that SCS's own project evaluation manuals The Watershed Protection Handbook (Handbook) and the Economics Guide for Watershed Protection and Flood Prevention (Guide) mandate separate evaluations for individual dams. SCS vigorously disputes this interpretation of its own economic evaluation procedures.

The Guide and the Handbook each establish a three-step project formulation and evaluation procedure. The required steps, as described in the Handbook, are as follows:

1. "Reach agreement with the sponsoring local organization on a minimum level of protection to solve the flood problems of the entire watershed community."

2. "For each hydrologic unit having an essentially separate flood plain, develop the least costly system of measures . . . which are needed to achieve the agreed upon levels of protection . . ."

3. "Evaluate the benefits that will accrue to each system."16

It is apparent from the language quoted above that a group of separate measures may be treated as a unitary "system of measures" and may be evaluated together as long as the individual dams do not have "essentially separate flood plain(s)" and are designed to meet agreed upon levels of protection. This point is made particularly in the Guide which discusses the concept of "Evaluation Units" and states:

"In project formulation consideration needs to be given to the interrelationship of structures. Such an interrelationship exists when a group of structures protect a common flood plain. Here the most economical system for attaining the agreed upon level of protection may be evaluated as a unit."17

In reaching its conclusion that the Guide requires a separate cost/ benefit justification for each dam in a multidam watershed project, the district court relied exclusively on a passage from Chapter 1 of the Guide. However, SCS interprets the language of that passage, which we have set out in the margin,18 in a manner that is consistent with the principle that interrelated structures may be evaluated together. And, in our view, that interpretation is reasonable and must therefore be accepted. See, e. g., Lucas Coal Co. v. Interior Bd. of Mine Operations App., 522 F.2d 581, 584 (3d Cir. 1975).

To parry the thrust of the foregoing view of the evaluation procedures established in the Guide and the Handbook, plaintiffs point out that the Guide articulates a principle of "maximizing net benefits" which appears to be applicable to all project evaluations.19 As explained in the Guide:

"From an economic viewpoint, the optimum scale of project development is the point at which net benefits are at a maximum. Net benefits are maximized when the benefits added by the last increment of scale or scope of project development are equal to the cost of adding that increment."

Guide, Chap. 2, section II-c. In plaintiffs' view, the principle of maximizing net benefits requires SCS to evaluate each dam in a multi-dam project. Unless each dam has a favorable cost/benefit ratio, they contend, the benefit to cost ratio of the project is reduced and the maximization principle is violated. We disagree.

The principal difficulty with plaintiffs' interpretation of the maximization principle is that it is seriously inconsistent with other sections of the Guide. By requiring SCS to make separate and independent evaluations of the component parts of a project of interrelated measures, plaintiffs' construction would undermine the apparent purpose of permitting single evaluations of groups of measures.20 As we have stated in an analogous context, if plaintiffs' view were adopted it would be "difficult to see how the number of separate (evaluations) could be kept to a minimum." McCullough v. Redevelopment Authority of Wilkes-Barre, 522 F.2d 858, 869 (3d Cir. 1975). In addition, under plaintiffs' interpretation, SCS might be required to reduce the desired level of protection whenever it became apparent that reaching the desired level would involve construction of a cost-inefficient dam. But nothing in the Guide suggests that the principle of maximizing net benefits invariably takes precedence over the goal of attaining agreed upon levels of protection.

In SCS's view, there is a limited exception to the policy of maximizing net benefits where dams are interrelated in the sense that they protect a common flood plain and meet the level of protection established by SCS and the appropriate local authorities in their work plan. Although the Guide is not a model of clarity in this regard, we think this construction is more reasonable than the interpretation offered by plaintiffs, since it avoids the difficulties described above. As long as the benefits of a group of structures exceed the costs of the group and the various structures are "interrelated" in the sense that they serve to protect a common area, no purpose required by Congress or the SCS procedures would be served by requiring that the structures be located and built in such a way that the benefits of each would exceed its costs. Further, the Guide explicitly permits an exception to the maximization principle where intangible benefits, incapable of monetary evaluation, are associated with the project. Guide, Chap. 1, section I-C-2. And the 1961 Work Plan indicates that such intangible benefits were considered in the process of project formulation (471a).

Since we are dealing here with materials prepared by SCS officials and used by them on a continuing basis, their interpretation of the Handbook and the Guide should be given "controlling weight" under the circumstances of this case. McCullough v. Redevelopment Authority of Wilkes-Barre, 522 F.2d 858, 870 n. 32 (3d Cir. 1975); Budd Co. v. Occupational Safety & Health Rev. Com'n,513 F.2d 201, 205 (3d Cir. 1975). As we stated in Lucas Coal Co. v. Interior Bd. of Mine Operations App., 522 F.2d 581, 584 (3d Cir. 1975):

"An agency's explication of its regulations if reasonable, . . . is controlling despite the existence of other interpretations that may seem even more reasonable."

Moreover, plaintiffs seem to believe that SCS is bound to follow the principles and procedures developed in the Guide in every case. However, as we have already noted, P.L. 566 itself does not mandate such evaluations. And since the Guide and the Handbook are merely internal operating procedures, rather than regulations officially promulgated under the APA or otherwise, they do not prescribe any rule of law binding on the agency. Estrada v. Hill, 401 F. Supp. 429, 437-38 (N.D.Ill.1975); Brown v. Lynn, 392 F. Supp. 559 (N.D.Ill.1975); see McCullough v. Redevelopment Authority of Wilkes-Barre, supra, 522 F.2d at 867 & n. 27.21 In fact, the Guide itself commences with the observation that "no single procedure can be used in every watershed" and provides that alternative procedures may therefore be used.22 Thus, deviations from established procedures may be tolerated.

For purposes of review, the Guide and the Handbook serve only as indicia of whether the evaluation procedures adopted in a particular case are " arbitrary and capricious." See Estrada v. Hill, supra. And on this record, we cannot say that the decision to construct Dam Pa-466 without making a separate cost/benefit determination was "arbitrary and capricious." The two Work Plans clearly establish that Dam Pa-466 was in fact treated as part of a single "evaluation unit."23 In addition, it appears that all of the dams protect "the common damage center of Canadensis" and thus are interrelated.24

In the second paragraph of Chapter 2-E of the Guide (670a), this wording is used:

"A simple approach is to select first a group or combination of structures that will give the desired level of protection at the least cost, then test the total group for economic feasibility. . . . If the ratio of benefits to costs for the selected group is favorable, and if mutually desired, the feasibility of adding successive increments to the system up to the point where benefits are maximized may be undertaken."^24a

It is noted that this language discusses one approach and not the exclusive approach. The level of protection agreed upon in this case was determined after discussion and consideration of the cost of protection. Although the record does not state positively that any minimum level of protection was ever agreed upon, we think the record supports the contention of SCS that the level of protection provided by the three dams established in the amended 1971 Work Plan is the level of protection ultimately agreed upon by SCS and the local sponsors of the project.25

Accordingly, the order of December 31, 1974, will be vacated to the extent it enjoins construction of Dam Pa-466 as a violation of P.L. 566, and the case will be remanded for further proceedings consistent with this opinion. We believe the trial judge should feel free to reconsider his determination that satisfactory assurances were not given prior to December 31, 1969 (see page ----, above, including note 14).