Opinion for the Court filed by Senior Circuit Judge McGOWAN.
McGOWAN, Senior Circuit Judge:This petition for attorneys’ fees under section' 307(f) of the Clean Air Act, 42 U.S.C. § 7607(f) (1982), seeks compensation for the petitioners' attorneys’ work in Sierra Club v. Environmental Protection Agency, 719 F.2d 436 (D.C.Gir.1983), cert. denied, — U.S. -, 104 S.Ct. 3571, 82 L.Ed.2d 870 (1984). The petitioners there sought from this court direct review of regulations proposed by the Environmental Protection Agency (EPA) to implement Congress’s 1977 amendments to the Clean Air Act, Clean. Air Act Amendments of 1977, Pub.L. No. 95-95, 91 Stat. 685 (codified at 42 U.S.C. § 7403 et seq. (1982)). The petitioners met with varying degrees of success on the various issues they brought before this court. The petitioners also successfully opposed the efforts of an intervenor to persuade the Supreme Court to grant a writ of certiorari. The government did not participate in the litigation before the Supreme Court. The petitioners here seek fees for their work before both this court and the Supreme Court, from the government and the intervenors, respectively. We grant the petitioners $51,360.52 in fees and costs from the government for their work before this court, but deny them any fees from the intervenors.
We treat first the request for fees against the government. The starting point for any calculation of attorneys’ fees is the “lodestar,” the product of a reasonable hourly rate and the number of hours reasonably expended on substantive issues on which the petitioner met the statutory threshold of success. See Copeland v. Marshall, 641 F.2d 880, 891 (D.C.Cir.1980) (en banc). Next, we discuss whether there should be any enhancement of the lodestar in this case. We then decide the amount of fees due for the attorneys’ work on the fee petition. Finally, we discuss the inappropriateness of awarding the petitioners fees against the intervenors in this case.
I
Section 307(f) of the Clean Air Act provides in full:
In any judicial proceeding under this section, the court may award costs of litigation (including reasonable attorney and expert witness fees) whenever it determines that such award is appropriate.
42 U.S.C. § 7607(f) (1982). There are two judicial glosses on this broad provision. First, the party awarded fees must have met with at least some success on the merits. See Ruckelshaus v. Sierra Club, 463 U.S. 680,103 S.Ct. 3274, 77 L.Ed.2d 938 *111(1983). Second, the party must have served the public interest by assisting in the proper implementation of the statute. See Alabama Power Co. v. Gorsuch, 672 F.2d 1, 3 (D.C.Cir.1982) (per curiam) (public-interest criterion is the “dominant” one in determining fee awards); Carson-Truckee Water Conservancy District v. Secretary of the Interior, 748 F.2d 523, 525-26 (9th Cir.1984) (discussing D.C. Circuit cases and concluding that even prevailing parties must meet public-interest criterion), cert. denied, — U.S. -, 105 S.Ct. 2139, 85 L.Ed.2d 497 (1985); see also H.R.Rep. No. 294, 95th Cong., 1st Sess. 337 (1977), U.S. Code Cong. & Admin. News 1977, pp. 1077, 1416 (section 307(f) included to help “assure proper implementation and administration of the act”). There is little question in this case that the petitioners’ extensive and partly successful challenges to a complex and important set of EPA regulations satisfy this second criterion. We therefore focus our attention upon the degree of success requisite to an award of attorneys’ fees.
Most fee statutes limit the beneficiaries to “prevailing” parties. See, e.g., 42 U.S.C. § 1988 (1982) (allowing award of fees to “prevailing” plaintiffs in civil-rights cases); 5 U.S.C. § 552(a)(4)(E) (allowing award of fees to complainant in Freedom of Information Act case who has “substantially prevailed”). A significant body of jurisprudence has been developed to determine when a party “prevails” for the purposes of a fee award. See, e.g., Maher v. Gagne, 448 U.S. 122, 129-30, 100 S.Ct. 2570, 2574-75, 65 L.Ed.2d 653 (1980) (party may “prevail” through consent decree); Hanrahan v. Hampton, 446 U.S. 754, 100 S.Ct. 1987, 64 L.Ed.2d 670 (1980) (per curiam) (interlocutory dispositions insufficient grounds for party to be considered “prevailing”).
Congress, however, specifically stated that awards under section 307(f) are not to be limited to “prevailing” parties:
The committee did not intend that the court’s discretion to award fees under this provision should be restricted to cases in which the party seeking fees was the “prevailing party.”
H.R.Rep. No. 294, at 337, U.S.Code Cong. & Admin. News 1977 at 1416. The Supreme Court has held that Congress did not thereby intend to allow the award of fees to parties wholly unsuccessful on the merits. Ruckelshaus v. Sierra Club, 463 U.S. at 693-94, 103 S.Ct. at 3281-82. Congress simply intended by section 307(f) to make eligible for fee awards a set of litigants intermediate in size between the group of litigants eligible as prevailing parties and the group of all litigants.
Because the Ruckelshaus Court was faced with a petitioner wholly unsuccessful on the merits, id. at 681, 103 S.Ct. at 3278, it did not reach the question of precisely what showing would be necessary for an award of fees to a partly successful litigant. The Court did, however, provide some guidance. The requisite success had to be more than “trivial” or “purely procedural,” id. at 688 n. 9, 103 S.Ct. at 3278 n. 9, but the court awarding the fees should not become mired in details such as whether the success is “substantial” or on a “central issue.” Id. Congress has also provided some guidance. As a matter of policy, we know that the award of fees under section 307(f) should help to “assure proper implementation and administration of the act,” H.R.Rep. No. 294, at 337, U.S.Code Cong. & Admin. News 1977 at 1416. Moreover, in discussing a similar fee provision in the Clean Air Act of 1970, Congress stated that the underlying action must be “legitimate.” S. Rep. No. 1196, 91st Cong., 2d Sess. 38 (1970) (discussing section 304(d)).1 We believe that it is a fair characterization of the intent of Congress and the directive of the Supreme Court to state that a party who is awarded fees under section 307(f) must meet with a modicum of success on the merits before it is eligible for an award of fees, and we apply that standard here.
