The Clean Air Amendments of 19701 established a national program of air pollution control administered jointly by the states and the Environmental Protection Agency (EPA). A realistic approach to the problem of effective enforcement of that statute by reduction of air pollution compels recognition of the fact that national air quality standards cannot be attained except at significant cost to industry. This case deals with the extent to which the EPA must accept that fact, yet may allow some equitable accommodation for companies earlier committed to a tall smokestack policy on the false assumption that tall smokestacks dispersing pollution over a broad area are an acceptable substitute for emission control of industrial pollution.
The petitioners in this case, the Natural Resources Defense Council (NRDC) and the State of Georgia, intervenor, filed separate motions for an order to show cause why the Environmental Protection Agency Administrator and certain other designated EPA officials should not be held in contempt for failing to comply with an order of this Court issued February 8, 1974.2 That order arose out of litigation begun almost two years earlier, when the NRDC challenged EPA approval of the State of Georgia’s plan for implementing the goals of the Clean Air Act.
BACKGROUND
In the original suit, the NRDC challenged EPA approval of Georgia’s State Implementation Plan (SIP) for achieving compliance with the National Ambient Air Quality Standards (NAAQS)3 of the Clean Air Act. The Georgia plan sought to reduce the ground level intensity of sulphur dioxides and particulate matter emitted from its power plants by the use of a “tall stack” strategy of “dispersion enhancement”.4 Under this strategy, Georgia power plants could, by constructing tall smoke stacks, disperse harmful pollutants over a wider area, thereby reducing the ground level concentration of pollutants in the immediate vicinity of the facility to comply with the NAAQS.5 We found that this meth*757od of dispersion enhancement, which reduced the ground level intensity of pollutants at a given source but did nothing to reduce the overall amount of pollutants in the atmosphere, was inconsistent with the nondegradation policy of the Clean Air Act. 489 F.2d at 408. We interpreted § 1857c-5(a)(2)(B) to require that states utilize all possible emission limitation techniques6 to achieve national ambient air standards. We found that Congress intended emission reduction (as opposed to mere dispersion) to be the preferred control method,7 and intended to permit “such other measures” as specified in the Act8 “only if emission reduction sufficient to [achieve the national standards] in the time specified [were] unavailable or infeasible — or, in the words of the Act, only if . ‘necessary’.”9
Accordingly, we held that control strategies such as Georgia’s tall stack dispersion strategy, may be included in a state’s plan only “(1) if it is demonstrated that emission limitation regulations included in the plan are sufficient standing alone, without the dispersion strategy, to attain the standards; or (2) if it is demonstrated that emission limitations sufficient to meet the standard are unachievable or infeasible, and that the state has adopted regulations which will attain the maximum degree of emission limitation achievable.”10
Noting that the EPA Administrator had never suggested that the second condition could be applied to validate the Georgia plan, we ordered him to make a determination whether the Georgia regulations were, standing alone, sufficiently stringent to guarantee attainment of the national standards. The EPA’s original approval of the Georgia plan in May 1972 stated merely that the state plan, with its combination of emission limitations and tall stack dispersion strategy, would guarantee attainment of the national standards. The Administrator’s second letter of May 1973 established only that the tall stack regulations had since been determined inappropriate under the Agency’s emerging policy against dispersion enhancement as a substitute for emission limitations. What had never been determined however, was whether the emission limitations prescribed in the Georgia plan, considered independently of the tall stack dispersion *758strategy, were sufficient to attain the national standards.
To resolve this issue, we directed the Administrator to make an explicit determination on the question “as promptly as is administratively feasible”.11 It is this order (and the Administrator’s failure to respond after more than a year) which prompted the motions for contempt now before the Court.
THE INSTANT SUIT
In March and April 1975, both the State of Georgia and the NRDC sought orders holding the EPA Administrator and certain designated officials 12 in contempt for their failure to comply with the Court’s order of February 8, 1974. The NRDC charged that the Administrator not had only failed to submit the required evaluation to the Court, but had, in the interim, issued proposed guidelines for evaluating the Georgia plan which would have effectively undermined the clear import of this Court’s directive. The State of Georgia asserted that the Administrator had, in fact, evaluated and approved the Georgia plan following this Court’s decision, but had failed to so inform the Court.
