COFFIN, Chief Judge.

This case returns to us challenging the Administrator's Remand Opinion reaffirming his approval of the proposed once-through cooling system for the Seabrook Nuclear Power Plant. We had remanded his first decision on procedural grounds. Seacoast Anti-Pollution League v. Costle, 572 F.2d 872, cert. denied, ___ U.S. ___, 99 S.Ct. 94, 58 L.Ed.2d 117 (1978). The petitioners, Seacoast Anti-Pollution League and Audubon Society of New Hampshire, now seek review of a procedural aspect of the Administrator's hearing on remand, the substance of the decision, and the adequacy of the Administrator's explanation of his decision.

I. The Procedure on Remand

We explained the procedural history of this case and the design of the Seabrook facility in our earlier opinion, supra. We will not repeat ourselves. That opinion left the Administrator with several options for correcting his errors. He issued a "Response to the Remand Order in the Seabrook Case" choosing to hold a new hearing at which the technical experts who had advised him before would testify and be subject to cross-examination. He also decided to allow the parties to introduce evidence not offered at the prior hearings. Though petitioners suggest they may somehow have been injured by the breadth of the remand hearing, they do not argue that these aspects of the procedural order were erroneous. They challenge, however, the Administrator's instructions to his staff:

"I am directing my Staff not to appear at this hearing as proponents of any particular result, and avoid to the extent possible taking an adversary position in it. The Staff of Region I shall prepare a technical summary and analysis of the evidence submitted . . . . This report shall be non-adversary in nature, but shall contain specific conclusions and recommendations."

The hearing was held, and the staff appeared and cross-examined witnesses. The staff replied to proposed findings and conclusions of the applicant Public Service Company of New Hampshire (PSCO), an intervenor before us, but did not file proposals of its own. The staff's technical summary and analysis did not make specific conclusions, apparently in part because the staff experts were not unanimous.

Petitioners spend a good deal of space arguing that such an order is unprecedented in agency practice. If so, we might be curious why the Administrator issued such a novel order, but we would not, for that reason, have any basis to hold the order illegal. Absent law to the contrary, agencies enjoy wide latitude in fashioning their procedural rules. See Vermont Yankee Nuclear Power Corp. v. NRDC, 435 U.S. 519, 543-44, 98 S.Ct. 1197, 55 L.Ed.2d 460 (1978). The only law petitioners cite to support their position that the Administrator is compelled to allow his staff to participate as an adversary is unpersuasive. Office of Communications of United Church of Christ v. FCC, 138 U.S.App.D.C. 112, 425 F.2d 543 (1969), is not on point because it was concerned only with the impact agency neutrality can have on the allocation of the burden of proof. The regulation, 40 CFR § 125.36(a)(1), in which the Administrator defines "party" to include "officers or employees of the Environmental Protection Agency" does not answer the question. Petitioners argue, by reference to Black's Law Dictionary, that a party is someone "who takes sides" and that, therefore, the regulation requires the Agency to take a side. In fact, of course, the word "Party" has a far broader meaning and can refer to anyone "concerned or having or taking part in any affair, matter, transaction, or proceeding", to quote again from Black's. One can participate in a proceeding without taking a side. In short, merely labeling the Agency's staff a party tells us nothing about the duties of that party.

Finally, petitioners suggest that EPA must take a side to meet its Congressional mandate to protect the environment. Certainly EPA need not always side against applicants for permits to discharge pollutants. The staff was an active participant in the remand hearing, both in building and testing the record. Staff experts offered views on both sides of the issues. The only thing the staff did not do was advocate a particular outcome. The Administrator found that the record would not have been more complete had the staff taken sides.

We are not sure how petitioners can show that they are aggrieved by the Agency's neutrality given the significant likelihood that the staff would have chosen, if put to it, to support the Administrator's earlier decision (the substance of which we had not addressed in our opinion). But petitioners have failed to point us to any law requiring EPA's staff not to be neutral. Therefore this challenge fails.

II. Substantive Challenges

PSCO, as a point source applying for permission to discharge heat, is required to "demonstrate to the satisfaction of the Administrator" that the design of the Seabrook plant "will assure the protection and propagation of a balanced, indigenous population of shellfish, fish, and wildlife . . ." 33 U.S.C. § 1326(a). Moreover, "the location, design, construction, and capacity of cooling water intake structures [must] reflect the best technology available for minimizing adverse environmental impact." 33 U.S.C. § 1326(b). We must affirm the Administrator's decisions under these sections if they are supported by substantial evidence in the record. 5 U.S.C. § 706(2)(E).