The petitioners clearly achieved a modicum of success — indeed, a good deal of *112success — in the litigation taken as a whole, see infra pp. 803-07 (discussing petitioners’ degree of success on merits), and are therefore eligible for some fees. To decide that a party is eligible for some fees, however, is not necessarily to determine that it is eligible for fees on all claims it made before the court. In the context of statutes awarding fees to prevailing parties, for example, the court must examine all the distinct issues in the case and determine on an issue-by-issue basis whether the petitioning party is eligible for fees on each issue. See, e.g., Martin v. Lauer, 740 F.2d 36 (D.C.Cir.1984).
The Supreme Court has spoken on this issue most recently in Hensley v. Eckerhart, 461 U.S. 424, 103 S.Ct. 1933, 76 L.Ed.2d 40 (1983). When a prevailing party succeeds on only some of its claims, at least two questions must be addressed:
First, did the plaintiff fail to prevail on claims that were unrelated to the claims on which he succeeded? Second, did the plaintiff achieve a level of success that makes the hours reasonably expended a satisfactory basis for making a fee award?
Id. at 434, 103 S.Ct. at 1939. Where a plaintiff presents in one lawsuit “distinctively different claims for relief that are based on different facts and legal theories,” id., the court cannot allow a plaintiff to recover fees on the unsuccessful claims: “The congressional intent to limit awards to prevailing parties requires that these unrelated claims be treated as if they had been raised in separate lawsuits, and therefore no fee may be awarded for services on the unsuccessful claim.” Id. at 435, 103 S.Ct. at 1940 (footnote omitted). A plaintiff should not be able to force his opponent to pay for the legal services involved in bringing groundless claims simply because those unsuccessful claims were brought in a lawsuit that included successful claims.
We believe that the rationale of the Hensley Court in requiring an issue-by-issue consideration of success is applicable to this case as well.2 In Hensley, the Court noted that a party should not be able to “piggyback” fees for unsuccessful claims upon unrelated, successful claims. Requiring an issue-by-issue analysis prevents parties who failed to meet the statutory' threshold of success on some issues from receiving fees for their work on those issues.
The Ruckelshaus Court emphasized that Congress had concerns in enacting section 307(f) similar to those involved in statutes awarding fees to “prevailing” parties: Congress did not want courts to award fees under section 307(f) to parties wholly unsuccessful on the merits. The Ruckelshaus Court considered whether a party who has been completely unsuccessful on all the issues should be eligible for fees. Although that case is therefore factually distinguishable from the instant case, its rationale is applicable to the question of whether a party completely unsuccessful on a particular issue should be eligible for fees on that particular issue. To award fees to a completely unsuccessful party goes against the grain of both the general American rule against any award of fees and the statutory exceptions allowing fees only to prevailing parties. Ruckelshaus, 463 U.S. at 682-86, 103 S.Ct. at 3276-77. As Hensley shows, this rationale applies to the consideration of individual issues as well as to the consideration of whether any fees are to be awarded at all. Hensley, 461 U.S. at 435, 103 S.Ct. at 1940. The Ruckelshaus Court was also clearly impressed by the potential inequity of forcing a defendant who had shown that he had not violated the law in any way to pay for his opponent's efforts. Ruckelshaus, 463 U.S. at 692, 103 S.Ct. at 3280. We believe that it would be an inequity similar in kind, if lesser in magnitude, to force a defendant who has completely vindicated his own position on a particular issue to pay for his *113opponent’s efforts on that issue. We therefore believe that courts awarding fees under section 307(f) must ask the first question of the Hensley court: Did the plaintiffs lose on any issues unrelated to the successful claims? If the answer to that question is “yes,” then we must be extremely reluctant to award fees for time expended on those unsuccessfully raised issues.
The second, related question that the Hensley Court requires us to ask is whether the success obtained by the petitioner is proportional to the effort reasonably expended on the case. Where a petitioner obtains excellent results, his attorneys’ fees should be “fully compensatory” for the hours worked; where the petitioner has obtained only partial or limited success, his attorneys should not be fully compensated. 461 U.S. at 435-36, 103 S.Ct. at 1940-41.
In the instant case, the petitioners achieved a modicum of success on some but not all of the issues that they raised, and thus were granted some but not all of the relief that they requested. The petitioners therefore deserve attorneys’ fees for some but not all of the time that they have expended in this litigation. Because they achieved no success at all on some claims unrelated to the claims on which they did succeed, the result of our addressing the first Hensley question is that we need not award the petitioners fees on all claims. Because the petitioners had only partial or limited success, the result of our addressing the second Hensley question is similarly that we need not award them the full fees that they seek.
The petitioners are of course entitled to fees'for litigating the claims on which they obtained at least a modicum of success. After a brief statement of the subject matter of the case, we turn to an issue-by-issue consideration of the petitioners’ success.3
II
In 1970, Congress passed the Clean Air Act to regulate the emissions from, inter alia, the smokestacks of electric plants and other major sources of air pollution. Such stationary sources were required to modify their emissions so that local air quality met certain minimum standards. Many sources chose to ensure that those standards were met by using taller stacks rather than decreasing the quantity of pollution emitted. Taller stacks disperse a given amount of pollutants over a greater area and thus increase local air quality, although at the expense of air quality in distant areas. In 1977, Congress amended the Clean Air Act to discourage stationary sources from making this tradeoff. Congress prohibited the EPA from regulating the actual heights of the stacks, 42 U.S.C. § 7423(c) (1982), but Congress mandated that emissions limitations for sources be calculated as if the stacks were no higher than that dictated by “good engineering practice” (GEP), id. § 7423(a). When EPA proposed the regulations implementing this statutory directive, the petitioners mounted a wide-ranging challenge to the legality of the regulations. This court affirmed some of the EPA’s regulations, reversed some as beyond the EPA’s statutory authority, and remanded some for further consideration by the EPA in light of the court’s opinion. Sierra Club v. EPA, 719 F.2d at 470.
