FEINBERG, Circuit Judge:

In March 1974, the Environmental Protection Agency (EPA) promulgated final regulations requiring curtailment of water pollution by crystalline cane sugar refineries in the United States. 1 Although four refining companies — Amstar Corporation, California & Hawaiian Sugar Company (C&H), Sucrest Corporation, and National Sugar Refining Company — filed petitions for review 2 of the new “guidelines for effluent limitations,” only C&H’s petition is still before us. 3 C&H challenges the EPA’s action as arbitrary and capricious in various respects, and urges us to vacate the regulations and remand for further consideration by the agency. For the reasons set forth below, we conclude that the agency acted reasonably. We therefore uphold the regulations.

Crystalline Cane Sugar Refining

Refining is the final step in the transformation of sugar cane into pure, white crystalline sucrose. 4 Before refining can begin, the cane must first be crushed in a cane sugar factory. The resulting liquid is then crystallized, producing raw sugar. This “intermediate product consisting of crystals of high purity covered with a film of low quality syrup” 5 is then shipped to a cane sugar refinery. The raw sugar unlike its refined descendant, sucrose, is not considered to be a foodstuff, and can therefore be transported without expensive sanitary safeguards. For this reason, refineries tend to be located in heavily populated areas, close to the retail markets. 6

The refining process consists of four basic steps: (1) washing the raw sugar crystals; (2) adding water to the crystals to form a solution; (3) clarifying and decolorizing the solution; and (4) recrystallizing and finishing the sucrose. Two types of waste water result — process water and condenser water. Process water includes all waste waters from the refining process except condenser water. The primary component of process water is water used in decolorization, but it also includes such miscellaneous streams as floor and filter washings. The principal pollutants in process water are suspended solids (TSS), Biochemical Oxygen Demand (BOD), 7 and pH, a factor of either acidity or *283 alkalinity of the discharge. 8 Condenser water is used to condense the water vapors that boil out of the sugar solution during the recrystallization step. The vapor that mixes with the condenser water contains sugar (in an amount the refiner desires to minimize for business reasons), which contributes substantially to BOD. Condenser water is by volume the greatest portion of the waste from cane sugar refining, but process water has a higher concentration of pollutants.

Statutory Regulation

Disposal of the waste water is governed by the Federal Water Pollution Control Act, as amended, 33 U.S.C. § 1251 et seq. (Supp. IV 1974). Since its amendment in 1972, Pub.L. No. 92-500, the Act has focused attention on the discharges from each “point source” of pollutants rather than on the pollutant levels in the public waterways themselves. All discharges are prohibited, under § 301 of the Act, 33 U.S.C. § 1311, unless authorized by a permit issued pursuant to another section of the Act. The permit section relevant here, § 402, 33 U.S.C. § 1342, directs the EPA to transfer authority for issuance of permits to the states as soon as they develop programs that meet the Act’s requirements. Section 301(b)(1)(A) requires that by July 1, 1977, each point source must comply with effluent limitations fixed by the EPA on the basis of the “best practicable control technology currently available” (BPT). By 1983, the sources must meet a different, presumably stiffer, standard based on the “best available technology economically achievable” (BAT). In setting the standards, the EPA is directed to consider six factors, all but one of which are phrased in identical terms for both BPT and BAT: age of equipment and facilities involved, the process employed, the engineering aspects of the application of various types of control techniques, process changes, and non-water quality environmental impact (including energy demands). The remaining factor involves costs, and the phrasing for the two standards differs. For BPT, the 1977 standard, the Act refers the EPA to “total cost of application of technology in relation to the effluent reduction benefits to be achieved by such application . ..” Section 304(b)(1)(B) of the Act, 33 U.S.C. § 1314(b)(1)(B). For BAT, the 1983 standard, the Act mandates consideration of “the cost of achieving such effluent reduction.” Section 304(b)(2)(B).