The EPA thereupon submitted to the Court a document entitled “Evaluation of the Georgia Control Strategy”, bearing an original date of June 1974, and a revised date of March 31, 1975 (four days after the State filed its motion for contempt). The EPA report substantially confirmed the State’s contentions that, based on a reevaluation allegedly undertaken according to the criteria outlined in the Court’s order, the EPA now approved the Georgia Implementation Plan. The EPA report was later supplemented by two additional reports on June 10 and July 10, 1975, but the original conclusion remained the same: “With three exceptions, the analysis performed by EPA indicates that the State SO2 emission limiting requirements standing alone are adequate to obtain the National Ambient Air quality Standards.” 13
The NRDC’s response to the revised EPA evaluation reiterated the objections raised in its original motion for contempt. NRDC alleged that the EPA had ignored the order of this Court to evaluate the ability of the Georgia plan to achieve and maintain National Ambient Air Quality Standards independent of the reduction achieved by the use of tall stacks. Instead, the NRDC charged, the EPA had given “credit” for the dispersive effect of certain tall stacks begun before this Court’s decision. The result of the EPA’s use of such “grandfather clauses” was, according to the NRDC, to validate the tall stack strategy expressly disapproved by this Court. NRDC then proposed its own set of guidelines for evaluating the Georgia plan in the light of the criteria set forth in the Court’s original decision.
STACK CREDIT
In order to appreciate the NRDC’s objections to the EPA guidelines, it is necessary to understand the concept of stack “credit”. As noted above, our order of February 8, 1974 directed the Administrator to reevaluate the Georgia plan without taking into account the dispersive effects achieved by the use of tall stacks. This involved the construction of new “diffusion models” to calculate the ground level intensity of pollu*759tants emitted at heights lower than those originally contemplated by the tall stack strategy.
Difficulties arose, however, when the Administrator was faced with stacks which were either (1) already under construction or subject to binding contracts before the State plan was filed in January 1972, or (2) under construction at the time our decision was rendered in February 1974.14 Some accommodation had to be made for sources which had planned and commenced construction on taller than average stacks before the tall stack strategy was disapproved. Thus, in assigning a height to be factored into the calculations for measuring pollution levels at such facilities, the EPA felt compelled to give some “credit” for the dispersive effects of stacks at sources which had undertaken construction and incurred considerable expense before this Court’s decision in February 1974.
EPA CREDIT FORMULA
The EPA concluded that for sources which had committed themselves to tall stacks at the time the State plan was filed, it would be inequitable to apply the Court’s ruling. Such stacks had been contracted for or commenced two years before the date of this Court’s ruling disapproving tall stacks, and before the State’s tall stack strategy was ever attacked. Thus, for stacks in existence, under construction, or subject to binding contracts as of January 31, 1972, the EPA gave full credit for proposed stack height. That is, in determining whether Georgia’s regulations were sufficiently stringent to achieve compliance with national standards, EPA officials factored into their calculations the ground level intensity which would result if pollutants were emitted at the actual proposed stack height.
For stacks which were contracted for after the State plan was filed, but were already under construction by the date of our decision, EPA officials gave less than full credit. These stacks, though planned and commenced before our decision publicly disapproved the tall stack strategy, were undertaken during the pendency of protracted and widely publicized litigation; presumably even non-parties were on notice of the possible import of our decision. Thus, for stacks contracted for after the State plan was filed in January 1972, but under construction as of the date of our decision in February 1974, the EPA proposed to give limited credit. In constructing its diffusion models for such stacks, the EPA took the proposed stack height, but only up to 2.5 times the height of the powerhouse building in the facility. The EPA declared that this 2.5 rule conformed to historical practice in that it was representative of the median stack height in the power industry before the Clean Air Act.15
As noted earlier, on the basis of diffusion models constructed according to these criteria, the EPA concluded that the regulations in the State plan, with three exceptions, were sufficiently stringent to guarantee compliance with National Ambient Air Quality Standards within the prescribed period.