The petitioners challenged regulations governing two of the three methods that EPA proposed for determining the permissible GEP stack height. The overall objec*114tive of the various methods was to produce a GEP height sufficient to allow dispersal of pollutants to prevent excessive “down-wash,” the phenomenon whereby structures very close to the stack disrupt airflow so as to produce excessively high concentrations of pollutants in the immediate vicinity of the structures. The source was allowed to pick whichever method produced the highest GEP height.
The unchallenged “method” was simply to build a stack sixty-five or fewer meters high. The first challenged method (the “formula method”) was to determine GEP height by using one of two simple mathematical formulae dependent only upon the height and/or width of “nearby” structures. The second challenged method (the “demonstration method”) required the source to demonstrate, using either a complicated mathematical model or an actual field study, that use of the simple mathematical formula would result in “excessive” concentrations of pollutants in the area. The source could then use as the GEP height the lowest height shown by the demonstration to be sufficient to avoid excessive downwash. See Sierra Club v. EPA, 719 F.2d at 442-43 (describing three methods in more detail).
We divide the case into its component issues along the same lines as did the merits panel. While all the issues in the case are related in the sense that they arise from the same set of regulations and the same administrative record, we do not believe that they are therefore inseparable for the purposes of attorneys’ fees. Each issue involves a particular substantive concern of the petitioners with a particular aspect of EPA’s regulations. While the petitioners’ legal theories are similar on many issues in the sense that they are grounded in the standard of our review prescribed in the Administrative Procedure Act, the different policy rationales and statutory provisions set forth by the agency as support for its decisions on different issues make the different claims legally distinct. The petitioners could be granted relief on any one issue without necessarily obtaining their desired relief on any other issue.
A. The Regulations’ Definition of “Nearby”
This case involved two questions concerning EPA's definition of “nearby.” The first was whether the definition of “nearby” for the purposes of the formula method was permissibly set at one-half mile or less from the source. 719 F.2d at 443-44. This court upheld EPA’s proposed regulations on this matter.
The second question involving “nearby” was whether EPA could fail to provide any uniform numerical definition of “nearby” for the purposes of using the demonstration method, and rely instead upon the results of the demonstration for determining which structures were near enough to the source to cause excessive downwash. 719 F.2d at 444-46. This court found the EPA’s proposed definition of “nearby” for purposes of the demonstration method to be inconsistent with congressional intent.
The petitioners were completely successful on the second question and unsuccessful on the first. Both questions were of significant import to the implementation of the Act, since the definition of “nearby” is crucial to determining GEP stack height and the determination of the GEP stack height is the core of the statutory solution to the problem of tall stacks. We believe that the petitioners’ complete success on the definition of “nearby” in the demonstration method is sufficient to demonstrate a modicum of success on the general issue of defining “nearby,” despite their lack of success in contesting the EPA’s definition of “nearby” in the formula method. The petitioner is therefore entitled to fees for hours reasonably expended on this issue.
B. Definition of “Excessive Concentrations”
The proposed EPA regulations allowed use of the demonstration method only when the source owner could show that the formula method would result in *115downwash causing “excessive concentrations” of pollutants in the vicinity of the plant. EPA proposed to define “excessive concentrations” as an increase of forty percent or more in pollutants compared to air quality in the absence of the downwash-creating obstacle. The petitioners argued that such a standard violated Congress’s intent to define “excessive concentrations” at least partly in terms of their threat to human health. 719 F.2d at 446-50. This court agreed with petitioners and remanded the proposed regulation to the agency for further consideration, although in the course of reaching that decision this court disagreed with the petitioners’ precise reading of the proper base to use in calculating the 40% increase resulting in an “excessive” concentration, id. at 449.
The petitioners met with more than a modicum of success on this issue, and thus are entitled to fees for hours expended challenging this aspect of the regulations. They successfully challenged the premise underlying EPA’s definition of “excessive concentrations”—that the EPA need take into account only current sound engineering practice. The petitioners were unsuccessful in challenging the particular base that EPA proposed to use in calculating the percentage increase sufficient to result in an “excessive concentration” of pollutants. We believe that their overall level of success, while mixed, is sufficient to meet the relatively generous standard of section 307(f).
C.Failure to Consider Plume Rise
Air propelled upwards has a tendency to continue rising as a result of its momentum. Hot air rises as well. The heated gas propelled upwards through a smokestack therefore has a tendency to rise, a phenomenon known as “plume rise.” The formula method did not take plume rise into account in determining GEP height.
Becauste plume rise makes it less likely that the plume and its accompanying pollutants will fall back to the ground in the immediate vicinity of the stack, plume rise makes a stack effectively “taller” than the simple height of the stack would indicate. The petitioners argued that plume rise was generally significant enough that EPA could not ignore its effects without granting source owners credit for stacks of. greater height than that necessary to avoid downwash. EPA argued that plume rise was insignificant and thus could be excluded from the formulae for calculating GEP height. We agreed with EPA and upheld the portion of the regulations excluding plume rise from GEP calculations. 719 F.2d at 450-52.
The petitioners failed to achieve any success in challenging the EPA’s treatment of plume rise and we therefore hold that they are ineligible for fees on this issue.
D. Inclusion of Plume Impaction
When an exhaust plume fails to disperse sufficiently before reaching a taller terrain obstacle in its path, the plume’s pollutants may collect against the hillside or mountainside, a phenomenon known as “plume impaction.” EPA proposed to give stack-height credit to sources sufficient for them to avoid causing plume impaction, even though the relevant obstacles were not “nearby.” We agreed with the petitioners that EPA’s proposed regulations violated Congress’s intent to limit stack-height credit to nearby obstacles and reversed EPA’s proposed regulations on this score as beyond its statutory authority. 719 F.2d at 452-56.