Administrative Action

Charged with establishing effluent limitations for all industrial subcategories, the EPA issued an “Advance Notice of Public Review Procedures.” 38 Fed.Reg. 21202 (1973). In accordance with that methodology, the agency had contracted with a private consultant to study waste water flows, pollutant constituents, and treatment technologies for cane sugar refining. The firm submitted a “Draft Development Document *284 for Effluent Limitations Guidelines and New Source Performance Standards” for the cane sugar processing industry, which the EPA published for public comment in July 1973. The EPA twice supplemented the draft, and received a number of comments, including one from the United States Cane Sugar Refining Association (USCSRA). In the following December, the agency published its proposed guidelines. 38 Fed.Reg. 33846 (1973). The proposal limited BOD effluents to 30 milligrams per liter (mg/1) and limited TSS concentration to 40 mg/1. 9 The agency proposed requiring reduction of these concentrations by 1983 to 18 mg/1 and 15 mg/1, respectively. A second Development Document accompanied the proposed guidelines, and recommended that the 1977 standard could be met by “activated sludge” treatment plants. In fixing the 1983 standard, the agency contemplated, among other things, construction of a new system employing cooling towers and sand filtration. The agency again received various public comments, including another from the USCSRA and one from C&H itself. After considering these comments, the agency on March 20, 1974 promulgated final regulations, 39 Fed.Reg. 10522 (1974), and issued a final Development Document. 10 The agency relaxed the 1977 standards for both BOD and TSS to 60 mg/1, and similarly changed the standard that would take effect in 1983 for BOD to 40 mg/1 from the 18 mg/1 level originally proposed.

C&H does not claim any procedural defect in this administrative proceeding. In C&H’s view, however, various agency determinations were arbitrary and capricious in light of all the evidence in the record. We will proceed to consideration of these challenges, but first we note that C&H questions the jurisdiction of this court. It argues that the regulations at issue here are authorized by § 304 of the Act, and are not regulations issued under § 301. Section 304 guidelines are not expressly made directly reviewable in the Court of Appeals by § 509(b), as are § 301 regulations. We rejected this argument in Hooker Chemicals & Plastics Corp. v. Train, 537 F.2d 620, 624-29 (2d Cir. 1976), and the Supreme Court has recently confirmed this position. E. I. duPont de Nemours & Co. v. Train, - U.S. -, 97 S.Ct. 965, 51 L.Ed.2d 204 (1977).

C&H challenges the regulations in four respects: It argues (1) the agency created an impermissibly inflexible regulatory scheme by imposing specific number limitations, rather than defining a range of limitations, and by failing to specify factors to be considered for individual permit applications; (2) the EPA acted arbitrarily and capriciously in concluding that certain treatment technology could be borrowed from other industries to enable the cane sugar refiners to comply with both the 1977 and the 1983 standards; (3) the agency failed adequately to consider the adverse environmental effects of the recommended effluent treatment; and (4) the high costs of the required measures are not justified by the environmental benefits achieved.

Flexibility of the Regulations

C&H contends that the agency should not have fixed precise numerical limitations for the discharge of pollutants from every point source. According to C&H, the Act’s authors intended the agency to define a range of permissible discharge levels, so that the limit for each individual plant could be set after consideration of the various factors enumerated in § 304. Application of this flexible approach would eventually be entrusted to the states, thus facili *285 tating further accommodation of the different situations of the refineries. C&H tells us that the uncertainty associated with transplanting technology from other industries makes this flexibility especially important.

This argument cannot stand in the face of our decision in Hooker Chemicals & Plastic Corp. v. Train, supra. In that case, chemical companies challenged effluent limitation guidelines for the phosphate-manufacturing industry on various bases, including failure to establish ranges and to specify factors to be considered for point sources in each category and subcategory. We concluded:

[W]henever Congress spoke of “ranges” in the debates over the Act, it meant only the spectrum comprised of varying discharge levels on a subcategorical, rather than individual, basis. [E. I. dupont de Nemours & Co. v. Train, 4 Cir., 541 F.2d 1018] de Nemours II, supra at 1029. Although variances are conceivable at the permit-granting stage (see our accompanying opinion Natural Resources Defense Council v. Environmental Protection Agency, 2 Cir., 537 F.2d 642), Congress intended that the regulations establish a single discharge level for a given subcategory. This is implicit in the Congressional choice of the superlative form in the statutory language requiring achievement of the degree of effluent reduction attainable by application of “best" technology.