NRDC OBJECTIONS
Petitioner NRDC acknowledged the desirability of a “grandfather” clause, but objected to the guidelines used by the EPA to give “credit” for stacks planned or under construction before the date of our decision, on the ground that the effect of giving such credit would be to permit certain sources to reap the benefits of the disapproved tall stack strategy. They also charged that the 2.5 times rule would permit stacks to in*760crease their height beyond that in existence at the time of our decision; that half of all the facilities in the country could substitute some increase in stack height for emission control.
The objections are serious. We have no intention of permitting either the EPA or the State to circumvent the import of the congressional mandate, as we read it in our February decision, through the use of overly generous stack credit. Tall stacks have been expressly disapproved by this Court as a substitute for emission limitations and may be included in a state’s plan only after all other available techniques of emission limitation have been exhausted. We cannot, however, apply our February 1974 decision retroactively. Sources with stacks in existence, under construction, or subject to binding contracts as of the time the State plan was filed were committed to build stacks of a certain height long before our decision and, in fact, before the tall stack strategy came under attack. Moreover, companies having stacks under construction as of the date of our decision had also invested considerable resources in a method of pollution control which, until less than a year before our decision, had been approved by the EPA.16 While some notice may be imputed to sources which commenced construction during the course of the litigation, we cannot deny that equity requires some credit to be given for expenditures incurred before our decision disapproved the tall stacks.
As to the NRDC’s objection to the formula used by the EPA for giving limited credit to stacks contracted for after the State plan was filed, but under construction as of the date of our decision, we. cannot say that the 2.5 rule is an arbitrary one. Although NRDC has urged that credit for such stacks be given only up to “a height conforming to historical practice in the industry”, the EPA asserts that the 2.5 rule does, in fact, represent the median stack height in the power industry prior to the advent of the Clean Air Act.17 Absent evidence to the contrary, we cannot dispute that administrative determination.
We find, therefore, that the EPA guidelines used to reevaluate the Georgia plan were not inconsistent with this Court’s order of February 1974, and that the Administrator’s approval of the plan must stand. We deny the motions of the NRDC and the State of Georgia to . hold the EPA and its designated officials in contempt, noting however, that such motions apparently were necessary to compel the Administrator to respond to our directive issued more than a year before.
The motion to intervene filed by Georgia Power Co. is denied.
. 42 U.S.C. §§ 1857a-58 (1970), amending 42 U.S.C. §§ 1857a-571 (1967). The Act provides that the Administrator of the Environmental Protection Agency promulgate National Ambient Air Quality Standards (NAAQS) and that states submit to the EPA State Implementation Plans (SIPs) establishing emissions limitations for stationary sources that will guarantee attainment of the NAAQS. Petitioner NRDC has challenged EPA approval of several state plans on bases similar to those raised in the original suit. See, e. g., NRDC v. EPA, 1 Cir. 1973, 478 F.2d 875 (Rhode Island and Massachusetts SIPs); NRDC v. EPA, 9 Cir. 1974, 507 F.2d 905 (Arizona SIP); NRDC v. EPA, 2 Cir. 1974, 494 F.2d 519 (New York SIP); NRDC v. EPA, 8 Cir. 1973, 483 F.2d 690 (Iowa SIP).
. NRDC v. EPA, 5 Cir. 1974, 489 F.2d 390, rev’d in part, on other grounds, sub nom. Train v. NRDC, 1975, 421 U.S. 60, 95 S.Ct. 1470, 43 L.Ed.2d 731.
. “Ambient Air” has been defined by federal regulation to mean that portion of ithe atmosphere (external to buildings) to which the general public has access. See 40 C.F.R. § 50.-1(e) (1971).
. With respect to human health, the scientific community increasingly agrees that damage is related more clearly to the levels of and acid sulfates than to concentrations of sulfur dioxide. Acid sulfates, unlike sulfur dioxide, are not emitted by polluting sources. Rather, they are the result of chemical transformation of sulfur dioxide in the atmosphere, occurring over a period of several days. Since in that period the air mass over a very large land area undergoes considerable mixing, high acid sulfate concentrations, unlike high sulfur dioxide concentrations, blanket entire regions relatively uniformly, depending on the total quantity of emissions of the precursor pollutants. Measures such as dispersion techniques that reduce peak sulfur dioxide concentrations without reducing the total atmospheric loading with the pollutant have virtually no effect on acid sulfate concentrations. Thus any system that does not reduce total sulfur dioxide emissions (as well as emissions of particulates) will be unavailing to protect the public health.