The petitioners were completely successful in challenging the EPA’s treatment of plume impaction and are clearly entitled to fees for their efforts in connection with this issue.
E. Requiring Demonstrations only for Stacks Above Formula Height
Believing the formula method to be a highly accurate indicator of GEP height, EPA proposed to require sources to use the demonstration method only when the source sought a stack-height credit greater than that indicated by the formula method. The petitioners argued that EPA *116should also require demonstrations whenever local or federal pollution authorities believed the formula method overly generous or whenever a source sought to raise the height of an existing stack. This court held that EPA’s faith in the accuracy of the formula method as an indicator of GEP height was sufficiently excessive to be an abuse of discretion. 719 F.2d at 456-60. EPA had insufficient factual evidence to conclude that the formula rarely gave sources more stack-height credit than they deserved. Id. at 459. In addition, EPA claimed that the formula was acceptable because it accurately reflected current engineering practice, but this court held that Congress intended the GEP height to take into account not merely current engineering practice but also threats to human health. See supra p. 804 (discussing definition of “excessive concentrations” of pollutants).
This court remanded the relevant portion of the regulations to EPA for the agency to determine whether, in light of the proper understanding of “excessive concentration,” demonstrations should be required whenever the stack-height credit of an existing stack is to be raised. 719 F.2d at 459-60. While we did not adopt wholesale the petitioners’ arguments on this issue, we granted them a good portion of the relief that they requested and rejected all of EPA’s justifications for the regulations. We believe that the petitioners thereby achieved at least a modicum of success on this issue, and we therefore hold that an award of fees to the petitioners on this issue is appropriate.
F.Definition of “Stack” to Exclude Flares
“A flare is a pipe used in the oil, natural gas, and chemical industries to vent combustible gases by burning them at the top.” 719 F.2d at 460. EPA initially proposed to include flares in the definition of “stack,” but changed its mind and excluded them from the final proposal. This court rejected the petitioners’ contention that this exclusion was contrary to law and affirmed EPA’s decision.
The petitioners met with no success on this issue and thus are not entitled to attorneys’ fees compensating them for their challenge to these provisions.
G. Definition of “Dispersion Techniques”
Congress barred emission credit for tall stacks and “any other dispersion technique.” 42 U.S.C. § 7423(a)(2) (1982). EPA’s proposed regulations defined “other dispersion techniques” rather narrowly. 719 F.2d at 461. The petitioners listed a number of additional techniques that they believed were covered by the statutory definition. This court held that, in light of the language of the statute and its general tone, the definition of “other dispersion techniques” was of a “sweep broad[ ] enough to encompass at least the meaning urged by petitioners.” Id. at 462. EPA had not shown that its definition stemmed from administrative necessity or excluded only de minimis threats to the statute’s mandate. Id. at 462-64. This court therefore overturned EPA’s proposed rules on this matter with instructions for the agency either to broaden its definition to make it consistent with the statute, or to show that a broader definition was either impossible or of de minimis benefit.
The petitioners were successful in challenging the EPA’s definition of dispersion techniques and are entitled to fees for their efforts on this issue.
H. Definition of “Stack Height in Existence”
Congress exempted “stack heights in existence before December 31, 1970” from many of the limits of the Clean Air Act. 42 U.S.C. § 7423(a) (1982). EPA de- . fined “in existence” to mean construction on the stack had begun or binding contracts difficult to cancel had been signed. 47 Fed.Reg. 5864, 5868 (1982). The petitioners argued that “in existence” meant “already constructed.” This court upheld the EPA’s definition. 719 F.2d at 466.
*117The petitioners also argued that the exemption for stack heights in existence before 1971 was subject to abuse unless EPA prohibited new stacks from tying into old stacks. EPA did not respond to petitioners’ comments on this issue, and we remanded that issue to the agency for an explanation. 719 F.2d at 466.
We believe that petitioners have on this issue so far failed to demonstrate the requisite modicum of success for an award of fees under section 307(f). They failed completely in their challenge to the EPA’s definition of “in existence.” If the petitioners are to receive a fee award on this issue, the requisite modicum of success would have to flow from their challenge to the agency’s failure to respond to the petitioners’ comments about tying new stacks into grandfathered stacks. Although petitioners were successful in this latter challenge, we granted them no substantive relief. The agency may be able to justify its position with a simple response containing no reformulation of the challenged portion of the rules. We believe that the current situation falls into the category of a “purely procedural” victory, since we have not found wanting any action of the agency but merely instructed the agency to provide us with its rationale. Since the Ruckelshaus Court warned that “purely procedural” victories are insufficient for a fee award under section 307(f), 463 U.S. at 688 n. 9, 103 S.Ct. at 3279 n. 9, we must deny the petitioners fees on this issue at this time.
I. Prospective Application of 1 +1.5 Rule
The formula method of calculating GEP height gave one formula for stacks in existence before January 12, 1979 (the “2.5 Rule”) and a slightly different formula for stacks coming into existence after that date (the “1 + 1.5 Rule”). The 1 + 1.5 Rule, first proposed on January 12, 1979, sometimes results in a lower GEP height than the 2.5 Rule. The petitioners argued that the 2.5 Rule did not give the minimum GEP height and thus should not be available to any sources.
This court held that “the statute does not prevent EPA from allowing its past rule to be applied to stacks built before its new formula was proposed, but that the agency has erred in allowing sources that did not rely on the old formula to use it.” 719 F.2d at 467. We remanded the provision to the agency to “formulate its rule to take actual reliance into account.” Id. at 468.