The plant's design requires 824,000 gallons of cooling water per minute. The design calls for cooling water to be drawn from the ocean beyond the Hampton-Seabrook Estuary about 7,000 feet offshore from Hampton Beach. The water there is 58 feet deep. The intake structures will be between 10 and 17 feet above the ocean bottom and will pull water into the system at a velocity of about one foot per second. The water is ultimately discharged at another location through a diffuser system and is about 39 ° F warmer than the ambient water temperature. Living organisms sucked into the cooling system are not expected to survive.

The Administrator found that most mature fish likely to encounter the intake flow would be able to detect that flow and would have sufficient swimming speed to avoid being trapped. During their larval stages, however, many fish are at the mercy of the tides and currents and would have no way to escape. Even some juvenile fish might be too small, lacking the necessary speed to escape the intake's pull. The Administrator decided, though, that no species would exist in sufficient numbers at the intake location to be endangered. Though the intake would act as an additional large predator, there are other more dangerous threats in the natural environment, which fish are able to survive because they are highly fecund. Remand Opinion at 51. For instances, individual female winter flounder produce about 500,000 eggs annually and rainbow smelt produce between 25,000 and 50,000. Id. at 35.

Against this background, petitioners raise three relatively narrow issues: that the Administrator erred in concluding (a) that the impact on juvenile rainbow smelt would be small enough to assure the protection and propagation of the species; (b) that the impact on winter flounder larvae would also be small enough; and (c) that the cooling water intake location "reflect[s] the best technology available for minimizing adverse environmental impact."

A. Juvenile Smelt

The Administrator found:

"Although there is some doubt whether most juvenile smelt possess the requisite swimming speed to avoid the intake, the plant's effects on them will not, in any event, be significant because most smelt will not be found as far out as the intake structure or in water as deep as its opening. Although in its most recent sampling, Public Service Company's consultant found some smelt in water at least as far out as, and, perhaps, as deep as the intake structure and its opening, this does not alter the conclusion based upon almost 100 years of experience that smelt is in general an inshore fish found a mile or so offshore inhabiting the upper 3 fathoms of the water column. This was clearly the view of well qualified experts who considered both this general experience and the more recent sampling. The conclusion of these experts is consistent with the general experience of professional fishermen in the area." (Footnote omitted.) (Emphasis added.)

Petitioners challenge this conclusion as unsupported. Their theory is that the Administrator relied on a 1953 book which set the outer boundary of smelt populations at about one mile; whereas the same authors ten years later said that smelt ranged as far as six miles out to sea. Petitioners also point to the PSCO sampling referred to by the Administrator as proof that "smelts were fairly numerous at an area near the proposed intake." The fact is that the Administrator was fully aware that some smelt could be found at the intake location and that juvenile smelt could be entrapped and killed. These facts do not undermine his finding that "most" smelt will not come near the intake and, thus, will not be endangered.

The record certainly supports the Administrator's finding. The experts who testified at the remand hearing did not base their testimony — that smelt are generally inshore fish — solely on any one document or one study. They testified from their own expertise and from assembled reports of others' experience stretching back over a century. As they testified, a particular sampling that found a number of smelt in a particular location does not necessarily disprove the collected theories about smelt. The sampling did not address two critical questions — the relative number of smelt in the intake vicinity compared to elsewhere and the location of the smelt in the water column. Without having some idea of the number of smelt closer to shore, the absolute number found near the intake is not very meaningful. It does not help one determine whether the intake will affect the ability of the smelt to propagate and survive. In short, we cannot say that the Administrator was wrong, considering the entire record, to rely on the expert testimony.

B. Flounder Larvae

The Administrator noted that he could not be certain from the record whether the winter flounder found near the intake were part of a "local population" distinct from other populations of flounder found elsewhere. If the flounder are a local population, then any impact on them from the plant would be magnified, at least potentially, because the local flounders would not interbreed with and recruit from others of the species to augment the population or replace those killed by the intake. Nor, the Administrator found, did the record conclusively demonstrate whether the flounder spawned close to shore or further out to sea. If they spawned closer to shore, it might be more likely that they constituted a distinct population confined to the Hampton-Seabrook Estuary.

Petitioners rely on these failures of proof to argue that the Administrator's conclusion that intake would not significantly endanger winter flounder is unsupported by the record. Resolution of these issues, however, was not essential to the Administrator's conclusion.