537 F.2d at 630. The Supreme Court, in E. I. duPont de Nemours & Co. v. Train, supra, recently considered this question, and in response to an argument based on the same language that C&H cites in the legislative history of the Act, said:

If construed to be consistent with the legislative history we have already discussed, and with what we have found to be the clear statutory language, this language can be fairly read to allow the use of subcategories based on factors such as size, age, and unit processes, with effluent limitations for each subcategory normally based on the performance of the best plants in that subcategory.

- U.S. at -, n.21, 97 S.Ct. at 976.

Thus, the EPA’s use of specific number limitations for the whole subcategory is permissible if accomplished on the basis of the appropriate factors. And the regulations as published recite the fact that

[i]n establishing the limitations set forth in this section, EPA took into account all information it was able to collect, develop and solicit with respect to factors (such as age and size of plant, raw materials, manufacturing processes, products produced, treatment technology available, energy requirements and costs) which can affect the industry subcategorization and effluent levels established.

39 Fed.Reg. 10525 (1974). The Development Document bears out this assertion. Furthermore, the agency expressly adverted to the possibility of adjusting the limitations for individual plants with “fundamentally different” attributes. Id. We therefore conclude that the agency properly fixed precise limitations on the discharge of pollutants, took sufficient account of the factors prescribed in the Act and allowed adequate tailoring of the limitations to peculiarities of specific point sources.

Transferability of Technology

The EPA established BOD and TSS limitations for both 1977 (BPT) and 1983 (BAT) on the basis of its determination that treatment techniques employed by other industries could work effectively on sugar refining waste. Neither the Act nor the agency requires the use of any particular treatment method at any point source; the regulations speak only of the quantity and quality of pollutants discharged. But the stringent limitations require substantial reduction of the discharges produced by the present refining process. To accomplish this, the EPA recommended that for the 1977 standard the refiners treat the process water stream by exposing it to active biota while it undergoes aeration. The EPA felt that this “activated sludge” process was “currently available” even though no cane sugar refiner employs it. The agency explained:

*286 However, the technology itself is widely available and practiced in other industries with similar raw waste characteristics— for example, the grain milling and the citrus and potato industries. There are no characteristics of the refinery waste waters that would render them unbeatable by the biological treatment system described.

39 Fed.Reg. 10522 (1974). The Development Document stated that the experiences of the beet sugar and fruit and vegetable industries contributed additional support for the EPA’s determination.

Congress clearly intended the concept of BPT to encompass transferred technology. See Congressional Research Service, A Legislative History of the Water Pollution Control Act Amendments of 1972, at 169 — 70 (1973). Restricting the EPA to treatment technology already in use in a particular industry would only insulate industrial categories in which present practices might be uniformly inadequate. C&H argues, however, that the efficacy of activated sludge treatment for cane sugar refining waste is so uncertain that it cannot be considered for the 1977 standard. C&H points out that the Development Document refers to the agency’s “feelings” and assumptions based on other industries. C&H also argues that even though the EPA analogized to beet sugar waste processing, most beet sugar wastes are not common to cane sugar refineries and the EPA’s own study conceded difficulty in assessing the efficacy of the treatment of beet sugar wastes. 11 Similarly, C&H challenges the EPA’s reliance on several other industries. 12 Finally, C&H challenges the EPA’s reliance on the experience of municipal treatment systems in processing waste from cane sugar refineries. Refinery waste is but a small percentage of the total volume they process, so that special problems that may arise in connection with sugar wastes are not present.