The petitioners achieved more than a modicum of success on this issue and are thus entitled to fees compensating them for their labors in challenging the retroactive application of the rule. They did not prevail on their claim that no sources could use the 2.5 Rule, but they did prevent all sources in existence as of January of 1979 from automatically being able to use the 2.5 Rule. This court found “persuasive evidence that many sources built stacks in the relevant period without in fact relying on the 2.5 Rule, but instead built their stacks tall in order to obtain credit for dispersion.” Id. In light of the significance of this finding by the court and of the relatively generous standard of section 307(f), we believe that it is proper to award fees to the petitioners for their efforts to challenge the proper time and manner in which to compel sources to apply the 1 + 1.5 Rule rather than the 2.5 Rule. We would note, however, that the situation before us on this issue approaches the dividing line between instances in which a fee award is appropriate under section 307(f) and those in which an award is inappropriate.
J. Timetable for State Implementation
The 1977 amendments provided a timetable under which states would implement the new statute. EPA’s statement of how it intended to implement the amendments was “plainly contrary” to Congress’s intent in passing the timetable provisions. 719 F.2d at 469. We reversed its implementation plan.
*118The petitioners won a clean victory on this issue and are entitled to fees for their labors on it.
K. Timetable on Remand
The petitioners requested that the court direct EPA to promulgate “without delay” new regulations. This court thought it “best to quantify that obligation” and directed that EPA promulgate new regulations within six months of the issuance of the court’s mandate. 719 F.2d at 470.
The petitioners obtained significant relief from this court of the kind that they sought, and are thus entitled to fees for those efforts.
L. Summary
The petitioners met with the requisite modicum of success in their challenges to the EPA’s regulations on the definitions of “nearby,” “excessive concentrations,” and “dispersion techniques”; on plume impaction; on the conditions under which use of the demonstration method is necessary; on the prospective application of the 1 + 1.5 Rule; and on the proper implementation of Congress’s timetable. They also succeeded in obtaining relief from this court on the timing of the EPA’s issuance of new regulations. The petitioners did not meet with the requisite modicum of success on the issues of plume rise and the definitions of “flare” and “stack heights in existence.”
Ill
Having determined on which issues the petitioners met with sufficient success for an award of fees on the merits to obtain, we must calculate how many hours were reasonably expended on those issues.
The petitioners seek from the government compensation on merits work for 80 hours of work by Richard Ayres, 675 hours for Howard Fox, and 45 hours for Frederick S. Middleton.
The government contends that the petitioners have devoted an unreasonable amount of time to the case. The government notes that it spent a total of about 400 hours on the merits of the case. The petitioners, in contrast, request compensation for 800 hours. We note, however, that the petitioners in a case involving direct review of administrative regulations obviously face the burden of deciding which provisions to challenge and on what grounds, while the respondents need only react to those challenges. In this particular suit, the regulations involved were “detailed and somewhat complex,” 719 F.2d at 439, and the resulting opinion consumed over thirty pages in the Federal Reporter. This complexity seems likely to us to make even more disproportionate the petitioners’ burden.
Several other factors make more reasonable the petitioners’ request. We are compensating the petitioners only for hours expended on those issues on which they achieved a modicum of success, and thus for hours that produced at least some results. In addition, the petitioners do not seek compensation for the 200 hours expended by law clerks and paralegals on the case. Cf Laffey v. Northwest Airlines, 746 F.2d 4, 9, 30 (D.C.Cir.1984) (allowing compensation of such hours under Title VII), cert. denied, — U.S.-, 105 S.Ct. 3488, 87 L.Ed.2d 622 (1985). Finally, the petitioners deducted some forty hours from those actually expended in the exercise of billing judgment. See Copeland v. Marshall, 641 F.2d 880, 891 (D.C.Cir.1980) (en banc). Taking into consideration all these circumstances, we do not believe that the petitioners have been unreasonable in their calculation of hours attributed to their labors in this case.
We also reject the government's suggestions that the petitioners unwisely allocated their resources among attorneys of varying levels of experience. Junior attorneys are not to be left unsupervised, and senior attorneys are not to perform tasks in which their experience is of little value. See Copeland v. Marshall, 641 F.2d at 903 n. 50 (agreeing with District Court that inadequate time by partners had been spent on case); In re Continental Illinois Securi*119ties Litigation, 572 F.Supp. 931, 933 (N.D. 111.1983) (warning attorneys restructuring representation in multi-lawyer case that partners should not devote time to work that could be assigned to associates). We believe that the petitioners here have done reasonably well in adhering to the most efficient use of attorneys of different levels of experience. Cf. Laffey v. Northwest Airlines, 746 F.2d at 26 (petitioners efficiently allocated tasks among paralegals, associates, and partners in light of fact that case was staffed much as it would have been for a paying client).
The petitioners, of course, are entitled to fees only for hours expended on issues on which they met with a modicum of success. See supra pp. 799-802. Mr. Fox performed the lion’s share of the petitioners’ work on the case. In Mr. Fox’s summary of his work, about 200 hours are specifically attributed to a particular issue in the case clearly raised in the brief (such as “plume impaction” or “grandfather clause”), as opposed to more general work (such as “review of administrative record” or “research into Clean Air cases”). See Exhibits in Support of Petitioners’ Supplement to Motion for Award of Costs of Litigation, Including Reasonable Attorneys’ Fees, at Tab B. Of those approximately 200 hours, some 60 hours — roughly 30% — are attributed to issues on which the petitioners did not achieve a modicum of success. We must obviously deny the petitioners fees in connection with those 60 hours of Mr. Fox’s work. We grant fees for the other 140 hours devoted to specific issues. We deduct 30% of Mr. Fox’s hours not allocated to any specific issues to represent some estimate of Mr. Fox’s nonspecific time allocated to issues on which the petitioners did not achieve a modicum of success. We similarly reduce by 30% the requested hours of compensation for Mr. Ayres and Mr. Middleton, which are set forth only in terms of general work.
We thus award petitioners fees for 140 hours of Mr. Fox’s time specifically allocated to various issues, and for 332 hours of Mr. Fox’s time not allocated to specific issues,4 for a total of 472 hours. We also compensate the petitioners for 56 hours of Mr. Ayres’s work and 31 hours for Mr. Middleton’s work.
IV
We move now from the determination of how many hours of labor are compensable to the determination of the reasonable hourly rate.