Ayres, Enforcement of Air Pollution Controls on Stationary Sources Under the Clean Air Amendments of 1970, -Vol. 4, Ecology Law Quarterly 441, 453 (1975).
.Since NAAQS measure the concentration of a pollutant at ground level, any control method which effects greater dispersion of the pollutant necessarily diminishes the ground level intensity at its source, thereby facilitating *757compliance with the NAAQS. Recognizing the effect of tall stacks on the ground level concentration of sulphur dioxides and particulate matter in the area of a power plant, Georgia adopted regulations which would have permitted sources to emit greater quantities of pollutants, provided that the height of the stacks were increased. For a contrary view of the ameliorative effects of tall stacks, see Ayres, Enforcement of Air Pollution Controls on Stationary Sources Under the Clean Air Amendments of 1970, 4 Ecology Law Quarterly 441, 453 (1975). The author suggests that the damage caused by high acid sulfate concentrates depends on the total quantity of emissions introduced into the atmosphere, and is in no way reduced by the use of dispersion techniques. Id at 453. Mr. Avres. it must be pointed out, is an attorney for the NRDC.
.Emission limitation techniques, as opposed to dispersion enhancement techniques, are control methods which effect a reduction in the total amount of pollutant emitted into the atmosphere, usually through a specific quantitative limitation on the amount any given source may emit. Emission limitation techniques differ from dispersion enhancement controls, in that the latter merely spread the pollutant over a broader area without reducing the concentration of the pollutant at its source or reducing the overall amount of pollutant emitted. See NRDC v. EPA, 5 Cir. 1974, 489 F.2d at 394 n. 2.
. 489 F.2d at 406-408.
. 42 U.S.C. § 1857c-5(a)(2)(B).
. Monitoring and Data Analysis division, Office of Air Quality Planning and Standards, Office of Air and Waste Pollution, Environmental Protection Agency, Staff Paper — Intermittent Control Systems, 119 Cong.Rec. 10948, 10955-56 (Daily ed. June 12, 1973), cited in NRDC v. EPA, 5 Cir. 1974, 489 F.2d at 406.
. 489 F.2d at 410. Support for our interpretation of the Act has since been given by the Sixth and Ninth Circuits, respectively, in Big Rivers Electric Corp., et al., v. EPA, 6 Cir. 1975, 523 F.2d 16, and Kennecott Copper Corp. v. Train, 9 Cir. 1975, 526 F.2d 1149 (continuous emission reduction technology). See also S. Bleicher, Economic and Technical Feasibility in Clean Air Act Enforcement Against Stationary Sources, 89 Harv.L.Rev. 316 (1975).
. 489 F.2d at 411.
. The individual respondents were Russell Train, EPA Administrator; Bernard Steiger-wald, Director, Office of Air Quality Planning and Standards; and Roger Strelow, Assistant Administrator of Air and Waste Management.
. Air Programs Office, Environmental Protection Agency, Region IV, Atlanta, Georgia, Evaluation of the Georgia Control Strategy, June 21, 1974, revised March 31, 1975, p. 3. The exceptions were Georgia Power Company’s Yates, Hammond, and Atkinson Generating Plants. Additional tests under the EPA guidelines have since determined that two of the three plants can, with existing controls, achieve compliance with the national standards. The third, Atkinson, has since installed measures to further limit emissions.
. These categories are not mutually exclusive. Some stacks contracted for before January 1972 were still under construction when our decision was rendered in February 1974. The latter category, however, is intended to refer only to those stacks under construction on February 8, 1974, for which binding contracts had not been negotiated as of January 31, 1972.
. EPA Evaluation of the Georgia Control Strategy, supra note 13, at 3.
. See text at notes 10-11 supra.
. See note 15 supra.