On the same basis, C&H also attacks the EPA’s recommendation that the 1983 (BAT) standard be met by construction of a sand filtration system to treat further the effluent from the biological treatment plant. Sand filtration has not been tested in a cane sugar refining context, but the agency asserted that sand filtration had worked well in the oil refining, grain milling, and soap and detergent industries. To be economical, the system depends on continuous operation of the filters at reasonable flow rates. But if the waste stream contains too high a concentration of suspended matter, the filters quickly become blocked. C&H says this problem may be particularly acute with cane sugar refining wastes, and points out that even the agency’s Development Document conceded that the dimensions of these filtration problems were unknown because of “uncertainty at present of the ratio of soluble to insoluble BOD in the effluent from the biological treatment system.” 13 Nor does the EPA’s reference to the grain milling industry buttress the case for sand filtration, according to C&H, because the EPA has admitted that sand filtration had not been proven in that industry. 14

In the light of this alleged insufficient support for the EPA’s conclusion that the transferred technology will enable cane sugar refiners to comply with the BPT and BAT limitations, C&H urges us to vacate the regulations as we did with other regulations in Hooker Chemicals & Plastics Corp. v. Train, supra, and as other circuits have done. See Tanners’ Council v. Train, 540 F.2d 1188 (4th Cir. 1976); FMC Corp. v. Train, 539 F.2d 973 (4th Cir. 1976); American Iron & Steel Institute v. Train, 526 F.2d *287 1027 (3d Cir. 1975); American Meat Institute v. EPA, 526 F.2d 442, 463 (7th Cir. 1975). The Fourth Circuit in FMC found only conclusory statements from the EPA in the record, and noted that the Court of Appeals “cannot decide such questions on blind faith.” 539 F.2d at 985. The same court in Tanners’ Council made clear that while transfer technology is acceptable in principle, more analysis of the transfer is required of the agency than mere arbitrary adjustment of discharge limitations. In that case, the agency relied on experience of the meat-packing industry to establish limitations for leather tanning. When the EPA learned that tanning discharges differed from meat-packing discharges in an important respect, it simply raised the pollutant limit for tanners. The court rejected this quick solution as “guesswork.” Our decision in Hooker similarly required the agency to act only with a firm foundation set out in the record for our review. We said:

But even if technology which is not presently in use can be treated as available and achievable, there must be some indication in the administrative record of the reasons for concluding that such technology is feasible and may reasonably be expected to yield the effluent reduction mandated when applied to the particular industry. American Meat Institute, supra, at 463; de Nemours II, supra, at 1032. Agencies must present sufficient support for their conclusions in order that the Court can properly review their decisions.

537 F.2d at 636.

With this background in mind, we have looked closely at the EPA’s basis for concluding that activated sludge treatment and sand filtration are transferable to cane sugar refining. We find adequate support for the agency’s determination. The record indicates that the EPA collected data on the use of biological treatment in various industries with high carbonaceous wastes similar to cane sugar refining wastes, in municipal treatment systems that receive cane sugar process water, in combined cane sugar factory-refineries, and in bench and pilot scale tests. While it may be that none of these completely duplicated the actual conditions in a cane sugar refinery, they were all relevant sources for the EPA to consult for information.

The factory-refineries and the municipal facilities indicated that cane sugar wastes could respond well to biological treatment. Although the factory-refineries did not use activated sludge, their treatment methods relied on the same natural biological processes. And although the municipal facilities had not had to deal with high concentrations, it appeared that the systems had dealt effectively with “other types of high strength organic wastes.”