The petitioners are what are commonly referred to as “public interest law firms.” The petitioners do not have paying clients in the traditional sense, but rather bring suit on behalf of their members, who have paid annual dues not specifically earmarked for litigation. We cannot therefore simply refer to the petitioners’ ordinary hourly rates in the market to determine the proper hourly rates for the purposes of determining fees. Cf. Laffey v. Northwest Airlines, Inc., 746 F.2d at 24 (where firm receiving fees is a for-profit partnership with a long history of billings to private customers, the reasonable hourly rate is the firm’s historical billing rate).
In the case of not-for-profit law firms, the objective is still to determine the prevailing market rate for that firm's services; but because such firms have no past billing history, a proxy for the actual market rate of the firm’s services must be found by examining the rates prevailing in the community for similar services by lawyers of reasonably comparable skill, experience, and reputation.
Laffey v. Northwest Airlines, 746 F.2d at 24; see also Blum v. Stenson, 465 U.S. 886, 104 S.Ct. 1541, 79 L.Ed.2d 891 (1984) (discussing proper determination of the reasonable hourly rates of the New York Legal Aid Society, a not-for-profit firm with non-paying clients and thus no billing histo*120ry). The petitioners have made submissions here that allow us to determine that proxy. They have produced both affidavits from their own attorneys and external evidence of prevailing rates. See Blum v. Stenson, 104 S.Ct. at 1547 n. 11. After reviewing petitioners’ affidavits on their attorneys’ experience and the range of billing rates prevalent among many of the larger Washington firms, we believe that the petitioners’ proposed hourly rate of $75.00 an hour for Mr. Fox is reasonable, but that its proposed rate of $150.00 an hour for Messrs. Ayres and Middleton is somewhat higher than is reasonable.
We therefore award Mr. Fox $75.00 an hour, Mr. Ayres $125.00 an hour, and Mr. Middleton $125.00 an hour. The rates for Messrs. Ayres and Middleton are slightly higher than those awarded a few years ago to attorneys with similar experience, an increase we consider warranted in light of the upwards trend in the prevailing market rate. See Environmental Defense Fund v. EPA, 672 F.2d 42, 51, 58 (D.C.Cir.1982) ($110/hour for attorney with 11 years of experience); Sierra Club v. Gorsuch, 684 F.2d 972, 975 (D.C.Cir.1982) (same), rev’d on other grounds sub nom. Ruckelshaus v. Sierra Club, 463 U.S. 680, 103 S.Ct. 3274, 77 L.Ed.2d 938 (1983).
For Mr. Fox’s time on the merits, then, we award the petitioners $35,400.00 ($75.00/hour X 472 hours). For Mr. Ayres’s time we award them $7,000.00 ($125.00/hour x 56 hours). For Mr. Middleton’s time, we award them $3,875.00 ($125.00/hour X 31 hours). The total award to the petitioners and against the government for work on the merits is therefore $46,275.00.
V
Although we accept the petitioners’ estimates of the hours reasonably expended on the regulations successfully challenged in the case and are not sharply in disagreement with their estimate of the reasonable hourly rates, we must reject the petitioners’ contentions that the resulting lodestar should be increased to compensate the petitioners for their risk of losing the case. In rejecting an identical contention, the Laffey court first emphasized that to compensate attorneys fully for the risk of losing a case is in the long run to reward attorneys equally for bringing meritorious and frivolous lawsuits. 746 F.2d at 26-28. The Laffey court noted that, even if one only attempts to compensate attorneys in part for their risk of losing, such an enhancement of the lodestar is “permissible only in the exceptional case.” 746 F.2d at 29 (quoting Murray v. Weinberger, 741 F.2d 1423, 1432 (D.C.Cir.1984)). In Laffey, the suggested increase for risk of loss was 100%, based upon the contention that the petitioners’ initial chances of success were 50%. Laffey, 746 F.2d at 29. Since this is approximately the average probability of initial success in any case that goes to litigation, the case before the Laffey court could hardly be “exceptional” as required by Murray. An enhancement of the lodestar for risk of loss was therefore clearly inappropriate. We are here faced with the same rationale for a 50% enhancement of the lodestar. See Petitioners’ Supplement to Motion for Award of Costs of Litigation, Including Reasonable Attorneys’ Fees at 27-28 (D.C.Cir. filed July 17, 1984) (petitioners believe themselves entitled to 100% risk-of-loss enhancement but seek only 50%). We can see no relevant differences between this case and Laffey for these purposes, and we must deny the petitioners’ request for an enhancement of the lodestar to compensate them for their risks of losing the case. See also Environmental Defense Fund v. EPA, 672 F.2d 42, 60 (D.C.Cir.1982) (enhancement for risk of loss inappropriate when petitioner did not have to “prevail” to obtain fees and so risks of obtaining no fees at all was minimal).
The petitioners also seek an enhancement of the lodestar for the risk of delayed payment of fees that they sought. Petitioners chose to present evidence of hourly rates prevailing at the time of their first fee petition rather than at the time of the resolution of the merits of the case. We have accepted those rates as the basis for our hourly rates, and have therefore al*121ready compensated for the risk of delay in payment. We believe that further enhancement of the lodestar for that delay in payment is inappropriate on the facts of this case. See Murray v. Weinberger, 741 F.2d 1423, 1433 (D.C.Cir.1984); National Association of Concerned Veterans v. Secretary of Defense, 675 F.2d 1319, 1329 (D.C.Cir.1982) (per curiam).
VI
After our initial decision on October 11, 1983, the federal government essentially dropped out of the case. The intervenors, however, filed several post-decisional motions in this court as well as a petition for a writ of certiorari in the Supreme Court. The intervenors met with some success in their efforts to delay the final resolution of the case, but did not persuade either this court or the Supreme Court to modify the substance of our opinion on the merits. This litigation was effectively concluded on July 2, 1984, when the Supreme Court denied certiorari. The petitioners seek attorneys’ fees against intervenors for petitioners’ work between October 12, 1983, and July 2, 1984.