The Administrator reasoned that if the flounder were a local population and spawned in the estuary, then few of the larvae would reach the areas of the intake or the diffuser (where the heated water being expelled might endanger the larvae). That is so because the tidal movement of the water in the estuary, the tidal plume, rarely, if ever, reaches those areas. Moreover, the younger larvae, which are most at the mercy of the water movement, would likely be high in the estuary, close to freshwater, and therefore, unlikely to be swept out of the harbor by the tidal plume. Remand Opinion at 41. "In addition, if flounder larvae leave the estuary, they will tend to disperse over a relatively broad area as the water in which they are transported mixes with the coastal water; and although some flounder larvae contained in this mixed coastal water may ultimately be carried northward toward the intake and diffuser, instead of southward with the generally-prevailing currents, only a small portion of these larvae will be affected by the intake and the diffuser." Id. at 42.

If, on the other hand, the flounder are a localized population but spawn offshore, "a much greater area or volume of water must be considered than the area of the diffuser and intake and the waters impacted thereby. There is a vast expanse of coastal area available for spawning. The area around the intake and diffuser does not appear to be unique with respect to flounder spawning and winter flounder larvae population should inhabit areas to the north and south and inshore and offshore from the intake and diffuser. There are no physiographic structures restricting the spawning area. Neither is there any reason to believe that the coastal area off the Hampton-Seabrook harbor entrance and in the immediate vicinity of the intake and diffuser is different or a more attractive spawning ground than other areas along or off the coast. Nor does any biological reason suggest itself for spawning limited to the area of concern." Id. at 42-43.

The Administrator was satisfied that the winter flounder larvae would not be endangered even if the uncertainties pointed to by petitioners were resolved against PSCO. The Administrator's logic strikes us as compelling, and petitioners give us no reason to doubt either the reasoning or the particular facts upon which it is based. Therefore, we must uphold the Administrator's conclusion that the protection and propagation of the flounder is assured. We see no reason to delve into factual problems that are not essential to the decision.

C. Intake Location

Petitioners' final substantive challenge, to the Administrator's approval of the intake location and design, is not a model of clarity. The Administrator decided that moving the intake further offshore might further minimize the entrainment of some plankton, but only slightly, and that the costs would be "wholly disproportionate to any environmental benefit". Remand Opinion at 49-50. Apparently petitioners read the cost figure considered by the Administrator, $20 million, as including the estimated costs of delay and reengineering as well as additional tunnelling. They suggest that the cost of delay is an improper consideration. The record is clear, however, that $20 million is the cost of the tunnelling alone. Petitioners, wisely, do not argue that the cost may not be considered, and no harm is done by noting that there would be other costs. The legislative history clearly makes cost an acceptable consideration in determining whether the intake design "reflect[s] the best technology available".

Petitioners also assign error to the Administrator's approval of the one-foot-per-second intake velocity. They allege that the Administrator neglected information indicating juvenile fish would be less able to detect and avoid the intake. He plainly did not ignore that information. His decision acknowledges that some juvenile fish will be killed, id. at 48, and, as we have already discussed, devotes some space to the special problems of juvenile rainbow smelt. Moreover, there is expert testimony that the intake velocity is unusually low and is the optimum intake velocity for the intake site.

III. The Administrator's Decision

Petitioners' last set of arguments is their weakest. They assert that the Administrator violated 5 U.S.C. § 557 by failing to explain adequately a number of procedural orders and factual or legal conclusions. As to most of these, however, they concede that the questioned order of finding was within the Administrator's discretion. As to the rest, it is plainly apparent either that there is sufficient record support or that the particular finding was not crucial to the decision. And, indeed, petitioners, with one exception, see note 3, supra, do not attack any of these orders or conclusions on their merits. Rather, they suggest that the failure to fully explain and provide record citations for each and every conclusions constitutes reversible error in and of itself.

Petitioners are wrong. The Administrative Procedure Act requires a reviewing court to take "due account . . . of the rule of prejudicial error." 5 U.S.C. § 706; cf. Seacoast Anti-Pollution League v. Costle, 572 F.2d 872, 881 n. 19 (1st Cir. 1978) (use of extra-record evidence not fatal unless it substantially prejudices petitioner). Moreover, the rule petitioners would have us apply would not serve the primary purpose of the explanation requirement, which, as petitioners recognize, is to facilitate appellate review of administrative decisions. See Northeast Airlines, Inc. v. CAB, 331 F.2d 579, 586 (1st Cir. 1964). We need no explanation of a decision that we are not asked to review.

This does not mean that petitioners do not have the right to a statement of reasons or that agencies do not err when they fail to explain decisions. It only means that we do not sit to resolve abstract arguments about whether particular explanations sufficed.

The Petition for review is dismissed.