To appraise the efficacy of activated sludge treatment in the industrial context, the EPA looked to the various industries whose discharges resembled cane sugar refining discharges in important respects. It is true, as C&H argues, that some beet sugar wastes do not appear in cane sugar waste streams. These are usually either recovered as byproducts, disposed of on land, or screened or sedimented out of the waste streams; in any case, they do not undergo activated sludge treatment. Nevertheless, the beet sugar streams that are treated with activated sludge have dissolved sugars as their main organic constituents, and are thus quite useful for comparison to cane sugar refining wastes. As to C&H’s point, see note 11 supra, that the EPA has not recommended activated sludge treatment for the beet sugar processors, the EPA explains that this is because other methods that work even better are feasible for beet sugar processors. Since most of that group are located in rural areas, disposal on land is more economical and better suited for their highly seasonal business. The EPA also looked to apple and citrus fruit processing plants, because their wastes included natural sugars, and to potato processing plants because high quantities of carbohydrates appear in their waste streams. The agency also drew support from the dairy industry, whose wastes C&H claims are dissimilar to cane sugar refining *288 wastes. But the record indicates that dairy processing wastes are sufficiently similar to justify the comparison, so experience with them is at least relevant.

These numerous sources of data provided sufficient support for the agency’s decision on the transferability of activated sludge technology. 15 And it appears that the EPA acted reasonably in setting the particular numerical limits contained in the guidelines. For example, in answer to C&H’s charge that greater efficiency of treatment is demanded of cane sugar refiners than activated sludge treatment has produced for the apple and citrus processors (84-94 percent), see note 12, supra, the EPA says that its regulations actually require only 76.8 percent efficiency if both process water and condenser water are considered, or 89 percent for process water alone. The discrepancy between these figures and C&H’s 95 percent figure apparently results from the unusually high BOD concentration in C&H process water. The agency based its computation on average industry figures, which is entirely appropriate. In fact, in Hooker Chemicals & Plastics Corp. v. Train, supra, we approved averaging of figures taken only from “exemplary” plants. 537 F.2d at 632. See American Meat Institute v. EPA, supra, 526 F.2d at 453; American Iron & Steel Institute v. Train, supra, 526 F.2d at 1057. C&H also argues that the EPA’s relaxation of the standards for BOD and TSS in the final guidelines demonstrated that the figures selected were arbitrary. The EPA acted after receiving comments that the standards were too rigid, and explained the increase in permissible levels of pollutants in the final regulations, as compared with the earlier proposed levels, as taking “into account operational problems,” 39 Fed.Reg. 10524 (1974), that may attend adapting the technology to a new industry. We believe that this approach was not arbitrary. The agency’s action in Tanners’ Council was quite different; there, the agency made no reference to information from the tanning industry “or a comparable industry.” 540 F.2d at 1194 n.14.

Turning to the sand filtration system recommended for use by 1983, we also find that the EPA acted reasonably on the basis of the record. Implicit in the distinction between BPT and BAT is Congress’ faith that technological progress can turn present hopes into effective waste treatment plants by 1983. We must therefore allow the EPA considerable latitude in extrapolating from today’s technology. See Hooker Chemicals & Plastics Corp. v. Train, supra, 537 F.2d at 634. C&H plans to use sand filtration on the effluent from its biological treatment plant currently under construction. Moreover, other comparable industries now use sand filtration. 16 These indications of the system’s usefulness and practicability adequately support the EPA’s determination that by 1983 it can be used for cane sugar refining wastes. The EPA calculated the BAT standard on the assumption that sand filtration would reduce TSS levels, but the agency did not credit sand filtration with any reduction in BOD levels. Thus, the uncertainty C&H raises about the ratio of soluble to insoluble BOD, see note 13, supra, and accompanying text, is no reason to suspect the validity of the BAT standards.

In sum, we conclude that the EPA did not act arbitrarily in determining that activated sludge, treatment and sand filtration technology could be transferred. We do not have here the sorts of infirmities in the agency’s analysis that have led to remands in other cases. E. g., FMC Corp. v. Train, supra, 539 F.2d at 980-82 (unsupported as *289 sumptions of waste water flow); Hooker Chemicals & Plastics Corp. v. Train, supra, 537 F.2d at 633-35 (complete failure to take account of relevant factors) and at 636 (complete lack of explanation in record); American Meat Institute v. EPA, supra, 526 F.2d at 459-60 (groundless disregard of relevant data).