A court that awards fees under section 307(f) must hold that such an award is both generally authorized by the statute and specifically appropriate in the case before it. See Sierra Club v. Gorsuch, 672 F.2d 33, 38 n. 8 (D.C.Cir.1982) (per curiam), reversed on other grounds sub nom. Ruckelshaus v. Sierra Club, 463 U.S. 680, 103 S.Ct. 3274, 77 L.Ed.2d 938 (1983). We here hold that fees against the intervenors would be inappropriate, and therefore we. do not need to reach the issue of whether the statute ever allows such an award.5
We note first that the Supreme Court has urged particular caution when the fee award is against private parties rather than the government:
Differing abilities to bear the cost of legal fees and differing notions of responsibility for fulfilling the goals of the Clean Air Act likely would justify exercising special care regarding the award of fees against private parties.
Ruckelshaus v. Sierra Club, 463 U.S. at 692 n. 12, 103 S.Ct. at 3281 n. 12. Congress provided for the award of fees to parties to encourage the effective interpretation and implementation of the Act. We believe it is inappropriate to award fees against parties who have met that same standard. Because we believe the intervenors advanced a reasonable position in their work before this court and in their petition for a writ of certiorari from the Supreme Court, we refuse to award the petitioners any fees against the intervenors.6
*122To evaluate the reasonableness of the position taken by the intervenors before the Supreme Court must involve some speculation on our part. The reasonableness of seeking a petition for certiorari depends in some measure upon the chances that the petition will be granted. Yet the overwhelming number of such petitions are denied, see Feinberg, The Office of Chief Judge of a Federal Court of Appeals, 53 Fordham L.Rev. 369, 369 n. 2 (1984) (collecting statistics), and we obviously cannot read the collective mind of the Supreme Court as to the relative merits of the denied petitions. Here, while it is true that the intervenors’ petition for a writ of certiorari was eventually denied, we believe that the intervenors’ decision to seek certiorari was based on a position that was far from frivolous.
The litigation on the merits involved complex and important issues. The petitioners estimate the likelihood of the intervenors’ success in obtaining certiorari was at least one in two. Petitioners’ Second Supplement to Motion for Award of Costs of Litigation, Including Reasonable Attorneys’ Fees, at 17 (D.C.Cir. filed Aug. 1, 1984). And the Supreme Court delayed its denial of the writ of certiorari until after it had decided Chevron U.S.A., Inc. v. Natural Resources Defense Council, 467 U.S. 837, 104 S.Ct. 2778, 81 L.Ed.2d 694 (1984), a case raising issues on the proper standards for direct review similar to those at issue here. We believe that these circumstances provide sufficient indicia that the intervenors took a position that reasonably attempted to advance the implementation of the Act. The intervenors’ lack of success in obtaining certiorari is not, particularly in light of the “special care” urged upon us by the Supreme Court in the consideration of awards against private parties under the Clean Air Act, enough for us to award fees against the intervenors for petitioners’ work before the Supreme Court.7
We also deny the petitioners fees for their work in connection with the intervenors’ post-decisional motions before this court. The intervenors were unsuccessful in these efforts as well, but we do not believe their positions sufficiently unfounded to award fees against them in light of the importance of the regulations at issue in interpreting the Act and the “special care” that we must take in awarding fees against private parties under section 307(f). We emphasize again, however, that we do not reach the issue of whether section 307(f) ever authorizes fees against intervenors.
VII
The petitioners also request fees for the time spent in preparing the fee petition. The hours reasonably expended on such a petition are compensable. See Laffey v. Northwest Airlines, 746 F.2d at 29; see also Environmental Defense Fund v. EPA, 672 F.2d 42, 62 (D.C.Cir. 1982). An outside attorney was the lead worker in preparing the petition, assisted by petitioners’ associate. The petitioners exercised some billing judgment in their submission. In addition, they do not seek compensation for hours expended on the fee petition by a legal intern, nor do they seek any enhancements of the lodestar.
We find the petitioners’ request for fees in connection with preparing the fee petition to be reasonable, except for their request that Mr. Wilson, the outside attorney used in preparation of the fee petitions, be compensated at the rate of $150 per hour. This is the rate that Mr. Wilson generally charges to commercial customers. Affida*123vit of Ronald J. Wilson at Tab D, Exhibits in Support of Petitioners’ Supplement to Motion for Award of Costs of Litigation, Including Reasonable Attorneys’ Fees, at 3 (D.C.Cir. dated Feb. 21, 1984). His rate to non-profit entities, however, is generally between $50 and $60 an hour. Id. In light of the discussion in Laffey v. Norwest Airlines, 746 F.2d at 14 n. 69, we hold that $60 an hour is a reasonable estimate of the hourly compensation due Mr. Wilson for his work in this case.
The petitioners request compensation for 24.25 hours by Mr. Fox at the same $75 an hour that we found reasonable for his work on the merits. We grant this request for an award of $1818.75. The petitioners request compensation for 44.75 hours by Mr. Wilson, which we grant at an hourly rate of $60 for an award of $2,685.00.
We therefore grant the petitioners $4,503.75 for their work in connection with their petition seeking fees against the government. We note that the hours sought seem perhaps excessive for a fee petition of relatively ordinary difficulty, but the government does not contest the hours requested. We therefore assume that some factor not apparent to this court justifies the request. We do not wish, however, to influence future panels should they be called upon to make similar determinations.
VIII
The petitioners also request compensation for costs of $581.77 for photocopying, postage, and telephone expenses. The government does not oppose this request. Respondents’ Memorandum in Opposition to Petitioners’ Motion for Award of Attorney Fees at 27. We grant this request in full.