Other Environmental Effects

C&H complains that the EPA failed to account for a number of non-water quality environmental problems that will flow from the new regulations. The activated sludge treatment plants will require additional land, possibly a problem for refineries in urban areas. Moreover, disposal of the resulting bacterial sludge as landfill may not endear the refiners to local public health authorities. The EPA, however, did consider that the many activated sludge treatment plants already in industrial and municipal use are able to dispose of their sludge without undue difficulty, and an agency pilot scale test indicated that drying the sludge was possible. C&H also complains of the agency’s “cavalier” consideration of energy consumption required by the new regulations. We see no merit to this argument on the record of this proceeding. Finally, C&H informs us that the cooling towers may cause serious fogging and noise and that the fog may imperil traffic on roads near the refineries. The EPA correctly points out that the comments it received on this point were based more on speculation than facts. Nonetheless, the agency acknowledged that these problems might develop. Development Document at 142. The record contains no evidence that any of the five refineries presently using cooling towers has encountered them, however. Moreover, the agency took into consideration the likelihood that design modification and careful placement could abate the noise and fogging problems. Finally, C&H argues in its reply brief that some other approach may efficiently reduce pollutant levels without the need for cooling towers; if so, C&H is free to use that approach. Under these circumstances, the EPA’s determination was not arbitrary.

Costs

As indicated above, § 304 directs the EPA to consider the cost of treatment technologies recommended for use as BPT and BAT. C&H complains that in doing so the EPA used 1971 cost data, understated the cost of capital and land, and overstated industry income, investment and cash flow, with the result that the EPA’s cost estimate was one-third of the real figure for BPT, and one-half the real price of BAT. C&H also emphasizes that the EPA should not just compute the price of the new treatment systems, but also must ascertain the specific resulting environmental benefits. Instead, C&H says, the agency merely parroted the boilerplate statement that “[i]t is not feasible to quantify in economic terms . the costs resulting from the discharge of these pollutants to our Nation’s Waterways.” 39 Fed.Reg. 10524 (1974). According to C&H, the expensive cooling towers will treat water containing only “minute” concentrations of BOD, which the EPA has not shown to threaten aquatic life or suitability for human use.

The EPA, however, need not document specifically the benefits to society from the curtailment of pollutants from a particular point source. Congress has established as a national goal the complete elimination of pollutant discharges by 1985. 33 U.S.C. § 1251(a)(1) (Supp. IV 1974). The EPA must lead industry toward that goal through the 1977 and 1983 standards, and the agency’s discretion is necessarily broad. See FMC Corp. v. Train, supra, 539 F.2d at 978-79. In its consideration in this proceeding, the EPA apparently did use the most recent cost information available, and focused its economic analysis narrowly on cane sugar refining, instead of the entire food processing industry, which C&H apparently used in reaching its higher cost estimates. As to C&H’s numerous other charges of specific inaccuracies in the EPA cost figures, the agency has presented data to support its estimates. We believe that the EPA estimates fall within the realm of reason on the basis of information in the *290 record. See id. at 979. In short, this is not a case like Hooker, upon which C&H relies, where there was an “absence of any practical consideration of costs.” 537 F.2d at 635.

We have considered all of C&H’s arguments, and we uphold the effluent limitations guidelines as promulgated by the EPA. The petition for review is denied.

Notes

. The Federal Water Pollution Control Act Amendments of 1972 refer to the Administrator of the Environmental Protection Agency as the official responsible for issuing the guidelines. 33 U.S.C. §§ 1251(d), 1314(b) (Supp. IV 1974). He acts through the agency, which is the respondent here and has filed the brief in this court. We will therefore refer hereafter only to the EPA or the agency.