We deny the petitioners’ request for $1616 to. compensate a Mr. Lazaro, a “technical consultant” who aided the petitioners in preparing their case. Section 307(f) allows compensation of the “costs of litigation (including reasonable attorney and expert witness fees).” 42 U.S.C. § 7607(f) (1982). Mr. Lazaro is obviously not an attorney. He is also not an “expert witness,” since review of the regulations was undertaken in this case on the administrative record without any new hearings before this or any other court. Cf. Asarco, Inc. v. EPA, 616 F.2d 1153, 1157-61 (9th Cir.1980) (allowing limited use of expert testimony before District Court in reviewing order by EPA). This leaves only the general rubric of “costs” under which to compensate petitioners for making use of Mr. Lazaro’s services. Those services do not fall under the traditional concept of costs, however, which ordinarily encompasses items such as filing fees or other court costs. While we are willing to compensate the petitioners for some additional expenses in phone and photocopying costs, we do not believe an award of fees for Mr. Lazaro’s services is allowed by the statute. Section 307(f) authorizes compensation for professional services of attorneys and expert witnesses. Mr. Lazaro is neither. We do not read section 307(f)’s waiver of sovereign immunity so broadly as to allow for fees in connection with the services of outside, nontestifying experts. See Ruckelshaus v. Sierra Club, 463 U.S. at 685, 103 S.Ct. at 3277 (waivers of sovereign immunity are to be construed narrowly).8
IX
We award the petitioners $46,275.00 from the government for petitioners’ efforts on the merits. We award petitioners $4,503.75 from the government for their efforts on the fee petition. We award petitioners $581.77 from the government for costs of photocopying and postage. The total award against the government is therefore $51,360.52. We deny the petitioners fees for the services of the technical consultant, and deny the petitioners any *124award of fees against the intervenors m this case.
It is so ordered.
. The Supreme Court has stated that “the history of § 304(d) is relevant to a construction of § 307(f).” Ruckelshaus v. Sierra Club, 463 U.S. at 692 n. 13, 103 S.Ct. at 3281 n. 13.
. The party seeking fees in Hensley had to "prevail” for a claim to be "successful,” while here the petitioners face a less demanding criterion of success. We therefore of course apply a different standard from the Hensley Court’s in determining success once we have used the Hensley Court’s analysis to divide the case into its component issues. See infra p. 802.
. We are aware that this court did not undertake an issue-by-issue analysis in awarding fees under section 307(f) in Alabama Power Co. v. Gorsuch, 672 F.2d 1 (D.C.Cir.1982) (per curiam). There, the court found that the "number, complexity and frequent interrelationships of the legal problems presented by these consolidated cases as a whole obviates the need for an issue-by-issue explication of our decision to award attorneys’ fees." Id. at 4 n. 16. This court decided Alabama Power Co. before the Supreme Court’s rulings in either Hensley or Ruckelshaus. The latter decision of the Supreme Court reversed Sierra Club v. Gorsuch, 672 F.2d 33 (D.C. Cir.1982), a case significantly relied upon in Alabama Power. In light of the Supreme Court’s decisions in Hensley and in Ruckelshaus, we believe that an issue-by-issue explication is proper in this case.
. The petitioners request approximately 675 hours for Mr. Fox’s compensable time. Of those hours, about 200 are allocated to specific issues, leaving approximately 475 hours devoted to general issues. We subtract 30% of those 475 hours (143) as devoted to issues on which the petitioners did not substantially prevail, leaving 332 compensable hours for general tasks.
. The courts have so far been skeptical of their authority to award fees against intervenors advancing non-frivolous positions. The Ninth Circuit has stated that section 307(f) "simply evinces a willingness on the part of Congress to have the Government bear the cost of litigation brought to promote air quality,” and denied efforts by the government to have an intervenor contribute to the payment of legal fees. See Northern Plains Resource Council v. EPA, 670 F.2d 847, 849 n. 4 (9th Cir.1982) (per curiam), vacated and remanded, 464 U.S. 806, 104 S.Ct. 54, 78 L.Ed.2d 73 (1983), on remand 734 F.2d 408 (9th Cir.1984) (per curiam) (holding that in light of Ruckelshaus v. Sierra Club, 463 U.S. 680, 103 S.Ct. 3274, 77 L.Ed.2d 938 (1983), fees could not be awarded even against the government). Our own District Court has held that section 505(d) of the Clean Water Act — a provision similar to that at issue here, compare 33 U.S.C. § 1365(d) (1982) with 42 U.S.C. § 7607(f) (1982) —does not authorize the award of fees against intervenors. Natural Resources Defense Council v. Administrator, Civ. Action Nos. 75-1698 et al., slip op. at 4-8 (D.D.C. June 20, 1984), appeal filed, Nos. 84-5566 et al. (D.C.Cir. Aug. 21, 1984).
. We realize that when one private party seeks fees from another, one of them will be deterred from future litigation no matter what the decision of the court. If fees are awarded, the party liable for those fees will be deterred from future litigation. If no fees are awarded, the party seeking fees will be deterred.
One could argue that,' when both the party seeking and the party potentially liable for fees have advanced the implementation of the statute, the balance for determining whether to award fees under a statute entitled a "Clean Air Act" should tip towards the "pro-environment" party. In this case, the balance would then tip in favor of the petitioners towards an award of fees. This court has stated, however, that the Clean Air Act is a compromise among a number *122of competing interests and is not biased in favor of "pro-environment” groups. Alabama Power Co. v. Gorsuch, 672 F.2d 1, 5 (D.C.Cir.1982). We cannot therefore use such a criterion in making our decision.
. We therefore need not treat the intervenors’ contentions that this court is not even authorized to make awards for work performed before the Supreme Court. Similarly, we do not reach the intervenors’ assertions that a petition for certiorari is a purely procedural matter that does not give rise to an action "under” section 307 and thus does not qualify for an award of fees under section 307(f). See 42 U.S.C. § 7607(f) (1982) (allowing award of fees for "any judicial proceeding under this section").
. We note that petitioners do not request compensation for the hours expended of paralegals, law clerks, or law interns. We therefore need not reach the question of whether hours expended by such non-attorneys closely associated with the firm are compensable under section 307(f).