. C&H filed its petition originally in the Ninth Circuit, and National Sugar Refining filed in the Third Circuit. Amstar and Sucrest both filed in this court. The first two petitions were transferred and the four petitions were subsequently consolidated.

. National Sugar Refining, Amstar, and Sucrest have stipulated to dismissals without prejudice of their petitions, the latter two only after oral argument in this court. The EPA agreed to allow Amstar and Sucrest to submit additional information to the agency for consideration in connection with its ongoing review of the effluent limitations guidelines.

. Cane sugar is also refined into liquid sugar. Two refineries in the United States produce both liquid and crystalline sugar, and five others produce liquid sugar only. Separate effluent limitations guidelines have been promulgated for the liquid cane sugar refining subcategory, 40 C.F.R. § 409.30-.36, and have not been challenged.

. EPA, Development Document for Effluent Limitations Guidelines and New Source Performance Standards for the Cane Sugar Refining Segment of the Sugar Processing Point Source Category, at 170 (1974) (hereinafter referred to as Development Document).

. C&H operates two cane sugar refineries, one near Honolulu, Hawaii, and the other in the San Francisco Bay area.

. Biochemical Oxygen Demand is not strictly speaking a pollutant at all. It is a measure of the oxygen consuming capabilities of organic matter. The undesirability of discharges with *283 high oxygen demands are explained in the Development Document at 67-68, as follows:

The BOD does not in itself cause direct harm to a water system, but the matter which it measures may exert an indirect effect by depressing the oxygen content of the water. Sewage and other organic effluents during their processes of decomposition exert a BOD, which can have a catastrophic effect on the ecosystem by depleting the oxygen supply. Conditions are reached frequently where all of the oxygen is used and the continuing decay process causes the production of noxious gases such as hydrogen sulfide and methane. Water with a high BOD indicates the presence of decomposing organic matter and subsequent high bacterial counts that degrade its quality and potential uses. Dissolved oxygen (DO) is a water quality constituent that, in appropriate concentrations, is essential not only to keep organisms living but also to sustain species reproduction, vigor, and the development of populations. Organisms undergo stress at reduced DO concentrations that make them less competitive and able to sustain their species within the aquatic environment. . If a high BOD is present, the quality of the water is usually visually degraded by the presence of decomposing materials and algae blooms due to the uptake of degraded materials that form the foodstuffs of the algal populations.

. The guidelines set pH limits for discharges, but those limits have not been questioned in the petition before us.

. The effluent limitations are actually expressed in terms of pounds of pollutants discharged per ton of sugar melt processed. Thus, what is a permissible level of pollutants in the final effluent from a refinery will vary from plant to plant depending on water usage. The figures cited in the text are the concentrations arrived at for plants with average water consumption.

. See note 5, supra.

. C&H points out that the EPA has not recommended activated sludge treatment for beet sugar wastes.

. Specifically, C&H argues that the biological treatment of com wet milling wastes has failed to keep pollutants consistently below the required limits. With respect to fruit and vegetable processing, C&H contends that the BOD removal level required is only 84-94 percent, whereas the cane sugar refiners are required to remove 95 percent or more.

. Development Document at 152.

. See CPC International, Inc. v. Train, 515 F.2d 1032, 1047 (8th Cir. 1975).

. The EPA has also directed our attention to post-promulgation developments that corroborate the agency’s reasonableness. Plants now operating in South Africa and Japan apparently are effectively employing biological treatment of cane sugar wastes. Although we have not relied on these developments, we note that the Tenth Circuit has permitted such confirmation. Cf. American Petroleum Institute v. Train, 540 F.2d 1023, 1034 (10th Cir. 1976).

. After remand in CPC International, Inc. v. Train, 515 F.2d 1032 (8th Cir. 1975), the Eighth Circuit accepted an EPA recommendation for BOD removal based on evidence that sand filtration was working well in an existing corn wet milling plant. CPC International, Inc. v. Train, 540 F.2d 1329, 1334-35 (8th Cir. 1976). See note 14, supra, and accompanying text.